Quantum Field Theory A Diagrammatic Approach 1St Edition Ronald Kleiss Online Ebook Texxtbook Full Chapter PDF

Quantum Field Theory A Diagrammatic
Approach 1st Edition Ronald Kleiss

Visit to download the full and correct content document:
https://ebookmeta.com/product/quantum-field-theory-a-diagrammatic-approach-1st-e
dition-ronald-kleiss/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Quantum Field Theory A Diagrammatic Approach 2019 draft

2019 Draft Edition Ronald Kleiss
https://ebookmeta.com/product/quantum-field-theory-a-
diagrammatic-approach-2019-draft-2019-draft-edition-ronald-
kleiss/
Statistical Approach to Quantum Field Theory Andreas

Wipf
https://ebookmeta.com/product/statistical-approach-to-quantum-
field-theory-andreas-wipf/
Statistical Approach to Quantum Field Theory 2nd

Edition Andreas Wipf
field-theory-2nd-edition-andreas-wipf/
Statistical Approach to Quantum Field Theory An

Introduction 2nd Edition Andreas Wipf
field-theory-an-introduction-2nd-edition-andreas-wipf/
Nonequilibrium Quantum Field Theory 2nd Edition Esteban
A. Calzetta
https://ebookmeta.com/product/nonequilibrium-quantum-field-
theory-2nd-edition-esteban-a-calzetta/
Relativistic Quantum Field Theory Michael T. Strickland
https://ebookmeta.com/product/relativistic-quantum-field-theory-
michael-t-strickland/
Quantum Field Theory 2nd Edition Michael V. Sadovskii
https://ebookmeta.com/product/quantum-field-theory-2nd-edition-
michael-v-sadovskii/
Quantum Field Theory An Introduction 1st Edition Gordon

Walter Semenoff
https://ebookmeta.com/product/quantum-field-theory-an-
introduction-1st-edition-gordon-walter-semenoff/
Factorization Algebras in Quantum Field Theory 1st

Edition Kevin Costello
https://ebookmeta.com/product/factorization-algebras-in-quantum-
field-theory-1st-edition-kevin-costello/
Quantum Field Theory
Quantum Field Theory (QFT), the language of particle physics, is crucial to a physicist’s
graduate education. Based on lecture notes for courses taught for many years at Radboud
University in the Netherlands, this book presents an alternative approach to teaching QFT
using Feynman diagrams. A diagrammatic approach to understanding QFT exposes young
physicists to an orthogonal introduction to the theory, bringing new ways to understand
challenges in the field. Diagrammatic techniques using Feynman diagrams are used didac-
tically, starting from simple discussions in lower dimensions to more complex topics in
the Standard Model. Worked-out examples and exercises help the reader develop a deep
understanding and intuition that enhance their problem-solving skills and understanding of
QFT. Classroom tested, this accessible textbook is valuable for physics graduates and for
researchers.
Ronald Kleiss has been a CERN staff member and is a full professor at Radboud University
in the Netherlands. Working in particle physics for more than 40 years, he was one of the
first theorists to develop Monte Carlo event generators for high-energy experiments.
Quantum Field Theory
A Diagrammatic Approach
Ronald Kleiss
Radboud 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 University’s 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/9781108486217
DOI: 10.1017/9781108665209
© Cambridge University Press 2021
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 2021
Printed in the United Kingdom by TJ Books Limited, Padstow Cornwall
A catalogue record for this publication is available from the British Library.
Library of Congress Cataloging-in-Publication Data
Names: Kleiss, Ronald, author.
Title: Quantum field theory : a diagrammatic approach / Ronald Kleiss.
Description: New York : Cambridge University Press, 2021. | Includes
bibliographical references and index.
Identifiers: LCCN 2020056956 (print) | LCCN 2020056957 (ebook) | ISBN
9781108486217 (hardback) | ISBN 9781108665209 (ebook)
Subjects: LCSH: Quantum field theory.
Classification: LCC QC174.45 .K58 2021 (print) | LCC QC174.45 (ebook) |
DDC 530.14/3–dc23
LC record available at https://lccn.loc.gov/2020056956
LC ebook record available at https://lccn.loc.gov/2020056957
ISBN 978-1-108-48621-7 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.
The Ancients were wont to draw Diagrams & thus divine Predictions for future Happen-
ings, by Arts magickal or conjectural. . . . In like Wise will the Savants of Times Future
learn to employ Diagrams, yet not by Arts magickal, but rather by the Arithmetick & Arts
algebraickal & by Geometrie and the Quadrature will they seek to foretell the Events of
Nature.
Inspired by Šimon Partlic

(Trest, 1590 – Trencin, 1640)
Astronomer and Mathematician
Contents
Preface page xiii
1 QFT in Zero Dimensions 1

1.1 Introduction 1
1.2 Probabilistic Considerations 1
1.3 Diagrammatics 7
1.4 Synopsis of Feynman Rules 15
1.5 Exercises 15
2 The Loop Expansion and the Effective Action 19

2.1 Planck’s Constant 19
2.2 The Effective Action 23
2.3 The Wetterich Equation 29
2.5 Exercises 31
3 On Renormalization 36
3.1 Doing Physics: Mentality Against Reality 36
3.2 A Handle on Loop Divergences 39
3.3 Scale Dependence 42
3.4 The Method of Counterterms 48
3.5 Asymptotics of Renormalization 49
3.6 Exercises 53
4 More Fields in Zero Dimensions 56

4.1 The Action and the Path Integral 56
4.2 Connected Green’s Functions and Field Functions 57
4.3 The Schwinger–Dyson Equations 57
4.4 The Sum Rules Revisited 58
4.5 A Zero-Dimensional Toy for QED 59
4.6 The Wetterich Equation Revisited 62
4.7 Exercises 62
5 QFT in Euclidean Spaces 66

5.1 One-Dimensional Discrete Theory 66
vii
viii Contents
5.2 One-Dimensional Continuum Theory 70

5.3 The Momentum Representation 76
5.4 Doing It in Momentum Space 77
5.5 More-Dimensional Theories 79
5.7 Exercises 82
6 QFT in Minkowski Space 85

6.1 Marking Time 85
6.2 Moving in Minkowski Space: Particles 93
6.4 Exercises 102
7 Scattering Processes 105

7.1 Introduction 105
7.2 Incursion into the Scattering Process 105
7.3 How to Build Predictions 108
7.4 Unitarity Issues 114
7.5 Some Example Calculations 120
7.6 Calculations at the Threshold 124
7.7 Unitary Amplitudes and Resonances 127
7.9 Exercises 131
8 Introduction to Loop Calculations 134

8.1 The One-Loop Cook Book 134
8.2 Dispersion Relations 140
8.3 Exercises 143
9 More on Renormalization 147

9.1 Renormalizability, or Non-, or Super- 147
9.2 One-Loop Renormalization 152
9.3 The Callan–Symanzik Equation 158
9.4 Exercises 159
10 Dirac Particles 162

10.1 Pimp That Propagator 162
10.2 The Dirac Algebra 164
10.3 Dirac Spinors 174
10.4 Dirac Particles 180
10.5 The Fermi Sign for Dirac Particles 188
10.6 The Dirac Equation 191
10.8 Exercises 193
ix Contents
11 Helicity Techniques for Dirac Particles 198

11.1 The Standard Form for Spinors 198
11.2 Summary of Tools for Spinor Techniques 203
11.3 Fermionic Decays: The Fermi Model 203
11.4 Charged Pion Decay 209
11.5 Exercises 210
12 Vector Particles 213

12.1 Massive Vector Particles 213
12.2 The Spin-Statistics Theorem 218
12.3 Massless Vector Particles 220
12.5 Exercises 225
13 Quantum Electrodynamics 228

13.1 Introduction 228
13.2 Constructing QED 228
13.3 Some QED Processes 238
13.4 Scalar Electrodynamics 253
13.5 The Coulomb Potential 257
13.6 Electrons in External Fields: g = 2 258
13.7 Selected Topics in QED 261
13.8 Synopsis of the Vertices for QED 270
13.9 Exercises 271
14 Higher-Order Effects in QED 279

14.1 The Photon Self-Energy 279
14.2 The Fermion Self-Energy 288
14.3 Soft Bremsstrahlung 293
14.4 The Vertex Correction 304
14.5 The Infrared Sector 307
14.6 The Nonrelativistic Limit: g − 2 Revisited 309
14.7 The Running Coupling for QED 310
14.8 Light-by-Light: Fear, Loathing, and Relief 313
14.9 Exercises 316
15 Quantum Chromodynamics 321

15.1 Introduction: Colored Quarks and Gluons 321
15.2 Quarks and Gluons: First Feynman Rules 322
15.3 The Three-Gluon Interaction 326
15.4 The Four-Gluon Interaction 336
15.5 Current Conservation in QCD 338
15.6 Selected Topics in QCD 340
x Contents
15.7 Synopsis of the Vertices for QCD 344

15.8 Exercises 345
16 Higher-Order Effects in QCD 348

16.1 The Feynman Gauge 348
16.2 Picking Propagators 351
16.3 The Gluonic Self-Energy 353
16.4 The Ward Identity for QCD 356
16.5 The Running QCD Coupling 357
16.6 Exercises 360
17 Electroweak Theory 362

17.1 Muon Decay 362
17.2 The W Particle 364
17.3 The Z Particle 371
17.4 The Higgs Sector 379
17.5 Archeology of the Higgs Sector 391
17.6 Running Masses and the Top 392
17.7 About Anomalies 396
17.8 Remarks on What We Did 397
17.9 Toward a Nonminimal Higgs Sector 398
17.10 Quartic Divergences 401
17.11 Toward the Rξ Gauge 406
17.12 Synopsis of the Electroweak Vertices 413
17.13 Exercises 414
18 More Example Computations 419

18.1 Neutrino Production in e+ e- Scattering 419
18.2 W Pair Production in e+ e- Scattering 422
18.3 Higgs Coupling to Massless Vectors 428
18.4 Neutral Pion Decay 432
18.5 “Naturalness” for the Higgs Propagator 434
18.6 Exercises 437
Appendices 439
A Perturbative (Non)convergence Issues 439
B More on Symmetry Factors 442
C Derivation of the Diagrammatic Sum Rules 445
D Scale Dependence for Theories with More Parameters 447
E An Album of Diagrams 449
F Alternative Solutions to the SDe 455
G Diagram Counting 459
H Concavity of the Effective Action 467
I Functional Derivatives 469
xi Contents
J Cruel and Unusual Actions 470

K Newton’s First Law Revisited 474
L Unitarity Bounds 477
M Ambiguities in Pauli–Villars Regularization 480
N The Fundamental Theorem for Dirac Matrices 483
O A Ward Identity for the Three-Gluon Amplitude 485
P States of Higher Spin 487
Q Multiparticle Phase Space 502
R The CPT Theorem 507
S Factorially Divergent Series and Renormalization 511
T Mathematical Miscellanies 520
Bibliography 536
Index 538
Preface
Why This Book?
This book grew out of lectures that I have been giving on quantum field theory (QFT)
for elementary-particle physics over the last 30 years or so. An internet search using the
terms “introduction” and “quantum field theory” yields millions of hits: what could I pos-
sibly add? Only this: that I present the subject in the way that, after all those years, and
with hindsight, pleases me most. Thus this text is a very personal one, and in some ways
my treatment may appear more or less orthogonal to more canonical introductions.1 The
moment I learned about Feynman diagrams I was struck by their power and beauty, and
in this book I use them as much as possible; that is well suited to its goal, to be an intro-
duction into a way that theoretical particle phenomenology can be done. In what follows,
whatever is correct I owe to many other people, but that which is wrong I managed on
my own. Physics is, like all serious science must be, a social undertaking; I have greatly
profited from discussions with many friends and colleagues! Thank you, Ernestos Argyres,
Dima Bardin, Wim Beenakker, Frits Berends, Alain Blondel, Oscar Boher Luna, Michael
Borinsky, Helmut Burkhardt, Dirk van Buul, Sascha Caron, Patrick de Causmaecker, Chris
Dams, Piet Hein Daverveldt, Petros Draggiotis, Helmut Eberl, Steve Ellis, John Ellis,
Frank Filthaut, Karel Gaemers, Raymond Gastmans, Edward Gibbon, Walter Giele, Nigel
Glover, Jeroen de Groot, André van Hameren, Lisa Hartgring, Wolfgang Hollik, Gerard
’t Hooft, Staszek Jadach, Frederick James, Tim Janssen, Sijbrand de Jong, Martijn Jon-
gen, Jos de Heren, Marcel van Kessel, Mila Keijer, Jochem Kip, Stefan Krieg, Hans Kühn,
Hans Kuijf, Kolja Kuijpers, Zoltan Kunszt, Achilleas Lazopoulos, Joep Leenaarts, Yan-
nis Malamos, Michelangelo Mangano, John March-Russell, Melvin Meijer, Ilija Milutin,
Mark Netjes, Harald Niederreiter, Gijs van der Oord, Kostas Papadopoulos, Simon Partlic2 ,
Giampiero Passarino, Roberto Pittau, Marcel Raas, Chris Ripken, Tom Rijken, Frank
Saueressig, Frank Schmidt, Bert Schellekens, James Stirling (much missed!), John Swain,
Oleg Teryaev, Theodor Todorov, Tini Veltman, Rob Verheyen, Jos Vermaseren, Bernard de
Wit, Stefan de Wit, Tai Tsun Wu, and Sjoerd Ypma, for allowing me to learn from you,
and together with you. And my gratitude to the people at Cambridge University Press, who
have allowed me to share what I have learned.
1 Indeed, canonical quantization will not appear at all.

2 Inspirator of various quotations.
xiii
xiv Preface
What Is in This Book?
It may be useful to briefly describe the structure of the book. Quantum field theory is not
trivial, not least because, in its fully fledged form, it uses a lot of technical tools.3 A real
understanding of any field can only grow from handling its tools very often,4 but it is advis-
able to start with the simplest tools first and then to let your skills develop gradually. We
therefore start with QFT in its simplest possible setting, a zero-dimensional universe. Here
we already encounter many of the objects of interest, especially the Feynman diagrams and
the Feynman rules, but in a setting that avoids much of the technicalities. Even the sub-
ject of renormalization can be discussed in zero dimensions, as we shall see. After having
become familiar with the fundamentals we can move toward more realism. We can care-
fully extend our universe, moving from zero to one and then to more dimensions,5 and after
that make a switch to Minkowski space by assigning to time its own special rôle; and then
the way is open to discussing particle dynamics, that is, scattering processes. The Feynman
rules will grow in complexity until we can describe processes involving scalar particles. At
that point we can afford to introduce particles with additional (spin) properties: Dirac par-
ticles and vector particles. The door is then open to the real world of particle physics: we
shall describe quantum electrodynamics, quantum chromodynamics, and finally the elec-
troweak theory. Thus we arrive at what we fondly call the (Minimal) Standard Model.
Throughout, we aim simply to find Feynman rules that work, without deep thoughts about
gauge symmetry.6 For instance, we shall from the outset acknowledge the fact that the W
and Z particles have masses and, rather than pining for a Platonic realm where they are
“really” massless, simply construct a viable theory for them, warts and all. The fact that
our theories turn out to be examples of gauge theories, broken or not, is a bonus in the
sense that it makes clear that you may argue that gauge symmetry is not a fundamental
organizing principle but that our world simply can’t help being that way if it is to function
at all.7 As a consequence, you will not find much “model building” here. In our search for
Feynman rules “that work” we will very often use handlebars. These are operations on
Feynman diagrams8 that allow us to study some of their properties. Putting a handlebar on
diagrams is, here, a purely mechanical operation, leading to things like current conserva-
tion and Ward identities. I think it is very useful to see how they come about from diagrams
rather than from symmetry arguments (which is how they are usually presented).
One of the largest portions of this book is the appendices. I have there collected the
material that, when included in the main text, would interfere with the narrative flow. But
that material is not secondary! The appendices (especially the mathematical one) contain
much that I would like to see at many a theorist’s fingertips.
3 For instance, those necessary for Dirac particles, or the calculational machinery of loop integrals.
4 As everyone who has tried cooking, or open-heart surgery, can attest.
5 And come to the surprising realization that the concepts of space and time may not be so fundamental after all.
6 You will find that I go very lightly on group theory – an example of orthogonality.
7 I am comforted in this opinion by [7].
8 In more technical language, computing the divergence of currents.
xv Preface
Simply reading about QFT is no substitute for actually sitting down and doing it yourself.
This book is intended to encourage you to dirty your hands and compute things,9 and
therefore the text is riddled with exercises, some almost trivial, and others far from trivial.
The exercises apposite to a topic are indicated by a box like ℵ in the margin, where “ℵ” is
0 the number of the exercise, situated at the end of the chapter. Doing the exercises is really
the best way to come to understand the marvellous structure that is QFT.10
Looking at the bibliography, you may be surprised to find how short it is. This is delib-
erate. I have included the necessary minimum (the sources that I actually used to write this
book), finding that in these days of Google, Wikipedia, and InSpire11 it is much more use-
ful for you to find your own way than to stare at a several-hundred-item list of publications.
This book is an introduction, not a comprehensive review.
Finally, a caveat: this is a book for physicists, not for mathematicians. We shall worry
about mathematical rigor but not so much as to become paralyzed.12
Basic Tools
Constants and Units

Classical physics essentially uses three basic SI units: the kilogram, the meter, and the sec-
ond. In addition, there exist fundamental constants of nature. The fundamental constants
of relativistic quantum field theory13 are the speed of light in vacuo,14
m
c = 299792458 , (1)
sec
and Planck’s constant,15
h = 6.62607015 × 10−34 kg m2 /sec. (2)
We shall use the reduced Planck constant (also called the Dirac constant):
= h/(2π) ≈ 1.054571817 × 10−34 kg m2 /sec

= 6.582119569 × 10−25 GeV sec. (3)
Using and c, we can exchange energies for inverse lengths:
9 Manus mundissimus fabricae inutilis (inspired by S. Partlic).

10 “[They] were constantly trained in both the morning and the evening, nor was age or knowledge allowed to
excuse the veterans from the daily repetition of what they had completely learned” (E. Gibbon, The History
of The Decline and Fall of the Roman Empire (W. Strahan and T. Cadell, in the Strand, 1776)).
11 InSpire uses extremely useful backward referencing: find it at https://inspirehep.net.
12 Rigor mathesis protractus mortis rigorem inducit (inspired by S. Partlic). In fact you might be surprised by
how much of mathematics was developed just to make physics more rigorous.
13 Their values are taken from [1]. Throughout this book I will use this as a source for all other empirical data
that I quote.
14 This is exact since it establishes the definition of the meter as of October 21, 1983.
15 This is exact since it establishes the definition of the kilogram as of May 20, 2019.
xvi Preface
c = 197.3269804 MeV fm, (4)

where fm stands for “Fermi” or “femtometer”: 1 fm = 10−15 meter.
Fundamental HEP Units

Compared to the scales of our everyday experiences, is miniscule and c is huge; in the
world of elementary particles, they are just about right. We can see this as follows. It is
useful to replace our human-scale meters,16 kilograms,17 and seconds18 by what may be
called fundamental HEP units of mass, length, and time:19
Mf ≈ 1.78 10−27 kg, Lf ≈ 1.97 10−16 m, Tf ≈ 6.58 10−25 sec. (5)
In terms of these units, we have precisely
= Mf Lf 2 /Tf , c = Lf /Tf , (6)
so that both and c have the numerical value 1; and the unit of energy turns out to be
Mf Lf 2 /Tf 2 = 1 GeV ≈ 1.602176634 10−10 Joule. (7)
The mass and size of the proton are of the same order as Mf and Lf , respectively, and Tf
is roughly the time scale of strong interactions. The use of fundamental units is attractive
since you won’t have to write factors of c and and can express both length and time in
inverse GeV and mass in GeV. This is customarily one of the first statements in advanced
courses: “we shall adopt units such that c = = 1.” This usage denies the various objects
their dimensionality,20 and I have decided to try to retain the s and cs where they belong;
after all, it is much easier to erase them from formulas than to put them back in.21
Planck Units
Along with and c, there exists a third fundamental constant of nature, namely. Newton’s
(or rather Cavendish’s) gravitational constant:22
GN = 6.67430(15) 10−11 m3 /(kg sec2 ). (8)
The truly, ultimately fundamental units of mass, length, and time that can be constructed
from c, , and GN are then the Planck units:
16 Typically, how far you can reach.

17 Typically, what you would not mind lifting.
18 Typically, how fast you can count things if you are not a teller.
19 I give these numbers with only a few digits, for simplicity. What matters is of course not the precise value, but
rather that such units exist.
20 And has the unfortunate consequence of making it impossible to check the (at least) grammatical correctness
of an expression by dimensional analysis. It also leads to erroneous statements about “classical limits” and
the like: see our discussion of particle masses in Chapter 6.
21 You would be surprised how difficult it is for even well-trained physicists to put the and c back, when asked
to do so in a hurry!
22 The uncertainty in the last two digits is bracketed.
xvii Preface

c GN
MP = , LP = , TP = LP /c. (9)
GN c3
These values evaluate to MP ∼ 2.2 10−8 kg, LP ∼ 1.6 10−35 m, and TP ∼ 5.4 10−44 sec,
outrageously far removed from the typical scales of particle phenomenology. The Planck
energy EP = MP c2 ∼ 1.2 1019 GeV tells us the same thing. We may interpret this as an
indication that in what follows the gravitational interaction will not play any part. In fact,
in any case we do not (yet) have a satisfactory quantum theory of gravitation leading to
specific and falsifiable predictions for particle phenomenology. To bring the Planck units
close to the fundamental units we need to increase the value of G, thus the strength of
gravity, by a factor of about 1038 . Let us finally realize that we have, basically, only the
three measures, of length, of time, and of mass, to do physics with; and there are precisely
three constants of nature that we know of. Suppose a fourth one were discovered.23 It
would be only natural that we would try very hard to relate it to the other three in some
way, and this is how fundamental physics must work.
Charges
The electrostatic charge is adopted to the Gaussian system, so as to have no truck with the
“permittivity of the vacuum” and suchlike: that is, two static charges Q1 and Q2 separated
by a distance r feel a mutual Coulomb force F , with
1 | Q1 Q2 |
|=
|F . (10)
4π r2
√
This implies that the charge has the dimensionality of c. It follows that, if we choose
the proton charge as the unit charge e, the fine structure constant
e2
α= (11)
4π c
is a dimensionless number.24 Experimentally,
1/αe = 137.035999084(21), (12)
which yields the result
√
e ≈ 0.303 c ≈ 5.38 10−14 (kg m3 )1/2 /sec. (13)
Conventions
Irrelevant Phases and Factors
If two quantities A and B are equal up to a disregarded overall complex phase or other
factor we shall write
A B or A ∼ B.
23 The cosmological constant, for instance; this requires a factor of about 10120 to be explained away.
24 Meaning that it has the same value in all possible systems of units! An alien civilization in outer space will
find the same value.
xviii Preface
The Step Function

The Whittaker (or step) function is a function of a real number:

1 if x ≥ 1
θ (x) = . (14)
0 if x < 0
We extend this to a logical step function of a statement P:

1 if P is true
θ (P) = . (15)
0 if P is false
The Metric
By convention, the Minkowski metric25 has the form
gμν = gμν = diag(1, −1, −1, −1). (16)
In many textbooks the metric tensor is introduced as a diagonal matrix. This is of course
misleading since the covariant metric tensor has only lower indices, whereas a matrix has
one upper and one lower index. Unfortunately, the “correct” matrix form of the metric,
which would be gμ ν , equals the unit matrix whatever the metric!
Kronecker’s Symbol
The Kronecker symbol is defined by

1 if α = μ,
δα μ = (17)
0 if α = μ.
Note: in Minkowski space, Kronecker symbols tend to carry one upper and one lower
index. Kronecker symbols with two upper or two lower indices are slightly suspect unless
they refer to some other label, such as color.
The Antisymmetrizer
Using Kronecker symbols, we can build the following antisymmetric objects:

μ1 μ1 μ1 μ2
= δ ν1 , = δ μ1 ν1 δ μ2 ν2 − δ μ1 ν2 δ μ2 ν1 ,
ν1 ν1 ν2

μ1 μ2 μ3
= δ μ1 ν1 δ μ2 ν2 δ μ3 ν3 + δ μ1 ν2 δ μ2 ν3 δ μ3 ν1 + δ μ1 ν3 δ μ2 ν1 δ μ3 ν2
ν1 ν2 ν3
− δ μ1 ν1 δ μ2 ν3 δ μ3 ν2 − δ μ1 ν2 δ μ2 ν1 δ μ3 ν3 − δ μ1 ν3 δ μ2 ν2 δ μ3 ν1 ,
(18)
25 In the usual Cartesian coordinate systems. Since we are not considering curved spacetime in this book, we
shall adhere to this simplest of coordinate system throughout, except when discussing phase space integration
where polar coordinates often come in handy. But there we shall not refer to the metric.
xix Preface
and so on. Here we encounter all signed permutations26 of the lower indices. These are
computationally handy as we see below.
The Levi–Civita Symbol

The totally antisymmetric Levi–Civita symbol is defined by
0123 = + 1 hence 0123 = − 1. (19)
This implies the following identities:

ν1 ν2 ν3 ν4 ν1 ν2 ν3 ν4
μ1 μ2 μ3 μ4 =− ,
μ1 μ2 μ3 μ4

ν1 ν2 ν3
μ1 μ2 μ3 μ4 ν1 ν2 ν3 μ4 = − ,
μ1 μ2 μ3

ν1 ν2 μ3 μ4 ν1 ν2
μ1 μ2 μ3 μ4 = −2 ,
μ1 μ2

ν1
μ1 μ2 μ3 μ4 ν1 μ2 μ3 μ4 = −6 = −6 δ ν1 μ1 ,
μ1
μ1 μ2 μ3 μ4 μ1 μ2 μ3 μ4 = −24. (20)
In order not to lumber ourselves with too many explicit Lorentz indices, we shall use the
following shorthand:
μ (a, b, c) means μνρσ aν bρ cσ , and so on.
Finally, although it falls outside our scope here, it is useful to remember that the Kronecker
and antisymmetrizer symbols are fully fledged tensors in a much wider class of spaces
than Minkowski space, but the Levi–Civita symbols themselves are not. For instance, if we
move from Cartesian to polar coordinates, say, the Kronecker symbol remains unaffected
but the Levi–Civita is changed.
Minus the Momentum

Here is a subtlety – the contravariant partial derivative contains a possibly surprising minus
sign:

μ ∂ 1∂
∂ = = , −∇ . (21)
∂xμ c ∂t
This explains why in nonrelativistic quantum mechanics the momentum operator is p =

whereas in the relativistic theory we use pμ = i ∂ μ .
−i ∇,
26 Even permutations occur with a + and odd permutations with a – sign.

xx Preface
The Polyparenthetophobe Rules

μ
This deals with the notation for compound vector products. If p1,2,3,4 are 4-vectors, then
the expression
(p1 + p2 · p3 + p4 )
must be understood to mean
(p1 · p3 ) + (p2 · p3 ) + (p1 · p4 ) + (p2 · p4 )
The more rigorously correct form

(p1 + p2 ) · (p3 + p4 )
is in my opinion less easily readable, unless by true parenthetophiliacs. In the same spirit,
a whiff of periodophobia will lead us to write, e.g., (pq) as shorthand for (p · q) where no
risk of confusion is likely. Also, k2 occurs when it is clear that (k · k) is intended rather than
the second spacelike component of kμ .
In Favor of Loose Terminology

Particle physicists tend to be sloppy with some terms, in particular “mass,” “momentum,”
and “energy.” Strictly speaking, in the Feynman rules that we shall use the “mass” m and
the “momentum” kμ have dimensions of inverse length and therefore cannot be the same as
the notions of the classical mass M, expressed in kg, the classical momentum pμ , expressed
in (kg m/sec), and the classical energy given in joules. As discussed in Section 6.2.5, these
various notions are related by
m = Mc/, kμ = pμ /. (22)
Pedantry is unbecoming, and I shall use “mass” and “momentum” insouciantly, and call k0
the energy. In the same spirit, a “massless” vector will mean the same as a “lightlike” one,
while “massive” and “timelike” are likewise equivalent. You will easily get used to this.
Exercise
Exercise 0 Almost an exercise

This might have been an exercise. If you really want to turn it into one, use the internet to
try to count the number of books and lecture notes that have “Introduction” and “Quantum
Field Theory” in their titles.
1 QFT in Zero Dimensions
1.1 Introduction
It makes no sense to talk about elementary particles without either special relativity
or quantum mechanics. For now, we concentrate on the quantum-mechanical nature of
nature1 . The fundamental object associated with particles is a quantum field.2 Such a field
assigns one or more numbers to every point in spacetime. So it is a pretty complicated
thing; its behavior cannot be described trivially, especially since it also undergoes quantum
fluctuations. It is a good idea to first build up some expertise, in a more controllable situ-
ation. Therefore we shall simplify the whole four-dimensional spacetime arena of particle
physics. We shall reduce spacetime to a single point, a zero-dimensional arena.3 Now we
have only a single point to assign numbers to, and the simplest quantum field is a single
stochastic, or random, number. We can already learn many of the techniques of quantum
field theory studying this simple case! We shall meet path integrals, Green’s functions, the
Schwinger–Dyson equation, Feynman diagrams and the effective action, in a quite natural
way.
1.2 Probabilistic Considerations
1.2.1 Green’s Functions and the Path Integral

Let us imagine a quantum field ϕ that can take on all real values from −∞ to +∞. Since
it is a random variable, the most we can hope to specify about it is its probability density
P(ϕ),4 which we write as5

P(ϕ) = N exp − S(ϕ) , N −1 = exp − S(ϕ) dϕ. (1.1)
1 In a moment you will understand why relativity does not enter – yet.
2 In many treatments quantum fields are considered to be distribution-valued operators. In this book I am not
interested in the internal details of quantum states, but rather in the scattering amplitudes; we shall adopt
Feynman’s approach and use what are called c-number fields.
3 That’s why.
4 This is what it means to be a random variable.
5 If not explicitly indicated otherwise, integrals run from −∞ to +∞.
1
The function S(ϕ) is called the action of the particular quantum field theory; it defines the
theory. For the probability density to be acceptable, S(ϕ) must go to infinity sufficiently fast
as |ϕ| → ∞.6 Since the quantum field is a random variable, the most that can be known
about it7 is the collection of its moments, in the jargon called Green’s functions:8

Gn ≡ ϕ n ≡ N exp − S(ϕ) ϕ n dϕ, n = 0, 1, 2, 3, . . . . (1.2)
We shall assume that Gn exists for all n. By construction, we must always have

G0 = ϕ 0 = 1 = 1. (1.3)
The most fruitful way9 of discussing the set of all Green’s functions is in terms of a
generating function:
1
Z(J) = J n Gn . (1.4)
n!
n≥0
This is called the path integral, for reasons that will become clear later. It can be written as

Z(J) = N dϕ PJ (ϕ), PJ (ϕ) = exp − S(ϕ) + Jϕ . (1.5)
The number J, which here serves purely as a device to distinguish the various Green’s
functions, is called a source, again for reasons that will become apparent later. Once Z(J)
is known, an individual Green’s function is extracted by differentiation:
n
∂
Gn = Z(J) . (1.6)
(∂J)n J=0
The path integral Z(J) contains all the information about the Green’s functions, and hence
about the probability density P(ϕ). The same information is, therefore, also contained in
its logarithm. We write
1
W(J) = log Z(J) ≡ J n Cn . (1.7)
n!
n≥1
6 Barring elaborate unnatural counter examples, of course; see the remarks on paralysis in the introduction.
7 You are here approaching a career decision! You may decide simply to measure the value of ϕ: in that case
you have decided to become an experimentalist rather than a theorist.
8 We need to clear up, in advance, a possible confusion. In this book, the Green’s functions are simply defined
to be expectation values. This may appear to contrast with the use of Green’s functions in the solution of
inhomogeneous linear differential equations such as are encountered in classical electrodynamics, where we
use them to compute the electromagnetic field configurations for given sources. The difference is only apparent
since, as we shall recognize, the latter type of Green’s functions are in our treatment simply the two-point
connected Green’s functions; and for theories such as electrodynamics, where the electromagnetic fields do
not undergo self-interaction, the two-point functions are in fact the only nonzero connected Green’s functions.
Be not, therefore, misled into thinking that there are somehow two sorts of Green’s functions. The Green’s
function formulation of electrodynamics will in fact appear as the classical limit of the Schwinger–Dyson
equation discussed below.
9 Kids! Do this at home. Whenever an infinite collection of objects with some kind of relation between them
occurs, generating functions are always a good idea.
3 Probabilistic Considerations
The quantities Cn (with, obviously C0 = 0 since G0 = 1) are called the connected Green’s
functions of the theory, and will play an important rôle in what follows. The connected
1 Green’s functions can be recognized to be the cumulants of the probability density:
C0 =0 by normalization,
C1 = ϕ the mean,
(1.8)
C2 = (ϕ − ϕ)2 the variance,
C3 = (ϕ − ϕ)3 the skewness,
and so on. Since W(0) = C0 = 0, all information about the probability density is also
contained in its derivative, the field function:
∂ 1
φ(J) ≡ W(J) = J n Cn+1 . (1.9)
∂J n!
n≥0
Since from its definition, we have

−1
φ(J) = dϕ ϕ PJ (ϕ) dϕ PJ (ϕ) , (1.10)
we can say that φ(J) is the expectation value of the quantum field ϕ in the presence of
sources: to denote this, we might write
φ(J) = ϕJ , (1.11)
which explains the similar typographies for the quantum field and the field function.10 We
should not, however, forget the difference in status of these objects! ϕ is the physical entity,
an unknowable, fluctuating random field; but φ(J) is a perfectly well-defined function that
contains all the information about the probability density of ϕ and is computable once the
action is given.11
1.2.2 The Free Theory

The simplest probability density is probably12 the Gaussian one, given by the action
1
μ ϕ2,
S(ϕ) = (1.12)
2
with μ a positive real number. For any action, we shall call the part quadratic in the fields
(or bilinear in the case of several fields) the kinetic part. This action, called the free action,
consists of only a kinetic part. The path integral is now simply computed by

1 2
Z(J) = N exp − μϕ + Jϕ dϕ
2
2
1 J 2 J2 J
=N exp − μ ϕ − + dϕ = exp . (1.13)
2 μ 2μ 2μ
10 Try this out: “phi of J equals phi with J”.

11 In principle, if not in practice easily or completely.
12 A uniform density may be thought even simpler, but then it cannot run from ϕ = −∞ to ϕ = +∞. As a
matter of fact, ask a mathematician or physicist to name you a nice probability density over the whole real
line, and she will almost certainly suggest the Gaussian.
It is not even necessary13 to actually calculate the value of N. By Taylor expansion of the
exponential, we immediately find that
(2n)!
G2n = , G2n+1 = 0, n = 0, 1, 2, . . . . (1.14)
(2 μ)n n!
The connected Green’s functions follow from
J2 J
W(J) = log Z(J) = , φ(J) = , (1.15)
2μ μ
so that the only nonvanishing connected Green’s function is C2 = 1/μ. The fact that here
only the two-point connected Green’s function is nonvanishing is the reason for calling
this model the free theory. Again, things will become clearer later on, in a more realistic
spacetime – after all, what does “freedom” mean if you are confined to a single point?
1.2.3 The ϕ 4 Model and Perturbation Theory

An action S(ϕ) may contain other terms than just the quadratic one. Such terms are called
interaction terms: they may be linear, but more usually they are of higher power in the field
ϕ. The simplest acceptable interacting theory in our probabilistic setting is therefore given
by the action
1 1
S(ϕ) = μϕ 2 + λ4 ϕ 4 . (1.16)
2 4!
The (nonnegative!) real number λ4 is called a coupling constant: this model is called the
2 ϕ 4 theory.14 Computing the path integral is now a much less trivial matter. A possible
approach is to assume that, in some sense, the ϕ 4 theory is close to a free theory, that is, in
the same some sense, λ4 is a small number. We can then expand the probability density in
powers of λ4 :

1 1 λ4 k 4k
exp(−S(ϕ)) = exp − μϕ 2 − ϕ . (1.17)
2 k! 24
k≥0
This procedure is called perturbation theory. Having thus reduced the problem to the pre-
vious case of the free theory, we cavalierly15 interchange the series expansion in λ4 with
the integration over ϕ and arrive at the following expression for the Green’s functions:
G2n = H2n /H0 ,

1 (4k + 2n)! λ4 k
H2n = n − . (1.18)
μ 22k+n (2k + n)! k! 24μ2
k≥0
13 Because we must always have Z(0) = 1.

14 An action in which ϕ 3 is the highest power does not lead to a convergent integral over the real axis (see,
however, Appendix F). Of course, an action of the form S(ϕ) = μϕ 2 /2 + λ3 ϕ 3 /3! +λ4 ϕ 4 /4! is perfectly
acceptable, and we shall consider this “ϕ 3/4 model” later on.
15 This interchange does not come without its price: see Appendix A. Nemo me impune lacessit.
5 Probabilistic Considerations
For example, we have

1 35 2 385 3
H0 = 1 − u + u − u + ···,
8 384 3072
1 29 2 107 3
1/H0 = 1 + u − u + u + ···, (1.19)
8 384 1024
3 with u ≡ λ4 /μ2 . In this theory, also the normalization N has to be treated perturbatively,
which explains the expression for 1/H0 . For the first few nonvanishing Green’s functions
we find
G0 = 1,

1 1 2 11
G2 = 1 − u + u2 − u3 + · · · ,
μ 2 3 8

1 33 2 68 3
G4 = 2 3 − 4u + u − u + · · · ,
μ 4 3

1 75 445 2 1585 3
G6 = 3 15 − u + u − u + ··· . (1.20)
μ 2 4 4
The corresponding nonzero connected Green’s functions are given by

1 1 2 11
C2 = 1 − u + u2 − u3 + · · · ,
μ 2 3 8

1 7 149 3
C4 = 2 −u + u2 − u + ··· ,
μ 2 12
1 2
C6 = 3 10u − 80u3 + · · · . (1.21)
μ
Whereas the Green’s functions all have a perturbation expansion starting with terms con-
taining no λ4 , the connected Green’s functions of increasing order are also of increasingly
high order in λ4 : the higher connected Green’s functions need more interactions than the
lower ones.
1.2.4 The Schwinger–Dyson Equation

Although the path integral is, generally, a very complicated function of J, we can easily
find an equation that describes it completely. This is the Schwinger–Dyson equation (SDe),
which we construct as follows. Let the action be given by the general expression
1
S(ϕ) = λk ϕ k , (1.22)
k!
k≥1
where λ2 = μ.16 Now, from the observation that

∂p
Z(J) = N exp − S(ϕ) + Jϕ ϕ p dϕ, p = 0, 1, 2, 3, . . . , (1.23)
(∂J) p
16 The sum starts at 1 since a constant, ϕ-independent term in the action is always immediately swallowed up by
the normalization factor N.
we immediately deduce that

⎡ ⎤
λk+1 ∂ k
⎣−J + ⎦ Z(J)
k! (∂J)k
k≥0
⎡ ⎤
λk+1
=N exp − S(ϕ) + Jϕ ⎣−J + ϕ k ⎦ dϕ
k!
k≥0

=N exp − S(ϕ) + Jϕ S (ϕ) − J dϕ = 0, (1.24)
where in the last lemma we have recognized a total derivative, and used the fact that the
integrand vanishes at the endpoints. Symbolically, we may write the SDe as

∂ ∂
S(ϕ) Z(J) = S Z(J) = JZ(J). (1.25)
∂ϕ ϕ=∂/∂J ∂J
4 For our sample model, the ϕ 4 theory, the SDe reads17

1
λ4 Z (J) + μZ (J) − JZ(J) = 0. (1.26)
6
Using the series expansion of the path integral, we can express this as a relation between
different Green’s functions:
λ4
Gn+3 + μGn+1 − nGn−1 = 0, n ≥ 1. (1.27)
6
This relation may usefully be rewritten as follows:

1 λ4
Gn = (n − 1)Gn−2 − Gn+2 , n ≥ 2. (1.28)
μ 6
If we start by assigning to the Green’s functions the values Gn = δ0,n , then repeated
applications of Eq. (1.28) will precisely reproduce the Green’s functions of Eq. (1.20).18
1.2.5 The Schwinger–Dyson Equation for the Field Function
From the definition of φ(J) as the derivative of the logarithm of the path integral, we can
infer that

1 ∂p ∂ p
Z(J) = φ(J) + e(J). (1.29)
Z(J) (∂J)p ∂J
Here, e(J) is the unit function: e(J) ≡ 1. We immediately arrive at the form of the SDe for
5 the field function:
17 The SD equation is, in general, of higher than the first order. It therefore has several independent solutions,
only one of which corresponds to the usual perturbative expansion. The nature of the other solutions is
discussed in Appendix F.
18 The correct way to do this is to subsequently evaluate G , G , G , . . .. On the first iteration, the lowest-order
2 4 6
expressions are obtained. Each subsequent iteration gives one higher order in perturbation theory. If we want
to obtain the kth order term in Gn , the (k + 1)th order term in Gn+2 is needed, and so on. It is therefore
necessary to compute the lower-order terms for more Gn s.
7 Diagrammatics

∂
S φ(J) + e(J) = J. (1.30)
∂J
For the ϕ 3/4 theory, it reads

J λ3 ∂
φ(J) = − φ(J)2 + φ(J)
μ 2μ ∂J

λ4 ∂ ∂2
− φ(J)3 + 3φ(J) φ(J) + φ(J) . (1.31)
6μ ∂J (∂J)2
6 Although this leads to very nonlinear relations between the various connected Green’s
functions this form of the SD equation is actually even simpler to apply with φ(J) = 0 as
7 a starting point, iterating the assignment (1.31) then results19 in the correct form of φ(J),
8 giving the connected Green’s functions of Eq. (1.21).
1.3 Diagrammatics
1.3.1 Feynman Diagrams
There exists an extremely useful toolbox for computing Green’s functions and connected
Green’s functions: Feynman diagrams. In this section we shall first introduce these dia-
grams and their concomitant Feynman rules. Only after that shall we prove that these
diagrams do, indeed, correctly describe Green’s functions.
Feynman diagrams are constructs of lines and vertices. A vertex is a meeting point for
one or more lines.20 Diagrams are allowed in which one or more lines do not end in a
vertex but, in a sense move off toward infinity: such lines are called external lines. Lines
that are not external lines, and end up at vertices at both ends, are called internal lines.
Diagrams may be connected, in which case one can move between any two points in
the diagram following lines of that diagram; or they may be disconnected, in which case it
consists of two or more disjoint pieces that are themselves connected. Any graph21 consists
of a finite number of connected subgraphs. The “empty” graph E, containing no lines or
vertices whatsoever, also exists; it does not count as connected.22 Diagrams containing one
or more closed loops are perfectly allowed. Diagrams with no closed loops are called tree
diagrams. A few examples are
, ,
19 For this approach to work in practice, it turns out to be useful to truncate φ(J) as a power series in J, the
truncation order increasing by one with each iteration. If you don’t do this, each iteration triples the highest
power in J, leading to very unwieldy expressions with only the first few terms being actually correct.
20 In the mathematical world of graph theory, lines are often called edges for some reason.
21 For us, the terms “diagram” and “graph” are interchangeable.
22 Sophistry alert: it has no points between which to move.
where you see, respectively, a connected graph, a disconnected graph, and a connected tree
graph. The precise shape of the lines and the precise position of the vertices are irrelevant.
The important thing is the way in which the lines are connected to the vertices.23
1.3.2 Feynman Rules

The noteworthy thing about Feynman diagrams is that they have an algebraic interpreta-
tion; that is, they correspond to numbers that may be added and multiplied. The assignment
of a number to a Feynman diagram is governed by the Feynman rules, which postulate a
numerical object for every ingredient of a Feynman graph. In the simple zero-dimensional
theories that we consider here the Feynman rules are just numbers. We will use the
following rules:24
1
↔ , ↔ −λ3 , ↔ −λ4 , ↔ +J. (1.32)
μ
A vertex at which a single line ends (and which carries a Feynman rule factor +J) is called a
source vertex. A disconnected diagram evaluates to the product of the values of its disjunct
connected pieces. Because of this multiplicative rule, the value of the empty diagram E is
taken to be unity. In addition, we assign to every Feynman diagram a symmetry factor. The
symmetry factor is the single most nontrivial ingredient of the diagrammatic approach, so
it deserves its own section.
1.3.3 Symmetries and Multiplicities

Feynman diagrams have, in general, an “inner” and an “outer” part. The “inner” part con-
sists of the various vertices and internal lines: the “outer” part is made up from the external
lines (if any). The inner part concomitates with the symmetry factor of the diagram, and
for the outer part we have what may be called the multiplicity, to be discussed below. Let
us first turn to the symmetry factor. The rules are as follows:
• for every set of k lines that may be permuted without changing the diagram, there will
be a factor 1/k!;
• for every set of m vertices that may be permuted without changing the diagram, there
will be a factor 1/m!;
• for every set of p disjunct connected pieces that maybe interchanged without changing
the diagram, there will be a factor 1/p!;
• a factor 1/k for every k-fold rotational symmetry;25
• a factor 1/2 for every mirror symmetry.
23 Throughout this book I try to avoid drawing Feynman diagrams with straight lines, or to draw blobs or
closed loops as circles. Many texts do employ only straight lines and circles. This not only leads to awfully
unæsthetic-looking pictures, but is also deeply misleading. There is a (natural) tendency to look at Feynman
diagrams with the idea that the lines represent “particles moving freely through space’ so that the lines “ought”
to be straight according to Newton’s first law. This is completely wrong! In the zero-dimensional world we
are dealing with for now, there cannot be any notion of movement yet, let alone any Newton to pronounce on
it. In fact, Newton’s first law ought to be derived from our theory, and we shall do so in due course.
24 These will make the diagrams do just what we want, see below.
25 Note: 1/k, not 1/(k! ).
9 Diagrammatics
External lines cannot be permuted without changing the diagram.26 Therefore only vacuum
diagrams, that is diagrams without any external lines, can have a rotational symmetry. The
symmetry factor cannot be read off from the individual components of the diagram, but
depends on the topology of the whole diagram.27 As our universe grows from zero to more
dimensions, and as the particles considered acquire more properties, the Feynman rules
will grow in complication; but the symmetry factors remain the same.28
Let us look at a few examples of diagram values. First, consider the diagram
λ3 2
= . (1.33)
μ5
In this case, the symmetry factor is 1, since for a tree diagram, no internal lines or vertices
can be interchanged with impunity. The similar-looking diagram
1 λ3 2 3
= J . (1.34)
2 μ5
has a symmetry factor 1/2! since the upper two one-point vertices are interchangeable.
Then, there is the graph
1 λ4
= − .
2 μ3
Here, there is a symmetry factor 1/2 because the “leaf” can be flipped over without
changing the diagram.29 The diagram
1 λ4 2
=
6 μ5
carries a symmetry factor of 1/3! because the three internal lines are interchangeable. The
graph
1 λ4 3
= −
4 μ7
carries a symmetry factor (1/2!)(1/2!) since there are now only two interchangeable internal
lines, and a single “leaf.” Finally, the diagram
1 λ4 2
=
48 μ4
has a symmetry factor (1/4!)(1/2!) since there are four equivalent internal lines, and
9 moreover the diagram can be “flipped over” without changing it.
26 Think of them as anchored somewhere very far away.

27 This is what makes the automated evaluation of diagrams a nontrivial task: component factors of diagrams
can be easily assigned, but working out the symmetry factor of a diagram calls for very complicated computer
algorithms indeed.
28 This is only modified if we include lines of different types, or oriented lines, that is lines that are deemed to
run in a particular direction. Then again, the more-dimensional diagrams have the same symmetry factors as
their zero-dimensional siblings.
29 This is due to the fact that the line in the loop is not oriented: for oriented lines it will no longer hold.
Next, we address the multiplicity. This is the number of different ways the external lines
(that each have their own “individuality”) can be attached. To determine the multiplicity we
must imagine that the whole diagram, or a part of it, can be “flipped over” while retaining
the same attachment of the external lines. To illustrate this, we temporarily denote the
external lines with a letter, and then notice that the two diagrams
a c d a
and b
b d c
are, in fact, identical; the multiplicity of this graph is therefore 3, since there are three ways
to group four letters into two pairs without regard to ordering:
a c a b a b
= d
+ + c
. (1.35)
b c d d
We shall often use the product of the symmetry and multiplicity, which factor we shall
denote by sm.
We see that the diagram of Eq. (1.33) has, also, multiplicity 3, while that of Eq. (1.34)
has multiplicity 1. We see that replacing p external lines with p one-point source vertices
reduces sm by a factor of 1/p!; that will become important later on.30
The determination of symmetry factors may appear somewhat fanciful,31 but of course it
has a solid and unambiguous basis; the symmetry factor (and the multiplicity) can always
be computed. The procedure is somewhat involved and is outlined in Appendix B.2.
1.3.4 Vacuum Bubbles

There are Feynman diagrams that contain neither external lines nor source vertices. These
are called vacuum bubbles; the empty graph E is obviously a vacuum bubble. We may
consider the set of all vacuum bubbles, which we denote by H0 . Let us assume that only
four-point vertices occur. Then, H0 , given by
H0 = E + + + + + ··· (1.36)
(where the ellipsis denotes diagrams with more four-vertices) evaluates to

1 λ4 1 1 λ4 2 1 λ4 2 1 λ4 2
H0 = 1 − + + + + ···
8 μ2 2 8 μ2 16 μ4 48 μ4
1 λ4 35 λ4 2
=1− + + · · ·, (1.37)
8 μ2 384 μ4
which, indeed, looks suspiciously like H0 for the ϕ 4 theory.
30 In higher dimensions the symmetry factor is unchanged, but you may wonder what happens to the multiplicity.
Essentially, that is also unchanged although less visible: the multiplicity tells us how many diagrams of a
certain type there are, as in Eq. (1.35). Of course the values of these diagrams are generally no longer the
same since the external lines carry their own momentum.
31 The discussion of symmetry factors of Feynman diagrams goes, in practice, with a lot of remarks like “so
you flip over this leaf, you wriggle this set of internal lines, you shove these vertices back and forth . . . see?”
Although the symmetry factor is totally unambiguous, the arguments for a symmetry factor often come with a
lot of prestidigitatorial arguments accompanied by hand-waving and finger-wriggling in front of a blackboard.
11 Diagrammatics
1.3.5 An Equation for Connected Graphs

We shall now construct an equation for a special set of diagrams. We do this for the set of
Feynman rules of Section 1.3.2. First, let us denote by Cn the set of all connected graphs
with no source vertices and precisely n external lines. Clearly this is a enumerably infinite
set. Next, we define the object (J), denoted by the symbol
(J) ≡ (1.38)
to be the set of all connected diagrams with precisely one external line, and any number of
source vertices. The shading indicates that all the diagrams in the blob must be connected.
Clearly, then, we have
1
(J) = J n Cn+1 , (1.39)
n!
n≥0
where the extra factor 1/n! is the additional sm factor for n source vertices.
Let us now consider what can happen if we enter the blob of Eq. (1.38) along the single
external line. In the first place, we can simply encounter a source vertex, so that we have
Game Over!, and the diagram is
= J/μ. (1.40)
Alternatively, we may encounter another vertex. If this is a three-point vertex, the line splits
into two. Taking one of these branches, we may be able to come back to the vertex via the
other branch. In that case, the diagram has the form
On the other hand, it may happen that the two branches end up in disjunct connected pieces
of the diagram, which then looks like
These two alternative cases can be unambiguously distinguished because we have restricted
ourselves to using only connected graphs. Another important insight is that, in this last
diagram, the two final blobs (with their attached lines) are both exactly identical to the
original (J) of Eq. (1.38), and therefore also to each other.32 In contrast, the closed-
loop blob of the first alternative is not equal to (J) since it has not one but two lines
sticking out; but then again these two lines are completely equivalent. If we encounter a
four-point rather than a three-point vertex, the line splits into three, with three alternatives:
no branches meeting again further on, all three meeting again, or only two out of the three.
32 This is of course only possible because the blobs represent infinite sets of diagrams
We find the diagrammatic equation
= + + +
+ + . (1.41)
Now, realize that

J n Cn+2 ∂(J) J n Cn+3 ∂ 2 (J)
= = , = = , (1.42)
n! ∂J n! (∂J)2
n≥0 n≥0
so that we can translate the diagrammatic equation (1.41) into an algebraic equation for
(J) by carefully implementing the correct Feynman rules, including the symmetry factors
for equivalent blobs and lines:

J λ3 1 1 ∂
(J) = − (J)2 + (J)
μ μ 2 2 ∂J

λ4 1 1 ∂ 1 ∂2
− (J)3 + (J) (J) + (J) . (1.43)
μ 6 2 ∂J 6 (∂J)2
Now Eq. (1.43), obtained from the Feynman diagrams via the Feynman rules, has exactly
the same form as Eq. (1.31), valid for the field function φ(J) – now you see how important
the symmetry factors are! Moreover, the iterative solution for φ(J) starts with φ(J) = J/μ,
also identical to the diagrammatic starting point . We therefore conclude that (J) =
10 φ(J), in other words Cn = Cn (n ≥ 1). This proves that connected Green’s functions can be
11 obtained by the following recipe: to obtain Cn (n ≥ 1), write out all connected Feynman
diagrams with no source vertices and precisely n external lines. Evaluate the diagrams
12
using the Feynman rules, and sum them.33
1.3.6 Semiconnected Graphs and the SDe

A useful notion, which allows us to write SDes more compactly, is that of semi-connected
graphs. We shall denote these with a lightly shaded blob, and they are defined as follows: a
semiconnected graph with n ≥ 1 lines at the left is a general graph with n lines on the left
(and any number of other external lines), with the constraint that each connected piece of
the semiconnected graph is attached to at least one of the lines indicated on the left. This
may sound more intimidating that is actually is: an example is
1 1 1 1
2 = 2
+ 2 + 3
3 3 3 2
33 Of course we can also obtain the G , by allowing disconnected diagrams.

n
13 Diagrammatics
1
2
+ 3 + 2 . (1.44)
1
3
A single semiconnected graph with n indicated lines stands for B(n) diagrams with explicit
connected graphs, where B(n) is the so-called Bell number: the number of ways to divide
n distinct objects into nonempty groups.34 For ϕ 3/4 theory, the SDe then becomes simply
= + + . (1.45)
You should keep in mind that a semiconnected graph acts as a symmetrizer: the lines enter-
ing it on the left are to be connected to whatever happens inside the blob in all admissible
ways. We shall use semiconnected diagrams to good effect in later chapters. Note that the
sum of the symmetry factors of all connected diagrams arising from a ϕ p vertex must be
equal to B(p − 1)/(p − 1)!, which may serve as a check on your SDes.
1.3.7 The Path Integral as a Set of Diagrams

By affixing a source vertex to the single external line of (J), we immediately have the
result that the generating function W(J) is the sum of all connected Feynman diagrams
without external lines and at least one source vertex. If we explicitly indicate the source
vertices, and recall that n source vertices in a diagram imply a factor 1/n!, we can write
W(J) = + + + + · · ·, (1.46)
where the ellipsis contains connected contributions with more source vertices. Vacuum
bubbles do not contribute to W(J). By taking careful account of the symmetry factor
assigned to identical connected parts of a disconnected diagram, we can see that
1
W(J)2 = + +
2!
+ + +
+ (lots of other diagrams). (1.47)

Similar arguments hold for higher powers of W(J). In addition, W(J)0 = E = 1. From
this it easy to see that the path integral Z(J) consists of all Feynman diagrams without
external lines, and without vacuum bubbles, but including the empty diagram.
34 For small n we have B(0) = 1, B(1) = 1, B(2) = 2, B(3) = 5, B(4) = 15, and B(5) = 52 ; more general
values can be obtained from the generating function derived in Appendix T.7.
Why are the vacuum bubbles so conspicuously absent? Suppose that we would allow
the inclusion of arbitrary numbers of vacuum bubbles in Z(J). Then the Green’s function
G0 = 1 would be represented not by the single empty graph but by the whole set H0
discussed before: indeed, H0 is proportional to H0 . In fact, any Green’s function Gn would
acquire exactly the same additional factor H0 . The normalization factor N, that must be
chosen such as to make G0 equal to unity, therefore extracts exactly the factor H0 from
any Green’s function. In the jargon, the vacuum bubbles “disappear into the normalization
of the path integral.” This is not to say that vacuum diagrams are never important; but in
our approach to computing Green’s functions and connected Green’s functions they are
indeed irrelevant. Another way of seeing this is very simple: if we take our diagrammatic
prescription of Z(J) and then take J = 0, all diagrams disappear except the empty one, and
we find Z(0) = E = 1, just as we must.
1.3.8 Dyson Summation

Why is the Feynman rule for lines, stemming from the quadratic part of the action, so
different from those for the vertices, that come from the nonquadratic terms? To see that
our treatment is actually a consistent one, let us consider the action
1 2 1 1
S(ϕ) = μϕ + λ2 ϕ 2 + λ4 ϕ 4 . (1.48)
2 2 4!
If we wish, we may treat the λ2 term as an interaction, described by a vertex with two legs.
the SDe is then seen to be
= + + , (1.49)
corresponding to

J λ2 λ4 ∂ ∂2
φ(J) = − φ(J) − φ(J)3 + 3φ(J) φ(J) + φ(J) . (1.50)
μ μ 6μ ∂J (∂J)2
Multiplying the equation by μ and transposing the λ2 term to the left, we obtain

J λ4 ∂ ∂2
φ(J) = − φ(J) + 3φ(J) φ(J) +
3
φ(J) , (1.51)
μ + λ2 6(μ + λ2 ) ∂J (∂J)2
precisely what we would have obtained by taking the combination (μ + λ2 ) as the kinetic
part from the start! This procedure, by which the effect of two-point (effective) vertices is
subsumed in a redefinition of the kinetic part, is called Dyson summation. In the present
example, the summation is of course trivial; but we shall see that two-point interactions
can also arise from more complicated Feynman diagrams corresponding to higher orders
in perturbation theory. The manner in which Dyson summation is usually treated is by
explicitly writing out the propagator, “dressed” with two-point vertices in all possible ways:
+ + + + ···
1 1 1 1 1 1 1 1 1 1
= − λ2 + λ2 λ 2 − λ 2 λ2 λ 2 + ···
μ μ μ μ μ μ μ μ μ μ
15 Exercises

1 λ2 k 1 1 1
= − = = , (1.52)
μ μ μ 1 + λ2 /μ μ + λ2
k≥0
where, no surprise, we cheerfully ignore all issues about convergence, in the spirit of per-
14 turbation theory. Every propagator line can (and must) be dressed in this way once any
two-point vertex (either elementary, or effective, that is, as the result of a collection of
closed loops with two legs sticking out) occurs.
1.4 Synopsis of Feynman Rules
The Feynman rules for zero-dimensional ϕ 3/4 theory found in this chapter are
1
↔ , ↔ −λ3 ,
μ
↔ −λ4 , ↔ +J. (1.53)
1.5 Exercises
Exercise 1 Green’s functions and connected Green’s functions

We have
Jn Jn
Z(J) = Gn , W(J) = Cn , W(J) = log (Z(J))
n! n!
n≥0 n≥0
1. Using that G0 = 1, and the expansion

xn x2 x3 x4 x5
− log(1 − x) = =x+ + + + − · · ·,
n 2 3 4 5
n≥1
express C0,...,5 in terms of the G’s.

2. From Z (J) = W (J)Z(J), prove the recursion relation
n−1

n−1
Gn = Cn + Cm Gn−m , n ≥ 1
m−1
m=1
n−1
3. Prove this last result using Feynman diagrams. Show how the factor m−1 avoids double
counting.
Exercise 2 The problem with ϕ 4 theory

Prove that in ϕ 4 theory any positive real value for λ4 leads to a well-defined theory, while
a negative value of λ4 leads to an undefined theory. The limit λ4 → 0 is, actually, an
essentially singular situation!
Exercise 3 Actually doing it for ϕ 4

This exercise shows how to express the path integral for the ϕ 4 theory in terms of “known
functions.” We consider

μ λ 4
H = dϕ exp − ϕ 2 − ϕ .
2 24
1. Show that the combination

λ
g=
μ2
is dimensionless.
2. Perform the substitution

3 1/4
ϕ= ψ − ψ −1/4
μg
to derive
∞
3 3/(4g) 3 1
H= e dψ ψ −3/4 exp − ψ+ .
4μg 8g ψ
0
3. The so-called modified Bessel functions of the second kind, denoted by Kν (z), are
discussed in Appendix T.8. Use this to express H in terms of these Bessel functions.
4. The function Kν (z) has two expansions: one, an unattractive-looking but convergent
series in positive powers of z, and a nonconvergent, asymptotic series in terms of
negative powers of z, given in Eq. (T.54). Show that H therefore has a regular series
expansion in 1/g and an asymptotic one in g. Show that the leading term in the asymp-
totic expansion is independent of g (but not of μ), and compute the next two terms.
Compare your result with Eq. (1.19).
Exercise 4 The SDe for Z in another action

Find the SDe for the path integral Z(J) for the action
μ 2 λk k
6
S(ϕ) = ϕ + ϕ .
2! k!
k=3
Exercise 5 Writing the SDe for φ

Prove Eq. (1.29). Do this using the fact that

1
Z(J) = exp dJ φ(J) ,

and then considering Z , Z , and so on.
Exercise 6 The SDe for φ in another action

For the action of exercise 4, derive the SDe for φ(J).
17 Exercises
Exercise 7 A programming exercise

For this exercise you have to employ computer-algebra code that is at least capable of
series expansions35 (or, if you want to do it by hand, you have to really want to). Starting
with φ(J) = 0, iterate the SDe for ϕ 4 theory a number of times, judiciously truncating
the series expansions at every iteration. I find it useful to replace the first term, J/μ, by
zJ/μ and expand in powers of z. That way, high powers of both J and are truncated
simultaneously. Try to at least reproduce Eq. (1.21).
Exercise 8 Stepping equations

Consider the path integral for the ϕ 3/4 theory:

μ λ3 λ4
Z(J) = Z(J, μ, λ3 , λ4 ) = N dϕ exp − ϕ 2 − ϕ 3 − ϕ 4 + Jϕ ,
2 6 24
where we have indicated all parameters this once.
1. Show that Z obeys the equation

∂ ∂ ∂
μ Z − λ3 Z − λ4 Z − JZ = 0.
∂J ∂μ ∂λ3
2. Differentiating this equation with respect to J, show that the field function φ =
φ(J, μ, λ3 , λ4 ) obeys the stepping equation
∂ 1 λ3 ∂ λ4 ∂
φ= + φ+ φ.
∂J μ μ ∂μ μ ∂λ3
If we are given C1 = φ(0, μ, λ3 , λ4 ), the stepping equation allows us to compute the
higher connected Green’s functions.
3. Prove the stepping equation diagrammatically.
Exercise 9 The symmetry factor of life, the universe, and everything

Devise a diagram (connected or disconnected) that has a symmetry factor of 1/42.
Exercise 10 Diagrammatic SDe for ϕ 6 theory

Give the diagrammatic SDe for ϕ 6 theory, and write it out algebraically.
Exercise 11 Actually doing it

Consider the diagrammatic SDe of Eq. (1.41). Starting with the empty diagram, iterate this
diagrammatic equation two or three times, and write down the resulting set of diagrams.
Exercise 12 Some diagrams to consider

Of the following 12 diagrams, determine the multiplicity factor, the symmetry factor, and
the number of loops:
35 I usually rely on Waterloo’s MAPLE package, but most other algebraic codes will serve as well.
Exercise 13 A multi-loop, multi-leg diagram

Consider the following diagram:
As before, determine the number of loops of this diagram, its symmetry factor, and its
multiplicity factor.
Exercise 14 The One Ring

The One Ring diagram
might be argued to evaluate to zero because of its ∞-fold rotational symmetry. In this
exercise we shall try to be more careful.
1. We consider a theory without any sources, with action
μ0 2 δμ 2
S(ϕ) = ϕ + ϕ , δμ = μ − μ0 ,
2 2
where the second term is, for the moment, considered an interaction. The Feynman rules
are therefore
= 1/μ0 , = −δμ.
Prove that the Dyson summation gives for the full propagator:
= + + + · · · = 1/μ.
2. Let us denote the two One Ring diagrams by
R(μ0 ) = , R(μ) =
From
= + + + + ···
prove that

1 μ 1 c
R(μ) = R(μ0 ) − log ⇒ R(μ) = log
2 μ0 2 μ
for some μ-independent constant c.
3. Introduce a ϕ 3 interaction, and show that the stepping equation (exercise 8) gives the
correct one-loop tadpole.
4. Choose c = 2π. Show that for the sum of all vacuum diagrams:
E + + + + ···
we obtain precisely the correct unnormalized free-action path integral.

2 The Loop Expansion and the Effective Action
2.1 Planck’s Constant
2.1.1 The Loop Expansion and Reverse Engineering

As we have seen, Green’s functions can be computed in a perturbative expansion in which
the coupling constant λ4 is in some sense a small number. Now consider doing perturbation
theory in the ϕ 3/4 theory. We then have to decide on the relative order of magnitude of the
two coupling constants λ3 and λ4 : are they of the same order, or should we take, say, λ4 to
be of the same order as (λ3 )5 ? And what if even more coupling constants are involved? We
shall adopt the approach that the order of magnitude of the various diagrams should depend
not on their coupling-constant content but rather on their complexity, in particular on the
number of closed loops. That is, the more closed loops a diagram contains, the smaller it
is considered to be, and perturbation theory then prescribes the perturbation expansion to
be truncated at a given number of closed loops. To quantify these ideas, we shall assign to
every closed loop a factor , where is a (small) number.1 That is, we define the following
order-of-magnitude ratios:
∼ , ∼ 2 , (2.1)
et cetera. This implies, of course, a modification of the Schwinger–Dyson equation from

the form (1.31) into

J λ3 ∂
φ(J) = − φ(J)2 + φ(J)
μ 2μ ∂J

λ4 ∂ 2 ∂
2
− φ(J) + 3φ(J) φ(J) +
3
φ(J) . (2.2)
6μ ∂J (∂J)2
Thus we have to reverse-engineer everything else as well: we must redefine
∂
φ(J) = log Z(J), (2.3)
∂J
1 As the notation suggests, it will develop into Planck’s (or Dirac’s) constant as our universe increases in
complexity.
19
so that the SDe for the path integral must read

∂
S Z(J) = JZ(J). (2.4)
∂J
The path integral must therefore be redefined with inclusion of ,

1
Z(J) = N exp − S(ϕ) − Jϕ dϕ, (2.5)

and for the Green’s functions we have

∂n ∂n
Gn = n Z(J) , Cn = n log Z(J) . (2.6)
(∂J)n J=0 (∂J)n J=0
The Feynman rules must, therefore, take the form
λ3 λ4 J
↔ , ↔ − , ↔ − , ↔ + . (2.7)
μ
The introduction of as the perturbation expansion parameter allows us to determine the
15 relative orders of magnitude of coupling constants. Since under our decree all tree diagrams
16 are of the same order, the two graphs
and
tell us that λ4 is of the same order as λ3 2 /μ.2 Similarly, a k-point coupling constant λk is of
the same order as λ3 k−2 /μk−3 . As a last point, you may note that the introduction of does
not influence the Dyson summation of Section 1.3.8, since every extra two-point vertex
(with 1/) also gives an extra propagator (with ). In the Feynman rules, appears all over
the place, so we had better check that the -behavior of the Feynman graphs is indeed as
desired. Consider an arbitrary given diagram. Let E be the number of its external lines, I
the number of internal lines, Vq the number of vertices of q-point type, L the number of
closed loops, and P the number of disjunct connected pieces. There are precisely two “sum
rules” that involve linear combinations of these quantities. They are derived in Appendix
C and read

Vq = I + P − L, qVq = 2I + E. (2.8)
q q
We are now able to read off the power of associated with an arbitrary connected diagram
(with P = 1). From the Feynman rules, we infer that every line contributes a factor and
every vertex a factor 1/. The total power of is therefore

E+I− Vq = E + L − 1. (2.9)
q
Independently of its precise form, the power of of any connected diagram depends only
on the number of its external lines and the number of loops, and indeed each extra loop
17 leads to an additional factor .
18 2 In zero dimensions. In general, we should formulate it slightly differently, by looking for dimensionless cou-
plings. In the present case, these are g4 ≡ λ4 /μ2 and g3 = λ3 (/μ3 )1/2 . Then g4 ∼ g3 2 , and this persists in
any dimension: one four-point coupling is worth two three-point couplings.
21 Planck’s Constant
2.1.2 The Classical Limit

Since in perturbation theory is taken to be an infinitesimally small quantity, the limit
→ 0 is of automatic interest. This limit has to be taken with some care, since = 0
strictly would imply that only Green’s functions with E + L = 1 would survive.3 Instead,
the classical limit → 0 is meant to be the result of leaving out diagrams containing
closed loops. The diagrammatic SDe will, for the ϕ 3/4 theory, then take the form
= + + . (2.10)
The corresponding solution will be denoted by φc (J) (with c for “classical”), and the
classical SDe is written as
J λ3 λ4
φc (J) = − φc (J)2 − φc (J)3 . (2.11)
μ 2μ 6μ
The classical field function is exclusively built up from tree diagrams: this is called the
tree approximation. It obeys an algebraic rather than a differential equation, which can be
written as
S (φc (J)) = J. (2.12)
This is called the classical field equation. This is not to be confused with equations from
classical, nonquantum physics. In fact, the classical field equations will turn out to be the
Klein–Gordon, Dirac, Proca, and Maxwell equations. Of these, only the Maxwell equations
can be considered classical, since they do not contain a particle mass. Such equations have,
in general, more than a single solution. Here, however, we are interested in that solution
that vanishes as J → 0, which may be written out using Lagrange expansion:4
n
J 1 n−1 ∂ n−1 J
φc (J) = + μ S . (2.13)
μ n! (∂J)n−1 μ
n≥1
19 Let us now look at the path-integral picture of the classical limit. When becomes small,
the fluctuations in the path integrand

1
exp − S(ϕ) − Jϕ

become extremely exaggerated. The main contribution to ϕ therefore comes from that
value ϕc of ϕ for which the probability distribution attains its maximum, that is,
ϕJ ≈ ϕc , where S (ϕc ) = J, S (ϕc ) > 0. (2.14)
Also in the classical limit, we therefore have φc (J) = ϕc .
3 As we shall see, when we compute actual cross sections, the results for processes described by tree diagrams
will, indeed, turn out to be independent of (see, e.g., Eq. (13.56)). But that calls for more machinery than we
have at this moment.
4 see Appendix T.11.
2.1.3 On Second Quantization

The “classical” approximations of our quantum field theory are5 quantum equations. In
fact, this is not so very surprising. In ordinary quantum mechanics, the classical vari-
ables, such as position and momentum, are identified with the expectation values of their
quantum-mechanical counterparts and considered a useful approximation of reality as long
as they are reasonably well defined.6 So here as well: the field-generating function φ(J) is
considered as the expectation value of the quantum field ϕ, and it is identified with the
quantum-mechanical wave function of whatever object it is we are studying. In this sense,
to go from ϕ to a classical observable, we have to “classicize” two times. The transition
from ordinary quantum mechanics to what we are doing here is therefore dubbed second
quantization. Of course, from the point of view we have taken here, this is simply a mat-
ter of taking limits (expectation value upon expectation value), but if one comes in from
the classical side, it may look quite mysterious. This is another reminder that one should
not try to build a more fundamental theory from a limiting case. Limiting cases are only
hints.
2.1.4 Instanton Contributions

As mentioned, for a nonfree action S(ϕ), the equation (2.14) has, of course, more than
(0) (1)
a single solution.7 Suppose that we have several such solutions, denoted by ϕc , ϕc ,
(2) (0)
ϕc , . . . , and that the minimal value of S(ϕ) − Jϕ is attained for ϕc . Then, the other
classical solutions will give contributions that, relative to the dominant one, are suppressed
by exponential factors of order

1
exp − S(ϕc(k) ) − S(ϕc(0) ) − Jϕc(k) + Jϕc(0) , k = 1, 2, . . . .

In Figure 2.1 we plot the (normalized) form of exp(−S(φ)/) for an action with two
minima, for two values of different by a factor 6. It is seen how the lowest minimum
of S(ϕ) starts to dominate the integral as becomes small; the contribution from the sub-
leading maximum decreases nonperturbatively fast. Such subdominant solutions to the
classical field equations are called instantons. Their contribution to Green’s functions do
not have a series expansion around = 0. Such effects are therefore not accessible using
Feynman diagrams: they are called nonpertubative. This is not to say that they are irrel-
evant. Indeed, we usually have a finite value for ; more dramatically, if we let J vary
(1) (0)
as a parameter, ϕc , say, may for some value of J take over from ϕc as the true maxi-
mum position of the probability density, causing a sudden shift in the value of φc (J) from
(0) (1)
ϕc to ϕc .
5 Will be found to be; see the later chapters.

6 With small uncertainty, that is, the variance of their statistical distribution around the expectation value.
7 Since the action is at least of order ϕ 3 , the classical field equation is at least quadratic.
23 The Effective Action
0.8
0.6
0.4
0.2
0
–3 –2 –1 0 1 2 3 4
t
Figure 2.1 The developing of instanton effects.
2.2 The Effective Action
2.2.1 The Effective Action as a Legendre Transform

Since perturbation theory more or less assumes that higher orders in the loop expansion
are small compared to lower orders, the following question suggests itself: is it possible to
find, for a given action S(ϕ), another action, called the effective action, with the property
that its tree approximation reproduces the full field function of the original action S? If
such an effective action, denoted by (φ), exists, we must have
(φ(J)) = J, (2.15)
where φ(J) is the full solution to the SDe belonging with S(ϕ). We can use partial
integration to find
(φ) = J dφ = J φ − φ dJ = J φ − W, (2.16)
where J is now to be interpreted as a function of φ. The transition from W(J) to (φ) is

called the Legendre transform. In classical mechanics, we have the same situation: there,
W would be the Lagrangian with J as the velocity and φ as the momentum, and then the
effective action would turn out to be the Hamiltonian.
2.2.2 Concavity of the Effective Action

The zero-dimensional effective action has an interesting concavity property. Let us con-
sider the derivative of φ(J). If we denote the probability density (including the sources) of
the quantum field ϕ by PJ (ϕ), that is,
4
3
3
2
1
1
0
0
–2 –1 0 1 2 –3 –2 –1 0 1 2 3
S(φ) Γ(φ) (0.3) S(φ) Γ(φ) (0.3) Γ(φ) (0.01)
t
Figure 2.2 Action and effective action for μ > 0 (left) and μ < 0 (right).

A(ϕ) 1
PJ (ϕ) = , A(ϕ) = exp − (S(ϕ) − Jϕ) , (2.17)
dϕA(ϕ)
we can write this derivative as
2
dφ d PJ (ϕ) ϕ dϕ PJ (ϕ) ϕ 2 dϕ PJ (ϕ) ϕ dϕ
= = − 2
dJ dJ PJ (ϕ) dϕ PJ (ϕ) dϕ PJ (ϕ) dϕ

PJ (ϕ1 )PJ (ϕ2 ) (ϕ1 2 − ϕ1 ϕ2 ) dϕ1 dϕ2
= 2 . (2.18)
PJ (ϕ) dϕ
By symmetry, we can replace the factor (ϕ1 2 − ϕ1 ϕ2 ) by (ϕ1 − ϕ2 )2 /2, so as to see that
dφ(J)/dJ is positive. This implies that
∂2 dJ
(φ) = > 0. (2.19)
(∂φ) 2 dφ
In other words, the effective action is concave8 everywhere.9 Whereas one would assume
that the effective action would differ only slightly from the original action S, this can
obviously no longer hold in situations where the action S is not concave. In Figure 2.2 we
illustrate this. We have used λ4 = 1 and μ = 0.4 for the left-hand plot and μ = −0.4
for the right-hand plot, where the action is not concave in the middle. The value of is
0.3 in the left-hand plot and both 0.3 and 0.01 in the right-hand plot. The effective action
is always concave, and as decreases, the effective action becomes flatter in the central
region. For → 0 it becomes completely flat.
8 What to call convex and what to call concave is disputable. In mathematics, when a function f (x) has f (x) > 0,
it is called convex or convex downward. I prefer the wording concave upward: the working surface of a concave
mirror lying on the ground. If it makes you feel better, substitute convex for concave everywhere.
9 This concavity persists in case there are more fields than just a single one involved. By extension, it also holds
for Euclidean theories in more dimensions; see also Appendix H.
25 The Effective Action
2.2.3 Diagrams for the Effective Action

A tree approximation consists of tree diagrams only. To see how the loop effects of the
action S end up in , we define a new concept, that of a one-particle irreducible (1PI)
diagram. A connected Feynman graph is 1PI if it contains no internal line such that cutting
that line makes the diagram disconnected:
1PI diagrams a non-1PI diagram
20 External lines, of course, do not enter in the 1PI criterion.10 Note that a diagram built from
only external lines and a single vertex also counts as 1PI, since it does not have any internal
lines to be cut. A typical one-loop 1PI diagram looks like this:
Let us denote the set of all 1PI graphs with precisely n external lines by −γn /, where the
convention is that the Feynman factors for the external lines are not included. Consider,
now, what happens if we enter the field function by way of its single external leg, as in the
SDe. If we encounter a vertex, that vertex is part of a 1PI subdiagram (possibly consisting
only of the vertex itself). Indicating the 1PI property with dark crosshatches, we therefore
obtain the diagrammatic equation11
= + + + + · · ·. (2.20)
Algebraically, it reads

J 1 1 1
φ(J) = − γ1 + γ2 φ(J) + γ3 φ(J)2 + γ4 φ(J)3 + · · · ; (2.21)
μ μ 2! 3!
in other words,
(φ) = J, (2.22)
where
1 1 1
(ϕ) = γ1 ϕ + (γ2 + μ)ϕ 2 + γ3 ϕ 3 + γ4 ϕ 4 + · · ·. (2.23)
2! 3! 4!
21 We conclude that the vertices of the effective action are determined by the 1PI diagrams.
22 In general, the effective action contains vertices with an unlimited number of legs, even if
the original action S goes up only to ϕ 3 or ϕ 4 , say.
10 Including them would be silly, since any diagram falls apart if we cut an external line.
11 This is again nothing but the SDe, but now the diagrams are grouped under different criteria.
Another random document with
no related content on Scribd:
for they knew all the printed ones, and Ida needed a rest sometimes,
too. Kai’s stories were short and simple at first, but they expanded
and grew bolder and more complicated with time. The interesting
thing about them was that they never stood quite in the air, but were
based upon a reality which he presented in a new and mysterious
light. Hanno particularly liked the one about the wicked enchanter
who tortured all human beings by his malignant art; who had
captured a beautiful prince named Josephus and turned him into a
green-and-red parrot, which he kept in a gilded cage. But in a far
distant land the chosen hero was growing up, who should one day
fearlessly advance at the head of an invincible army of dogs,
chickens, and guinea-pigs and slay the base enchanter with a single
sword-thrust, and deliver all the world—in particular, Hanno
Buddenbrook—from his clutches. Then Josephus would be restored
to his proper form and return to his kingdom, in which Kai and Hanno
would be appointed to high offices.
Senator Buddenbrook saw the two friends together now and then, as
he passed the door of the play-room. He had nothing against the
intimacy, for it was clear that the two lads did each other good.
Hanno gentled, tamed, and ennobled Kai, who loved him tenderly,
admired his white hands, and, for his sake, let Ida Jungmann wash
his own with soap and a nail-brush. And if Hanno could absorb some
of his friend’s wild energy and spirits, it would be welcome, for the
Senator realized keenly the constant feminine influence that
surrounded the boy, and knew that it was not the best means for
developing his manly qualities.
The faithful devotion of the good Ida could not be repaid with gold.
She had been in the family now for more than thirty years. She had
cared for the previous generations with self-abnegation; but Hanno
she carried in her arms, lapped him in tender care, and loved him to
idolatry. She had a naïve, unshakable belief in his privileged station
in life, which sometimes went to the length of absurdity. In whatever
touched him she showed a surprising, even an unpleasant effrontery.
Suppose, for instance, she took him with her to buy cakes at the
pastry-shop: she would poke among the sweets on the counter and
select a piece for Hanno, which she would coolly hand him without
paying for it—the man should feel himself honoured, indeed! And
before a crowded show-window she would ask the people in front, in
her west-Prussian dialect, pleasantly enough, but with decision, to
make a place for her charge. He was so uncommon in her eyes that
she felt there was hardly another child in the world worthy to touch
him. In little Kai’s case, the mutual preference of the two children had
been too strong for her. Possibly she was a little taken by his name,
too. But if other children came up to them on the Mill-wall, as she sat
with Hanno on a bench, Fräulein Jungmann would get up almost at
once, make some excuse or other—it was late, or there was a
draught—and take her charge away. The pretexts she gave to little
Johann would have led him to believe that all his contemporaries
were either scrofulous of full of “evil humours,” and that he himself
was a solitary exception; which did not tend to increase his already
deficient confidence and ease of manner.
Senator Buddenbrook did not know all the details; but he saw
enough to convince him that his son’s development was not taking
the desired course. If he could only take his upbringing in his own
hands, and mould his spirit by daily and hourly contact! But he had
not the time. He perceived the lamentable failure of his occasional
efforts: he knew they only strained the relations between father and
son. In his mind was a picture which he longed to reproduce: it was a
picture of Hanno’s great-grandfather, whom he himself had known as
a boy: a clear-sighted man, jovial, simple, sturdy, humorous—why
could not little Johann grow up like that? If only he could suppress or
forbid the music, which was surely not good for the lad’s physical
development, absorbed his powers, and took his mind from the
practical affairs of life! That dreamy nature—did it not almost, at
times, border on irresponsibility?
One day, some three quarters of an hour before dinner, Hanno had
gone down alone to the first storey. He had practised for a long time
on the piano, and now was idling about in the living-room. He half
lay, half sat, on the chaise-longue, tying and untying his sailor’s knot,
and his eyes, roving aimlessly about, caught sight of an open
portfolio on his mother’s nut-wood writing-table. It was the leather
case with the family papers. He rested his elbow on the sofa-
cushion, and his chin in his hand, and looked at the things for a while
from a distance. Papa must have had them out after second
breakfast, and left them there because he was not finished with
them. Some of the papers were sticking in the portfolio, some loose
sheets lying outside were weighted with a metal ruler, and the large
gilt-edged notebook with the motley paper lay there open.
Hanno slipped idly down from the sofa and went to the writing-table.
The book was open at the Buddenbrook family tree, set forth in the
hand of his various forbears, including his father; complete, with
rubrics, parentheses, and plainly marked dates. Kneeling with one
knee on the desk-chair, leaning his head with its soft waves of brown
hair on the palm of his hand, Hanno looked at the manuscript
sidewise, carelessly critical, a little contemptuous, and supremely
indifferent, letting his free hand toy with Mamma’s gold-and-ebony
pen. His eyes roved all over these names, masculine and feminine,
some of them in queer old-fashioned writing with great flourishes,
written in faded yellow or thick black ink, to which little grains of sand
were sticking. At the very bottom, in Papa’s small, neat handwriting
that ran so fast over the page, he read his own name, under that of
his parents: Justus, Johann, Kaspar, born April 15, 1861. He liked
looking at it. He straightened up a little, and took the ruler and pen,
still rather idly; let his eye travel once more over the whole
genealogical host; then, with absent care, mechanically and
dreamily, he made with the gold pen a beautiful, clean double line
diagonally across the entire page, the upper one heavier than the
lower, just as he had been taught to embellish the page of his
arithmetic book. He looked at his work with his head on one side,
and then moved away.
After dinner the Senator called him up and surveyed him with his
eyebrows drawn together.
“What is this? Where did it come from? Did you do it?”
Hanno had to think a minute, whether he really had done it; and then
he answered “Yes.”
“What for? What is the matter with you? Answer me! What
possessed you, to do such a mischievous thing?” cried the Senator,
and struck Hanno’s cheek lightly with the rolled-up notebook.
And little Johann stammered, retreating, with his hand to his cheek,
“I thought—I thought—there was nothing else coming.”
CHAPTER VIII
Nowadays, when the family gathered at table on Thursdays, under
the calmly smiling gaze of the immortals on the walls, they had a
new and serious theme. It called out on the faces of the female
Buddenbrooks, at least the Broad Street ones, an expression of cold
restraint. But it highly excited Frau Permaneder, as her manner and
gestures betrayed. She tossed back her head, stretched out her
arms before her, or flung them above her head as she talked; and
her voice showed by turns anger and dismay, passionate opposition
and deep feeling. She would pass over from the particular to the
general, and talk in her throaty voice about wicked people,
interrupting herself with the little cough that was due to poor
digestion. Or she would utter little trumpetings of disgust: Teary
Trietschke, Grünlich, Permaneder! A new name had now been
added to these, and she pronounced it in a tone of indescribable
scorn and hatred: “The District Attorney!”
But when Director Hugo Weinschenk entered—late, as usual, for he
was overwhelmed with work; balancing his two fists and weaving
about more than ever at the waist of his frock-coat—and sat down at
table, his lower lip hanging down with its impudent expression under
his moustaches, then the conversation would come to a full stop,
and heavy silence would brood over the table until the Senator came
to the rescue by asking the Director how his affair was going on—as
if it were an ordinary business dealing.
Hugo Weinschenk would answer that things were going very well,
very well indeed, they could not go otherwise; and then he would
blithely change the subject. He was much more sprightly than he
used to be; there was a certain lack of restraint in his roving eye, and
he would ask ever so many times about Gerda Buddenbrook’s fiddle
without getting any reply. He talked freely and gaily—only it was a
pity his flow of spirits prevented him from guarding his tongue; for he
now and then told anecdotes which were not at all suited to the
company. One, in particular, was about a wet-nurse who prejudiced
the health of her charge by the fact that she suffered from flatulence.
Too late, or not at all, he remarked that his wife was flushing rosy
red, that Thomas, the Frau Consul and Gerda were sitting like
statues, and the Misses Buddenbrook exchanging glances that were
fairly boring holes in each other. Even Riekchen Severin was looking
insulted at the bottom of the table, and old Consul Kröger was the
single one of the company who gave even a subdued snort.
What was the trouble with Director Weinschenk? This industrious,
solid citizen with the rough exterior and no social graces, who
devoted himself with an obstinate sense of duty to his work alone—
this man was supposed to have been guilty, not once but repeatedly,
of a serious fault: he was accused of, he had been indicted for,
performing a business manœuvre which was not only questionable,
but directly dishonest and criminal. There would be a trial, the
outcome of which was not easy to guess. What was he accused of?
It was this: certain fires of considerable extent had taken place in
different localities, which would have cost his company large sums of
money. Director Weinschenk was accused of having received private
information of such accidents through his agents, and then, in
wrongful possession of this information, of having transferred the
back insurance to another firm, thus saving his own the loss. The
matter was now in the hands of the State Attorney, Dr. Moritz
Hagenström.
“Thomas,” said the Frau Consul in private to her son, “please explain
it to me. I do not understand. What do you make of the affair?”
“Why, my dear Mother,” he answered, “what is there to say? It does
not look as though things were quite as they should be—
unfortunately. It seems unlikely to me that Weinschenk is as guilty as
people think. In the modern style of doing business, there is a thing
they call usance. And usance—well, imagine a sort of manœuvre,
not exactly open and above-board, something that looks dishonest to
the man in the street, yet perhaps quite customary and taken for
granted in the business world: that is usance. The boundary line
between usance and actual dishonesty is extremely hard to draw.
Well—if Weinschenk has done anything he shouldn’t, he has
probably done no more than a good many of his colleagues who will
not get caught. But—I don’t see much chance of his being cleared.
Perhaps in a larger city he might be, but here everything depends on
cliques and personal motives. He should have borne that in mind in
selecting his lawyer. It is true that we have no really eminent lawyer
in the whole town, nobody with superior oratorical talent, who knows
all the ropes and is versed in dubious transactions. All our jurists
hang together; they have family connections, in many cases; they
eat together; they work together, and they are accustomed to
considering each other. In my opinion, it would have been clever to
take a town lawyer. But what did Weinschenk do? He thought it
necessary—and this in itself makes his innocence look doubtful—to
get a lawyer from Berlin, a Dr. Breslauer, who is a regular rake, an
accomplished orator and up to all the tricks of the trade. He has the
reputation of having got so-and-so many dishonest bankrupts off
scot-free. He will conduct this affair with the same cleverness—for a
consideration. But will it do any good? I can see already that our
town lawyers will band together to fight him tooth and nail, and that
Dr. Hagenström’s hearers will already be prepossessed in his favour.
As for the witnesses: well, Weinschenk’s own staff won’t be any too
friendly to him, I’m afraid. What we indulgently call his rough exterior
—he would call it that, himself, too—has not made him many friends.
In short, Mother, I am looking forward to trouble. It will be a pity for
Erica, if it turns out badly; but I feel most for Tony. You see, she is
quite right in saying that Hagenström is glad of the chance. The thing
concerns all of us, and the disgrace will fall on us too; for
Weinschenk belongs to the family and eats at our table. As far as I
am concerned, I can manage. I know what I have to do: in public, I
shall act as if I had nothing whatever to do with the affair. I will not go
to the trial—although I am sorry not to, for Breslauer is sure to be
interesting. And in general I must behave with complete indifference,
to protect myself from the imputation of wanting to use my influence.
But Tony? I don’t like to think what a sad business a conviction will
be for her. She protests vehemently against envious intrigues and
calumniators and all that; but what really moves her is her anxiety
lest, after all her other troubles, she may see her daughter’s
honourable position lost as well. It is the last blow. She will protest
her belief in Weinschenk’s innocence the more loudly the more she
is forced to doubt it. Well, he may be innocent, after all. We can only
wait and see, Mother, and be very tactful with him and Tony and
Erica. But I’m afraid—”
It was under these circumstances that the Christmas feast drew
near, to which little Hanno was counting the days, with a beating
heart and the help of a calendar manufactured by Ida Jungmann,
with a Christmas tree on the last leaf.
The signs of festivity increased. Ever since the first Sunday in
Advent a great gaily coloured picture of a certain Ruprecht had been
hanging on the wall in grandmama’s dining-room. And one morning
Hanno found his covers and the rug beside his bed sprinkled with
gold tinsel. A few days later, as Papa was lying with his newspaper
on the living-room sofa, and Hanno was reading “The Witch of
Endor” out of Gerock’s “Palm Leaves,” an “old man” was announced.
This had happened every year since Hanno was a baby—and yet
was always a surprise. They asked him in, this “old man,” and he
came shuffling along in a big coat with the fur side out, sprinkled with
bits of cotton-wool and tinsel. He wore a fur cap, and his face had
black smudges on it, and his beard was long and white. The beard
and the big, bushy eyebrows were also sprinkled with tinsel. He
explained—as he did every year—in a harsh voice, that this sack (on
his left shoulder) was for good children, who said their prayers (it
contained apples and gilded nuts); but that this sack (on his right
shoulder) was for naughty children. The “old man” was, of course,
Ruprecht; perhaps not actually the real Ruprecht—it might even be
Wenzel the barber, dressed up in Papa’s coat turned fur side out—
but it was as much Ruprecht as possible. Hanno, greatly impressed,
said Our Father for him, as he had last year—both times interrupting
himself now and again with a little nervous sob—and was permitted
to put his hand into the sack for good children, which the “old man”
forgot to take away.
The holidays came, and there was not much trouble over the report,
which had to be presented for Papa to read, even at Christmas-time.
The great dining-room was closed and mysterious, and there were
marzipan and gingerbread to eat—and in the streets, Christmas had
already come. Snow fell, the weather was frosty, and on the sharp
clear air were borne the notes of the barrel-organ, for the Italians,
with their velvet jackets and their black moustaches, had arrived for
the Christmas feast. The shop-windows were gay with toys and
goodies; the booths for the Christmas fair had been erected in the
market-place; and wherever you went you breathed in the fresh,
spicy odour of the Christmas trees set out for sale.
The evening of the twenty-third came at last, and with it the present-
giving in the house in Fishers’ Lane. This was attended by the family
only—it was a sort of dress rehearsal for the Christmas Eve party
given by the Frau Consul in Meng Street. She clung to the old
customs, and reserved the twenty-fourth for a celebration to which
the whole family group was bidden; which, accordingly, in the late
afternoon, assembled in the landscape-room.
The old lady, flushed of cheek, and with feverish eyes, arrayed in a
heavy black-and-grey striped silk that gave out a faint scent of
patchouli, received her guests as they entered, and embraced them
silently, her gold bracelets tinkling. She was strangely excited this
evening— “Why, Mother, you’re fairly trembling,” the Senator said
when he came in with Gerda and Hanno. “Everything will go off very
easily.” But she only whispered, kissing all three of them, “For Jesus
Christ’s sake—and my blessed Jean’s.”
Indeed, the whole consecrated programme instituted by the
deceased Consul had to be carried out to the smallest detail; and the
poor lady fluttered about, driven by her sense of responsibility for the
fitting accomplishment of the evening’s performance, which must be
pervaded with a deep and fervent joy. She went restlessly back and
forth, from the pillared hall where the choir-boys from St. Mary’s
were already assembled, to the dining-room, where Riekchen
Severin was putting the finishing touches to the tree and the table-
full of presents, to the corridor full of shrinking old people—the “poor”
who were to share in the presents—and back into the landscape-
room, where she rebuked every unnecessary word or sound with
one of her mild sidelong glances. It was so still that the sound of a
distant hand-organ, faint and clear like a toy music-box, came across
to them through the snowy streets. Some twenty persons or more
were sitting or standing about in the room; yet it was stiller than a
church—so still that, as the Senator cautiously whispered to Uncle
Justus, it reminded one more of a funeral!
There was really no danger that the solemnity of the feast would be
rudely broken in upon by youthful high spirits. A glance showed that
almost all the persons in the room were arrived at an age when the
forms of expression are already long ago fixed. Senator Thomas
Buddenbrook, whose pallor gave the lie to his alert, energetic,
humorous expression; Gerda, his wife, leaning back in her chair, the
gleaming, blue-ringed eyes in her pale face gazing fixedly at the
crystal prisms in the chandelier; his sister, Frau Permaneder; his
cousin, Jürgen Kröger, a quiet, neatly-dressed official; Friederike,
Henriette, and Pfiffi, the first two more long and lean, the third
smaller and plumper than ever, but all three wearing their
stereotyped expression, their sharp, spiteful smile at everything and
everybody, as though they were perpetually saying “Really—it
seems incredible!” Lastly, there was poor, ashen-grey Clothilde,
whose thoughts were probably fixed upon the coming meal.—Every
one of these persons was past forty. The hostess herself, her brother
Justus and his wife, and little Therese Weichbrodt were all well past
sixty; while old Frau Consul Buddenbrook, Uncle Gotthold’s widow,
born Stüwing, as well as Madame Kethelsen, now, alas almost
entirely deaf, were already in the seventies.
Erica Weinschenk was the only person present in the bloom of
youth; she was much younger than her husband, whose cropped,
greying head stood out against the idyllic landscape behind him.
When her eyes—the light-blue eyes of Herr Grünlich—rested upon
him, you could see how her full bosom rose and fell without a sound,
and how she was beset with anxious, bewildered thoughts about
usance and book-keeping, witnesses, prosecuting attorneys,
defence, and judges. Thoughts like these, un-Christmaslike though
they were, troubled everybody in the room. They all felt uncanny at
the presence in their midst of a member of the family who was
actually accused of an offence against the law, the civic weal, and
business probity, and who would probably be visited by shame and
imprisonment. Here was a Christmas family party at the
Buddenbrooks’—with an accused man in the circle! Frau
Permaneder’s dignity became majestic, and the smile of the Misses
Buddenbrook more and more pointed.
And what of the children, the scant posterity upon whom rested the
family hopes? Were they conscious too of the slightly uncanny
atmosphere? The state of mind of the little Elisabeth could not be
fathomed. She sat on her bonne’s lap in a frock trimmed by Frau
Permaneder with satin bows, folded her small hands into fists,
sucked her tongue, and stared straight ahead of her. Now and then
she would utter a brief sound, like a grunt, and the nurse would rock
her a little on her arm. But Hanno sat still on his footstool at his
mother’s knee and stared up, like her, into the chandelier.
Christian was missing—where was he? At the last minute they
noticed his absence. The Frau Consul’s characteristic gesture, from
the corner of her mouth up to her temple, as though putting back a
refractory hair, became frequent and feverish. She gave an order to
Mamsell Severin, and the spinster went out through the hall, past the
choir-boys and the “poor” and down the corridor to Christian’s room,
where she knocked on the door.
Christian appeared straightway; he limped casually into the
landscape-room, rubbing his bald brow. “Good gracious, children,”
he said, “I nearly forgot the party!”
“You nearly forgot—” his mother repeated, and stiffened.
“Yes, I really forgot it was Christmas. I was reading a book of travel,
about South America.—Dear me, I’ve seen such a lot of
Christmases!” he added, and was about to launch out upon a
description of a Christmas in a fifth-rate variety theatre in London—
when all at once the church-like hush of the room began to work
upon him, and he moved on tip-toe to his place, wrinkling up his
nose.
“Rejoice, O Daughter of Zion!” sang the choir-boys. They had
previously been indulging in such audible practical jokes that the
Senator had to get up and stand in the doorway to inspire respect.
But now they sang beautifully. The clear treble, sustained by the
deeper voices, soared up in pure, exultant, glorifying tones, bearing
all hearts along with them: softening the smiles of the spinsters,
making the old folk look in upon themselves and back upon the past;
easing the hearts of those still in the midst of life’s tribulations, and
helping them to forget for a little while.
Hanno unclasped his hands from about his knees. He looked very
pale, and cold, played with the fringe of his stool, and twisted his
tongue about among his teeth. He had to draw a deep breath every
little while, for his heart contracted with a joy almost painful at the
exquisite bell-like purity of the chorale. The white folding doors were
still tightly closed, but the spicy poignant odour drifted through the
cracks and whetted one’s appetite for the wonder within. Each year
with throbbing pulses he awaited this vision of ineffable, unearthly
splendour. What would there be for him, in there? What he had
wished for, of course; there was always that—unless he had been
persuaded out of it beforehand. The theatre, then, the long-desired
toy theatre, would spring at him as the door opened, and show him
the way to his place. This was the suggestion which had stood
heavily underlined at the top of his list, ever since he had seen
Fidelio; indeed, since then, it had been almost his single thought.
He had been taken to the opera as compensation for a particularly
painful visit to Herr Brecht; sitting beside his mother, in the dress
circle, he had followed breathless a performance of Fidelio, and
since that time he had heard nothing, seen nothing, thought of
nothing but opera, and a passion for the theatre filled him and almost
kept him sleepless. He looked enviously at people like Uncle
Christian, who was known as a regular frequenter and might go
every night if he liked: Consul Döhlmann, Gosch the broker—how
could they endure the joy of seeing it every night? He himself would
ask no more than to look once a week into the hall, before the
performance: hear the voices of the instruments being tuned, and
gaze for a while at the curtain! For he loved it all, the seats, the
musicians, the drop-curtain—even the smell of gas.
Would his theatre be large? What sort of curtain would it have? A
tiny hole must be cut in it at once—there was a peep-hole in the
curtain at the theatre. Had Grandmamma, or rather had Mamsell
Severin—for Grandmamma could not see to everything herself—
been able to find all the necessary scenery for Fidelio? He
determined to shut himself up to-morrow and give a performance all
by himself, and already in fancy he heard his little figures singing: for
he was approaching the theatre by way of his music.
“Exult, Jerusalem!” finished the choir; and their voices, following one
another in fugue form, united joyously in the last syllable. The clear
accord died away; deep silence reigned in the pillared hall and the
landscape-room. The elders looked down, oppressed by the pause;
only Director Weinschenk’s eyes roved boldly about, and Frau
Permaneder coughed her dry cough, which she could not suppress.
Now the Frau Consul moved slowly to the table and sat among her
family. She turned up the lamp and took in her hands the great Bible
with its edges of faded gold-leaf. She stuck her glasses on her nose,
unfastened the two great leather hasps of the book, opened it to the
place where there was a bookmark, took a sip of eau sucrée, and
began to read, from the yellowed page with the large print, the
Christmas chapter.
She read the old familiar words with a simple, heart-felt accent that
sounded clear and moving in the pious hush. “‘And to men good
will,’” she finished, and from the pillared hall came a trio of voices:
“Holy night, peaceful night!” The family in the landscape-room joined
in. They did so cautiously, for most of them were unmusical, as a
tone now and then betrayed. But that in no wise impaired the effect
of the old hymn. Frau Permaneder sang with trembling lips; it
sounded sweetest and most touching to the heart of her who had a
troubled life behind her, and looked back upon it in the brief peace of
this holy hour. Madame Kethelsen wept softly, but comprehended
nothing.
Now the Frau Consul rose. She grasped the hands of her grandson
Johann and her granddaughter Elisabeth, and proceeded through
the room. The elders of the family fell in behind, and the younger
brought up the rear; the servants and poor joined in from the hall;
and so they marched, singing with one accord “Oh, Evergreen”—
Uncle Christian sang “Oh, Everblue,” and made the children laugh by
lifting up his legs like a jumping-jack—through the wide-open, lofty
folding doors, and straight into Paradise.
The whole great room was filled with the fragrance of slightly singed
evergreen twigs and glowing with light from countless tiny flames.
The sky-blue hangings with the white figures on them added to the
brilliance. There stood the mighty tree, between the dark-red
window-curtains, towering nearly to the ceiling, decorated with silver
tinsel and large white lilies, with a shining angel at the top and the
manger at the foot. Its candles twinkled in the general flood of light
like far-off stars. And a row of tiny trees, also full of stars and hung
with comfits, stood on the long white table, laden with presents, that
stretched from the window to the door. All the gas-brackets on the
wall were lighted too, and thick candles burned in all four of the
gilded candelabra in the corners of the room. Large objects, too
large to stand upon the table, were arranged upon the floor, and two
smaller tables, likewise adorned with tiny trees and covered with gifts
for the servants and the poor, stood on either side of the door.
Dazzled by the light and the unfamiliar look of the room, they
marched once around it, singing, filed past the manger where lay the
little wax figure of the Christ-child, and then moved to their places
and stood silent.
Hanno was quite dazed. His fevered glance had soon sought out the
theatre, which, as it stood there upon the table, seemed larger and
grander than anything he had dared to dream of. But his place had
been changed—it was now opposite to where he had stood last year,
and this made him doubtful whether the theatre was really his. And
on the floor beneath it was something else, a large, mysterious
something, which had surely not been on his list; a piece of furniture,
that looked like a commode—could it be meant for him?
“Come here, my dear child,” said the Frau Consul, “and look at this.”
She lifted the lid. “I know you like to play chorals. Herr Pfühl will
show you how. You must tread all the time, sometimes more and
sometimes less; and then, not lift up the hands, but change the
fingers so, peu à peu.”
It was a harmonium—a pretty little thing of polished brown wood,
with metal handles at the sides, gay bellows worked with a treadle,
and a neat revolving stool. Hanno struck a chord. A soft organ tone
released itself and made the others look up from their presents. He
hugged his grandmother, who pressed him tenderly to her, and then
left him to receive the thanks of her other guests.
He turned to his theatre. The harmonium was an overpowering
dream—which just now he had no time to indulge. There was a
superfluity of joy; and he lost sight of single gifts in trying to see and
notice everything at once. Ah, here was the prompter’s box, a shell-
shaped one, and a beautiful red and gold curtain rolled up and down
behind it. The stage was set for the last act of Fidelio. The poor
prisoners stood with folded hands. Don Pizarro, in enormous puffed
sleeves, was striking a permanent and awesome attitude, and the
minister, in black velvet, approached from behind with hasty strides,
to turn all to happiness. It was just as in the theatre, only almost
more beautiful. The Jubilee chorus, the finale, echoed in Hanno’s
ears, and he sat down at the harmonium to play a fragment which
stuck in his memory. But he got up again, almost at once, to take up
the book he had wished for, a mythology, in a red binding with a gold
Pallas Athene on the cover. He ate some of the sweetmeats from his
plate full of marzipan, gingerbread, and other goodies, looked
through various small articles like writing utensils and school-bag—
and for the moment forgot everything else, to examine a penholder
with a tiny glass bulb on it: when you held this up to your eye, you
saw, like magic, a broad Swiss landscape.
Mamsell Severin and the maid passed tea and biscuits; and while
Hanno dipped and ate, he had time to look about. Every one stood
talking and laughing; they all showed each other their presents and
admired the presents of others. Objects of porcelain, silver, gold,
nickel, wood, silk, cloth, and every other conceivable material lay on
the table. Huge loaves of decorated gingerbread, alternating with
loaves of marzipan, stood in long rows, still moist and fresh. All the
presents made by Frau Permaneder were decorated with huge satin
bows.
Now and then some one came up to little Johann, put an arm across
his shoulders, and looked at his presents with the overdone, cynical
admiration which people manufacture for the treasures of children.
Uncle Christian was the only person who did not display this grown-
up arrogance. He sauntered over to his nephew’s place, with a
diamond ring on his finger, a present from his mother; and his
pleasure in the toy theatre was as unaffected as Hanno’s own.
“By George, that’s fine,” he said, letting the curtain up and down, and
stepping back for a view of the scenery. “Did you ask for it? Oh, so
you did ask for it!” he suddenly said after a pause, during which his
eyes had roved about the room as though he were full of unquiet
thoughts. “Why did you ask for it? What made you think of it? Have
you been in the theatre? Fidelio, eh? Yes, they give that well. And
you want to imitate it, do you? Do opera yourself, eh? Did it make
such an impression on you? Listen, son—take my advice: don’t think
too much about such things—theatre, and that sort of thing. It’s no
good. Believe your old uncle. I’ve always spent too much time on
them, and that is why I haven’t come to much good. I’ve made great
mistakes, you know.”
Thus he held forth to his nephew, while Hanno looked up at him
curiously. He paused, and his bony, emaciated face cleared up as he
regarded the little theatre. Then he suddenly moved forward one of
the figures on the stage, and sang, in a cracked and hollow tremolo,
“Ha, what terrible transgression!” He sat down on the piano-stool,
which he shoved up in front of the theatre, and began to give a
performance, singing all the rôles and the accompaniment as well,
and gesticulating furiously. The family gathered at his back, laughed,
nodded their heads, and enjoyed it immensely. As for Hanno, his
pleasure was profound. Christian broke off, after a while, very
abruptly. His face clouded, he rubbed his hand over his skull and
down his left side, and turned to his audience with his nose wrinkled
and his face quite drawn.
“There it is again,” he said. “I never have a little fun without having to
pay for it. It is not an ordinary pain, you know, it is a misery, down all
this left side, because the nerves are too short.”
But his relatives took his complaints as little seriously as they had his
entertainment. They hardly answered him, but indifferently
dispersed, leaving Christian sitting before the little theatre in silence.
He blinked rapidly for a bit and then got up.
“No, child,” said he, stroking Hanno’s head: “amuse yourself with it,
but not too much, you know: don’t neglect your work for it, do you
hear? I have made a great many mistakes.—I think I’ll go over to the
club for a while,” he said to the elders. “They are celebrating there
to-day, too. Good-bye for the present.” And he went off across the
hall, on his stiff, crooked legs.
They had all eaten the midday meal earlier than usual to-day, and
been hungry for the tea and biscuits. But they had scarcely finished
when great crystal bowls were handed round full of a yellow, grainy
substance which turned out to be almond cream. It was a mixture of
eggs, ground almonds, and rose-water, tasting perfectly delicious;
but if you ate even a tiny spoonful too much, the result was an attack
of indigestion. However, the company was not restrained by fear of
consequences—even though Frau Consul begged them to “leave a
little corner for supper.” Clothilde, in particular, performed miracles
with the almond cream, and lapped it up like so much porridge, with
heart-felt gratitude. There was also wine jelly in glasses, and English
plum-cake. Gradually they all moved over to the landscape-room,
where they sat with their plates round the table.
Hanno remained alone in the dining-room. Little Elisabeth
Weinschenk had already been taken home; but he was to stop up for
supper, for the first time in his life. The servants and the poor folk
had had their presents and gone; Ida Jungmann was chattering with
Riekchen Severin in the hall—although generally, as a governess,
she preserved a proper distance between herself and the Frau
Consul’s maid.—The lights of the great tree were burnt down and
extinguished, the manger was in darkness. But a few candles still
burned on the small trees, and now and then a twig came within
reach of the flame and crackled up, increasing the pungent smell in
the room. Every breath of air that stirred the trees stirred the pieces
of tinsel too, and made them give out a delicate metallic whisper. It
was still enough to hear the hand-organ again, sounding through the
frosty air from a distant street.
Hanno abandoned himself to the enjoyment of the Christmas sounds
and smells. He propped his head on his hand and read in his
mythology book, munching mechanically the while, because that was
proper to the day: marzipan, sweetmeats, almond cream, and plum-
cake; until the chest-oppression caused by an over-loaded stomach
mingled with the sweet excitation of the evening and gave him a
feeling of pensive felicity. He read about the struggles of Zeus before
he arrived at the headship of the gods; and every now and then he
listened into the other room, where they were going at length into the
future of poor Aunt Clothilde.
Clothilde, on this evening, was far and away the happiest of them all.
A smile lighted up her colourless face as she received
congratulations and teasing from all sides; her voice even broke now
and then out of joyful emotion. She had at last been made a member
of the Order of St. John. The Senator had succeeded by
subterranean methods in getting her admitted, not without some
private grumblings about nepotism, on the part of certain gentlemen.
Now the family all discussed the excellent institution, which was
similar to the homes in Mecklenburg, Dobberthien, and Ribnitz, for
ladies from noble families. The object of these establishments was
the suitable care of portionless women from old and worthy families.
Poor Clothilde was now assured of a small but certain income, which
would increase with the years, and finally, when she had succeeded
to the highest class, would secure her a decent home in the cloister
itself.
Little Hanno stopped awhile with the grown-ups, but soon strayed
back to the dining-room, which displayed a new charm now that the
brilliant light did not fairly dazzle one with its splendours. It was an
extraordinary pleasure to roam about there, as if on a half-darkened
stage after the performance, and see a little behind the scenes. He
touched the lilies on the big fir-tree, with their golden stamens;
handled the tiny figures of people and animals in the manger, found
the candles that lighted the transparency for the star of Bethlehem
over the stable; lifted up the long cloth that covered the present-
table, and saw quantities of wrapping-paper and pasteboard boxes
stacked beneath.
The conversation in the landscape-room was growing less and less
agreeable. Inevitably, irresistibly, it had arrived at the one dismal
theme which had been in everybody’s mind, but which they had thus
far avoided, as a tribute to the festal evening. Hugo Weinschenk
himself dilated upon it, with a wild levity of manner and gesture. He
explained certain details of the procedure—the examination of
witnesses had now been interrupted by the Christmas recess—
condemned the very obvious bias of the President, Dr. Philander,
and poured scorn on the attitude which the Public Prosecutor, Dr.
Hagenström, thought it proper to assume toward himself and the
witnesses for the defence. Breslauer had succeeded in drawing the
sting of several of his most slanderous remarks; and he had assured
the Director that, for the present, there need be no fear of a
conviction. The Senator threw in a question now and then, out of
courtesy; and Frau Permaneder, sitting on the sofa with elevated
shoulders, would utter fearful imprecations against Dr. Moritz
Hagenström. But the others were silent: so profoundly silent that the
Director at length fell silent too. For little Hanno, over in the dining-
room, the time sped by on angels’ wings; but in the landscape-room
there reigned an oppressive silence, which dragged on till Christian
came back from the club, where he had celebrated Christmas with
the bachelors and good fellows.
The cold stump of a cigar hung between his lips, and his haggard
cheeks were flushed. He came through the dining-room and said, as
he entered the landscape-room, “Well, children, the tree was simply
gorgeous. Weinschenk, we ought to have had Breslauer come to
see it. He has never seen anything like it, I am sure.”
He encountered one of his mother’s quiet, reproachful side-glances,
and returned it with an easy, unembarrassed questioning look. At
nine o’clock the party sat down to supper.
It was laid, as always on these occasions, in the pillared hall. The
Frau Consul recited the ancient grace with sincere conviction:
“Come, Lord Jesus, be our guest,
And bless the bread thou gavest us”
—to which, as usual on the holy evening, she added a brief prayer,
the substance of which was an admonition to remember those who,
on this blessed night, did not fare so well as the Buddenbrook family.
This accomplished, they all sat down with good consciences to a
lengthy repast, beginning with carp and butter sauce and old Rhine
wine.
The Senator put two fish-scales into his pocket, to help him save
money during the coming year. Christian, however, ruefully remarked
that he hadn’t much faith in the prescription; and Consul Kröger had
no need of it. His pittance had long since been invested securely,
beyond the reach of fluctuations in the exchange. The old man sat
as far away as possible from his wife, to whom he hardly ever spoke
nowadays. She persisted in sending money to Jacob, who was still
roaming about, nobody knew where, unless his mother did. Uncle
Justus scowled forbiddingly when the conversation, with the advent
of the second course, turned upon the absent members of the family,
and he saw the foolish mother wipe her eyes. They spoke of the
Frankfort Buddenbrooks and the Duchamps in Hamburg, and of
Pastor Tibertius in Riga, too, without any ill-will. And the Senator and
his sister touched glasses in silence to the health of Messrs Grünlich
and Permaneder—for, after all, did they not in a sense belong to the
family too?
The turkey, stuffed with chestnuts, raisins, and apples, was
universally praised. They compared it with other years, and decided
that this one was the largest for a long time. With the turkey came
roast potatoes and two kinds of compote, and each dish held enough
to satisfy the appetite of a family all by itself. The old red wine came
from the firm of Möllendorpf.
Little Johann sat between his parents and choked down with
difficulty a small piece of white meat with stuffing. He could not begin
to compete with Aunt Tilda, and he felt tired and out of sorts. But it
was a great thing none the less to be dining with the grown-ups, and
to have one of the beautiful little rolls with poppy-seed in his
elaborately folded serviette, and three wine-glasses in front of his
place. He usually drank out of the little gold mug which Uncle Justus
gave him. But when the red, white, and brown meringues appeared,
and Uncle Justus poured some oily, yellow Greek wine into the
smallest of the three glasses, his appetite revived. He ate a whole
red ice, then half a white one, then a little piece of the chocolate, his

Quantum Field Theory A Diagrammatic Approach 1St Edition Ronald Kleiss Online Ebook Texxtbook 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

Quantum Field Theory A Diagrammatic Approach 1St Edition Ronald Kleiss Online Ebook Texxtbook Full Chapter PDF

Uploaded by

Copyright:

Available Formats

Quantum Field Theory A Diagrammatic

Approach 1st Edition Ronald Kleiss

Quantum Field Theory A Diagrammatic Approach 2019 draft

Statistical Approach to Quantum Field Theory Andreas

Statistical Approach to Quantum Field Theory 2nd

Statistical Approach to Quantum Field Theory An

Relativistic Quantum Field Theory Michael T. Strickland

Quantum Field Theory 2nd Edition Michael V. Sadovskii

Quantum Field Theory An Introduction 1st Edition Gordon

Factorization Algebras in Quantum Field Theory 1st

Cambridge University Press is part of the University of Cambridge.

Inspired by Šimon Partlic

Preface page xiii

1 QFT in Zero Dimensions 1

2 The Loop Expansion and the Effective Action 19

4 More Fields in Zero Dimensions 56

5 QFT in Euclidean Spaces 66

5.2 One-Dimensional Continuum Theory 70

6 QFT in Minkowski Space 85

7 Scattering Processes 105

8 Introduction to Loop Calculations 134

9 More on Renormalization 147

10 Dirac Particles 162

11 Helicity Techniques for Dirac Particles 198

12 Vector Particles 213

13 Quantum Electrodynamics 228

14 Higher-Order Effects in QED 279

15 Quantum Chromodynamics 321

15.7 Synopsis of the Vertices for QCD 344

16 Higher-Order Effects in QCD 348

17 Electroweak Theory 362

18 More Example Computations 419

J Cruel and Unusual Actions 470

Why This Book?

1 Indeed, canonical quantization will not appear at all.

What Is in This Book?

Constants and Units

h = 6.62607015 × 10−34 kg m2 /sec. (2)

= h/(2π) ≈ 1.054571817 × 10−34 kg m2 /sec

Using and c, we can exchange energies for inverse lengths:

9 Manus mundissimus fabricae inutilis (inspired by S. Partlic).

c = 197.3269804 MeV fm, (4)

Fundamental HEP Units

16 Typically, how far you can reach.

The Step Function

gμν = gμν = diag(1, −1, −1, −1). (16)

The Levi–Civita Symbol

0123 = + 1 hence  0123 = − 1. (19)

This implies the following identities:

 μ (a, b, c) means  μνρσ aν bρ cσ , and so on.

Minus the Momentum

This explains why in nonrelativistic quantum mechanics the momentum operator is p =

26 Even permutations occur with a + and odd permutations with a – sign.

The Polyparenthetophobe Rules

In Favor of Loose Terminology

Exercise 0 Almost an exercise

1.2 Probabilistic Considerations

1.2.1 Green’s Functions and the Path Integral

Since from its definition, we have

1.2.2 The Free Theory

10 Try this out: “phi of J equals phi with J”.

1.2.3 The ϕ 4 Model and Perturbation Theory

G2n = H2n /H0 ,

0123 = + 1 hence 0123 = − 1. (19)

μ (a, b, c) means μνρσ aν bρ cσ , and so on.

This explains why in nonrelativistic quantum mechanics the momentum operator is p =

and then considering Z , Z , and so on.

ϕJ ≈ ϕc , where S (ϕc ) = J, S (ϕc ) > 0. (2.14)

(φ) = J dφ = J φ − φ dJ = J φ − W, (2.16)

where J is now to be interpreted as a function of φ. The transition from W(J) to (φ) is