String Theory University of Cambridge Part III Mathematical Tripos January 2009

String Theory University of Cambridge Part III Mathematical Tripos January 2009
January 2009
Preprint typeset in JHEP style - HYPER VERSION
String Theory
University of Cambridge Part III Mathematical Tripos
Dr David Tong
Department of Applied Mathematics and Theoretical Physics,
Centre for Mathematical Sciences,
Wilberforce Road,
Cambridge, CB3 OWA, UK
http://www.damtp.cam.ac.uk/user/tong/string.html
[email protected]
–1–
Recommended Books and Resources
• J. Polchinski, String Theory
This two volume work is the standard introduction to the subject. Our lectures will
more or less follow the path laid down in volume one covering the bosonic string. The
book contains explanations and descriptions of many details that have been deliberately
(and, I suspect, at times inadvertently) swept under a very large rug in these lectures.
Volume two covers the superstring.
• M. Green, J. Schwarz and E. Witten, Superstring Theory
Another two volume set. It is now over 20 years old and takes a slightly old-fashioned
route through the subject, with no explicit mention of conformal field theory. However, it does contain much good material and the explanations are uniformly excellent.
Volume one is most relevant for these lectures.
• B. Zwiebach, A First Course in String Theory
This book grew out of a course given to undergraduates who had no previous exposure
to general relativity or quantum field theory. It has wonderful pedagogical discussions
of the basics of lightcone quantization. More surprisingly, it also has some very clear
descriptions of several advanced topics, even though it misses out all the bits in between.
• P. Di Francesco, P. Mathieu and D. Sénéchal, Conformal Field Theory
This big yellow book is affectionately known as the yellow pages. It’s a great way
to learn conformal field theory. At first glance, it comes across as slightly daunting
because it’s big. (And yellow). But you soon realise that it’s big because it starts at
the beginning and provides detailed explanations at every step. The material necessary
for this course can be found in chapters 5 and 6.
Further References: “String Theory and M-Theory” by Becker, Becker and Schwarz
and “String Theory in a Nutshell” (it’s a big nutshell) by Kiritsis both deal with the
bosonic string fairly quickly, but include more advanced topics that may be of interest.
The book “D-Branes” by Johnson has lively and clear discussions about the many joys
of D-branes. Links to several excellent online resources, including video lectures by
Shiraz Minwalla, are listed on the course webpage.
Contents
0. Introduction
0.1 Quantum Gravity
1
3
1. The Relativistic String
1.1 The Relativistic Point Particle
1.1.1 Quantization
1.1.2 Ein Einbein
1.2 The Nambu-Goto Action
1.2.1 Symmetries of the Nambu-Goto Action
1.2.2 Equations of Motion
1.3 The Polyakov Action
1.3.1 Symmetries of the Polyakov Action
1.3.2 Fixing a Gauge
1.4 Mode Expansions
1.4.1 The Constraints Revisited
9
9
11
13
14
17
18
18
20
22
25
26
2. The Quantum String
2.1 A Lightning Look at Covariant Quantization
2.1.1 Ghosts
2.1.2 Constraints
2.2 Lightcone Quantization
2.2.1 Lightcone Gauge
2.2.2 Quantization
2.3 The String Spectrum
2.3.1 The Tachyon
2.3.2 The First Excited States
2.3.3 Higher Excited States
2.4 Lorentz Invariance Revisited
2.5 A Nod to the Superstring
28
28
30
30
32
33
36
40
40
41
45
46
48
3. Open Strings and D-Branes
3.1 Quantization
3.1.1 The Ground State
3.1.2 First Excited States: A World of Light
50
53
54
55
–1–
3.1.3 Higher Excited States and Regge Trajectories
3.1.4 Another Nod to the Superstring
Brane Dynamics: The Dirac Action
Multiple Branes: A World of Glue
56
56
57
59
4. Introducing Conformal Field Theory
4.0.1 Euclidean Space
4.0.2 The Holomorphy of Conformal Transformations
4.1 Classical Aspects
4.1.1 The Stress-Energy Tensor
4.1.2 Noether Currents
4.1.3 An Example: The Free Scalar Field
4.2 Quantum Aspects
4.2.1 Operator Product Expansion
4.2.2 Ward Identities
4.2.3 Primary Operators
4.3 An Example: The Free Scalar Field
4.3.1 The Propagator
4.3.2 An Aside: No Goldstone Bosons in Two Dimensions
4.3.3 The Stress-Energy Tensor and Primary Operators
4.4 The Central Charge
4.4.1 c is for Casimir
4.4.2 The Weyl Anomaly
4.4.3 c is for Cardy
4.4.4 c has a Theorem
4.5 The Virasoro Algebra
4.5.1 Radial Quantization
4.5.2 The Virasoro Algebra
4.5.3 Representations of the Virasoro Algebra
4.5.4 Consequences of Unitarity
4.6 The State-Operator Map
4.6.1 Some Simple Consequences
4.6.2 Our Favourite Example: The Free Scalar Field
4.7 Brief Comments on Conformal Field Theories with Boundaries
61
62
63
63
64
66
67
68
68
70
73
77
77
79
80
82
84
86
89
91
92
92
94
96
98
99
101
102
105
5. The Polyakov Path Integral and Ghosts
5.1 The Path Integral
5.1.1 The Faddeev-Popov Method
108
108
109
3.2
3.3
–2–
5.2
5.3
5.4
5.1.2 The Faddeev-Popov Determinant
5.1.3 Ghosts
The Ghost CFT
The Critical “Dimension” of String Theory
5.3.1 The Usual Nod to the Superstring
5.3.2 An Aside: Non-Critical Strings
States and Vertex Operators
5.4.1 An Example: Closed Strings in Flat Space
5.4.2 An Example: Open Strings in Flat Space
5.4.3 More General CFTs
112
113
114
117
118
119
120
122
123
124
6. String Interactions
6.1 What to Compute?
6.1.1 Summing Over Topologies
6.2 Closed String Amplitudes at Tree Level
6.2.1 Remnant Gauge Symmetry: SL(2,C)
6.2.2 The Virasoro-Shapiro Amplitude
6.2.3 Lessons to Learn
6.3 Open String Scattering
6.3.1 The Veneziano Amplitude
6.3.2 The Tension of D-Branes
6.4 One-Loop Amplitudes
6.4.1 The Moduli Space of the Torus
6.4.2 The One-Loop Partition Function
6.4.3 Interpreting the String Partition Function
6.4.4 So is String Theory Finite?
6.4.5 Beyond Perturbation Theory?
6.5 Appendix: Games with Integrals and Gamma Functions
125
125
127
130
130
132
135
139
141
142
143
143
146
149
152
153
154
7. Low Energy Effective Actions
7.1 Einstein’s Equations
7.1.1 The Beta Function
7.1.2 Ricci Flow
7.2 Other Couplings
7.2.1 Charged Strings and the B field
7.2.2 The Dilaton
7.2.3 Beta Functions
7.3 The Low-Energy Effective Action
157
158
159
163
163
163
165
167
167
–3–
7.4
7.5
7.6
7.7
7.3.1 String Frame and Einstein Frame
7.3.2 Corrections to Einstein’s Equations
7.3.3 Nodding Once More to the Superstring
Some Simple Solutions
7.4.1 Compactifications
7.4.2 The String Itself
7.4.3 Magnetic Branes
7.4.4 Moving Away from the Critical Dimension
7.4.5 The Elephant in the Room: The Tachyon
D-Branes Revisited: Background Gauge Fields
7.5.1 The Beta Function
7.5.2 The Born-Infeld Action
The DBI Action
7.6.1 Coupling to Closed String Fields
The Yang-Mills Action
7.7.1 D-Branes in Type II Superstring Theories
8. Compactification and T-Duality
8.1 The View from Spacetime
8.1.1 Moving around the Circle
8.2 The View from the Worldsheet
8.2.1 Massless States
8.2.2 Charged Fields
8.2.3 Enhanced Gauge Symmetry
8.3 Why Big Circles are the Same as Small Circles
8.3.1 A Path Integral Derivation of T-Duality
8.3.2 T-Duality for Open Strings
8.3.3 T-Duality for Superstrings
8.3.4 Mirror Symmetry
8.4 Epilogue
–4–
168
170
171
173
174
175
177
180
183
183
184
187
188
189
191
195
197
197
199
200
202
202
203
204
206
207
208
208
209
Acknowledgements
These lectures are aimed at beginning graduate students. They assume a working
knowledge of quantum field theory and general relativity. The lectures were given over
one semester and are based broadly on Volume one of the book by Joe Polchinski. I
inherited the course from Michael Green whose notes were extremely useful. I also
benefited enormously from the insightful and entertaining video lectures by Shiraz
Minwalla.
I’m grateful to Anirban Basu, Niklas Beisert, Joe Bhaseen, Diego Correa, Nick Dorey,
Michael Green, Anshuman Maharana, Malcolm Perry and Martin Schnabl for discussions and help with various aspects of these notes. I’m also grateful to the students,
especially Carlos Guedes, for their excellent questions and superhuman typo-spotting
abilities. Finally, my thanks to Alex Considine for infinite patience and understanding
over the weeks these notes were written. I am supported by the Royal Society.
–5–
0. Introduction
String theory is an ambitious project. It purports to be an all-encompassing theory
of the universe, unifying the forces of Nature, including gravity, in a single quantum
mechanical framework.
The premise of string theory is that, at the fundamental level, matter does not consist
of point-particles but rather of tiny loops of string. From this slightly absurd beginning,
the laws of physics emerge. General relativity, electromagnetism and Yang-Mills gauge
theories all appear in a surprising fashion. However, they come with baggage. String
theory gives rise to a host of other ingredients, most strikingly extra spatial dimensions
of the universe beyond the three that we have observed. The purpose of this course is
to understand these statements in detail.
These lectures differ from most other courses that you will take in a physics degree.
String theory is speculative science. There is no experimental evidence that string
theory is the correct description of our world and scant hope that hard evidence will
arise in the near future. Moreover, string theory is very much a work in progress and
certain aspects of the theory are far from understood. Unresolved issues abound and
it seems likely that the final formulation has yet to be written. For these reasons, I’ll
begin this introduction by suggesting some answers to the question: Why study string
theory?
Reason 1. String theory is a theory of quantum gravity
String theory unifies Einstein’s theory of general relativity with quantum mechanics.
Moreover, it does so in a manner that retains the explicit connection with both quantum
theory and the low-energy description of spacetime.
But quantum gravity contains many puzzles, both technical and conceptual. What
does spacetime look like at the shortest distance scales? How can we understand
physics if the causal structure fluctuates quantum mechanically? Is the big bang truely
the beginning of time? Do singularities that arise in black holes really signify the end
of time? What is the microscopic origin of black hole entropy and what is it telling
us? What is the resolution to the information paradox? Some of these issues will be
reviewed later in this introduction.
Whether or not string theory is the true description of reality, it offers a framework
in which one can begin to explore these issues. For some questions, string theory
has given very impressive and compelling answers. For others, string theory has been
almost silent.
–1–
Reason 2. String theory may be the theory of quantum gravity
With broad brush, string theory looks like an extremely good candidate to describe the
real world. At low-energies it naturally gives rise to general relativity, gauge theories,
scalar fields and chiral fermions. In other words, it contains all the ingredients that
make up our universe. It also gives the only presently credible explanation for the value
of the cosmological constant although, in fairness, I should add that the explanation is
so distasteful to some that the community is rather amusingly split between whether
this is a good thing or a bad thing. Moreover, string theory incorporates several ideas
which do not yet have experimental evidence but which are considered to be likely
candidates for physics beyond the standard model. Prime examples are supersymmetry
and axions.
However, while the broad brush picture looks good, the finer details have yet to
be painted. String theory does not provide unique predictions for low-energy physics
but instead offers a bewildering array of possibilities, mostly dependent on what is
hidden in those extra dimensions. Partly, this problem is inherent to any theory of
quantum gravity: as we’ll review shortly, it’s a long way down from the Planck scale
to the domestic energy scales explored at the LHC. Using quantum gravity to extract
predictions for particle physics is akin to using QCD to extract predictions for how
coffee makers work. But the mere fact that it’s hard is little comfort if we’re looking
for convincing evidence that string theory describes the world in which we live.
While string theory cannot at present offer falsifiable predictions, it has nonetheless
inspired new and imaginative proposals for solving outstanding problems in particle
physics and cosmology. There are scenarios in which string theory might reveal itself
in forthcoming experiments. Perhaps we’ll find extra dimensions at the LHC, perhaps
we’ll see a network of fundamental strings stretched across the sky, or perhaps we’ll
detect some feature of non-Gaussianity in the CMB that is characteristic of D-branes
at work during inflation. My personal feeling however is that each of these is a long
shot and we may not know whether string theory is right or wrong within our lifetimes.
Of course, the history of physics is littered with naysayers, wrongly suggesting that
various theories will never be testable. With luck, I’ll be one of them.
Reason 3. String theory provides new perspectives on gauge theories
String theory was born from attempts to understand the strong force. Almost forty
years later, this remains one of the prime motivations for the subject. String theory
provides tools with which to analyze down-to-earth aspects of quantum field theory
that are far removed from high-falutin’ ideas about gravity and black holes.
–2–
Of immediate relevance to this course are the pedagogical reasons to invest time in
string theory. At heart, it is the study of conformal field theory and gauge symmetry.
The techniques that we’ll learn are not isolated to string theory, but apply to countless
systems which have direct application to real world physics.
On a deeper level, string theory provides new and very surprising methods to understand aspects of quantum gauge theories. Of these, the most startling is the AdS/CFT
correspondence, first conjectured by Juan Maldacena, which gives a relationship between strongly coupled quantum field theories and gravity in higher dimensions. These
ideas have been applied in areas ranging from nuclear physics to condensed matter
physics and have provided qualitative (and arguably quantitative) insights into strongly
coupled phenomena.
Reason 4. String theory provides new results in mathematics
For the past 250 years, the close relationship between mathematics and physics has
been almost a one-way street: physicists borrowed many things from mathematicians
but, with a few noticeable exceptions, gave little back. In recent times, that has
changed. Ideas and techniques from string theory and quantum field theory have been
employed to give new “proofs” and, perhaps more importantly, suggest new directions
and insights in mathematics. The most well known of these is mirror symmetry, a
relationship between topologically different Calabi-Yau manifolds.
The four reasons described above also crudely characterize the string theory community: there are “relativists” and “phenomenologists” and “field theorists” and “mathematicians”. Of course, the lines between these different sub-disciplines are not fixed
and one of the great attractions of string theory is its ability to bring together people
working in different areas — from cosmology to condensed matter to pure mathematics
— and provide a framework in which they can profitably communicate. In my opinion,
it is this cross-fertilization between fields which is the greatest strength of string theory.
0.1 Quantum Gravity
This is a starter course in string theory. Our focus will be on the perturbative approach
to the bosonic string and, in particular, why this gives a consistent theory of quantum
gravity. Before we leap into this, it is probably best to say a few words about quantum
gravity itself. Like why it’s hard. And why it’s important. (And why it’s not).
The Einstein Hilbert action is given by
SEH
1
=
16πGN
Z
–3–
√
d4 x −gR
Newton’s constant GN can be written as
8πGN =
~c
Mpl2
Throughout these lectures we work in units with ~ = c = 1. The Planck mass Mpl
defines an energy scale
Mpl ≈ 2 × 1018 GeV .
(This is sometimes referred to as the reduced
Planck mass, to distinguish it from the
p
scale without the factor of 8π, namely 1/GN ≈ 1 × 1019 GeV).
There are a couple of simple lessons that we can already take from this. The first is
that the relevant coupling in the quantum theory is 1/Mpl . To see that this is indeed
the case from the perspective of the action, we consider small perturbations around flat
Minkowski space,
gµν = ηµν +
1
hµν
Mpl
The factor of 1/Mpl is there to ensure that when we expand out the Einstein-Hilbert
action, the kinetic term for h is canonically normalized, meaning that it comes with no
powers of Mpl . This then gives the kind of theory that you met in your first course on
quantum field theory, albeit with an infinite series of interaction terms,
Z
1
1
h (∂h)2 + 2 h2 (∂h)2 + . . .
SEH = d4 x (∂h)2 +
Mpl
Mpl
Each of these terms is schematic: if you were to do this explicitly, you would find a
mess of indices contracted in different ways. We see that the interactions are suppressed
by powers of Mpl . This means that quantum perturbation theory is an expansion in
the dimensionless ratio E 2 /Mpl2 , where E is the energy associated to the process of
interest. We learn that gravity is weak, and therefore under control, at low-energies.
But gravitational interactions become strong as the energy involved approaches the
Planck scale. In the language of the renormalization group, couplings of this type are
known as irrelevant.
The second lesson to take away is that the Planck scale Mpl is very very large. The
LHC will probe the electroweak scale, MEW ∼ 103 GeV. The ratio is MEW /Mpl ∼ 10−15 .
For this reason, quantum gravity will not affect your daily life, even if your daily life
involves the study of the most extreme observable conditions in the universe.
–4–
Gravity is Non-Renormalizable
Quantum field theories with irrelevant couplings are typically ill-behaved at highenergies, rendering the theory ill-defined. Gravity is no exception. Theories of this
type are called non-renormalizable, which means that the divergences that appear in
the Feynman diagram expansion cannot be absorbed by a finite number of counterterms. In pure Einstein gravity, the symmetries of the theory are enough to ensure that
the one-loop S-matrix is finite. The first divergence occurs at two-loops and requires
the introduction of a counterterm of the form,
Z
1 1
Γ∼
Mpl4
√
d4 x −g Rµνρσ Rρσλκ Rλκµν
with = 4 − D. All indications point towards the fact that this is the first in an infinite
number of necessary counterterms.
Coupling gravity to matter requires an interaction term of the form,
Z
Sint =
d4 x
1
hµν T µν + O(h2 )
Mpl
This makes the situation marginally worse, with the first divergence now appearing at one-loop. The Feynman diagram in the
figure shows particle scattering through the exchange of two gravitons. When the momentum k running in the loop is large, the
diagram is badly divergent: it scales as
1
Mpl4
Z
∞
Figure 1:
d4 k
Non-renormalizable theories are commonplace in the history of physics, the most commonly cited example being Fermi’s theory of the weak interaction. The first thing to say
about them is that they are far from useless! Non-renormalizable theories are typically
viewed as effective field theories, valid only up to some energy scale Λ. One deals with
the divergences by simply admitting ignorance beyond this scale and treating Λ as a
UV cut-off on any momentum integral. In this way, we get results which are valid to an
accuracy of E/Λ (perhaps raised to some power). In the case of the
p weak interaction,
Fermi’s theory accurately predicts physics up to an energy scale of 1/GF ∼ 100 GeV.
In the case of quantum gravity, Einstein’s theory works to an accuracy of (E/Mpl )2 .
–5–
However, non-renormalizable theories are typically unable to describe physics at their
cut-off scale Λ or beyond. This is because they are missing the true ultra-violet degrees
of freedom which tame the high-energy behaviour. In the case of the weak force, these
new degrees of freedom are the W and Z bosons. We would like to know what missing
degrees of freedom are needed to complete gravity.
Singularities
Only a particle physicist would phrase all questions about the universe in terms of
scattering amplitudes. In general relativity we typically think about the geometry as
a whole, rather than bastardizing the Einstein-Hilbert action and discussing perturbations around flat space. In this language, the question of high-energy physics turns into
one of short distance physics. Classical general relativity is not to be trusted in regions
where the curvature of spacetime approaches the Planck scale and ultimately becomes
singular. A quantum theory of gravity should resolve these singularities.
The question of spacetime singularities is morally equivalent to that of high-energy
scattering. Both probe the ultra-violet nature of gravity. A spacetime geometry is
made of a coherent collection of gravitons, just as the electric and magnetic fields in a
laser are made from a collection of photons. The short distance structure of spacetime
is governed – after Fourier transform – by high momentum gravitons. Understanding
spacetime singularities and high-energy scattering are different sides of the same coin.
There are two situations in general relativity where singularity theorems tell us that
the curvature of spacetime gets large: at the big bang and in the center of a black hole.
These provide two of the biggest challenges to any putative theory of quantum gravity.
Gravity is Subtle
It is often said that general relativity contains the seeds of its own destruction. The
theory is unable to predict physics at the Planck scale and freely admits to it. Problems
such as non-renormalizability and singularities are, in a Rumsfeldian sense, known
unknowns. However, the full story is more complicated and subtle. On the one hand,
the issue of non-renormalizability may not quite be the crisis that it first appears. On
the other hand, some aspects of quantum gravity suggest that general relativity isn’t
as honest about its own failings as is usually advertised. The theory hosts a number of
unknown unknowns, things that we didn’t even know that we didn’t know. We won’t
have a whole lot to say about these issues in this course, but you should be aware of
them. Here I mention only a few salient points.
–6–
Firstly, there is a key difference between Fermi’s theory of the weak interaction and
gravity. Fermi’s theory
was unable to provide predictions for any scattering process
p
at energies above 1/GF . In contrast, if we scatter two objects at extremely highenergies in gravity — say, at energies E Mpl — then we know exactly what will
happen: we form a big black hole. We don’t need quantum gravity to tell us this.
Classical general relativity is sufficient. If we restrict attention to scattering, the crisis
of non-renormalizability is not problematic at ultra-high energies. It’s troublesome only
within a window of energies around the Planck scale.
Similar caveats hold for singularities. If you are foolish enough to jump into a black
hole, then you’re on your own: without a theory of quantum gravity, no one can tell you
what fate lies in store at the singularity. Yet, if you are smart and stay outside of the
black hole, you’ll be hard pushed to see any effects of quantum gravity. This is because
Nature has conspired to hide Planck scale curvatures from our inquisitive eyes. In the
case of black holes this is achieved through cosmic censorship which is a conjecture in
classical general relativity that says singularities are hidden behind horizons. In the
case of the big bang, it is achieved through inflation, washing away any traces from the
very early universe. Nature appears to shield us from the effects of quantum gravity,
whether in high-energy scattering or in singularities. I think it’s fair to say that no one
knows if this conspiracy is pointing at something deep, or is merely inconvenient for
scientists trying to probe the Planck scale.
While horizons may protect us from the worst excesses of singularities, they come
with problems of their own. These are the unknown unknowns: difficulties that arise
when curvatures are small and general relativity says “trust me”. The entropy of black
holes and the associated paradox of information loss strongly suggest that local quantum field theory breaks down at macroscopic distance scales. Attempts to formulate
quantum gravity in de Sitter space, or in the presence of eternal inflation, hint at similar
difficulties. Ideas of holography, black hole complimentarity and the AdS/CFT correspondence all point towards non-local effects and the emergence of spacetime. These are
the deep puzzles of quantum gravity and their relationship to the ultra-violet properties
of gravity is unclear.
As a final thought, let me mention the one observation that has an outside chance of
being related to quantum gravity: the cosmological constant. With an energy scale of
Λ ∼ 10−3 eV it appears to have little to do with ultra-violet physics. If it does have its
origins in a theory of quantum gravity, it must either be due to some subtle “unknown
unknown”, or because it is explained away as an environmental quantity as in string
theory.
–7–
Is the Time Ripe?
Our current understanding of physics, embodied in the standard model, is valid up to
energy scales of 103 GeV. This is 15 orders of magnitude away from the Planck scale.
Why do we think the time is now ripe to tackle quantum gravity? Surely we are like
the ancient Greeks arguing about atomism. Why on earth do we believe that we’ve
developed the right tools to even address the question?
The honest answer, I think, is hubris.
However, there is mild circumstantial evidence
that the framework of quantum field theory might
hold all the way to the Planck scale without anything very dramatic happening in between. The
main argument is unification. The three coupling
constants of Nature run logarithmically, meeting
miraculously at the GUT energy scale of 1015 GeV.
Just slightly later, the fourth force of Nature, gravity, joins them. While not overwhelming, this does
provide a hint that perhaps quantum field theory
Figure 2:
can be taken seriously at these ridiculous scales.
Historically I suspect this was what convinced large parts of the community that it was
ok to speak about processes at 1018 GeV.
Finally, perhaps the most compelling argument for studying physics at the Planck
scale is that string theory does provide a consistent unified quantum theory of gravity
and the other forces. Given that we have this theory sitting in our laps, it would be
foolish not to explore its consequences. The purpose of these lecture notes is to begin
this journey.
–8–
1. The Relativistic String
All lecture courses on string theory start with a discussion of the point particle. Ours
is no exception. We’ll take a flying tour through the physics of the relativistic point
particle and extract a couple of important lessons that we’ll take with us as we move
onto string theory.
1.1 The Relativistic Point Particle
We want to write down the Lagrangian describing a relativistic particle of mass m.
In anticipation of string theory, we’ll consider D-dimensional Minkowski space R1,D−1 .
Throughout these notes, we work with signature
ηµν = diag(−1, +1, +1, . . . , +1)
Note that this is the opposite signature to my quantum field theory notes.
If we fix a frame with coordinates X µ = (t, ~x) the action is simple:
q
Z
S = −m dt 1 − ~x˙ · ~x˙ .
(1.1)
To see that this is correct we can compute the momentum p~, conjugate to ~x, and the
energy E which is equal to the Hamiltonian,
m~x˙
p~ = p
1 − ~x˙ · ~x˙
,
E=
p
m2 + p~2 ,
both of which should be familiar from courses on special relativity.
Although the Lagrangian (1.1) is correct, it’s not fully satisfactory. The reason is
that time t and space ~x play very different roles in this Lagrangian. The position ~x is
a dynamical degree of freedom. In contrast, time t is merely a parameter providing a
label for the position. Yet Lorentz transformations are supposed to mix up t and ~x and
such symmetries are not completely obvious in (1.1). Can we find a new Lagrangian
in which time and space are on equal footing?
One possibility is to treat both time and space as labels. This leads us to the
concept of field theory. However, in this course we will be more interested in the other
possibility: we will promote time to a dynamical degree of freedom. At first glance,
this may appear odd: the number of degrees of freedom is one of the crudest ways we
have to characterize a system. We shouldn’t be able to add more degrees of freedom
–9–
at will without fundamentally changing the system that we’re talking about. Another
way of saying this is that the particle has the option to move in space, but it doesn’t
have the option to move in time. It has to move in time. So we somehow need a way
to promote time to a degree of freedom without it really being a true dynamical degree
of freedom! How do we do this? The answer, as we will now show, is gauge symmetry.
Consider the action,
X0
Z
S = −m
q
dτ −Ẋ µ Ẋ ν ηµν ,
(1.2)
where µ = 0, . . . , D − 1 and Ẋ µ = dX µ /dτ . We’ve introduced a
new parameter τ which labels the position along the worldline of
the particle as shown by the dashed lines in the figure. This action
R
has a simple interpretation: it is just the proper time ds along the
worldline.
X
Figure 3:
Naively it looks as if we now have D physical degrees of freedom rather than D − 1
because, as promised, the time direction X 0 ≡ t is among our dynamical variables:
X 0 = X 0 (τ ). However, this is an illusion. To see why, we need to note that the action
(1.2) has a very important property: reparameterization invariance. This means that
we can pick a different parameter τ̃ on the worldline, related to τ by any monotonic
function
τ̃ = τ̃ (τ ) .
Let’s check that the action is invariant under transformations of this type. The integration measure in the action changes as dτ = dτ̃ |dτ /dτ̃ |. Meanwhile, the velocities
change as dX µ /dτ = (dX µ /dτ̃ ) (dτ̃ /dτ ). Putting this together, we see that the action
can just as well be written in the τ̃ reparameterization,
r
Z
dX µ dX ν
S = −m dτ̃ −
ηµν .
dτ̃ dτ̃
The upshot of this is that not all D degrees of freedom X µ are physical. For example,
suppose you find a solution to this system, so that you know how X 0 changes with
τ and how X 1 changes with τ and so on. Not all of that information is meaningful
because τ itself is not meaningful. In particular, we could use our reparameterization
invariance to simply set
τ = X 0 (τ ) ≡ t
– 10 –
(1.3)
If we plug this choice into the action (1.2) then we recover our initial action (1.1). The
reparameterization invariance is a gauge symmetry of the system. Like all gauge symmetries, it’s not really a symmetry at all. Rather, it is a redundancy in our description.
In the present case, it means that although we seem to have D degrees of freedom X µ ,
one of them is fake.
The fact that one of the degrees of freedom is a fake also shows up if we look at the
momenta,
pµ =
mẊ ν ηµν
∂L
=q
∂ Ẋ µ
−Ẋ λ Ẋ ρ ηλρ
(1.4)
These momenta aren’t all independent. They satisfy
pµ pµ + m2 = 0
(1.5)
This is a constraint on the system. It is, of course, the mass-shell constraint for a
relativistic particle of mass m. From the worldline perspective, it tells us that the
particle isn’t allowed to sit still in Minkowski space: at the very least, it had better
keep moving in a timelike direction with (p0 )2 ≥ m2 .
One advantage of the action (1.2) is that the Poincaré symmetry of the particle is
now manifest, appearing as a global symmetry on the worldline
X µ → Λµν X ν + cµ
(1.6)
where Λ is a Lorentz transformation satisfying Λµν η νρ Λσρ = η µσ , while cµ corresponds
to a constant translation. We have made all the symmetries manifest at the price of
introducing a gauge symmetry into our system. A similar gauge symmetry will arise
in the relativistic string and much of this course will be devoted to understanding its
consequences.
1.1.1 Quantization
It’s a trivial matter to quantize this action. We introduce a wavefunction Ψ(X). This
satisfies the usual Schrödinger equation,
i
∂Ψ
= HΨ .
∂τ
But, computing the Hamiltonian H = Ẋ µ pµ − L, we find that it vanishes: H = 0. This
shouldn’t be surprising. It is simply telling us that the wavefunction doesn’t depend on
– 11 –
τ . Since the wavefunction is something physical while, as we have seen, τ is not, this is
to be expected. Note that this doesn’t mean that time has dropped out of the problem.
On the contrary, in this relativistic context, time X 0 is an operator, just like the spatial
coordinates ~x. This means that the wavefunction Ψ is immediately a function of space
and time. It is not like a static state in quantum mechanics, but more akin to the fully
integrated solution to the non-relativistic Schrödinger equation.
The classical system has a constraint given by (1.5). In the quantum theory, we
impose this constraint as an operator equation on the wavefunction, namely (pµ pµ +
m2 )Ψ = 0. Using the usual representation of the momentum operator pµ = −i∂/∂X µ ,
we recognize this constraint as the Klein-Gordon equation
∂
∂
µν
2
−
(1.7)
η + m Ψ(X) = 0
∂X µ ∂X ν
Although this equation is familiar from field theory, it’s important to realize that the
interpretation is somewhat different. In relativistic field theory, the Klein-Gordon equation is the equation of motion obeyed by a scalar field. In relativistic quantum mechanics, it is the equation obeyed by the wavefunction. In the early days of field theory,
the fact that these two equations are the same led people to think one should view
the wavefunction as a classical field and quantize it a second time. This isn’t correct, but nonetheless the language has stuck and it is common to talk about the point
particle perspective as “first quantization” and the field theory perspective as “second
quantization”.
So far we’ve considered only a free point particle. How can we
introduce interactions into this framework? We would have to first
decide which interactions are allowed: perhaps the particle can split
into two; perhaps it can fuse with other particles? Obviously, there is
a huge range of options for us to choose from. We would then assign
Figure 4:
amplitudes for these processes to happen. There would be certain
restrictions coming from the requirement of unitarity which, among
other things, would lead to the necessity of anti-particles. We could draw diagrams
associated to the different interactions — an example is given in the figure — and in
this manner we would slowly build up the Feynman diagram expansion that is familiar
from field theory. In fact, this was pretty much the way Feynman himself approached
the topic of QED. However, in practice we rarely construct particle interactions in
this way because the field theory framework provides a much better way of looking at
things. In contrast, this way of building up interactions is exactly what we will later
do for strings.
– 12 –
1.1.2 Ein Einbein
There is another action that describes the relativistic point particle. We introduce yet
another field on the worldline, e(τ ), and write
Z
1
−1 2
2
S=
dτ e Ẋ − em ,
(1.8)
2
where we’ve used the notation Ẋ 2 = Ẋ µ Ẋ ν ηµν . For the rest of these lectures, terms
like X 2 will always mean an implicit contraction with the spacetime Minkowski metric.
This form of the action makes it look as if we have coupled the worldline theory to
1d gravity, with the field e(τ ) acting as an einbein (in the sense of vierbeins that are
introduced in general relativity). To see this, note that we could change notation and
write this action in the more suggestive form
Z
√
1
S=−
dτ −gτ τ g τ τ Ẋ 2 + m2 .
(1.9)
2
√
where gτ τ = (g τ τ )−1 is the metric on the worldline and e = −gτ τ
Although our action appears to have one more degree of freedom, e, it can be easily
checked that it has the same equations of motion as (1.2). The reason for this is that
e is completely fixed by its equation of motion, Ẋ 2 + e2 m2 = 0. Substituting this into
the action (1.8) recovers (1.2)
The action (1.8) has a couple of advantages over (1.2). Firstly, it works for massless
particles with m = 0. Secondly, the absence of the annoying square root means that
it’s easier to quantize in a path integral framework.
The action (1.8) retains invariance under reparameterizations which are now written
in a form that looks more like general relativity. For transformations parameterized by
an infinitesimal η, we have
τ → τ̃ = τ − η(τ ) ,
δe =
d
(η(τ )e) ,
dτ
δX µ =
dX µ
η(τ )
dτ
(1.10)
The einbein e transforms as a density on the worldline, while each of the coordinates
X µ transforms as a worldline scalar.
– 13 –
1.2 The Nambu-Goto Action
A particle sweeps out a worldline in Minkowski space. A string
sweeps out a worldsheet. We’ll parameterize this worldsheet by
one timelike coordinate τ , and one spacelike coordinate σ. In this
section we’ll focus on closed strings and take σ to be periodic,
with range
σ ∈ [0, 2π) .
τ
σ
(1.11)
We will sometimes package the two worldsheet coordinates together as σ α = (τ, σ), α = 0, 1. Then the string sweeps out a
Figure 5:
surface in spacetime which defines a map from the worldsheet to
Minkowski space, X µ (σ, τ ) with µ = 0, . . . , D − 1. For closed strings, we require
X µ (σ, τ ) = X µ (σ + 2π, τ ) .
In this context, spacetime is sometimes referred to as the target space to distinguish it
from the worldsheet.
We need an action that describes the dynamics of this string. The key property
that we will ask for is that nothing depends on the coordinates σ α that we choose
on the worldsheet. In other words, the string action should be reparameterization
invariant. What kind of action does the trick? Well, for the point particle the action
was proportional to the length of the worldline. The obvious generalization is that the
action for the string should be proportional to the area, A, of the worldsheet. This
is certainly a property that is characteristic of the worldsheet itself, rather than any
choice of parameterization.
How do we find the area A in terms of the coordinates X µ (σ, τ )? The worldsheet is
a curved surface embedded in spacetime. The induced metric, γαβ , on this surface is
the pull-back of the flat metric on Minkowski space,
γαβ =
∂X µ ∂X ν
ηµν .
∂σ α ∂σ β
(1.12)
Then the action which is proportional to the area of the worldsheet is given by,
Z
p
S = −T d2 σ − det γ .
(1.13)
Here T is a constant of proportionality. We will see shortly that it is the tension of the
string, meaning the mass per unit length.
– 14 –
We can write this action a little more explicitly. The pull-back of the metric is given
by,
γαβ =
Ẋ 2
Ẋ · X 0
Ẋ · X 0 X 02
!
.
where Ẋ µ = ∂X µ /∂τ and X µ 0 = ∂X µ /∂σ. The action then takes the form,
Z
q
2
S = −T d σ −(Ẋ)2 (X 0 )2 + (Ẋ · X 0 )2 .
(1.14)
This is the Nambu-Goto action for a relativistic string.
Action = Area: A Check
If you’re unfamiliar with differential geometry, the arguτ
ment about the pull-back of the metric may be a bit slick.
Thankfully, there’s a more pedestrian way to see that the
dl 2
action (1.14) is equal to the area swept out by the worldσ
dl1
sheet. It’s slightly simpler to make this argument for a surface embedded in Euclidean space rather than Minkowski
Figure 6:
space. We choose some parameterization of the sheet in
terms of τ and σ, as drawn in the figure, and we write the
~
coordinates of Euclidean space as X(σ,
τ ). We’ll compute the area of the infinitesimal
shaded region. The vectors tangent to the boundary are,
~
~ 1 = ∂X
dl
∂σ
,
~
~ 2 = ∂X .
dl
∂τ
If the angle between these two vectors is θ, then the area is then given by
q
q
2
2
2
2
~
~
~ 1 · dl
~ 2 )2
ds = |dl1 ||dl2 | sin θ = dl1 dl2 (1 − cos θ) = dl21 dl22 − (dl
(1.15)
which indeed takes the form of the integrand of (1.14).
Tension and Dimension
Let’s now see that T has the physical interpretation of tension. We write Minkowski
coordinates as X µ = (t, ~x). We work in a gauge with X 0 ≡ t = Rτ , where R is a
constant that is needed to balance up dimensions (see below) and will drop out at the
end of the argument. Consider a snapshot of a string configuration at a time when
– 15 –
d~x/dτ = 0 so that the instantaneous kinetic energy vanishes. Evaluating the action for
a time dt gives
Z
S = −T
Z
p
2
dτ dσR (d~x/dσ) = −T dt (spatial length of string) .
(1.16)
But, when the kinetic energy vanishes, the action is proportional to the time integral
of the potential energy,
potential energy = T × (spatial length of string) .
So T is indeed the energy per unit length as claimed. We learn that the string acts
rather like an elastic band and its energy increases linearly with length. (This is different
from the elastic bands you’re used to which obey Hooke’s law where energy increased
quadratically with length). To minimize its potential energy, the string will want to
shrink to zero size. We’ll see that when we include quantum effects this can’t happen
because of the usual zero point energies.
There is a slightly annoying way of writing the tension that has its origin in ancient
history, but is commonly used today
T =
1
2πα0
(1.17)
where α0 is pronounced “alpha-prime”. In the language of our ancestors, α0 is referred
to as the “universal Regge slope”. We’ll explain why later in this course.
At this point, it’s worth pointing out some conventions that we have, until now,
left implicit. The spacetime coordinates have dimension [X] = −1. In contrast, the
worldsheet coordinates are taken to be dimensionless, [σ] = 0. (This can be seen in our
identification σ ≡ σ + 2π). The tension is equal to the mass per unit length and has
dimension [T ] = 2. Obviously this means that [α0 ] = −2. We can therefore associate a
length scale, ls , by
α0 = ls2
(1.18)
The string scale ls is the natural length that appears in string theory. In fact, in a
certain sense (that we will make more precise later in the course) this length scale is
the only parameter of the theory.
– 16 –
Actual Strings vs. Fundamental Strings
There are several situations in Nature where string-like objects arise. Prime examples
include magnetic flux tubes in superconductors and chromo-electric flux tubes in QCD.
Cosmic strings, a popular speculation in cosmology, are similar objects, stretched across
the sky. In each of these situations, there are typically two length scales associated to
the string: the tension, T and the width of the string, L. For all these objects, the
dynamics is governed by the Nambu-Goto action as long as the curvature of the string is
much greater than L. (In the case of superconductors, one should work with a suitable
non-relativistic version of the Nambu-Goto action).
However, in each of these other cases, the Nambu-Goto action is not the end of the
story. There will typically be additional terms in the action that depend on the width
of the string. The form of these terms is not universal, but often includes a rigidity
R
piece of form L K 2 , where K is the extrinsic curvature of the worldsheet. Other
terms could be added to describe fluctuations in the width of the string.
The string scale, ls , or equivalently the tension, T , depends on the kind of string that
we’re considering. For example, if we’re interested in QCD flux tubes then we would
take
T ∼ (1 GeV)2
(1.19)
In this course we will consider fundamental strings which have zero width. What this
means in practice is that we take the Nambu-Goto action as the complete description
for all configurations of the string. These strings will have relevance to quantum gravity
and the tension of the string is taken to be much larger, typically an order of magnitude
or so below the Planck scale.
T . Mpl2 = (1018 GeV)2
(1.20)
However, I should point out that when we try to view string theory as a fundamental
theory of quantum gravity, we don’t really know what value T should take. As we
will see later in this course, it depends on many other aspects, most notably the string
coupling and the volume of the extra dimensions.
1.2.1 Symmetries of the Nambu-Goto Action
The Nambu-Goto action has two types of symmetry, each of a different nature.
• Poincaré invariance of the spacetime (1.6). This is a global symmetry from the
perspective of the worldsheet, meaning that the parameters Λµν and cµ which label
– 17 –
the symmetry transformation are constants and do not depend on worldsheet
coordinates σ α .
• Reparameterization invariance, σ α → σ̃ α (σ). As for the point particle, this is a
gauge symmetry. It reflects the fact that we have a redundancy in our description
because the worldsheet coordinates σ α have no physical meaning.
1.2.2 Equations of Motion
To derive the equations of motion for the Nambu-Goto string, we first introduce the
momenta which we call Π because there will be countless other quantities that we want
to call p later,
Πτµ
(Ẋ · X 0 )Xµ0 − (X 0 2 )Ẋµ
∂L
=
= −T q
.
∂ Ẋ µ
(Ẋ · X 0 )2 − Ẋ 2 X 0 2
Πσµ
(Ẋ · X 0 )Ẋµ − (Ẋ 2 )Xµ0
∂L
= −T q
.
=
∂X 0 µ
(Ẋ · X 0 )2 − Ẋ 2 X 0 2
The equations of motion are then given by,
∂Πτµ ∂Πσµ
+
=0
∂τ
∂σ
These look like nasty, non-linear equations. In fact, there’s a slightly nicer way to write
these equations, starting from the earlier action (1.13). Recall that the variation of a
√
√
determinant is δ −γ = 12 −γ γ αβ δγαβ . Using the definition of the pull-back metric
γαβ , this gives rise to the equations of motion
p
(1.21)
∂α ( − det γ γ αβ ∂β X µ ) = 0 ,
Although this notation makes the equations look a little nicer, we’re kidding ourselves.
Written in terms of X µ , they are still the same equations. Still nasty.
1.3 The Polyakov Action
The square-root in the Nambu-Goto action means that it’s rather difficult to quantize
using path integral techniques. However, there is another form of the string action
which is classically equivalent to the Nambu-Goto action. It eliminates the square root
at the expense of introducing another field,
Z
√
1
S=−
d2 σ −gg αβ ∂α X µ ∂β X ν ηµν
(1.22)
0
4πα
where g ≡ det g. This is the Polyakov action. (Polyakov didn’t discover the action, but
he understood how to work with it in the path integral and for this reason it carries
his name. The path integral treatment of this action will be the subject of Chapter 5).
– 18 –
The new field is gαβ . It is a dynamical metric on the worldsheet. From the perspective
of the worldsheet, the Polyakov action is a bunch of scalar fields X coupled to 2d gravity.
The equation of motion for X µ is
√
∂α ( −gg αβ ∂β X µ ) = 0 ,
(1.23)
which coincides with the equation of motion (1.21) from the Nambu-Goto action, except
that gαβ is now an independent variable which is fixed by its own equation of motion. To
√
√
determine this, we vary the action (remembering again that δ −g = − 21 −ggαβ δg αβ =
√
+ 21 −gg αβ δgαβ ),
T
δS = −
2
Z
d2 σ δg αβ
√
−g ∂α X µ ∂β X ν −
1
2
√
−g gαβ g ρσ ∂ρ X µ ∂σ X ν ηµν = 0 .(1.24)
The worldsheet metric is therefore given by,
gαβ = 2f (σ) ∂α X · ∂β X ,
(1.25)
where the function f (σ) is given by,
f −1 = g ρσ ∂ρ X · ∂σ X
A comment on the potentially ambiguous notation: here, and below, any function f (σ)
is always short-hand for f (σ, τ ): it in no way implies that f depends only on the spatial
worldsheet coordinate.
We see that gαβ isn’t quite the same as the pull-back metric γαβ defined in equation
(1.12); the two differ by the conformal factor f . However, this doesn’t matter because,
rather remarkably, f drops out of the equation of motion (1.23). This is because the
√
−g term scales as f , while the inverse metric g αβ scales as f −1 and the two pieces
cancel. We therefore see that Nambu-Goto and the Polyakov actions result in the same
equation of motion for X.
In fact, we can see more directly that the Nambu-Goto and Polyakov actions coincide.
We may replace gαβ in the Polyakov action (1.22) with its equation of motion gαβ =
2f γαβ . The factor of f also drops out of the action for the same reason that it dropped
out of the equation of motion. In this manner, we recover the Nambu-Goto action
(1.13).
– 19 –
1.3.1 Symmetries of the Polyakov Action
The fact that the presence of the factor f (σ, τ ) in (1.25) didn’t actually affect the
equations of motion for X µ reflects the existence of an extra symmetry which the
Polyakov action enjoys. Let’s look more closely at this. Firstly, the Polyakov action
still has the two symmetries of the Nambu-Goto action,
• Poincaré invariance. This is a global symmetry on the worldsheet.
X µ → Λµν X ν + cµ .
• Reparameterization invariance, also known as diffeomorphisms. This is a gauge
symmetry on the worldsheet. We may redefine the worldsheet coordinates as
σ α → σ̃ α (σ). The fields X µ transform as worldsheet scalars, while gαβ transforms
in the manner appropriate for a 2d metric.
X µ (σ) → X̃ µ (σ̃) = X µ (σ)
∂σ γ ∂σ δ
gγδ (σ)
gαβ (σ) → g̃αβ (σ̃) =
∂ σ̃ α ∂ σ̃ β
It will sometimes be useful to work infinitesimally. If we make the coordinate
change σ α → σ̃ α = σ α − η α (σ), for some small η. The transformations of the
fields then become,
δX µ (σ) = η α ∂α X µ
δgαβ (σ) = ∇α ηβ + ∇β ηα
where the covariant derivative is defined by ∇α ηβ = ∂α ηβ − Γσαβ ησ with the LeviCivita connection associated to the worldsheet metric given by the usual expression,
Γσαβ = 21 g σρ (∂α gβρ + ∂β gρα − ∂ρ gαβ )
Together with these familiar symmetries, there is also a new symmetry which is novel
to the Polyakov action. It is called Weyl invariance.
• Weyl Invariance.
changes as
Under this symmetry, X µ (σ) → X µ (σ), while the metric
gαβ (σ) → Ω2 (σ) gαβ (σ) .
Or, infinitesimally, we can write Ω2 (σ) = e2φ(σ) for small φ so that
δgαβ (σ) = 2φ(σ) gαβ (σ) .
– 20 –
(1.26)
It is simple to see that the Polyakov action is invariant under this transformation:
the factor of Ω2 drops out just as the factor of f did in equation (1.25), canceling
√
between −g and the inverse metric g αβ . This is a gauge symmetry of the string,
as seen by the fact that the parameter Ω depends on the worldsheet coordinates
σ. This means that two metrics which are related by a Weyl transformation (1.26)
are to be considered as the same physical state.
Figure 7: An example of a Weyl transformation
How should we think of Weyl invariance? It is not a coordinate change. Instead it is
the invariance of the theory under a local change of scale which preserves the angles
between all lines. For example the two worldsheet metrics shown in the figure are
viewed by the Polyakov string as equivalent. This is rather surprising! And, as you
might imagine, theories with this property are extremely rare. It should be clear from
the discussion above that the property of Weyl invariance is special to two dimensions,
√
for only there does the scaling factor coming from the determinant −g cancel that
coming from the inverse metric. But even in two dimensions, if we wish to keep Weyl
invariance then we are strictly limited in the kind of interactions that can be added to
the action. For example, we would not be allowed a potential term for the worldsheet
scalars of the form,
Z
√
d2 σ −g V (X) .
These break Weyl invariance. Nor can we add a worldsheet cosmological constant term,
Z
√
µ d2 σ −g .
This too breaks Weyl invariance. We will see later in this course that the requirement
of Weyl invariance becomes even more stringent in the quantum theory. We will also
see what kind of interactions terms can be added to the worldsheet. Indeed, much of
this course can be thought of as the study of theories with Weyl invariance.
– 21 –
1.3.2 Fixing a Gauge
As we have seen, the equation of motion (1.23) looks pretty nasty. However, we can use
the redundancy inherent in the gauge symmetry to choose coordinates in which they
simplify. Let’s think about what we can do with the gauge symmetry.
Firstly, we have two reparameterizations to play with. The worldsheet metric has
three independent components. This means that we expect to be able to set any two of
the metric components to a value of our choosing. We will choose to make the metric
locally conformally flat, meaning
gαβ = e2φ ηαβ ,
(1.27)
where φ(σ, τ ) is some function on the worldsheet. You can check that this is possible
by writing down the change of the metric under a coordinate transformation and seeing
that the differential equations which result from the condition (1.27) have solutions, at
least locally. Choosing a metric of the form (1.27) is known as conformal gauge.
We have only used reparameterization invariance to get to the metric (1.27). We still
have Weyl transformations to play with. Clearly, we can use these to remove the last
independent component of the metric and set φ = 0 such that,
gαβ = ηαβ .
(1.28)
We end up with the flat metric on the worldsheet in Minkowski coordinates.
A Diversion: How to make a metric flat
The fact that we can use Weyl invariance to make any two-dimensional metric flat is
an important result. Let’s take a quick diversion from our main discussion to see a
different proof that isn’t tied to the choice of Minkowski coordinates on the worldsheet.
We’ll work in 2d Euclidean space to avoid annoying minus signs. Consider two metrics
0
related by a Weyl transformation, gαβ
= e2φ gαβ . One can check that the Ricci scalars
of the two metrics are related by,
p 0 √
g 0 R = g(R − 2∇2 φ) .
(1.29)
We can therefore pick a φ such that the new metric has vanishing Ricci scalar, R0 = 0,
simply by solving this differential equation for φ. However, in two dimensions (but
not in higher dimensions) a vanishing Ricci scalar implies a flat metric. The reason is
simply that there aren’t too many indices to play with. In particular, symmetry of the
Riemann tensor in two dimensions means that it must take the form,
Rαβγδ =
R
(gαγ gβδ − gαδ gβγ ) .
2
– 22 –
0
So R0 = 0 is enough to ensure that Rαβγδ
= 0, which means that the manifold is flat. In
equation (1.28), we’ve further used reparameterization invariance to pick coordinates
in which the flat metric is the Minkowski metric.
The equations of motion and the stress-energy tensor
With the choice of the flat metric (1.28), the Polyakov action simplifies tremendously
and becomes the theory of D free scalar fields. (In fact, this simplification happens in
any conformal gauge).
Z
1
S=−
d2 σ ∂ α X · ∂ α X ,
(1.30)
4πα0
and the equations of motion for X µ reduce to the free wave equation,
∂α ∂ α X µ = 0 .
(1.31)
Now that looks too good to be true! Are the horrible equations (1.23) really equivalent
to a free wave equation? Well, not quite. There is something that we’ve forgotten:
we picked a choice of gauge for the metric gαβ . But we must still make sure that the
equation of motion for gαβ is satisfied. In fact, the variation of the action with respect
to the metric gives rise to a rather special quantity: it is the stress-energy tensor, Tαβ .
With a particular choice of normalization convention, we define the stress-energy tensor
to be
Tαβ = −
2 1
∂S
√
.
T −g ∂g αβ
We varied the Polyakov action with respect to gαβ in (1.24). When we set gαβ = ηαβ
we get
Tαβ = ∂α X · ∂β X − 12 ηαβ η ρσ ∂ρ X · ∂σ X .
(1.32)
The equation of motion associated to the metric gαβ is simply Tαβ = 0. Or, more
explicitly,
T01 = Ẋ · X 0 = 0
T00 = T11 = 12 (Ẋ 2 + X 0 2 ) = 0 .
(1.33)
We therefore learn that the equations of motion of the string are the free wave equations
(1.31) subject to the two constraints (1.33) arising from the equation of motion Tαβ = 0.
– 23 –
Getting a feel for the constraints
Let’s try to get some intuition for these constraints. There is a simple
meaning of the first constraint in (1.33): we must choose our parameterization such that lines of constant σ are perpendicular to the lines
of constant τ , as shown in the figure.
But we can do better. To gain more physical insight, we need to make
use of the fact that we haven’t quite exhausted our gauge symmetry.
We will discuss this more in Section 2.2, but for now one can check that
there is enough remnant gauge symmetry to allow us to go to static
gauge,
Figure 8:
X 0 ≡ t = Rτ ,
so that (X 0 )0 = 0 and Ẋ 0 = R, where R is a constant that is needed on dimensional
grounds. The interpretation of this constant will become clear shortly. Then, writing
X µ = (t, ~x), the equation of motion for spatial components is the free wave equation,
¨ − ~x 00 = 0
~x
while the constraints become
~x˙ · ~x 0 = 0
~x˙ 2 + ~x 0 2 = R2
(1.34)
The first constraint tells us that the motion of the string must be perpendicular to the
string itself. In other words, the physical modes of the string are transverse oscillations.
There is no longitudinal mode. We’ll also see this again in Section 2.2.
From the second constraint, we can understand the meaning of the constant R: it is
related to the length of the string when ~x˙ = 0,
Z
p
dσ (d~x/dσ)2 = 2πR .
Of course, if we have a stretched string with ~x˙ = 0 at one moment of time, then it won’t
stay like that for long. It will contract under its own tension. As this happens, the
second constraint equation relates the length of the string to the instantaneous velocity
of the string.
– 24 –
1.4 Mode Expansions
Let’s look at the equations of motion and constraints more closely. The equations of
motion (1.31) are easily solved. We introduce lightcone coordinates on the worldsheet,
σ± = τ ± σ ,
in terms of which the equations of motion simply read
∂+ ∂− X µ = 0
The most general solution is,
X µ (σ, τ ) = XLµ (σ + ) + XRµ (σ − )
for arbitrary functions XLµ and XRµ . These describe left-moving and right-moving waves
respectively. Of course the solution must still obey both the constraints (1.33) as well
as the periodicity condition,
X µ (σ, τ ) = X µ (σ + 2π, τ ) .
The most general, periodic solution can be expanded in Fourier modes,
r
α0 X 1 µ −inσ+
µ
+
µ
0
µ
+
XL (σ ) = 21 x + 12 α p σ + i
α̃ e
,
2 n6=0 n n
r
α0 X 1 µ −inσ−
µ
α e
.
XR (σ − ) = 12 xµ + 12 α0 pµ σ − + i
2 n6=0 n n
(1.35)
(1.36)
This mode expansion will be very important when we come to the quantum theory.
Let’s make a few simple comments here.
• Various normalizations in this expression, such as the α0 and factor of 1/n have
been chosen for later convenience.
• XL and XR do not individually satisfy the periodicity condition (1.35) due to the
terms linear in σ ± . However, the sum of them is invariant under σ → σ + 2π as
required.
• The variables xµ and pµ are the position and momentum of the center of mass of
the string. This can be checked, for example, by studying the Noether currents
arising from the spacetime translation symmetry X µ → X µ + cµ . One finds that
the conserved charge is indeed pµ .
• Reality of X µ requires that the coefficients of the Fourier modes, αnµ and α̃nµ , obey
µ ?
αnµ = (α−n
)
,
– 25 –
µ ?
) .
α̃nµ = (α̃−n
(1.37)
1.4.1 The Constraints Revisited
We still have to impose the two constraints (1.33). In the worldsheet lightcone coordinates σ ± , these become,
(∂+ X)2 = (∂− X)2 = 0 .
(1.38)
These equations give constraints on the momenta pµ and the Fourier modes αnµ and α̃nµ .
To see what these are, let’s look at
r
α0 µ
α0 X µ −inσ−
µ
µ
∂ − X = ∂ − XR =
p +
α e
2
2 n6=0 n
r
α0 X µ −inσ−
α e
=
2 n n
where in the second line the sum is over all n ∈ Z and we have defined α0µ to be
r
α0 µ
µ
α0 ≡
p .
2
The constraint (1.38) can then be written as
α0 X
−
αm · αp e−i(m+p)σ
2 m,p
α0 X
−
αm · αn−m e−inσ
=
2 m,n
X
−
≡ α0
Ln e−inσ = 0 .
(∂− X)2 =
n
where we have defined the sum of oscillator modes,
1X
Ln =
αn−m · αm .
2 m
(1.39)
We can also do the same for the left-moving modes, where we again define an analogous
sum of operator modes,
1X
L̃n =
α̃n−m · α̃m .
(1.40)
2 m
with the zero mode defined to be,
r
α̃0µ ≡
α0 µ
p .
2
– 26 –
The fact that α̃0µ = α0µ looks innocuous but is a key point to remember when we come
to quantize the string. The Ln and L̃n are the Fourier modes of the constraints. Any
classical solution of the string of the form (1.36) must further obey the infinite number
of constraints,
Ln = L̃n = 0
n∈Z.
We’ll meet these objects Ln and L̃n again in a more general context when we come to
discuss conformal field theory.
The constraints arising from L0 and L̃0 have a rather special interpretation. This is
because they include the square of the spacetime momentum pµ . But, the square of the
spacetime momentum is an important quantity in Minkowski space: it is the square of
the rest mass of a particle,
pµ pµ = −M 2 .
So the L0 and L̃0 constraints tell us the effective mass of a string in terms of the excited
oscillator modes, namely
M2 =
4 X
4 X
α
·
α
=
α̃n · α̃−n
n
−n
α0 n>0
α0 n>0
(1.41)
p
Because both α0µ and α̃0µ are equal to α0 /2 pµ , we have two expressions for the invariant
mass: one in terms of right-moving oscillators αnµ and one in terms of left-moving
oscillators α̃nµ . And these two terms must be equal to each other. This is known as
level matching. It will play an important role in the next section where we turn to the
quantum theory.
– 27 –
2. The Quantum String
Our goal in this section is to quantize the string. We have seen that the string action
involves a gauge symmetry and whenever we wish to quantize a gauge theory we’re
presented with a number of different ways in which we can proceed. If we’re working
in the canonical formalism, this usually boils down to one of two choices:
• We could first quantize the system and then subsequently impose the constraints
that arise from gauge fixing as operator equations on the physical states of the
system. For example, in QED this is the Gupta-Bleuler method of quantization
that we use in Lorentz gauge. In string theory it consists of treating all fields X µ ,
including time X 0 , as operators and imposing the constraint equations (1.33) on
the states. This is usually called covariant quantization.
• The alternative method is to first solve all of the constraints of the system to
determine the space of physically distinct classical solutions. We then quantize
these physical solutions. For example, in QED, this is the way we proceed in
Coulomb gauge. Later in this chapter, we will see a simple way to solve the
constraints of the free string.
Of course, if we do everything correctly, the two methods should agree. Usually, each
presents a slightly different challenge and offers a different viewpoint.
In these lectures, we’ll take a brief look at the first method of covariant quantization.
However, at the slightest sign of difficulties, we’ll bail! It will be useful enough to
see where the problems lie. We’ll then push forward with the second method described
above which is known as lightcone quantization in string theory. Although we’ll succeed
in pushing quantization through to the end, our derivations will be a little cheap and
unsatisfactory in places. In Section 5 we’ll return to all these issues, armed with more
sophisticated techniques from conformal field theory.
2.1 A Lightning Look at Covariant Quantization
We wish to quantize D free scalar fields X µ whose dynamics is governed by the action
(1.30). We subsequently wish to impose the constraints
Ẋ · X 0 = Ẋ 2 + X 0 2 = 0 .
(2.1)
The first step is easy. We promote X µ and their conjugate momenta Πµ = (1/2πα0 )Ẋµ
to operator valued fields obeying the canonical equal-time commutation relations,
[X µ (σ, τ ), Πν (σ 0 , τ )] = iδ(σ − σ 0 ) δ µν
µ
ν
0
0
,
[X (σ, τ ), X (σ , τ )] = [Πµ (σ, τ ), Πν (σ , τ )] = 0 .
– 28 –
We translate these into commutation relations for the Fourier modes xµ , pµ , αnµ and
α̃nµ . Using the mode expansion (1.36) we find
[xµ , pν ] = iδ µν
ν
ν
] = n η µν δn+m, 0 ,
] = [α̃nµ , α̃m
and [αnµ , αm
(2.2)
with all others zero. The commutation relations for xµ and pµ are expected for operators governing the position and momentum of the center of mass of the string. The
commutation relations of αnµ and α̃nµ are those of harmonic oscillator creation and annihilation operators in disguise. And the disguise isn’t that good. We just need to define
(ignoring the µ index for now)
αn
α−n
an = √
, a†n = √
n>0
(2.3)
n
n
Then (2.2) gives the familiar [an , a†m ] = δmn . So each scalar field gives rise to two infinite
towers of creation and annihilation operators, with αn acting as a rescaled annihilation
operator for n > 0 and as a creation operator for n < 0. There are two towers because
we have right-moving modes αn and left-moving modes α̃n .
With these commutation relations in hand we can now start building the Fock space
of our theory. We introduce a vacuum state of the string |0i, defined to obey
αnµ |0i = α̃nµ |0i = 0
for n > 0
(2.4)
The vacuum state of string theory has a different interpretation from the analogous
object in field theory. This is not the vacuum state of spacetime. It is instead the
vacuum state of a single string. This is reflected in the fact that the operators xµ
and pµ give extra structure to the vacuum. The true ground state of the string is |0i,
tensored with a spatial wavefunction Ψ(x). Alternatively, if we work in momentum
space, the vacuum carries another quantum number, pµ , which is the eigenvalue of the
momentum operator. We should therefore write the vacuum as |0; pi, which still obeys
(2.4), but now also
p̂µ |0; pi = pµ |0; pi
(2.5)
where (for the only time in these lecture notes) we’ve put a hat on the momentum
operator p̂µ on the left-hand side of this equation to distinguish it from the eigenvalue
pµ on the right-hand side.
We can now start to build up the Fock space by acting with creation operators αnµ
and α̃nµ with n < 0. A generic state comes from acting with any number of these
creation operators on the vacuum,
µ2 nµ2
µ1 nµ1
ν1 nν1
ν2 nν2
(α−1
) (α−2
) . . . (α̃−1
) (α̃−2
) . . . |0; pi
– 29 –
Each state in the Fock space is a different excited state of the string. Each has the
interpretation of a different species of particle in spacetime. We’ll see exactly what
particles they are shortly. But for now, notice that because there’s an infinite number
of ways to excite a string there are an infinite number of different species of particles
in this theory.
2.1.1 Ghosts
There’s a problem with the Fock space that we’ve constructed: it doesn’t have positive
norm. The reason for this is that one of the scalar fields, X 0 , comes with the wrong sign
kinetic term in the action (1.30). From the perspective of the commutation relations,
this issue raises its head in presence of the spacetime Minkowski metric in the expression
ν†
[αnµ , αm
] = n η µν δn,m .
This gives rise to the offending negative norm states, which come with an odd number
of timelike oscillators excited, for example
0
hp0 ; 0|α10 α−1
|0; pi ∼ −δ D (p − p0 )
This is the first problem that arises in the covariant approach to quantization. States
with negative norm are referred to as ghosts. To make sense of the theory, we have
to make sure that they can’t be produced in any physical processes. Of course, this
problem is familiar from attempts to quantize QED in Lorentz gauge. In that case,
gauge symmetry rides to the rescue since the ghosts are removed by imposing the gauge
fixing constraint. We must hope that the same happens in string theory.
2.1.2 Constraints
Although we won’t push through with this programme at the present time, let us briefly
look at what kind of constraints we have in string theory. In terms of Fourier modes,
the classical constraints can be written as Ln = L̃n = 0, where
Ln =
1X
αn−m · αm
2 m
and similar for L̃n . As in the Gupta-Bleuler quantization of QED, we don’t impose all
of these as operator equations on the Hilbert space. Instead we only require that the
operators Ln and L̃n have vanishing matrix elements when sandwiched between two
physical states |physi and |phys0 i,
hphys0 |Ln |physi = hphys0 |L̃n |physi = 0
– 30 –
Because L†n = L−n , it is therefore sufficient to require
Ln |physi = L̃n |physi = 0
for n > 0
(2.6)
However, we still haven’t explained how to impose the constraints L0 and L̃0 . And
these present a problem that doesn’t arise in the case of QED. The problem is that,
unlike for Ln with n 6= 0, the operator L0 is not uniquely defined when we pass to the
quantum theory. There is an operator ordering ambiguity arising from the commutation relations (2.2). Commuting the αnµ operators past each other in L0 gives rise to
extra constant terms.
Question: How do we know what order to put the αnµ operators in the quantum
operator L0 ? Or the α̃nµ operators in L̃0 ?
Answer: We don’t! Yet. Naively it looks as if each different choice will define a
different theory when we impose the constraints. To make this ambiguity manifest, for
now let’s just pick a choice of ordering. We define the quantum operators to be normal
ordered, with the annihilation operators αni , n > 0, moved to the right,
∞
X
1
α−m · αm + α02
L0 =
2
m=1
,
L̃0 =
∞
X
α̃−m · α̃m +
m=1
1 2
α̃
2 0
Then the ambiguity rears its head in the different constraint equations that we could
impose, namely
(L0 − a)|physi = (L̃0 − a)|physi = 0
(2.7)
for some constant a.
As we saw classically, the operators L0 and L̃0 play an important role in determining
the spectrum
pof the µstring because they include a term quadratic in the momentum
µ
µ
α0 = α̃0 = α0 /2 p . Combining the expression (1.41) with our constraint equation
for L0 and L̃0 , we find the spectrum of the string is given by,
!
!
∞
∞
X
X
4
4
2
M = 0 −a +
α−m · αm = 0 −a +
α̃−m · α̃m
α
α
m=1
m=1
We learn therefore that the undetermined constant a has a direct physical effect: it
changes the mass spectrum of the string. In the quantum theory, the sums over αnµ
modes are related to the number operators for the harmonic oscillator: they count the
number of excited modes of the string. The level matching in the quantum theory
tells us that the number of left-moving modes must equal the number of right-moving
modes.
– 31 –
Ultimately, we will find that the need to decouple the ghosts forces us to make a
unique choice for the constant a. (Spoiler alert: it turns out to be a = 1). In fact, the
requirement that there are no ghosts is much stronger than this. It also restricts the
number of scalar fields that we have in the theory. (Another spoiler: D = 26). If you’re
interested in how this works in covariant formulation then you can read about it in the
book by Green, Schwarz and Witten. Instead, we’ll show how to quantize the string
and derive these values for a and D in lightcone gauge. However, after a trip through
the world of conformal field theory, we’ll come back to these ideas in a context which
is closer to the covariant approach.
2.2 Lightcone Quantization
We will now take the second path described at the beginning of this section. We will
try to find a parameterization of all classical solutions of the string. This is equivalent
to finding the classical phase space of the theory. We do this by solving the constraints
(2.1) in the classical theory, leaving behind only the physical degrees of freedom.
Recall that we fixed the gauge to set the worldsheet metric to
gαβ = ηαβ .
However, this isn’t the end of our gauge freedom. There still remain gauge transformations which preserve this choice of metric. In particular, any coordinate transformation
σ → σ̃(σ) which changes the metric by
ηαβ → Ω2 (σ)ηαβ ,
(2.8)
can be undone by a Weyl transformation. What are these coordinate transformations?
It’s simplest to answer this using lightcone coordinates on the worldsheet,
σ± = τ ± σ ,
(2.9)
where the flat metric on the worldsheet takes the form,
ds2 = −dσ + dσ −
In these coordinates, it’s clear that any transformation of the form
σ + → σ̃ + (σ + )
,
σ − → σ̃ − (σ − ) ,
(2.10)
simply multiplies the flat metric by an overall factor (2.8) and so can be undone by
a compensating Weyl transformation. Some quick comments on this surviving gauge
symmetry:
– 32 –
• Recall that in Section 1.3.2 we used the argument that 3 gauge invariances (2
reparameterizations + 1 Weyl) could be used to fix 3 components of the worldsheet metric gαβ . What happened to this argument? Why do we still have some
gauge symmetry left? The reason is that σ̃ ± are functions of just a single variable,
not two. So we did fix nearly all our gauge symmetries. What is left is a set of
measure zero amongst the full gauge symmetry that we started with.
• The remaining reparameterization invariance (2.10) has an important physical
implication. Recall that the solutions to the equations of motion are of the form
XLµ (σ + ) + XRµ (σ − ) which looks like 2D functions worth of solutions. Of course,
we still have the constraints which, in terms of σ ± , read
(∂+ X)2 = (∂− X)2 = 0 ,
(2.11)
which seems to bring the number down to 2(D − 1) functions. But the reparameterization invariance (2.10) tells us that even some of these are fake since we
can always change what we mean by σ ± . The physical solutions of the string are
therefore actually described by 2(D − 2) functions. But this counting has a nice
interpretation: the degrees of freedom describe the transverse fluctuations of the
string.
• The above comment reaches the same conclusion as the discussion in Section
1.3.2. There, in an attempt to get some feel for the constraints, we claimed that
we could go to static gauge X 0 = Rτ for some dimensionful parameter R. It
is easy to check that this is simple to do using reparameterizations of the form
(2.10). However, to solve the string constraints in full, it turns out that static
gauge is not that useful. Rather we will use something called “lightcone gauge”.
2.2.1 Lightcone Gauge
We would like to gauge fix the remaining reparameterization invariance (2.10). The best
way to do this is called lightcone gauge. In counterpoint to the worldsheet lightcone
coordinates (2.9), we introduce the spacetime lightcone coordinates,
r
1 0
±
X =
(X ± X D−1 ) .
(2.12)
2
Note that this choice picks out a particular time direction and a particular spatial
direction. It means that any calculations that we do involving X ± will not be manifestly
Lorentz invariant. You might think that we needn’t really worry about this. We could
try to make the following argument: “The equations may not look Lorentz invariant
– 33 –
but, since we started from a Lorentz invariant theory, at the end of the day any physical
process is guaranteed to obey this symmetry”. Right?! Well, unfortunately not. One
of the more interesting and subtle aspects of quantum field theory is the possibility of
anomalies: these are symmetries of the classical theory that do not survive the journey
of quantization. When we come to the quantum theory, if our equations don’t look
Lorentz invariant then there’s a real possibility that it’s because the underlying physics
actually isn’t Lorentz invariant. Later we will need to spend some time figuring out
under what circumstances our quantum theory keeps the classical Lorentz symmetry.
In lightcone coordinates, the spacetime Minkowski metric reads
2
−
+
ds = −2dX dX +
D−2
X
dX i dX i
i=1
This means that indices are raised and lowered with A+ = −A− and A− = −A+ and
Ai = Ai . The product of spacetime vectors reads A · B = −A+ B − − A− B + + Ai B i .
Let’s look at the solution to the equation of motion for X + . It reads,
X + = XL+ (σ + ) + XR+ (σ − ) .
We now gauge fix. We use our freedom of reparameterization invariance to choose
coordinates such that
XL+ = 12 x+ + 12 α0 p+ σ +
,
XR+ = 12 x+ + 12 α0 p+ σ − .
You might think that we could go further and eliminate p+
and x+ but this isn’t possible because we don’t quite have
the full freedom of reparameterization invariance since all
functions should remain periodic in σ. The upshot of this
choice of gauge is that
X + = x+ + α 0 p + τ .
X+
(2.13)
This is lightcone gauge. Notice that, as long as p+ 6= 0, we
can always shift x+ by a shift in τ .
Figure 9:
There’s something a little disconcerting about the choice
(2.13). We’ve identified a timelike worldsheet coordinate with a null spacetime coordinate. Nonetheless, as you can see from the figure, it seems to be a good parameterization of the worldsheet. One could imagine that the parameterization might break
if the string is actually massless and travels in the X − direction, with p+ = 0. But
otherwise, all should be fine.
– 34 –
Solving for X −
The choice (2.13) does the job of fixing the reparameterization invariance (2.10). As
we will now see, it also renders the constraint equations trivial. The first thing that we
have to worry about is the possibility of extra constraints arising from this new choice
of gauge fixing. This can be checked by looking at the equation of motion for X + ,
∂+ ∂− X − = 0
But we can solve this by the usual ansatz,
X − = XL− (σ + ) + XR− (σ − ) .
We’re still left with all the other constraints (2.11). Here we see the real benefit of
working in lightcone gauge (which is actually what makes quantization possible at all):
X − is completely determined by these constraints. For example, the first of these reads
−
+
D−2
X
∂+ X i ∂+ X i
(2.14)
∂+ XL−
D−2
1 X
∂+ X i ∂+ X i .
= 0 +
α p i=1
(2.15)
∂− XR−
D−2
1 X
∂− X i ∂− X i .
= 0 +
α p i=1
(2.16)
2∂+ X ∂+ X =
i=1
which, using (2.13), simply becomes
Similarly,
So, up to an integration constant, the function X − (σ + , σ − ) is completely determined
−
in terms of the other fields. If we write the usual mode expansion for XL/R
r
α0 X 1 − −inσ+
α̃ e
,
XL− (σ + ) = 12 x− + 12 α0 p− σ + + i
2 n6=0 n n
r
α0 X 1 − −inσ−
.
XR− (σ − ) = 12 x− + 12 α0 p− σ − + i
α e
2 n6=0 n n
then x− is the undetermined integration constant, while p− , αn− and α̃n− are all fixed by
the constraints (2.15) and (2.16). For example, the oscillator modes αn− are given by,
r
+∞ D−2
1 1 X X i
−
α
αi ,
(2.17)
αn =
2α0 p+ m=−∞ i=1 n−m m
– 35 –
A special case of this is the α0− =
p
α0 /2 p− equation, which reads
D−2
α 0 p−
1 X
= +
2
2p i=1
!
1
2
α 0 pi pi +
X
i
αni α−n
.
(2.18)
n6=0
We also get another equation for p− from the α̃0− equation arising from (2.15)
D−2
α 0 p−
1 X
= +
2
2p i=1
!
1
2
α 0 pi pi +
X
i
α̃ni α̃−n
.
(2.19)
n6=0
From these two equations, we can reconstruct the old, classical, level matching conditions (1.41). But now with a difference:
2
+ −
M = 2p p −
D−2
X
i=1
D−2
D−2
4 XX i i
4 XX i i
α α =
α̃ α̃ .
pp = 0
α i=1 n>0 −n n α0 i=1 n>0 −n n
i i
(2.20)
The difference is that now the sum is over oscillators αi and α̃i only, with i = 1, . . . , D −
2. We’ll refer to these as transverse oscillators. Note that the string isn’t necessarily
living in the X 0 -X D−1 plane, so these aren’t literally the transverse excitations of the
string. Nonetheless, if we specify the αi then all other oscillator modes are determined.
In this sense, they are the physical excitation of the string.
Let’s summarize the state of play so far. The most general classical solution is
described in terms of 2(D − 2) transverse oscillator modes αni and α̃ni , together with
a number of zero modes describing the center of mass and momentum of the string:
xi , pi , p+ and x− . But x+ can be absorbed by a shift of τ in (2.13) and p− is constrained
to obey (2.18) and (2.19). In fact, p− can be thought of as (proportional to) the
lightcone Hamiltonian. Indeed, we know that p− generates translations in x+ , but this
is equivalent to shifts in τ .
2.2.2 Quantization
Having identified the physical degrees of freedom, let’s now quantize. We want to
impose commutation relations. Some of these are easy:
[xi , pj ] = iδ ij
,
[x− , p+ ] = −i
j
j
] = nδ ij δn+m,0 .
[αni , αm
] = [α̃ni , α̃m
– 36 –
(2.21)
all of which follow from the commutation relations (2.2) that we saw in covariant
quantization1 .
What to do with x+ and p− ? We could implement p− as the Hamiltonian acting on
states. In fact, it will prove slightly more elegant (but equivalent) if we promote both
x+ and p− to operators with the expected commutation relation,
[x+ , p− ] = −i .
(2.22)
This is morally equivalent to writing [t, H] = −i in non-relativistic quantum mechanics,
which is true on a formal level. In the present context, it means that we can once again
choose states to be eigenstates of pµ , with µ = 0, . . . , D, but the constraints (2.18) and
(2.19) must still be imposed as operator equations on the physical states. We’ll come
to this shortly.
The Hilbert space of states is very similar to that described in covariant quantization:
we define a vacuum state, |0; pi such that
p̂µ |0; pi = pµ |0; pi
,
αni |0; pi = α̃ni |0; pi = 0
for n > 0
(2.23)
i
i
with
and α̃−n
and we build a Fock space by acting with the creation operators α−n
n > 0. The difference with the covariant quantization is that we only act with transverse
oscillators which carry a spatial index i = 1, . . . , D − 2. For this reason, the Hilbert
space is, by construction, positive definite. We don’t have to worry about ghosts.
1
Mea Culpa: We’re not really supposed to do this. The whole point of the approach that we’re
taking is to quantize just the physical degrees of freedom. The resulting commutation relations are
not, in general, inherited from the larger theory that we started with simply by closing our eyes
and forgetting about all the other fields that we’ve gauge fixed. We can see the problem by looking
at (2.17), where αn− is determined in terms of αni . This means that the commutation relations for
αni might be infected by those of αn− which could potentially give rise to extra terms. The correct
procedure to deal with this is to figure out the Poisson bracket structure of the physical degrees of
freedom in the classical theory. Or, in fancier language, the symplectic form on the phase space which
schematically looks like
Z
ω ∼ dσ − dẊ + ∧ dX − − dẊ − ∧ dX + + 2dẊ i ∧ dX i ,
The reason that the commutation relations (2.21) do not get infected is because the α− terms in the
symplectic form come multiplying X + . Yet X + is given in (2.13). It has no oscillator modes. That
means that the symplectic form doesn’t pick up the Fourier modes of X − and so doesn’t receive any
corrections from αn− . The upshot of this is that the naive commutation relations (2.21) are actually
right.
– 37 –
The Constraints
Because p− is not an independent variable in our theory, we must impose the constraints
(2.18) and (2.19) by hand as operator equations which define the physical states. In the
classical theory, we saw that these constraints are equivalent to mass-shell conditions
(2.20).
But there’s a problem when we go to the quantum theory. It’s the same problem
that we saw in covariant quantization: there’s an ordering ambiguity in the sum over
oscillator modes on the right-hand side of (2.20). If we choose all operators to be normal
ordered then this ambiguity reveals itself in an overall constant, a, which we have not
yet determined. The final result for the mass of states in lightcone gauge is:
!
!
D−2
D−2
X
X
X
X
4
4
i
i
M2 = 0
α−n
α̃−n
αni − a = 0
α̃ni − a
α i=1 n>0
α i=1 n>0
Since we’ll use this formula quite a lot in what follows, it’s useful to introduce quantities
related to the number operators of the harmonic oscillator,
N=
D−2
XX
i
α−n
αni
,
Ñ =
D−2
XX
i
α̃−n
α̃ni .
(2.24)
i=1 n>0
i=1 n>0
√
These are not quite number operators because of the factor of 1/ n in (2.3). The value
of N and Ñ is often called the level. Which, if nothing else, means that the name “level
matching” makes sense. We now have
M2 =
4
4
(N − a) = 0 (Ñ − a) .
0
α
α
(2.25)
How are we going to fix a? Later in the course we’ll see the correct way to do it. For
now, I’m just going to give you a quick and dirty derivation.
The Casimir Energy
What follows is a heuristic derivation of the normal ordering constant a. Suppose that
we didn’t notice that there was any ordering ambiguity and instead took the naive
classical result directly over to the quantum theory, that is
1X i i
1X i i
1X i i
α−n αn =
α−n αn +
α α .
2 n6=0
2 n<0
2 n>0 −n n
where we’ve left the sum over i = 1, . . . , D − 2 implicit. We’ll now try to put this in
normal ordered form, with the annihilation operators αni with n > 0 on the right-hand
– 38 –
side. It’s the first term that needs changing. We get
X
1X i i
1 X i i
D−2X
i
αn α−n − n(D − 2) +
α−n αn =
α−n
αni +
n .
2 n<0
2 n>0
2 n>0
n>0
The final term clearly diverges. But it at least seems to have a physical interpretation:
it is the sum of zero point energies of an infinite number of harmonic oscillators. In
fact, we came across exactly the same type of term in the course on quantum field
theory where we learnt that, despite the divergence, one can still extract interesting
physics from this. This is the physics of the Casimir force.
Let’s recall the steps that we took to derive the Casimir force. Firstly, we introduced
an ultra-violet cut-off 1, probably muttering some words about no physical plates
being able to withstand very high energy quanta. Unfortunately, those words are no
longer available to us in string theory, but let’s proceed regardless. We replace the
divergent sum over integers by the expression,
∞
X
n=1
n −→
∞
X
∞
ne
n=1
−n
∂ X −n
=−
e
∂ n=1
∂
(1 − e− )−1
∂
1
1
= 2−
+ O()
12
=−
Obviously the 1/2 piece diverges as → 0. This term should be renormalized away.
In fact, this is necessary to preserve the Weyl invariance of the Polyakov action since it
contributes to a cosmological constant on the worldsheet. After this renormalization,
we’re left with the answer
∞
X
n=1
n=−
1
.
12
While heuristic, this argument does predict the correct physical Casimir energy measured in one-dimensional systems. For example, this effect is seen in simulations of
quantum spin chains.
What does this mean for our string? It means that we should take the unknown
constant a in the mass formula (2.25) to be,
4
D−2
4
D−2
2
M = 0 N−
= 0 Ñ −
.
(2.26)
α
24
α
24
This is the formula that we will use to determine the spectrum of the string.
– 39 –
Zeta Function Regularization
I appreciate that the preceding argument is not totally convincing. We could spend
some time making it more robust at this stage, but it’s best if we wait until later in the
course when we will have the tools of conformal field theory at our disposal. We will
eventually revisit this issue and provide a respectable derivation of the Casimir energy
in Section 4.4.1. For now I merely offer an even less convincing argument, known as
zeta-function regularization.
The zeta-function is defined, for Re(s) > 1, by the sum
ζ(s) =
∞
X
n−s .
n=1
But ζ(s) has a unique analytic continuation to all values of s. In particular,
ζ(−1) = −
1
.
12
Good? Good. This argument is famously unconvincing the first time you meet it! But
it’s actually a very useful trick for getting the right answer.
2.3 The String Spectrum
Finally, we’re in a position to analyze the spectrum of a single, free string.
2.3.1 The Tachyon
Let’s start with the ground state |0; pi defined in (2.23). With no oscillators excited,
the mass formula (2.26) gives
M2 = −
1 D−2
.
α0 6
(2.27)
But that’s a little odd. It’s a negative mass-squared. Such particles are called tachyons.
In fact, tachyons aren’t quite as pathological as you might think. If you’ve heard
of these objects before, it’s probably in the context of special relativity where they’re
strange beasts which always travel faster than the speed of light. But that’s not the
right interpretation. Rather we should think more in the language of quantum field
theory. Suppose that we have a field in spacetime — let’s call it T (X) — whose quanta
will give rise to this particle. The mass-squared of the particle is simply the quadratic
term in the action, or
∂ 2 V (T ) 2
M =
∂T 2 T =0
– 40 –
So the negative mass-squared in (2.27) is telling us that we’re expanding around a
maximum of the potential for the tachyon field as shown in the figure. Note that from
this perspective, the Higgs field in the standard model at H = 0 is also a tachyon.
The fact that string theory turns out to sit at an unstable
V(T)
point in the tachyon field is unfortunate. The natural question is whether the potential has a good minimum elsewhere,
?
as shown in the figure to the right. No one knows the answer
to this! Naive attempts to understand this don’t work. We
T
know that around T = 0, the leading order contribution to
the potential is negative and quadratic. But there are furFigure 11:
ther terms that we can compute using techniques that we’ll
describe in Section 6. An expansion of the tachyon potential
around T = 0 looks like
1
V (T ) = M 2 T 2 + c3 T 3 + c4 T 4 + . . .
2
3
It turns out that the T term in the potential does give rise to a minimum. But the T 4
term destabilizes it again. Moreover, the T field starts to mix with other scalar fields
in the theory that we will come across soon. The ultimate fate of the tachyon in the
bosonic string is not yet understood.
The tachyon is a problem for the bosonic string. It may well be that this theory
makes no sense — or, at the very least, has no time-independent stable solutions. Or
perhaps we just haven’t worked out how to correctly deal with the tachyon. Either
way, the problem does not arise when we introduce fermions on the worldsheet and
study the superstring. This will involve several further technicalities which we won’t
get into in this course. Instead, our time will be put to better use if we continue to
study the bosonic string since all the lessons that we learn will carry over directly to the
superstring. However, one should be aware that the problem of the unstable vacuum
will continue to haunt us throughout this course.
Although we won’t describe it in detail, at several times along our journey we’ll make
an aside about how calculations work out for the superstring.
2.3.2 The First Excited States
j
We now look at the first excited states. If we act with a creation operator α−1
, then the
i
level matching condition (2.25) tells us that we also need to act with a α̃−1 operator.
This gives us (D − 2)2 particle states,
j
i
α̃−1
α−1
|0; pi ,
– 41 –
(2.28)
each of which has mass
4
M = 0
α
2
D−2
1−
24
.
But now we seem to have a problem. Our states have space indices i, j = 1, . . . , D − 2.
The operators αi and α̃i each transform in the vector representation of SO(D − 2) ⊂
SO(1, D − 1) which is manifest in lightcone gauge. But ultimately we want these states
to fit into some representation of the full Lorentz SO(1, D − 1) group. That looks as
if it’s going to be hard to arrange. This is the first manifestation of the comment that
we made after equation (2.12): it’s tricky to see Lorentz invariance in lightcone gauge.
To proceed, let’s recall Wigner’s classification of representations of the Poincaré
group. We start by looking at massive particles in R1,D−1 . After going to the rest
frame of the particle by setting pµ = (p, 0, . . . , 0), we can watch how any internal
indices transform under the little group SO(D − 1) of spatial rotations. The upshot
of this is that any massive particle must form a representation of SO(D − 1). But the
particles described by (2.28) have (D − 2)2 states. There’s no way to package these
states into a representation of SO(D − 1) and this means that there’s no way that the
first excited states of the string can form a massive representation of the D-dimensional
Poincaré group.
It looks like we’re in trouble. Thankfully, there’s a way out. If the states are massless,
then we can’t go to the rest frame. The best that we can do is choose a spacetime
momentum for the particle of the form pµ = (p, 0, . . . , 0, p). In this case, the particles
fill out a representation of the little group SO(D−2). This means that massless particles
get away with having fewer internal states than massive particles. For example, in four
dimensions the photon has two polarization states, but a massive spin-1 particle must
have three.
The first excited states (2.28) happily sit in a representation of SO(D − 2). We learn
that if we want the quantum theory to preserve the SO(1, D − 1) Lorentz symmetry
that we started with, then these states will have to be massless. And this is only the
case if the dimension of spacetime is
D = 26 .
This is our first derivation of the critical dimension of the bosonic string.
Moreover, we’ve found that our theory contains a bunch of massless particles. And
massless particles are interesting because they give rise to long range forces. Let’s look
– 42 –
more closely at what massless particles the string has given us. The states (2.28) transform in the 24 ⊗ 24 representation of SO(24). These decompose into three irreducible
representations:
traceless symmetric ⊕ anti-symmetric ⊕ singlet (=trace)
To each of these modes, we associate a massless field in spacetime such that the string
oscillation can be identified with a quantum of these fields. The fields are:
Gµν (X)
,
Bµν (X)
,
Φ(X)
(2.29)
Of these, the first is the most interesting and we shall have more to say momentarily.
The second is an anti-symmetric tensor field which is usually called the anti-symmetric
tensor field. It also goes by the names of the “Kalb-Ramond field” or, in the language
of differential geometry, the “2-form”. The scalar field is called the dilaton. These three
massless fields are common to all string theories. We’ll learn more about the role these
fields play later in the course.
The particle in the symmetric traceless representation of SO(24) is particularly interesting. This is a massless spin 2 particle. However, there are general arguments,
due originally to Feynman and Weinberg, that any theory of interacting massless spin
two particles must be equivalent to general relativity2 . We should therefore identify
the field Gµν (X) with the metric of spacetime. Let’s pause briefly to review the thrust
of these arguments.
Why Massless Spin 2 = General Relativity
Let’s call the spacetime metric Gµν (X). We can expand around flat space by writing
Gµν = ηµν + hµν (X) .
Then the Einstein-Hilbert action has an expansion in powers of h. If we truncate to
quadratic order, we simply have a free theory which we may merrily quantize in the
usual canonical fashion: we promote hµν to an operator and introduce the associated
creation and annihilation operators aµν and a†µν . This way of looking at gravity is
anathema to those raised in the geometrical world of general relativity. But from a
particle physics language it is very standard: it is simply the quantization of a massless
spin 2 field, hµν .
2
A very readable description of this can be found in the first few chapters of the Feynman Lectures
on Gravitation.
– 43 –
However, even on this simple level, there is a problem due to the indefinite signature
of the spacetime Minkowski metric. The canonical quantization relations of the creation
and annihilation operators are schematically of the form,
[aµν , a†ρσ ] ∼ ηµρ ηνσ + ηµσ ηνρ
But this will lead to a Hilbert space with negative norm states coming from acting with
time-like creation operators. For example, the one-graviton state of the form,
a†0i |0i
(2.30)
suffers from a negative norm. This should be becoming familiar by now: it is the usual
problem that we run into if we try to covariantly quantize a gauge theory. And, indeed,
general relativity is a gauge theory. The gauge transformations are diffeomorphisms.
We would hope that this saves the theory of quantum gravity from these negative norm
states.
Let’s look a little more closely at what the gauge symmetry looks like for small fluctuations hµν . We’ve butchered the Einstein-Hilbert action and left only terms quadratic
in h. Including all the index contractions, we find
Z
Mpl2
1
1
4
ρ
ρ µν
ν µ ρ
µν
ρ µν
SEH =
d x ∂µ h ρ ∂ν h − ∂ h ∂µ hρν + ∂ρ hµν ∂ h − ∂µ h ν ∂ h ρ + . . .
2
2
2
One can check that this truncated action is invariant under the gauge symmetry,
hµν −→ hµν + ∂µ ξν + ∂ν ξµ
(2.31)
for any function ξµ (X). The gauge symmetry is the remnant of diffeomorphism invariance, restricted to small deviations away from flat space. With this gauge invariance
in hand one can show that, just like QED, the negative norm states decouple from all
physical processes.
To summarize, theories of massless spin 2 fields only make sense if there is a gauge
symmetry to remove the negative norm states. In general relativity, this gauge symmetry descends from diffeomorphism invariance. The argument of Feynman and Weinberg
now runs this logic in reverse. It goes as follows: suppose that we have a massless, spin
2 particle. Then, at the linearized level, it must be invariant under the gauge symmetry
(2.31) in order to eliminate the negative norm states. Moreover, this symmetry must
survive when interaction terms are introduced. But the only way to do this is to ensure
that the resulting theory obeys diffeomorpism invariance. That means the theory of
any interacting, massless spin 2 particle is Einstein gravity, perhaps supplemented by
higher derivative terms.
– 44 –
We haven’t yet shown that string theory includes interactions for hµν but we will
come to this later in the course. More importantly, we will also explicitly see how
Einstein’s field equations arise directly in string theory.
A Comment on Spacetime Gauge Invariance
We’ve surreptitiously put µ, ν = 0, . . . , 25 indices on the spacetime fields, rather than
i, j = 1, . . . , 24. The reason we’re allowed to do this is because both Gµν and Bµν enjoy
a spacetime gauge symmetry which allows us to eliminate appropriate modes. Indeed,
this is exactly the gauge symmetry (2.31) that entered the discussion above. It isn’t
possible to see these spacetime gauge symmetries from the lightcone formalism of the
string since, by construction, we find only the physical states (although, by consistency
alone, the gauge symmetries must be there). One of the main advantages of pushing
through with the covariant calculation is that it does allow us to see how the spacetime
gauge symmetry emerges from the string worldsheet. Details can be found in Green,
Schwarz and Witten. We’ll also briefly return to this issue in Section 5.
2.3.3 Higher Excited States
We rescued the Lorentz invariance of the first excited states by choosing D = 26 to
ensure that they are massless. But now we’ve used this trick once, we still have to
worry about all the other excited states. These also carry indices that take the range
i, j = 1, . . . , D − 2 = 24 and, from the mass formula (2.26), they will all be massive
and so must form representations of SO(D − 1). It looks like we’re in trouble again.
Let’s examine the string at level N = Ñ = 2. In the right-moving sector, we now
j
i
i
|0i. The same is true for the left-moving
α−1
|0i and α−2
have two different states: α−1
sector, meaning that the total set of states at level 2 is (in notation that is hopefully
obvious, but probably technically wrong)
j
j
i
i
i
i
(α−1
α−1
⊕ α−2
) ⊗ (α̃−1
α̃−1
⊕ α̃−2
)|0; pi .
These states have mass M 2 = 4/α0 . How many states do we have? In the left-moving
sector, we have,
1
(D
2
− 2)(D − 1) + (D − 2) = 12 D(D − 1) − 1 .
But, remarkably, that does fit nicely into a representation of SO(D − 1), namely the
traceless symmetric tensor representation.
In fact, one can show that all excited states of the string fit nicely into SO(D − 1)
representations. The only consistency requirement that we need for Lorentz invariance
is to fix up the first excited states: D = 26.
– 45 –
Note that if we are interested in a fundamental theory of quantum gravity, then all
these excited states will have masses close to the Planck scale so are unlikely to be
observable in particle physics experiments. Nonetheless, as we shall see when we come
to discuss scattering amplitudes, it is the presence of this infinite tower of states that
tames the ultra-violet behaviour of gravity.
2.4 Lorentz Invariance Revisited
The previous discussion allowed to us to derive both the critical dimension and the
spectrum of string theory in the quickest fashion. But the derivation creaks a little in
places. The calculation of the Casimir energy is unsatisfactory the first time one sees
it. Similarly, the explanation of the need for massless particles at the first excited level
is correct, but seems rather cheap considering the huge importance that we’re placing
on the result.
As I’ve mentioned a few times already, we’ll shortly do better and gain some physical
insight into these issues, in particular the critical dimension. But here I would just like
to briefly sketch how one can be a little more rigorous within the framework of lightcone
quantization. The question, as we’ve seen, is whether one preserves spacetime Lorentz
symmetry when we quantize in lightcone gauge. We can examine this more closely.
Firstly, let’s go back to the action for free scalar fields (1.30) before we imposed
lightcone gauge fixing. Here the full Poincaré symmetry was manifest: it appears as a
global symmetry on the worldsheet,
X µ → Λµν X ν + cµ
(2.32)
But recall that in field theory, global symmetries give rise to Noether currents and
their associated conserved charges. What are the Noether currents associated to this
Poincaré transformation? We can start with the translations X µ → X µ + cµ . A quick
computation shows that the current is,
Pµα = T ∂ α Xµ
(2.33)
which is indeed a conserved current since ∂α Pµα = 0 is simply the equation of motion.
Similarly, we can compute the 21 D(D − 1) currents associated to Lorentz transformations. They are,
α
Jµν
= Pµα Xν − Pνα Xµ
α
It’s not hard to check that ∂α Jµν
= 0 when the equations of motion are obeyed.
– 46 –
The conserved charges arising from this current are given by Mµν =
the mode expansion (1.36) for X µ , these can be written as
M
µν
R
τ
dσJµν
. Using
∞
∞
X
X
1 ν µ
1 ν µ
µ
µ
ν
α̃nν
α−n αn − α−n αn − i
α̃−n α̃n − α̃−n
= (p x − p x ) − i
n
n
n=1
n=1
µ ν
ν µ
≡ lµν + S µν + S̃ µν
The first piece, lµν , is the orbital angular momentum of the string while the remaining
pieces S µν and S̃ µν tell us the angular momentum due to excited oscillator modes.
Classically, these obey the Poisson brackets of the Lorentz algebra. Moreover, if we
quantize in the covariant approach, the corresponding operators obey the commutation
relations of the Lorentz Lie algebra, namely
[Mρσ , Mτ ν ] = η στ Mρν − η ρτ Mσν + η ρν Mστ − η σν Mρτ
However, things aren’t so easy in lightcone gauge. Lorentz invariance is not guaranteed
and, in general, is not there. The right way to go about looking for it is to make sure
that the Lorentz algebra above is reproduced by the generators Mµν . It turns out that
the smoking gun lies in the commutation relation,
[Mi− , Mj− ] = 0
Does this equation hold in lightcone gauge? The problem is that it involves the operators p− and αn− , both of which are fixed by (2.17) and (2.18) in terms of the other
operators. So the task is to compute this commutation relation [Mi− , Mj− ], given the
commutation relations (2.21) for the physical degrees of freedom, and check that it vanishes. To do this, we re-instate the ordering ambiguity a and the number of spacetime
dimension D as arbitrary variables and proceed.
The part involving orbital angular momenta li− is fairly straightforward. (Actually,
there’s a small subtlety because we must first make sure that the operator lµν is Hermitian by replacing xµ pν with 21 (xµ pν + pν xµ )). The real difficulty comes from computing
the commutation relations [S i− , S j− ]. This is messy3 . After a tedious computation,
one finds,
2 X
D−2
1
D−2
j
i−
j−
i
[M , M ] = + 2
−1 n+
a−
(α−n
αnj − α−n
αni ) + (α ↔ α̃)
(p ) n>0
24
n
24
3
The original, classic, paper where lightcone quantization was first implemented is Goddard, Goldstone, Rebbi and Thorn “Quantum Dynamics of a Massless Relativistic String”, Nucl. Phys. B56
(1973). A pedestrian walkthrough of this calculation can be found in the lecture notes by Gleb Arutyunov. A link is given on the course webpage.
– 47 –
The right-hand side does not, in general, vanish. We learn that the relativistic string
can only be quantized in flat Minkowski space if we pick,
D = 26 and a = 1 .
2.5 A Nod to the Superstring
We won’t provide details of the superstring in this course, but will pause occasionally
to make some pertinent comments. Although what follows is nothing more than a list
of facts, it will hopefully be helpful in orienting you when you do come to study this
material.
The key difference between the bosonic string and the superstring is the addition of
fermionic modes on its worldsheet. The resulting worldsheet theory is supersymmetric.
(At least in the so-called Neveu-Schwarz-Ramond formalism). Hence the name “superstring”. Applying the kind of quantization procedure we’ve discussed in this section,
one finds the following results:
• The critical dimension of the superstring is D = 10.
• There is no tachyon in the spectrum.
• The massless bosonic fields Gµν , Bµν and Φ are all part of the spectrum of the
superstring. In this context, Bµν is sometimes referred to as the Neveu-Schwarz
2-form. There are also massless spacetime fermions, as well as further massless
bosonic fields. As we now discuss, the exact form of these extra bosonic fields
depends on exactly what superstring theory we consider.
While the bosonic string is unique, there are a number of discrete choices that one
can make when adding fermions to the worldsheet. This gives rise to a handful of
different perturbative superstring theories. (Although later developments reveal that
they are actually all part of the same framework which sometimes goes by the name of
M-theory). The most important of these discrete options is whether we add fermions in
both the left-moving and right-moving sectors of the string, or whether we choose the
fermions to move only in one direction, usually taken to be right-moving. This gives
rise to two different classes of string theory.
• Type II strings have both left and right-moving worldsheet fermions. The resulting spacetime theory in D = 10 dimensions has N = 2 supersymmetry, which
means 32 supercharges.
• Heterotic strings have just right-moving fermions. The resulting spacetime theory
has N = 1 supersymmetry, or 16 supercharges.
– 48 –
In each of these cases, there is then one further discrete choice that we can make.
This leaves us with four superstring theories. In each case, the massless bosonic fields
include Gµν , Bµν and Φ together with a number of extra fields. These are:
• Type IIA: In the type II theories, the extra massless bosonic excitations of the
string are referred to as Ramond-Ramond fields. For Type IIA, they are a 1form Cµ and a 3-form Cµνρ . Each of these is to be thought of as a gauge field.
The gauge invariant information lies in the field strengths which take the form
F = dC.
• Type IIB: The Ramond-Ramond gauge fields consist of a scalar C, a 2-form Cµν
and a 4-form Cµνρσ . The 4-form is restricted to have a self-dual field strength:
F5 = ? F5 . (Actually, this statement is almost true...we’ll look a little closer at
this in Section 7.3.3).
• Heterotic SO(32): The heterotic strings do not have Ramond-Ramond fields.
Instead, each comes with a non-Abelian gauge field in spacetime. The heterotic
strings are named after the gauge group. For example, the Heterotic SO(32)
string gives rise to an SO(32) Yang-Mills theory in ten dimensions.
• Heterotic E8 × E8 : The clue is in the name. This string gives rise to an E8 × E8
Yang-Mills field in ten-dimensions.
It is sometimes said that there are five perturbative superstring theories in ten dimensions. Here we’ve only mentioned four. The remaining theory is called Type I and
includes open strings moving in flat ten dimensional space as well as closed strings.
We’ll mention it in passing in the following section.
– 49 –
3. Open Strings and D-Branes
In this section we discuss the dynamics of open strings. Clearly
their distinguishing feature is the existence of two end points.
Our goal is to understand the effect of these end points. The
spatial coordinate of the string is parameterized by
τ
σ ∈ [0, π] .
σ
The dynamics of a generic point on a string is governed by local
physics. This means that a generic point has no idea if it is part
of a closed string or an open string. The dynamics of an open
Figure 12:
string must therefore still be described by the Polyakov action.
But this must now be supplemented by something else: boundary conditions to tell us
how the end points move. To see this, let’s look at the Polyakov action in conformal
gauge
Z
1
S=−
d2 σ ∂α X · ∂ α X .
4πα0
As usual, we derive the equations of motion by finding the extrema of the action. This
involves an integration by parts. Let’s consider the string evolving from some initial
configuration at τ = τi to some final configuration at τ = τf :
Z τf
Z π
1
δS = −
dτ
dσ ∂α X · ∂ α δX
2πα0 τi
0
Z
1
=
d2 σ (∂ α ∂α X) · δX + total derivative
2πα0
For an open string the total derivative picks up the boundary contributions
Z π
τ =τf
Z τ f
σ=π
1
1
0
dσ Ẋ · δX
−
dτ X · δX
2πα0 0
2πα0 τi
τ =τi
σ=0
The first term is the kind that we always get when using the principle of least action.
The equations of motion are derived by requiring that δX µ = 0 at τ = τi and τf and so
it vanishes. However, the second term is novel. In order for it too to vanish, we require
∂σ X µ δXµ = 0
at σ = 0, π
There are two different types of boundary conditions that we can impose to satisfy this:
– 50 –
• Neumann boundary conditions.
∂σ X µ = 0
at σ = 0, π
(3.1)
Because there is no restriction on δX µ , this condition allows the end of the string
to move freely. To see the consequences of this, it’s useful to repeat what we
did for the closed string and work in static gauge with X 0 ≡ t = Rτ , for some
dimensionful constant R. Then, as in equations (1.34), the constraints read
~x˙ · ~x 0 = 0 and ~x˙ 2 + ~x 0 2 = R2
But at the end points of the string, ~x 0 = 0. So the second equation tells us that
|d~x/dt| = 1. Or, in other words, the end point of the string moves at the speed
of light.
• Dirichlet boundary conditions
δX µ = 0
at σ = 0, π
(3.2)
This means that the end points of the string lie at some constant position, X µ =
cµ , in space.
At first sight, Dirichlet boundary conditions may
seem a little odd. Why on earth would the strings
be fixed at some point cµ ? What is special about
that point? Historically people were pretty hung
up about this and Dirichlet boundary conditions
were rarely considered until the mid-1990s. Then
everything changed due to an insight of Polchinski...
Dirichlet
Neumann
Figure 13:
Let’s consider Dirichlet boundary conditions for some coordinates and Neumann for
the others. This means that at both end points of the string, we have
∂σ X a = 0
I
X =c
I
for a = 0, . . . , p
for I = p + 1, . . . , D − 1
(3.3)
This fixes the end-points of the string to lie in a (p + 1)-dimensional hypersurface in
spacetime such that the SO(1, D − 1) Lorentz group is broken to,
SO(1, D − 1) → SO(1, p) × SO(D − p − 1) .
This hypersurface is called a D-brane or, when we want to specify its dimension, a
Dp-brane. Here D stands for Dirichlet, while p is the number of spatial dimensions
of the brane. So, in this language, a D0-brane is a particle; a D1-brane is itself a
string; a D2-brane a membrane and so on. The brane sits at specific positions cI in the
transverse space. But what is the interpretation of this hypersurface?
– 51 –
It turns out that the D-brane hypersurface should be thought of as a new, dynamical
object in its own right. This is a conceptual leap that is far from obvious. Indeed, it
took decades for people to fully appreciate this fact. String theory is not just a theory
of strings: it also contains higher dimensional branes. In Section 7.5 we will see how
these D-branes develop a life of their own. Some comments:
• We’ve defined D-branes that are infinite in space. However, we could just as well
define finite D-branes by specifying closed surfaces on which the string can end.
• There are many situations where we want to describe strings that have Neumann
boundary conditions in all directions, meaning that the string is free to move
throughout spacetime. It’s best to understand this in terms of a space-filling
D-brane. No Dirichlet conditions means D-branes are everywhere!
• The Dp-brane described above always has Neumann boundary conditions in the
X 0 direction. What would it mean to have Dirichlet conditions for X 0 ? Obviously
this is a little weird since the object is now localized at a fixed point in time. But
there is an interpretation of such an object: it is an instanton. This “D-instanton”
is usually referred to as a D(−1)-brane. It is related to tunneling effects in the
quantum theory.
Mode Expansion
We take the usual mode expansion for the string, with X µ = XLµ (σ + ) + XRµ (σ − ) and
r
α0 X 1 µ −inσ+
µ
0 µ +
+
1 µ
α̃ e
,
XL (σ ) = 2 x + α p σ + i
2 n6=0 n n
r
α0 X 1 µ −inσ−
µ
−
0 µ −
1 µ
α e
.
(3.4)
XR (σ ) = 2 x + α p σ + i
2 n6=0 n n
The boundary conditions impose relations on the modes of the string. They are easily
checked to be:
• Neumann boundary conditions, ∂σ X a = 0, at the end points require that
αna = α̃na
(3.5)
• Dirichlet boundary conditions, X I = cI , at the end points require that
x I = cI
,
pI = 0
,
αnI = −α̃nI
So for both boundary conditions, we only have one set of oscillators, say αn . The α̃n
are then determined by the boundary conditions.
– 52 –
It’s worth pointing out that there is a factor of 2 difference in the pµ term between
the open string (3.4) and the closed string (1.36). This is to ensure that pµ for the
open string retains the interpretation of the spacetime momentum of the string when
σ ∈ [0, π]. To see this, one needs to check the Noether current associated to translations
of X µ on the worldsheet: it was given in (2.33). The conserved charge is then
µ
π
Z
1
dσ (P ) =
2πα0
τ µ
P =
0
Z
π
dσ Ẋ µ = pµ
0
as advertised. Note that we’ve needed to use the Neumann conditions (3.5) to ensure
that the Fourier modes don’t contribute to this integral.
3.1 Quantization
To quantize, we promote the fields xa and pa and αnµ to operators. The other elements in the mode expansion are fixed by the boundary conditions. An obvious, but
important, point is that the position and momentum degrees of freedom, xa and pa ,
have a spacetime index that takes values a = 0, . . . , p. This means that the spatial
wavefunctions only depend on the coordinates of the brane not the whole spacetime.
Said another, quantizing an open string gives rise to states which are restricted to lie
on the brane.
To determine the spectrum, it is again simplest to work in lightcone gauge. The
spacetime lightcone coordinate is chosen to lie within the brane,
r
±
X =
1
(X 0 ± X p )
2
Quantization now proceeds in the same manner as for the closed string until we arrive
at the mass formula for states which is a sum over the transverse modes of the string.
1
M2 = 0
α
p−1
X
X
i=1 n>0
D−1
X
i
α−n
αni +
!
X
i
α−n
αni − a
i=p+1 n>0
The first sum is over modes parallel to the brane, the second over modes perpendicular
to the brane. It’s worth commenting on the differences with the closed string formula.
Firstly, there is an overall factor of 4 difference. This can be traced to the lack of the
factor of 1/2 in front of pµ in the mode expansion that we discussed above. Secondly,
there is a sum only over α modes. The α̃ modes are not independent because of the
boundary conditions.
– 53 –
Open and Closed
In the mass formula, we have once again left the normal ordering constant a ambiguous.
As in the closed string case, requiring the Lorentz symmetry of the quantum theory —
this time the reduced symmetry SO(1, p) × SO(D − p − 1) — forces us to choose
D = 26 and a = 1 .
These are the same values that we found for the closed string. This reflects an important
fact: the open string and closed string are not different theories. They are both different
states inside the same theory.
More precisely, theories of open strings necessarily contain closed strings. This is because, once we
consider interactions, an open string can join to form
a closed string as shown in the figure. We’ll look at
interactions in Section 6. The question of whether
Figure 14:
this works the other way — meaning whether closed
string theories require open strings — is a little more involved and is cleanest to state
in the context of the superstring. For type II superstrings, the open strings and Dbranes are necessary ingredients. For heterotic superstrings, there appear to be no
open strings and no D-branes. For the bosonic theory, it seems likely that the open
strings are a necessary ingredient although I don’t know of a killer argument. But since
we’re not sure whether the theory exists due to the presence of the tachyon, the point
is probably moot. In the remainder of these lectures, we’ll view the bosonic string in
the same manner as the type II string and assume that the theory includes both closed
strings and open strings with their associated D-branes.
3.1.1 The Ground State
The ground state is defined by
αni |0; pi = 0
n>0
The spatial index now runs over i = 1, . . . , p − 1, p + 1, . . . , D − 1. The ground state
has mass
M2 = −
1
α0
It is again tachyonic. Its mass is half that of the closed string tachyon. As we commented above, this time the tachyon is confined to the brane. In contrast to the closed
string tachyon, the open string tachyon is now fairly well understood and its potential
– 54 –
is of the form shown in the figure. The interpretation is that the
brane is unstable. It will decay, much like a resonance state in
field theory. It does this by dissolving into closed string modes.
The end point of this process – corresponding to the minimum
at T > 0 in the figure – is simply a state with no D-brane. The
difference between the value of the potential at the minimum
and at T = 0 is the tension of the D-brane.
V(T)
T
Figure 15:
Notice that although there is a minimum of the potential at
T > 0, it is not a global minimum. The potential seems to drop
off without bound to the left. This is still not well understood. There are suggestions
that it is related in some way to the closed string tachyon.
3.1.2 First Excited States: A World of Light
The first excited states are massless. They fall into two classes:
• Oscillators longitudinal to the brane,
a
α−1
|0; pi
a = 1, . . . , p − 1
The spacetime indices a lie within the brane so this state transforms under the
SO(1, p) Lorentz group. It is a spin 1 particle on the brane or, in other words, it
is a photon. We introduce a gauge field Aa with a = 0, . . . , p lying on the brane
whose quanta are identified with this photon.
• Oscillators transverse to the brane,
I
α−1
|0; pi
I = p + 1, . . . , D − 1
These states are scalars under the SO(1, p) Lorentz group of the brane. They can
be thought of as arising from scalar fields φI living on the brane. These scalars
have a nice interpretation: they are fluctuations of the brane in the transverse
directions. This is our first hint that the D-brane is a dynamical object. Note that
although the φI are scalar fields under the SO(1, p) Lorentz group of the brane,
they do transform as a vector under the SO(D − p − 1) rotation group transverse
to the brane. This appears as a global symmetry on the brane worldvolume.
– 55 –
3.1.3 Higher Excited States and Regge Trajectories
At level N , the mass of the string state is
M2 =
1
(N − 1)
α0
The maximal spin of these states arises from
the symmetric tensor. It is
Jmax = N = α0 M 2 + 1
Plotting the spin vs. the mass-squared, we find
straight lines. These are usually called Regge
trajectories. (Or sometimes Chew-Fraschuti traFigure 16:
jectories). They are seen in Nature in both the
spectrum of mesons and baryons. Some examples involving ρ-mesons are shown in the
figure. These stringy Regge trajectories suggest a naive cartoon picture of mesons as
two rotating quarks connected by a confining flux tube.
The value of the string tension required to match the hadron spectrum of QCD is
T ∼ 1 GeV. This relationship between the strong interaction and the open string was
one of the original motivations for the development of string theory and it is from here
that the parameter α0 gets its (admittedly rarely used) name “Regge slope”. In these
enlightened modern times, the connection between the open string and quarks lives on
in the AdS/CFT correspondence.
3.1.4 Another Nod to the Superstring
Just as supersymmetry eliminates the closed string tachyon, so it removes the open
string tachyon. Open strings are an ingredient of the type II string theories. The
possible D-branes are
• Type IIA string theory has stable Dp-branes with p even.
• Type IIB string theory has stable Dp-branes with p odd.
The most important reason that D-branes are stable in the type II string theories is that
they are charged under the Ramond-Ramond fields. (This was actually Polchinski’s
insight that made people take D-branes seriously). However, type II string theories
also contain unstable branes, with p odd in type IIA and p even in type IIB.
– 56 –
The fifth string theory (which was actually the first to be discovered) is called Type
I. Unlike the other string theories, it contains both open and closed strings moving
in flat ten-dimensional Lorentz-invariant spacetime. It can be thought of as the Type
IIB theory with a bunch of space-filling D9-branes, together with something called an
orientifold plane. You can read about this in Polchinski.
As we mentioned above, the heterotic string doesn’t have (finite energy) D-branes.
This is due to an inconsistency in any attempt to reflect left-moving modes into rightmoving modes.
3.2 Brane Dynamics: The Dirac Action
We have introduced D-branes as fixed boundary conditions for the open string. However, we’ve already seen a hint that these objects are dynamical in their own right,
since the massless scalar excitations φI have a natural interpretation as transverse fluctuations of the brane. Indeed, if a theory includes both open strings and closed strings,
then the D-branes have to be dynamical because there can be no rigid objects in a
theory of gravity. The dynamical nature of D-branes will become clearer as the course
progresses.
But any dynamical object should have an action which describes how it moves.
Moreover, after our discussion in Section 1, we already know what this is! On grounds
of Lorentz invariance and reparameterization invariance alone, the action must be a
higher dimensional extension of the Nambu-Goto action. This is
Z
p
SDp = −Tp
dp+1 ξ − det γ
(3.6)
where Tp is the tension of the Dp-brane which we will determine later, while ξ a , a =
0, . . . , p, are the worldvolume coordinates of the brane. γab is the pull back of the
spacetime metric onto the worldvolume,
γab =
∂X µ ∂X ν
ηµν .
∂ξ a ∂ξ b
This is called the Dirac action. It was first written down by Dirac for a membrane
some time before Nambu and Goto rediscovered it in the context of the string.
To make contact with the fields φI , we can use the reparameterization invariance of
the Dirac action to go to static gauge. For an infinite, flat Dp-brane we can choose
X a = ξa
a = 0, . . . , p .
– 57 –
The dynamical transverse coordinates are then identified with the fluctuations φI
through
X I (ξ) = 2πα0 φI (ξ)
I = p + 1, . . . , D − 1
However, the Dirac action can’t be the whole story. It describes the transverse fluctuations of the D-brane, but has nothing to say about the U (1) gauge field Aµ which
lives on the D-brane. There must be some action which describes how this gauge field
moves as well. We will return to this in Section 7.
What’s Special About Strings?
We could try to quantize the Dirac action (3.6) for a D-brane in the same manner that
we quantized the action for the string. Is this possible? The answer, at present, is
no. There appear to be both technical and conceptual obstacles . The technical issue
is just that it’s hard. Weyl invariance was one of our chief weapons in attacking the
string, but it doesn’t hold for higher dimensional objects.
The conceptual issue is that quantizing a membrane, or higher dimensional object,
would not give rise to a discrete spectrum of states which have the interpretation of
particles. In this way, they appear to be fundamentally different from the string.
Let’s get some intuition for why this is the case.
The energy of a string is proportional to its length.
This ensures that strings behave more or less like familiar elastic bands. What about D2-branes? Now
the energy is proportional to the area. In the back
Figure 17:
of your mind, you might be thinking of a rubber-like
sheet. But membranes, and higher dimensional objects, governed by the Dirac action
don’t behave as household rubber sheets. They are more flexible. This is because a
membrane can form many different shapes with the same area. For example, a tubular
membrane of length L and radius 1/L has the same area for all values of L; short and
stubby, or long and thin. This means that long thin spikes can develop on a membrane
at no extra cost of energy. In particular, objects connected by long thin tubes have
the same energy, regardless of their separation. After quantization, this property gives
rise to a continuous spectrum of states. A quantum membrane, or higher dimensional
object, does not have the single particle interpretation that we saw for the string. The
expectation is that the quantum membrane should describe multi-particle states.
– 58 –
3.3 Multiple Branes: A World of Glue
Consider two parallel Dp-branes. An open string now has
options. It could either end on the same brane, or stretch between the two branes. Let’s consider the string that stretches
between the two. It obeys
X I (0, τ ) = cI
and
X I (π, τ ) = dI
where cI and dI are the positions of the two branes. In terms of
the mode expansion, this requires
Figure 18:
(dI − cI )σ
+ oscillator modes
π
The classical constraints then read
|d~ − ~c|2
∂+ X · ∂+ X = α0 2 p2 +
+ oscillator modes = 0
4π 2
which means the classical mass-shell condition is
|d~ − ~c|2
2
M =
+ oscillator modes
(2πα0 )2
The extra term has an obvious interpretation: it is the mass of a classical string
stretched between the two branes. The quantization of this string proceeds as before. After we include the normal ordering constant, the ground state of this string is
only tachyonic if |d~ − ~c|2 < 4π 2 α0 . Or in other words, the ground state is tachyonic if
the branes approach to a sub-stringy distance.
X I = cI +
There is an obvious generalization of this to the case of N parallel branes. Each end
point of the string has N possible places on which to end. We can label each end point
with a number m, n = 1, . . . , N which tell us which brane it ends on. This label is
sometimes referred to as a Chan-Paton factor.
Consider now the situation where all branes lie at the same position in spacetime.
Each end point can lie on one of N different branes, giving N 2 possibilities in total.
Each of these strings has the mass spectrum of an open string, meaning that there are
now N 2 different particles of each type. It’s natural to arrange the associated fields to
sit inside N × N Hermitian matrices. We then have the open string tachyon T mn and
the massless fields
(φI )mn
,
(Aa )mn
(3.7)
Here the components of the matrix tell us which string the field came from. Diagonal
components arise from strings which have both ends on the same brane.
– 59 –
The gauge field Aa is particularly interesting. Written in this way, it looks like a
U (N ) gauge connection. We will later see that this is indeed the case. One can show
that as N branes coincide, the U (1)N gauge symmetry of the branes is enhanced to
U (N ). The scalar fields φI transform in the adjoint of this symmetry.
– 60 –
4. Introducing Conformal Field Theory
The purpose of this section is to get comfortable with the basic language of two dimensional conformal field theory4 . This is a topic which has many applications outside of
string theory, most notably in statistical physics where it offers a description of critical
phenomena. Moreover, it turns out that conformal field theories in two dimensions
provide rare examples of interacting, yet exactly solvable, quantum field theories. In
recent years, attention has focussed on conformal field theories in higher dimensions
due to their role in the AdS/CFT correspondence.
A conformal transformation is a change of coordinates σ α → σ̃ α (σ) such that the
metric changes by
gαβ (σ) → Ω2 (σ)gαβ (σ)
(4.1)
A conformal field theory (CFT) is a field theory which is invariant under these transformations. This means that the physics of the theory looks the same at all length scales.
Conformal field theories care about angles, but not about distances.
A transformation of the form (4.1) has a different interpretation depending on whether
we are considering a fixed background metric gαβ , or a dynamical background metric.
When the metric is dynamical, the transformation is a diffeomorphism; this is a gauge
symmetry. When the background is fixed, the transformation should be thought of as
an honest, physical symmetry, taking the point σ α to point σ̃ α . This is now a global
symmetry with the corresponding conserved currents.
In the context of string theory in the Polyakov formalism, the metric is dynamical and
the transformations (4.1) are residual gauge transformations: diffeomorphisms which
can be undone by a Weyl transformation.
In contrast, in this section we will be primarily interested in theories defined on
fixed backgrounds. Apart from a few noticeable exceptions, we will usually take this
background to be flat. This is the situation that we are used to when studying quantum
field theory.
4
Much of the material covered in this section was first described in the ground breaking paper by
Belavin, Polyakov and Zamalodchikov, “Infinite Conformal Symmetry in Two-Dimensional Quantum
Field Theory”, Nucl. Phys. B241 (1984). The application to string theory was explained by Friedan,
Martinec and Shenker in “Conformal Invariance, Supersymmetry and String Theory”, Nucl. Phys.
B271 (1986). The canonical reference for learning conformal field theory is the excellent review by
Ginsparg. A link can be found on the course webpage.
– 61 –
Of course, we can alternate between thinking of theories as defined on fixed or fluctuating backgrounds. Any theory of 2d gravity which enjoys both diffeomorphism and
Weyl invariance will reduce to a conformally invariant theory when the background
metric is fixed. Similarly, any conformally invariant theory can be coupled to 2d gravity where it will give rise to a classical theory which enjoys both diffeomorphism and
Weyl invariance. Notice the caveat “classical”! In some sense, the whole point of this
course is to understand when this last statement also holds at the quantum level.
Even though conformal field theories are a subset of quantum field theories, the
language used to describe them is a little different. This is partly out of necessity.
Invariance under the transformation (4.1) can only hold if the theory has no preferred
length scale. But this means that there can be nothing in the theory like a mass or a
Compton wavelength. In other words, conformal field theories only support massless
excitations. The questions that we ask are not those of particles and S-matrices. Instead
we will be concerned with correlation functions and the behaviour of different operators
under conformal transformations.
4.0.1 Euclidean Space
Although we’re ultimately interested in Minkowski signature worldsheets, it will be
much simpler and elegant if we work instead with Euclidean worldsheets. There’s no
funny business here — everything we do could also be formulated in Minkowski space.
The Euclidean worldsheet coordinates are (σ 1 , σ 2 ) = (σ 1 , iσ 0 ) and it will prove useful
to form the complex coordinates,
z = σ 1 + iσ 2
and
z̄ = σ 1 − iσ 2
which are the Euclidean analogue of the lightcone coordinates. Motivated by this
analogy, it is common to refer to holomorphic functions as “left-moving” and antiholomorphic functions as “right-moving”.
The holomorphic derivatives are
1
∂z ≡ ∂ = (∂1 − i∂2 )
2
and
1
∂z̄ ≡ ∂¯ = (∂1 + i∂2 )
2
¯ = 0. We will usually work in flat Euclidean
These obey ∂z = ∂¯z̄ = 1 and ∂ z̄ = ∂z
space, with metric
ds2 = (dσ 1 )2 + (dσ 2 )2 = dz dz̄
– 62 –
(4.2)
In components, this flat metric reads
gzz = gz̄z̄ = 0 and gzz̄ =
1
2
With this convention, the measure factor is dzdz̄ = 2dσ 1 dσ 2 . We define the deltaR
R
function such that d2 z δ(z, z̄) = 1. Notice that because we also have d2 σ δ(σ) = 1,
this means that there is a factor of 2 difference between the two delta functions. Vectors
naturally have their indices up: v z = (v 1 + iv 2 ) and v z̄ = (v 1 − iv 2 ). When indices are
down, the vectors are vz = 21 (v 1 − iv 2 ) and vz̄ = 21 (v 1 + iv 2 ).
4.0.2 The Holomorphy of Conformal Transformations
In the complex Euclidean coordinates z and z̄, conformal transformations of flat space
are simple: they are any holomorphic change of coordinates,
z → z 0 = f (z)
and
z̄ → z̄ 0 = f¯(z̄)
Under this transformation, ds2 = dzdz̄ → |df /dz|2 dzdz̄, which indeed takes the
form (4.1). Note that we have an infinite number of conformal transformations — in
fact, a whole functions worth f (z). This is special to conformal field theories in two
dimensions. In higher dimensions, the space of conformal transformations is a finite
dimensional group. For theories defined on Rp,q , the conformal group is SO(p+1, q +1)
when p + q > 2.
A couple of particularly simple and important examples of 2d conformal transformations are
• z → z + a: This is a translation.
• z → ζz: This is a rotation for |ζ| = 1 and a scale transformation (also known as
a dilatation) for real ζ 6= 1.
For many purposes, it’s simplest to treat z and z̄ as independent variables. In doing
this, we’re really extending the worldsheet from R2 to C2 . This will allow us to make
use of various theorems from complex methods. However, at the end of the day we
should remember that we’re really sitting on the real slice R2 ⊂ C2 defined by z̄ = z ? .
4.1 Classical Aspects
We start by deriving some properties of classical theories which are invariant under
conformal transformations (4.1).
– 63 –
4.1.1 The Stress-Energy Tensor
One of the most important objects in any field theory is the stress-energy tensor (also
known as the energy-momentum tensor). This is defined in the usual way as the matrix
of conserved currents which arise from translational invariance,
δσ α = α .
In flat spacetime, a translation is a special case of a conformal transformation.
There’s a cute way to derive the stress-energy tensor in any theory. Suppose for the
moment that we are in flat space gαβ = ηαβ . Recall that we can usually derive conserved
currents by promoting the constant parameter that appears in the symmetry to a
function of the spacetime coordinates. The change in the action must then be of the
form,
Z
δS = d2 σ J α ∂α (4.3)
for some function of the fields, J α . This ensures that the variation of the action vanishes
when is constant, which is of course the definition of a symmetry. But when the
equations of motion are satisfied, we must have δS = 0 for all variations (σ), not just
constant . This means that when the equations of motion are obeyed, J α must satisfy
∂α J α = 0
The function J α is our conserved current.
Let’s see how this works for translational invariance. If we promote to a function
of the worldsheet variables, the change of the action must be of the form (4.3). But
what is J α ? At this point we do the cute thing. Consider the same theory, but now
coupled to a dynamical background metric gαβ (σ). In other words, coupled to gravity.
Then we could view the transformation
δσ α = α (σ)
as a diffeomorphism and we know that the theory is invariant as long as we make the
corresponding change to the metric
δgαβ = ∂α β + ∂β α .
This means that if we just make the transformation of the coordinates in our original
theory, then the change in the action must be the opposite of what we get if we just
– 64 –
transform the metric. (Because doing both together leaves the action invariant). So
we have
Z
Z
∂S
∂S
2
δgαβ = −2 d2 σ
∂α β
δS = − d σ
∂gαβ
∂gαβ
Note that ∂S/∂gαβ in this expression is really a functional derivatives but we won’t be
careful about using notation to indicate this. We now have the conserved current arising
from translational invariance. We will add a normalization constant which is standard
in string theory (although not necessarily in other areas) and define the stress-energy
tensor to be
4π ∂S
Tαβ = − √
g ∂g αβ
(4.4)
If we have a flat worldsheet, we evaluate Tαβ on gαβ = δαβ and the resulting expression
obeys ∂ α Tαβ = 0. If we’re working on a curved worldsheet, then the energy-momentum
tensor is covariantly conserved, ∇α Tαβ = 0.
The Stress-Energy Tensor is Traceless
In conformal theories, Tαβ has a very important property: its trace vanishes. To see
this, let’s vary the action with respect to a scale transformation which is a special case
of a conformal transformation,
δgαβ = gαβ
(4.5)
Then we have
Z
δS =
∂S
1
dσ
δgαβ = −
∂gαβ
4π
2
Z
√
d2 σ g T αα
But this must vanish in a conformal theory because scaling transformations are a
symmetry. So
T αα = 0
This is the key feature of a conformal field theory in any dimension. Many theories
have this feature at the classical level, including Maxwell theory and Yang-Mills theory
in four-dimensions. However, it is much harder to preserve at the quantum level. (The
weight of the world rests on the fact that Yang-Mills theory fails to be conformal at the
quantum level). Technically the difficulty arises due to the need to introduce a scale
when regulating the theories. Here we will be interested in two-dimensional theories
– 65 –
which succeed in preserving the conformal symmetry at the quantum level.
Looking Ahead: Even when the conformal invariance survives in a 2d quantum
theory, the vanishing trace T αα = 0 will only turn out to hold in flat space. We will
derive this result in section 4.4.2.
The Stress-Tensor in Complex Coordinates
In complex coordinates, z = σ 1 + iσ 2 , the vanishing of the trace T αα = 0 becomes
Tzz̄ = 0
¯ z̄z̄ = 0. Or,
Meanwhile, the conservation equation ∂α T αβ = 0 becomes ∂T zz = ∂T
lowering the indices on T ,
∂¯ Tzz = 0
and
∂ Tz̄z̄ = 0
In other words, Tzz = Tzz (z) is a holomorphic function while Tz̄z̄ = Tz̄z̄ (z̄) is an antiholomorphic function. We will often use the simplified notation
Tzz (z) ≡ T (z) and Tz̄z̄ (z̄) ≡ T̄ (z̄)
4.1.2 Noether Currents
The stress-energy tensor Tαβ provides the Noether currents for translations. What are
the currents associated to the other conformal transformations? Consider the infinitesimal change,
z 0 = z + (z) ,
z̄ 0 = z̄ + ¯(z̄)
where, making contact with the two examples above, constant corresponds to a translation while (z) ∼ z corresponds to a rotation and dilatation. To compute the current,
we’ll use the same trick that we saw before: we promote the parameter to depend
on the worldsheet coordinates. But it’s already a function of half of the worldsheet
coordinates, so this now means (z) → (z, z̄). Then we can compute the change in the
action, again using the fact that we can make a compensating change in the metric,
Z
∂S
δS = − d2 σ
δg αβ
αβ
∂g
Z
1
=
d2 σ Tαβ (∂ α δσ β )
2π
Z
1
1
=
d2 z [Tzz (∂ z δz) + Tz̄z̄ (∂ z̄ δz̄)]
2π
2
Z
1
=
d2 z [Tzz ∂z̄ + Tz̄z̄ ∂z ¯]
(4.6)
2π
– 66 –
Firstly note that if is holomorphic and ¯ is anti-holomorphic, then we immediately
have δS = 0. This, of course, is the statement that we have a symmetry on our hands.
(You may wonder where in the above derivation we used the fact that the theory was
conformal. It lies in the transition to the third line where we needed Tzz̄ = 0).
At this stage, let’s use the trick of treating z and z̄ as independent variables. We
look at separate currents that come from shifts in z and shifts z̄. Let’s first look at the
symmetry
δz = (z) ,
δz̄ = 0
We can read off the conserved current from (4.6) by using the standard trick of letting
the small parameter depend on position. Since (z) already depends on position, this
¯ terms
means promoting → (z)f (z̄) for some function f and then looking at the ∂f
in (4.6). This gives us the current
J z = 0 and J z̄ = Tzz (z) (z) ≡ T (z) (z)
(4.7)
Importantly, we find that the current itself is also holomorphic. We can check that this
is indeed a conserved current: it should satisfy ∂α J α = ∂z J z + ∂z̄ J z̄ = 0. But in fact it
does so with room to spare: it satisfies the much stronger condition ∂z̄ J z̄ = 0.
Similarly, we can look at transformations δz̄ = ¯(z̄) with δz = 0. We get the anti¯
holomorphic current J,
J¯z = T̄ (z̄) ¯(z̄) and J¯z̄ = 0
(4.8)
4.1.3 An Example: The Free Scalar Field
Let’s illustrate some of these ideas about classical conformal theories with the free
scalar field,
Z
1
d2 σ ∂ α X ∂ α X
S=
4πα0
Notice that there’s no overall minus sign, in contrast to our earlier action (1.30). That’s
because we’re now working with a Euclidean worldsheet metric. The theory of a free
scalar field is, of course, dead easy. We can compute anything we like in this theory.
Nonetheless, it will still exhibit enough structure to provide an example of all the
abstract concepts that we will come across in CFT. For this reason, the free scalar field
will prove a good companion throughout this part of the lectures.
– 67 –
Firstly, let’s just check that this free scalar field is actually conformal. In particular,
we can look at rescaling σ α → λσ α . If we view this in the sense of an active transformation, the coordinates remain fixed but the value of the field at point σ gets moved
to point λσ. This means,
X(σ) → X(λ−1 σ)
and
∂X(σ)
∂X(λ−1 σ)
1 ∂X(σ̃)
→
=
α
α
∂σ
∂σ
λ ∂ σ̃
where we’ve defined σ̃ = λ−1 σ. The factor of λ−2 coming from the two derivatives
in the Lagrangian then cancels the Jacobian factor from the measure d2 σ = λ2 d2 σ̃,
leaving the action invariant. Note that any polynomial interaction term for X would
break conformal invariance.
The stress-energy tensor for this theory is defined using (4.4),
1
1
2
Tαβ = − 0 ∂α X∂β X − δαβ (∂X)
,
α
2
(4.9)
which indeed satisfies T αα = 0 as it should. The stress-energy tensor looks much simpler
in complex coordinates. It is simple to check that Tzz̄ = 0 while
T =−
1
∂X∂X
α0
and
T̄ = −
1 ¯ ¯
∂X ∂X
α0
¯ = 0. The general classical solution decomposes
The equation of motion for X is ∂ ∂X
as,
X(z, z̄) = X(z) + X̄(z̄)
When evaluated on this solution, T and T̄ become holomorphic and anti-holomorphic
functions respectively.
4.2 Quantum Aspects
So far our discussion has been entirely classical. We now turn to the quantum theory.
The first concept that we want to discuss is actually a feature of any quantum field
theory. But it really comes into its own in the context of CFT: it is the operator product
expansion.
4.2.1 Operator Product Expansion
Let’s first describe what we mean by a local operator in a CFT. We will also refer to
these objects as fields. There is a slight difference in terminology between CFTs and
more general quantum field theories. Usually in quantum field theory, one reserves the
– 68 –
term “field” for the objects φ which sit in the action and are integrated over in the
path integral. In contrast, in CFT the term “field” refers to any local expression that
we can write down. This includes φ, but also includes derivatives ∂ n φ or composite
operators such as eiφ . All of these are thought of as different fields in a CFT. It should
be clear from this that the set of all “fields” in a CFT is always infinite even though,
if you were used to working with quantum field theory, you would talk about only a
finite number of fundamental objects φ. Obviously, this is nothing to be scared about.
It’s just a change of language: it doesn’t mean that our theory got harder.
We now define the operator product expansion (OPE). It is a statement about what
happens as local operators approach each other. The idea is that two local operators
inserted at nearby points can be closely approximated by a string of operators at one
of these points. Let’s denote all the local operators of the CFT by Oi , where i runs
over the set of all operators. Then the OPE is
X
Oi (z, z̄) Oj (w, w̄) =
Cijk (z − w, z̄ − w̄) Ok (w, w̄)
(4.10)
k
Here Cijk (z − w, z̄ − w̄) are a set of functions which, on
grounds of translational invariance, depend only on the
x
x
separation between the two operators. We will write a lot
of operator equations of the form (4.10) and it’s important to clarify exactly what they mean: they are always
to be understood as statements which hold as operator
x
insertions inside time-ordered correlation functions,
X
Cijk (z − w, z̄ − w̄) hOk (w, w̄) . . . i
hOi (z, z̄) Oj (w, w̄) . . . i =
O1(w)
xx
x
O2(z)
Figure 19:
k
where the . . . can be any other operator insertions that we choose. Obviously it would
be tedious to continually write h. . .i. So we don’t. But it’s always implicitly there.
There are further caveats about the OPE that are worth stressing
• The correlation functions are always assumed to be time-ordered. (Or something
similar that we will discuss in Section 4.5.1). This means that as far as the OPE
is concerned, everything commutes since the ordering of operators is determined
inside the correlation function anyway. So we must have Oi (z, z̄) Oj (w, w̄) =
Oj (w, w̄) Oi (z, z̄). (There is a caveat here: if the operators are Grassmann objects,
then they pick up an extra minus sign when commuted, even inside time-ordered
products).
– 69 –
• The other operator insertions in the correlation function (denoted . . . above) are
arbitrary. Except they should be at a distance large compared to |z − w|. It turns
out — rather remarkably — that in a CFT the OPEs are exact statements and
have a radius of convergence equal to the distance to the nearest other insertion.
We will return to this in Section 4.6. The radius of convergence is denoted in the
figure by the dotted line.
• The OPEs have singular behaviour as z → w. In fact, this singular behaviour
will really be the only thing we care about! It will turn out to contain the
same information as commutation relations, as well as telling us how operators
transform under symmetries. Indeed, in many equations we will simply write the
singular terms in the OPE and denote the non-singular terms as + . . ..
4.2.2 Ward Identities
The spirit of Noether’s theorem in quantum field theories is captured by operator
equations known as Ward Identities. Here we derive the Ward identities associated to
conformal invariance. We start by considering a general theory with a symmetry. Later
we will restrict to conformal symmetries.
Games with Path Integrals
We’ll take this opportunity to get comfortable with some basic techniques using path
integrals. Schematically, the path integral takes the form
Z
Z = Dφ e−S[φ]
where φ collectively denote all the fields (in the path integral sense...not the CFT
sense!). A symmetry of the quantum theory is such that an infinitesimal transformation
φ0 = φ + δφ
leaves both the action and the measure invariant,
S[φ0 ] = S[φ]
and
Dφ0 = Dφ
(In fact, we only really need the combination Dφ e−S[φ] to be invariant but this subtlety
won’t matter in this course). We use the same trick that we employed earlier in the
classical theory and promote → (σ). Then, typically, neither the action nor the
measure are invariant but, to leading order in , the change has to be proportional to
– 70 –
∂. We have
Z
Dφ0 exp (−S[φ0 ])
Z
Z
1
α
=
Dφ exp −S[φ] −
J ∂α 2π
Z
Z
1
−S[φ]
α
1−
=
Dφ e
J ∂α 2π
R
R
√
where the factor of 1/2π is merely a convention and is shorthand for d2 σ g. Notice
that the current J α may now also have contributions from the measure transformation
as well as the action.
Z −→
Now comes the clever step. Although the integrand has changed, the actual value of
the partition function can’t have changed at all. After all, we just redefined a dummy
integration variable φ. So the expression above must be equal to the original Z. Or, in
other words,
Z
Z
−S[φ]
α
Dφ e
J ∂α = 0
Moreover, this must hold for all . This gives us the quantum version of Noether’s
theorem: the vacuum expectation value of the divergence of the current vanishes:
h∂α J α i = 0 .
We can repeat these tricks of this sort to derive some stronger statements. Let’s see
what happens when we have other insertions in the path integral. The time-ordered
correlation function is given by
Z
1
hO1 (σ1 ) . . . On (σn )i =
Dφ e−S[φ] O1 (σ1 ) . . . On (σn )
Z
We can think of these as operators inserted at particular points on the plane as shown
in the figure. As we described above, the operators Oi are any general expressions
that we can form from the φ fields. Under the symmetry of interest, the operator will
change in some way, say
Oi → Oi + δOi
We once again promote → (σ). As our first pass, let’s pick a choice of (σ) which
only has support away from the operator insertions as shown in the Figure 20. Then,
δOi (σi ) = 0
– 71 –
and the above derivation goes through in exactly the same
way to give
ε=0
O1(σ 1)
O2(σ 2)
x
α
h∂α J (σ) O1 (σ1 ) . . . On (σn )i = 0
x
for σ 6= σi
O3(σ 3)
x
Because this holds for any operator insertions away from
σ, from the discussion in Section 4.2.1 we are entitled to
write the operator equation
O4(σ 4)
x
Figure 20:
α
∂α J = 0
But what if there are operator insertions that lie at
ε=0
O1(σ 1)
the same point as J α ? In other words, what happens as
O2(σ 2)
x
σ approaches one of the insertion points? The resulting
x
formulae are called Ward identities. To derive these, let’s
O3(σ 3)
O4(σ 4)
x
take (σ) to have support in some region that includes the
x
point σ1 , but not the other points as shown in Figure 21.
The simplest choice is just to take (σ) to be constant inside
Figure 21:
the shaded region and zero outside. Now using the same
procedure as before, we find that the original correlation
function is equal to,
Z
Z
1
1
−S[φ]
α
Dφ e
1−
J ∂α (O1 + δO1 ) O2 . . . On
Z
2π
Working to leading order in , this gives
Z
1
−
∂α hJ α (σ) O1 (σ1 ) . . .i = hδO1 (σ1 ) . . .i
2π (4.11)
where the integral on the left-hand-side is only over the region of non-zero . This is
the Ward Identity.
Ward Identities for Conformal Transformations
Ward identities (4.11) hold for any symmetries. Let’s now see what they give when
applied to conformal transformations. There are two further steps needed in the derivation. The first simply comes from the fact that we’re working in two dimensions and
we can use Stokes’ theorem to convert the integral on the left-hand-side of (4.11) to a
line integral around the boundary. Let n̂α be the unit vector normal to the boundary.
For any vector J α , we have
Z
I
I
I
α
α
2
1
∂α J =
Jα n̂ =
(J1 dσ − J2 dσ ) = −i
(Jz dz − Jz̄ dz̄)
∂
∂
∂
– 72 –
where we have written the expression both in Cartesian coordinates σ α and complex
coordinates on the plane. As described in Section 4.0.1, the complex components of
the vector with indices down are defined as Jz = 12 (J1 − iJ2 ) and Jz̄ = 21 (J1 + iJ2 ). So,
applying this to the Ward identity (4.11), we find for two dimensional theories
I
I
i
i
dz hJz (z, z̄) O1 (σ1 ) . . .i −
dz̄ hJz̄ (z, z̄) O1 (σ1 ) . . .i = hδO1 (σ1 ) . . .i
2π ∂
2π ∂
So far our derivation holds for any conserved current J in two dimensions. At this stage
we specialize to the currents that arise from conformal transformations (4.7) and (4.8).
Here something nice happens because Jz is holomorphic while Jz̄ is anti-holomorphic.
This means that the contour integral simply picks up the residue,
I
i
dz Jz (z)O1 (σ1 ) = − Res [Jz O1 ]
2π ∂
where this means the residue in the OPE between the two operators,
Jz (z) O1 (w, w̄) = . . . +
Res [Jz O1 (w, w̄)]
+ ...
z−w
So we find a rather nice way of writing the Ward identities for conformal transformations. If we again view z and z̄ as independent variables, the Ward identities split into
two pieces. From the change δz = (z), we get
δO1 (σ1 ) = −Res [Jz (z)O1 (σ1 )] = −Res [(z)T (z)O1 (σ1 )]
(4.12)
where, in the second equality, we have used the expression for the conformal current
(4.7). Meanwhile, from the change δz̄ = ¯(z̄), we have
δO1 (σ1 ) = −Res [J¯z̄ (z̄)O1 (σ1 )] = −Res [¯(z̄)T̄ (z̄)O1 (σ1 )]
H
where the minus sign comes from the fact that the dz̄ boundary integral is taken in
the opposite direction.
This result means that if we know the OPE between an operator and the stresstensors T (z) and T̄ (z̄), then we immediately know how the operator transforms under
conformal symmetry. Or, standing this on its head, if we know how an operator transforms then we know at least some part of its OPE with T and T̄ .
4.2.3 Primary Operators
The Ward identity allows us to start piecing together some OPEs by looking at how
operators transform under conformal symmetries. Although we don’t yet know the
– 73 –
action of general conformal symmetries, we can start to make progress by looking at
the two simplest examples.
Translations: If δz = , a constant, then all operators transform as
O(z − ) = O(z) − ∂O(z) + . . .
The Noether current for translations is the stress-energy tensor T . The Ward identity
in the form (4.12) tells us that the OPE of T with any operator O must be of the form,
T (z) O(w, w̄) = . . . +
∂O(w, w̄)
+ ...
z−w
(4.13)
¯
∂O(w,
w̄)
+ ...
z̄ − w̄
(4.14)
Similarly, the OPE with T̄ is
T̄ (z̄) O(w, w̄) = . . . +
Rotations and Scaling: The transformation
z → z + z
and z̄ → z̄ + ¯z̄
(4.15)
describes rotation for purely imaginary and scaling (dilatation) for real. Not all
operators have good transformation properties under these actions. This is entirely
analogous to the statement in quantum mechanics that not all states transform nicely
under the Hamiltonian H and angular momentum operator L. However, in quantum
mechanics we know that the eigenstates of H and L can be chosen as a basis of the
Hilbert space provided, of course, that [H, L] = 0.
The same statement holds for operators in a CFT: we can choose a basis of local
operators that have good transformation properties under rotations and dilatations. In
fact, we will see in Section 4.6 that the statement about local operators actually follows
from the statement about states.
Definition: An operator O is said to have weight (h, h̃) if, under δz = z and δz̄ = ¯z̄,
O transforms as
¯
δO = −(hO + z ∂O) − ¯(h̃O + z̄ ∂O)
(4.16)
The terms ∂O in this expression would be there for any operator. They simply come
from expanding O(z − z, z̄ − ¯z̄). The terms hO and h̃O are special to operators which
are eigenstates of dilatations and rotations. Some comments:
– 74 –
• Both h and h̃ are real numbers. In a unitary CFT, all operators have h, h̃ ≥ 0.
We will prove this is Section 4.5.4.
• The weights are not as unfamiliar as they appear. They simply tell us how
operators transform under rotations and scalings. But we already have names
for these concepts from undergraduate days. The eigenvalue under rotation is
usually called the spin, s, and is given in terms of the weights as
s = h − h̃
Meanwhile, the scaling dimension ∆ of an operator is
∆ = h + h̃
• To motivate these definitions, it’s worth recalling how rotations and scale transformations act on the underlying coordinates. Rotations are implemented by the
operator
L = −i(σ 1 ∂2 − σ 2 ∂1 ) = z∂ − z̄ ∂¯
while the dilation operator D which gives rise to scalings is
D = σ α ∂α = z∂ + z̄ ∂¯
• The scaling dimension is nothing more than the familiar “dimension” that we
usually associate to fields and operators by dimensional analysis. For example, worldsheet derivatives always increase the dimension of an operator by one:
∆[∂] = +1. The tricky part is that the naive dimension that fields have in the
classical theory is not necessarily the same as the dimension in the quantum
theory.
Let’s compare the transformation law (4.16) with the Ward identity (4.12). The
Noether current arising from rotations and scaling δz = z was given in (4.7): it is
J(z) = zT (z). This means that the residue of the JO OPE will determine the 1/z 2
term in the T O OPE. Similar arguments hold, of course, for δz̄ = ¯z̄ and T̄ . So, the
upshot of this is that, for an operator O with weight (h, h̃), the OPE with T and T̄
takes the form
O(w, w̄) ∂O(w, w̄)
+
+ ...
(z − w)2
z−w
¯
O(w, w̄) ∂O(w,
w̄)
T̄ (z̄) O(w, w̄) = . . . + h̃
+
+ ...
(z̄ − w̄)2
z̄ − w̄
T (z) O(w, w̄) = . . . + h
– 75 –
Primary Operators
A primary operator is one whose OPE with T and T̄ truncates at order (z − w)−2 or
order (z̄ − w̄)−2 respectively. There are no higher singularities:
O(w, w̄) ∂O(w, w̄)
+ non-singular
+
(z − w)2
z−w
¯
O(w, w̄) ∂O(w,
w̄)
T̄ (z̄) O(w, w̄) = h̃
+ non-singular
+
2
(z̄ − w̄)
z̄ − w̄
T (z) O(w, w̄) = h
Since we now know all singularities in the T O OPE, we can reconstruct the transformation under all conformal transformations. The importance of primary operators is
that they have particularly simple transformation properties. Focussing on δz = (z),
we have
δO(w, w̄) = −Res [(z) T (z) O(w, w̄)]
O(w, w̄) ∂O(w, w̄)
+
+ ...
= −Res (z) h
(z − w)2
z−w
We want to look at smooth conformal transformations and so require that (z) itself
has no singularities at z = w. We can then Taylor expand
(z) = (w) + 0 (w) (z − w) + . . .
We learn that the infinitesimal change of a primary operator under a general conformal
transformation δz = (z) is
δO(w, w̄) = −h0 (w) O(w, w̄) − (w) ∂O(w, w̄)
(4.17)
There is a similar expression for the anti-holomorphic transformations δz̄ = ¯(z̄).
Equation (4.17) holds for infinitesimal conformal transformations. It is a simple
matter to integrate up to find how primary operators change under a finite conformal
transformation,
z → z̃(z)
and
¯
z̄ → z̃(z̄)
The general transformation of a primary operator is given by
−h −h̃
∂ z̃¯
¯ = ∂ z̃
O(z, z̄) → Õ(z̃, z̃)
O(z, z̄)
∂z
∂ z̄
(4.18)
It will turn out that one of the main objects of interest in a CFT is the spectrum of
weights (h, h̃) of primary fields. This will be equivalent to computing the particle mass
spectrum in a quantum field theory. In the context of statistical mechanics, the weights
of primary operators are the critical exponents.
– 76 –
4.3 An Example: The Free Scalar Field
Let’s look at how all of this works for the free scalar field. We’ll start by familiarizing
ourselves with some techniques using the path integral. The action is,
Z
1
S=
d2 σ ∂α X ∂ α X
(4.19)
0
4πα
The classical equation of motion is ∂ 2 X = 0. Let’s start by seeing how to derive the
analogous statement in the quantum theory using the path integral. The key fact that
we’ll need is that the integral of a total derivative vanishes in the path integral just as
it does in an ordinary integral. From this we have,
Z
Z
δ
1
−S
−S
2
0 = DX
e = DX e
∂ X(σ)
δX(σ)
2πα0
But this is nothing more than the Ehrenfest theorem which states that expectation
values of operators obey the classical equations of motion,
h∂ 2 X(σ)i = 0
4.3.1 The Propagator
The next thing that we want to do is compute the propagator for X. We could do this
using canonical quantization, but it will be useful to again see how it works using the
path integral. This time we look at,
Z
−S
R
1 2
δ
0
0
0 = DX
e X(σ 0 ) = DX e−S 2πα
0 ∂ X(σ) X(σ ) + δ(σ − σ )
δX(σ)
So this time we learn that
h∂ 2 X(σ) X(σ 0 )i = −2πα0 δ(σ − σ 0 )
(4.20)
Note that if we’d computed this in the canonical approach, we would have found the
same answer: the δ-function arises in this calculation because all correlation functions
are time-ordered.
We can now treat (4.20) as a differential equation for the propagator hX(σ)X(σ 0 )i.
To solve this equation, we need the following standard result
∂ 2 ln(σ − σ 0 )2 = 4πδ(σ − σ 0 )
– 77 –
(4.21)
Since this is important, let’s just quickly check that it’s true. It’s a simple application
R
of Stokes’ theorem. Set σ 0 = 0 and integrate over d2 σ. We obviously get 4π from the
right-hand-side. The left-hand-side gives
I
Z
Z
(σ1 dσ 2 − σ2 dσ 1 )
2σα
2
2
2
2
2
α
=2
d σ ∂ ln(σ1 + σ2 ) = d σ ∂
σ12 + σ22
σ12 + σ22
Switching to polar coordinates σ1 + iσ2 = reiθ , we can rewrite this expression as
Z 2
r dθ
2
= 4π
r2
confirming (4.21). Applying this result to our equation (4.20), we get the propagator
of a free scalar in two-dimensions,
hX(σ)X(σ 0 )i = −
α0
ln(σ − σ 0 )2
2
The propagator has a singularity as σ → σ 0 . This is an ultra-violet divergence and is
common to all field theories. It also has a singularity as |σ − σ 0 | → ∞. This is telling
us something important that we will mention in Section 4.3.2.
Finally, we could repeat our trick of looking at total derivatives in the path integral,
now with other operator insertions O1 (σ1 ), . . . On (σn ) in the path integral. As long
as σ, σ 0 6= σi , then the whole analysis goes through as before. But this is exactly our
criterion to write the operator product equation,
α0
X(σ)X(σ ) = − ln(σ − σ 0 )2 + . . .
2
0
(4.22)
¯ =
We can also write this in complex coordinates. The classical equation of motion ∂ ∂X
0 allows us to split the operator X into left-moving and right-moving pieces,
X(z, z̄) = X(z) + X̄(z̄)
We’ll focus just on the left-moving piece. This has the operator product expansion,
X(z)X(w) = −
α0
ln(z − w) + . . .
2
The logarithm means that X(z) doesn’t have any nice properties under the conformal
transformations. For this reason, the “fundamental field” X is not really the object of
interest in this theory! However, we can look at the derivative of X. This has a rather
nice looking OPE,
∂X(z) ∂X(w) = −
α0
1
+ non-singular
2 (z − w)2
– 78 –
(4.23)
4.3.2 An Aside: No Goldstone Bosons in Two Dimensions
The infra-red divergence in the propagator has an important physical implication. Let’s
start by pointing out one of the big differences between quantum mechanics and quantum field theory in d = 3 + 1 dimensions. Since the language used to describe these
two theories is rather different, you may not even be aware that this difference exists.
Consider the quantum mechanics of a particle on a line. This is a d = 0 + 1 dimensional theory of a free scalar field X. Let’s prepare the particle in some localized
state – say a Gaussian wavefunction Ψ(X) ∼ exp(−X 2 /L2 ). What then happens?
The wavefunction starts to spread out. And the spreading doesn’t stop. In fact, the
would-be ground state of the system is a uniform wavefunction of infinite width, which
isn’t a state in the Hilbert space because it is non-normalizable.
Let’s now compare this to the situation of a free scalar field X in a d = 3 + 1
dimensional field theory. Now we think of this as a scalar without potential. The physics
is very different: the theory has an infinite number of ground states, determined by the
expectation value hXi. Small fluctuations around this vacuum are massless: they are
Goldstone bosons for broken translational invariance X → X + c.
We see that the physics is very different in field theories in d = 0 + 1 and d = 3 + 1
dimensions. The wavefunction spreads along flat directions in quantum mechanics, but
not in higher dimensional field theories. But what happens in d = 1 + 1 and d = 2 + 1
dimensions? It turns out that field theories in d = 1 + 1 dimensions are more like
quantum mechanics: the wavefunction spreads. Theories in d = 2 + 1 dimensions and
higher exhibit the opposite behaviour: they have Goldstone bosons. The place to see
this is the propagator. In d spacetime dimensions, it takes the form
(
1/rd−2
d 6= 2
hX(r) X(0)i ∼
ln r
d=2
which diverges at large r only for d = 1 and d = 2. If we perturb the vacuum slightly
by inserting the operator X(0), this correlation function tells us how this perturbation
falls off with distance. The infra-red divergence in low dimensions is telling us that the
wavefunction wants to spread.
The spreading of the wavefunction in low dimensions means that there is no spontaneous symmetry breaking and no Goldstone bosons. It is usually referred to as the
Coleman-Mermin-Wagner theorem. Note, however, that it certainly doesn’t prohibit
massless excitations in two dimensions: it only prohibits Goldstone-like massless excitations.
– 79 –
4.3.3 The Stress-Energy Tensor and Primary Operators
We want to compute the OPE of T with other operators. Firstly, what is T ? We
computed it in the classical theory in (4.9). It is,
1
∂X∂X
(4.24)
α0
But we need to be careful about what this means in the quantum theory. It involves
the product of two operators defined at the same point and this is bound to mean
divergences if we just treat it naively. In canonical quantization, we would be tempted
to normal order by putting all annihilation operators to the right. This guarantees that
the vacuum has zero energy. Here we do something that is basically equivalent, but
without reference to creation and annihilation operators. We write
T =−
T =−
1
1
: ∂X∂X : ≡ − 0 limit (∂X(z)∂X(w) − h∂X(z)∂X(w)i)
0
α
α z→w
(4.25)
which, by construction, has hT i = 0.
With this definition of T , let’s start to compute the OPEs to determine the primary
fields in the theory.
Claim 1: ∂X is a primary field with weight h = 1 and h̃ = 0.
Proof: We need to figure out how to take products of normal ordered operators
T (z) ∂X(w) = −
1
: ∂X(z)∂X(z) : ∂X(w)
α0
The operators on the left-hand side are time-ordered (because all operator expressions
of this type are taken to live inside time-ordered correlation functions). In contrast,
the right-hand side is a product of normal-ordered operators. But we know how to
change normal ordered products into time ordered products: this is the content of
Wick’s theorem. Although we have defined normal ordering in (4.25) without reference
to creation and annihilation operators, Wick’s theorem still holds. We must sum over
all possible contractions of pairs of operators, where the term “contraction” means that
we replace the pair by the propagator,
z
}|
{
α0
1
∂X(z) ∂X(w) = −
2 (z − w)2
Using this, we have
2
T (z)∂X(w) = − 0 ∂X(z)
α
α0
1
−
+ non-singular
2 (z − w)2
– 80 –
Here the “non-singular” piece includes the totally normal ordered term : T (z)∂X(w) :.
It is only the singular part that interests us. Continuing, we have
T (z)∂X(w) =
∂X(z)
∂X(w)
∂ 2 X(w)
+ ...
+
.
.
.
=
+
(z − w)2
(z − w)2
z−w
This is indeed the OPE for a primary operator of weight h = 1.
Note that higher derivatives ∂ n X are not primary for n > 1. For example, ∂ 2 X has
weight (h, h̃) = (2, 0), but is not a primary operator, as we see from the OPE,
∂X(w)
2∂X(w) 2∂ 2 X(w)
2
T (z) ∂ X(w) = ∂w
+
.
.
.
=
+
+ ...
(z − w)2
(z − w)3
(z − w)2
The fact that the field ∂ n X has weight (h, h̃) = (n, 0) fits our natural intuition: each
derivative provides spin s = 1 and dimension ∆ = 1, while the field X does not appear
to be contributing, presumably reflecting the fact that it has naive, classical dimension
zero. However, in the quantum theory, it is not correct to say that X has vanishing
dimension: it has an ill-defined dimension due to the logarithmic behaviour of its OPE
(4.22). This is responsible for the following, more surprising, result
Claim 2: The field : eikX : is primary with weight h = h̃ = α0 k 2 /4.
This result is not what we would guess from the classical theory. Indeed, it’s obvious
that it has a quantum origin because the weight is proportional to α0 , which sits outside
the action in the same place that ~ would (if we hadn’t set it to one). Note also that
this means that the spectrum of the free scalar field is continuous. This is related to the
fact that the range of X is non-compact. Generally, CFTs will have a discrete spectrum.
Proof: Let’s first compute the OPE with ∂X. We have
∞
X
(ik)n
ikX(w)
∂X(z) : X(w)n :
∂X(z) : e
:=
n!
n=0
∞
X (ik)n
α0 1
n−1
=
: X(w)
: −
+ ...
(n
−
1)!
2
z
−
w
n=1
iα0 k : eikX(w) :
+ ...
2
z−w
From this, we can compute the OPE with T .
1
T (z) : eikX(w) : = − 0 : ∂X(z)∂X(z) : : eikX(w) :
α
α0 k 2 : eikX(w) :
: ∂X(z)eikX(w) :
=
+
ik
+ ...
4 (z − w)2
z−w
=−
– 81 –
(4.26)
where the first term comes from two contractions, while the second term comes from a
single contraction. Replacing ∂z by ∂w in the final term we get
T (z) : eikX(w) :=
α0 k 2 : eikX(w) : ∂w : eikX(w) :
+ ...
+
4 (z − w)2
z−w
(4.27)
showing that : eikX(w) : is indeed primary. We will encounter this operator frequently
later, but will choose to simplify notation and drop the normal ordering colons. Normal
ordering will just be assumed from now on.
.
Finally, lets check to see the OPE of T with itself. This is again just an exercise in
Wick contractions.
1
: ∂X(z) ∂X(z) : : ∂X(w) ∂X(w) :
α0 2
2
2
1
α0
4 α0 : ∂X(z) ∂X(w) :
= 02 −
+ ...
−
α
2 (z − w)2
α0 2 2
(z − w)2
T (z) T (w) =
The factor of 2 in front of the first term comes from the two ways of performing two
contractions; the factor of 4 in the second term comes from the number of ways of
performing a single contraction. Continuing,
1/2
2T (w)
2 ∂ 2 X(w) ∂X(w)
+
−
+ ...
(z − w)4 (z − w)2 α0
z−w
1/2
2T (w)
∂T (w)
=
+
+
+ ...
4
2
(z − w)
(z − w)
z−w
T (z) T (w) =
(4.28)
We learn that T is not a primary operator in the theory of a single free scalar field. It
is an operator of weight (h, h̃) = (2, 0), but it fails the primary test on account of the
(z − w)−4 term. In fact, this property of the stress energy tensor a general feature of
all CFTs which we now explore in more detail.
4.4 The Central Charge
In any CFT, the most prominent example of an operator which is not primary is the
stress-energy tensor itself.
For the free scalar field, we have already seen that T is an operator of weight (h, h̃) =
(2, 0). This remains true in any CFT. The reason for this is simple: Tαβ has dimension
∆ = 2 because we obtain the energy by integrating over space. It has spin s = 2
because it is a symmetric 2-tensor. But these two pieces of information are equivalent
– 82 –
to the statement that T is has weight (2, 0). Similarly, T̄ has weight (0, 2). This means
that the T T OPE takes the form,
T (z) T (w) = . . . +
2T (w)
∂T (w)
+ ...
+
2
(z − w)
z−w
and similar for T̄ T̄ . What other terms could we have in this expansion? Since each
term has dimension ∆ = 4, any operators that appear on the right-hand-side must be
of the form
On
(z − w)n
(4.29)
where ∆[On ] = 4 − n. But, in a unitary CFT there are no operators with h, h̃ < 0.
(We will prove this shortly). So the most singular term that we can have is of order
(z − w)−4 . Such a term must be multiplied by a constant. We write,
T (z) T (w) =
c/2
2T (w)
∂T (w)
+
+
+ ...
4
2
(z − w)
(z − w)
z−w
T̄ (z̄) T̄ (w̄) =
2T̄ (w̄)
∂¯T̄ (w̄)
c̃/2
+
+
+ ...
(z̄ − w̄)4 (z̄ − w̄)2
z̄ − w̄
and, similarly,
The constants c and c̃ are called the central charges. (Sometimes they are referred to as
left-moving and right-moving central charges). They are perhaps the most important
numbers characterizing the CFT. We can already get some intuition for the information
contained in these two numbers. Looking back at the free scalar field (4.28) we see that
it has c = c̃ = 1. If we instead considered D non-interacting free scalar fields, we would
get c = c̃ = D. This gives us a hint: c and c̃ are somehow measuring the number of
degrees of freedom in the CFT. This is true in a deep sense! However, be warned: c is
not necessarily an integer.
Before moving on, it’s worth pausing to explain why we didn’t include a (z − w)−3
term in the T T OPE. The reason is that the OPE must obey T (z)T (w) = T (w)T (z)
because, as explained previously, these operator equations are all taken to hold inside
time-ordered correlation functions. So the quick answer is that a (z − w)−3 term would
not be invariant under z ↔ w. However, you may wonder how the (z − w)−1 term
manages to satisfy this property. Let’s see how this works:
T (w) T (z) =
c/2
2T (z)
∂T (z)
+
+
+ ...
(z − w)4 (z − w)2 w − z
– 83 –
Now we can Taylor expand T (z) = T (w)+(z −w)∂T (w)+. . . and ∂T (z) = ∂T (w)+. . ..
Using this in the above expression, we find
T (w) T (z) =
c/2
2T (w) + 2(z − w)∂T (w) ∂T (w)
+ . . . = T (z) T (w)
+
−
4
(z − w)
(z − w)2
z−w
This trick of Taylor expanding saves the (z − w)−1 term. It wouldn’t work for the
(z − w)−3 term.
The Transformation of Energy
So T is not primary unless c = 0. And we will see shortly that all theories have c > 0.
What does this mean for the transformation of T ?
δT (w) = −Res [(z) T (z) T (w)]
c/2
2T (w)
∂T (w)
= −Res (z)
+
+
+ ...
(z − w)4 (z − w)2
z−w
If (z) contains no singular terms, we can expand
1
1
(z) = (w) + 0 (w)(z − w) + 00 (z − w)2 + 000 (w)(z − w)3 + . . .
2
6
from which we find
c
δT (w) = −(w) ∂T (w) − 20 (w) T (w) − 000 (w)
(4.30)
12
This is the infinitesimal version. We would like to know what becomes of T under the
finite conformal transformation z → z̃(z). The answer turns out to be
−2 h
i
∂ z̃
c
T̃ (z̃) =
T (z) −
S(z̃, z)
(4.31)
∂z
12
where S(z̃, z) is known as the Schwarzian and is defined by
3 −1
2 −2
∂ z̃
∂ z̃
3 ∂ 2 z̃
∂ z̃
S(z̃, z) =
−
3
2
∂z
∂z
2 ∂z
∂z
(4.32)
It is simple to check that the Schwarzian has the right infinitesimal form to give (4.30).
Its key property is that it preserves the group structure of successive conformal transformations.
4.4.1 c is for Casimir
Note that the extra term in the transformation (4.31) of T does not depend on T itself.
In particular, it will be the same evaluated on all states. It only affects the constant
term — or zero mode — in the energy. In other words, it is the Casimir energy of the
system.
– 84 –
Let’s look at an example that will prove to be useful later for the string. Consider
the Euclidean cylinder, parameterized by
w = σ + iτ
σ ∈ [0, 2π)
,
We can make a conformal transformation from the cylinder to the complex
plane by
z = e−iw
The fact that the cylinder and the plane
Figure 22:
are related by a conformal map means
that if we understand a given CFT on
the cylinder, then we immediately understand it on the plane. And vice-versa. Notice
that constant time slices on the cylinder are mapped to circles of constant radius. The
origin, z = 0, is the distant past, τ → −∞.
What becomes of T under this transformation? The Schwarzian can be easily calculated to be S(z, w) = 1/2. So we find,
Tcylinder (w) = −z 2 Tplane (z) +
c
24
(4.33)
Suppose that the ground state energy vanishes when the theory is defined on the plane:
hTplane i = 0. What happens on the cylinder? We want to look at the Hamiltonian,
which is defined by
Z
H≡
Z
dσ Tτ τ = −
dσ (Tww + T̄w̄w̄ )
The conformal transformation then tells us that the ground state energy on the cylinder
is
E=−
2π(c + c̃)
24
This is indeed the (negative) Casimir energy on a cylinder. For a free scalar field, we
have c = c̃ = 1 and the energy density E/2π = −1/12. This is the same result that we
got in Section 2.2.2, but this time with no funny business where we throw out infinities.
– 85 –
An Application: The Lüscher Term
If we’re looking at a physical system, the cylinder will have a radius L. In this case,
the Casimir energy is given by E = −2π(c + c̃)/24L. There is an application of this to
QCD-like theories. Consider two quarks in a confining theory, separated by a distance
L. If the tension of the confining flux tube is T , then the string will be stable as long
as T L . m, the mass of the lightest quark. The energy of the stretched string as a
function of L is given by
πc
+ ...
E(L) = T L + a −
24L
Here a is an undetermined constant, while c counts the number of degrees of freedom
of the QCD flux tube. (There is no analog of c̃ here because of the reflecting boundary
conditions at the end of the string). If the string has no internal degrees of freedom,
then c = 2 for the two transverse fluctuations. This contribution to the string energy
is known as the Lüscher term.
4.4.2 The Weyl Anomaly
There is another way in which the central charge affects the stress-energy tensor. Recall
that in the classical theory, one of the defining features of a CFT was the vanishing of
the trace of the stress tensor,
T αα = 0
However, things are more subtle in the quantum theory. While hT αα i indeed vanishes
in flat space, it will not longer be true if we place the theory on a curved background.
The purpose of this section is to show that
c
hT αα i = − R
(4.34)
12
where R is the Ricci scalar of the 2d worldsheet. Before we derive this formula, some
quick comments:
• Equation (4.34) holds for any state in the theory — not just the vacuum. This
reflects the fact that it comes from regulating short distant divergences in the
theory. But, at short distances all finite energy states look basically the same.
• Because hT αα i is the same for any state it must be equal to something that depends
only on the background metric. This something should be local and must be
dimension 2. The only candidate is the Ricci scalar R. For this reason, the
formula hT αα i ∼ R is the most general possibility. The only question is: what is
the coefficient. And, in particular, is it non-zero?
– 86 –
• By a suitable choice of coordinates, we can always put any 2d metric in the form
gαβ = e2ω δαβ . In these coordinates, the Ricci scalar is given by
R = −2e−2ω ∂ 2 ω
(4.35)
which depends explicitly on the function ω. Equation (4.34) is then telling us
that any conformal theory with c 6= 0 has at least one physical observable, hT αα i,
which takes different values on backgrounds related by a Weyl transformation ω.
This result is referred to as the Weyl anomaly, or sometimes as the trace anomaly.
• There is also a Weyl anomaly for conformal field theories in higher dimensions.
For example, 4d CFTs are characterized by two numbers, a and c, which appear
as coefficients in the Weyl anomaly,
a
c
hT µµ i4d =
Cρσκλ C ρσκλ −
R̃ρσκλ R̃ρσκλ
2
16π
16π 2
where C is the Weyl tensor and R̃ is the dual of the Riemann tensor.
• Equation (4.34) involves only the left-moving central charge c. You might wonder
what’s special about the left-moving sector. The answer, of course, is nothing.
We also have
c̃
hT αα i = − R
12
In flat space, conformal field theories with different c and c̃ are perfectly acceptable. However, if we wish these theories to be consistent in fixed, curved backgrounds, then we require c = c̃. This is an example of a gravitational anomaly.
• The fact that Weyl invariance requires c = 0 will prove crucial in string theory.
We shall return to this in Chapter 5.
We will now prove the Weyl anomaly formula (4.34). Firstly, we need to derive
an intermediate formula: the Tzz̄ Tww̄ OPE. Of course, in the classical theory we found
that conformal invariance requires Tzz̄ = 0. We will now show that it’s a little more
subtle in the quantum theory.
Our starting point is the equation for energy conservation,
∂Tzz̄ = −∂¯ Tzz
Using this, we can express our desired OPE in terms of the familiar T T OPE,
c/2
+ ...
(4.36)
∂z Tzz̄ (z, z̄) ∂w Tww̄ (w, w̄) = ∂¯z̄ Tzz (z, z̄) ∂¯w̄ Tww (w, w̄) = ∂¯z̄ ∂¯w̄
(z − w)4
– 87 –
Now you might think that the right-hand-side just vanishes: after all, it is an antiholomorphic derivative ∂¯ of a holomorphic quantity. But we shouldn’t be so cavalier
because there is a singularity at z = w. For example, consider the following equation,
∂¯z̄ ∂z ln |z − w|2 = ∂¯z̄
1
= 2πδ(z − w, z̄ − w̄)
z−w
(4.37)
We proved this statement after equation (4.21). (The factor of 2 difference from (4.21)
can be traced to the conventions we defined for complex coordinates in Section 4.0.1).
Looking at the intermediate step in (4.37), we again have an anti-holomorphic derivative
of a holomorphic function and you might be tempted to say that this also vanishes. But
you’d be wrong: subtle things happen because of the singularity and equation (4.37)
tells us that the function 1/z secretly depends on z̄. (This should really be understood
as a statement about distributions, with the delta function integrated against arbitrary
test functions). Using this result, we can write
1
1¯ ¯
π
1
2
¯
¯
∂z̄ ∂w̄
= ∂z̄ ∂w̄ ∂z ∂w
= ∂z2 ∂w ∂¯w̄ δ(z − w, z̄ − w̄)
4
(z − w)
6
z−w
3
Inserting this into the correlation function (4.36) and stripping off the ∂z ∂w derivatives
on both sides, we end up with what we want,
Tzz̄ (z, z̄) Tww̄ (w, w̄) =
cπ ¯
∂z ∂w̄ δ(z − w, z̄ − w̄)
6
(4.38)
So the OPE of Tzz̄ and Tww̄ almost vanishes, but there’s some strange singular behaviour
going on as z → w. This is usually referred to as a contact term between operators
and, as we have shown, it is needed to ensure the conservation of energy-momentum.
We will now see that this contact term is responsible for the Weyl anomaly.
We assume that hT αα i = 0 in flat space. Our goal is to derive an expression for hT αα i
close to flat space. Firstly, consider the change of hT αα i under a general shift of the
metric δgαβ . Using the definition of the energy-momentum tensor (4.4), we have
Z
α
δ hT α (σ)i = δ Dφ e−S T αα (σ)
Z
Z
1
−S
α
2 0√
βγ
0
=
Dφ e
T α (σ) d σ g δg Tβγ (σ )
4π
If we now restrict to a Weyl transformation, the change to a flat metric is δgαβ = 2ωδαβ ,
so the change in the inverse metric is δg αβ = −2ωδ αβ . This gives
Z
Z
1
β
α
−S
α
2 0
0
0
δ hT α (σ)i = −
Dφ e
T α (σ) d σ ω(σ ) T β (σ )
(4.39)
2π
– 88 –
Now we see why the OPE (4.38) determines the Weyl anomaly. We need to change
between complex coordinates and Cartesian coordinates, keeping track of factors of 2.
We have
T αα (σ) T ββ (σ 0 ) = 16 Tzz̄ (z, z̄) Tww̄ (w, w̄)
Meanwhile, using the conventions laid down in 4.0.1, we have 8∂z ∂¯w̄ δ(z − w, z̄ − w̄) =
−∂ 2 δ(σ − σ 0 ). This gives us the OPE in Cartesian coordinates
cπ
T αα (σ) T ββ (σ 0 ) = − ∂ 2 δ(σ − σ 0 )
3
We now plug this into (4.39) and integrate by parts to move the two derivatives onto
the conformal factor ω. We’re left with,
c
c
δ hT αα i = ∂ 2 ω ⇒ hT αα i = − R
6
12
where, to get to the final step, we’ve used (4.35) and, since we’re working infinitesimally,
we can replace e−2ω ≈ 1. This completes the proof of the Weyl anomaly, at least for
spaces infinitesimally close to flat space. The fact that R remains on the right-handside for general 2d surfaces follows simply from the comments after equation (4.34),
most pertinently the need for the expression to be reparameterization invariant.
4.4.3 c is for Cardy
The Casimir effect and the Weyl anomaly have a similar smell. In both, the central
charge provides an extra contribution to the energy. We now demonstrate a different
avatar of the central charge: it tells us the density of high energy states.
We will study conformal field theory on a Euclidean torus. We’ll keep our normalization σ ∈ [0, 2π), but now we also take τ to be periodic, lying in the range
τ ∈ [0, β)
The partition function of a theory with periodic Euclidean time has a very natural
interpretation: it is related to the free energy of the theory at temperature T = 1/β.
Z[β] = Tr e−βH = e−βF
(4.40)
At very low temperatures, β → ∞, the free energy is dominated by the lowest energy
state. All other states are exponentially suppressed. But we saw in 4.4.1 that the
vacuum state on the cylinder has Casimir energy H = −c/12. In the limit of low
temperature, the partition function is therefore approximated by
Z → ecβ/12
as β → ∞
– 89 –
(4.41)
Now comes the trick. In Euclidean space,
both directions of the torus are on equal
footing. We’re perfectly at liberty to de2π
cide that σ is “time” and τ is “space”.
This can’t change the value of the parβ
tition function. So let’s make the swap.
To compare to our original partition function, we want the spatial direction to have
Figure 23:
range [0, 2π). Happily, due to the conformal nature of our theory, we arrange this through the scaling
τ→
2π
τ
β
,
σ→
4π2
β
2π
2π
σ
β
Now we’re back where we started, but with the temporal direction taking values in
σ ∈ [0, 4π 2 /β). This tells us that the high-temperature and low-temperature partition
functions are related,
Z[4π 2 /β] = Z[β]
This is called modular invariance. We’ll come across it again in Section 6.4. Writing
β 0 = 4π 2 /β, this tells us the very high temperature behaviour of the partition function
Z[β 0 ] → ecπ
2 /3β 0
as β 0 → 0
But the very high temperature limit of the partition function is sampling all states in
the theory. On entropic grounds, this sampling is dominated by the high energy states.
So this computation is telling us how many high energy states there are.
To see this more explicitly, let’s do some elementary manipulations in statistical
mechanics. Any system has a density of states ρ(E) = e S(E) , where S(E) is the
entropy. The free energy is given by
Z
Z
−βF
−βE
e
= dE ρ(E) e
= dE eS(E)−βE
In two dimensions, all systems have an entropy which scales at large energy as
√
S(E) → N E
(4.42)
√
The coefficient N counts the number of degrees of freedom. The fact that S ∼ E is
equivalent to the fact that F ∼ T 2 , as befits an energy density in a theory with one
– 90 –
spatial dimension. To see this, we need only approximate the integral by the saddle
point S 0 (E? ) = β. From (4.42), this gives us the free energy
F ∼ N 2T 2
We can now make the statement about the central charge more explicit. In a conformal
field theory, the entropy of high energy states is given by
√
S(E) ∼ cE
This is Cardy’s formula.
4.4.4 c has a Theorem
The connection between the central charge and the degrees of freedom in a theory
is given further weight by a result of Zamalodchikov, known as the c-theorem. The
idea of the c-theorem is to stand back and look at the space of all theories and the
renormalization group (RG) flows between them.
Conformal field theories are special. They are the fixed points of the renormalization
group, looking the same at all length scales. One can consider perturbing a conformal
field theory by adding an extra term to the action,
Z
S → S + α d2 σ O(σ)
Here O is a local operator of the theory, while α is some coefficient. These perturbations
fall into three classes, depending on the dimension ∆ of O.
• ∆ < 2: In this case, α has positive dimension: [α] = 2 − δ. Such deformations
are called relevant because they are important in the infra-red. RG flow takes
us away from our original CFT. We only stop flowing when we hit a new CFT
(which could be trivial with c = 0).
• ∆ = 2: The constant α is dimensionless. Such deformations are called marginal.
The deformed theory defines a new CFT.
• ∆ > 2: The constant α has negative dimension. These deformations are irrelevant. The infra-red physics is still described by the original CFT. But the
ultra-violet physics is altered.
We expect information is lost as we flow from an ultra-violet theory to the infra-red.
The c-theorem makes this intuition precise. The theorem exhibits a function c on the
space of all theories which monotonically decreases along RG flows. At the fixed points,
c coincides with the central charge of the CFT.
– 91 –
4.5 The Virasoro Algebra
So far our discussion has been limited to the operators of the CFT. We haven’t said
anything about states. We now remedy this. We start by taking a closer look at the
map between the cylinder and the plane.
4.5.1 Radial Quantization
To discuss states in a quantum field theory we need to think
about where they live and how they evolve. For example, consider a two dimensional quantum field theory defined on the
plane. Traditionally, when quantizing this theory, we parameterize the plane by Cartesian coordinates (t, x) which we’ll
call “time” and “space”. The states live on spatial slices. The
Hamiltonian generates time translations and hence governs the
evolution of states.
t
x
Figure 24:
However, the map between the cylinder and the plane suggests a different way to
quantize a CFT on the plane. The complex coordinate on the cylinder is taken to be
ω, while the coordinate on the plane is z. They are related by,
ω = σ + iτ
,
z = e−iω
On the cylinder, states live on spatial slices of constant σ and evolve by the Hamiltonian,
H = ∂τ
After the map to the plane, the Hamiltonian becomes the dilatation operator
D = z∂ + z̄ ∂¯
If we want the states on the plane to remember their cylindrical roots, they should live
on circles of constant radius. Their evolution is governed by the dilatation operator D.
This approach to a theory is known as radial quantization.
Usually in a quantum field theory, we’re interested in time-ordered correlation functions. Time ordering on the cylinder becomes radial ordering on the plane. Operators
in correlation functions are ordered so that those inserted at larger radial distance are
moved to the left.
– 92 –
τ
σ
Figure 25: The map from the cylinder to the plane.
Virasoro Generators
Let’s look at what becomes of the stress tensor T (z) evaluated on the plane. On the
cylinder, we would decompose T in a Fourier expansion.
∞
X
c
Tcylinder (w) = −
Lm eimw +
24
m=−∞
After the transformation (4.33) to the plane, this becomes the Laurent expansion
∞
X
Lm
T (z) =
z m+2
m=−∞
As always, a similar statement holds for the right-moving sector
∞
X
L̃m
T̄ (z̄) =
z̄ m+2
m=−∞
We can invert these expressions to get Lm in terms of T (z). We need to take a suitable
contour integral
I
I
1
1
n+1
Ln =
dz z
T (z)
,
L̃n =
dz̄ z̄ n+1 T̄ (z̄)
(4.43)
2πi
2πi
where, if we just want Ln or L̃n , we must make sure that there are no other insertions
inside the contour.
In radial quantization, Ln is the conserved charge associated to the conformal transformation δz = z n+1 . To see this, recall that the corresponding Noether current, given
H
in (4.7), is J(z) = z n+1 T (z). Moreover, the contour integral dz maps to the integral
around spatial slices on the cylinder. This tells us that Ln is the conserved charge
where “conserved” means that it is constant under time evolution on the cylinder, or
under radial evolution on the plane. Similarly, L̃n is the conserved charge associated
to the conformal transformation δz̄ = z̄ n+1 .
– 93 –
When we go to the quantum theory, conserved charges become generators for the
transformation. Thus the operators Ln and L̃n generate the conformal transformations
δz = z n+1 and δz̄ = z̄ n+1 . They are known as the Virasoro generators. In particular,
our two favorite conformal transformations are
• L−1 and L̃−1 generate translations in the plane.
• L0 and L̃0 generate scaling and rotations.
The Hamiltonian of the system — which measures the energy of states on the cylinder
— is mapped into the dilatation operator on the plane. When acting on states of the
theory, this operator is represented as
D = L0 + L̃0
4.5.2 The Virasoro Algebra
If we have some number of conserved charges, the first thing that we should do is
compute their algebra. Representations of this algebra then classify the states of the
theory. (For example, think angular momentum in the hydrogen atom). For conformal
symmetry, we want to determine the algebra obeyed by the Ln generators. It’s a nice
fact that the commutation relations are actually encoded T T OPE. Let’s see how this
works.
H
We want to compute [Lm , Ln ]. Let’s write Lm as a contour integral over dz and
H
Ln as a contour integral over dw. (Note: both z and w denote coordinates on the
complex plane now). The commutator is
I
I
I
I
dz
dw
dw
dz
[Lm , Ln ] =
−
z m+1 wn+1 T (z) T (w)
2πi
2πi
2πi
2πi
What does this actually mean?! We need to remember that all operator equations
are to be viewed as living inside time-ordered correlation functions. Except, now we’re
working on the z-plane, this statement has transmuted into radially ordered correlation
functions: outies to the left, innies to the right.
z
w
z
w
So Lm Ln means
while Ln Lm means
– 94 –
.
The trick to computing the commutator is to first fix w and do the
The resulting contour is,
H
dz integrations.
z
z
z
w
w
In other words, we do the z-integration around a fixed point w, to get
I
I
dw
dz m+1 n+1
z
[Lm , Ln ] =
w
T (z) T (w)
2πi w 2πi
I
2T (w)
∂T (w)
dw
c/2
m+1 n+1
Res z
w
+
+
+ ...
=
2πi
(z − w)4 (z − w)2
z−w
To compute the residue at z = w, we first need to Taylor expand z m+1 about the point
w,
1
z m+1 = wm+1 + (m + 1)wm (z − w) + m(m + 1)wm−1 (z − w)2
2
1
2
m−2
+ m(m − 1)w
(z − w)3 + . . .
6
The residue then picks up a contribution from each of the three terms,
I
i
dw n+1 h m+1
c
m
2
m−2
[Lm , Ln ] =
w
w
∂T (w) + 2(m + 1)w T (w) + m(m − 1)w
2πi
12
To proceed, it is simplest to integrate the first term by parts. Then we do the wintegral. But for both the first two terms, the resulting integral is of the form (4.43)
and gives us Lm+n . For the third term, we pick up the pole. The end result is
[Lm , Ln ] = (m − n)Lm+n +
c
m(m2 − 1)δm+n,0
12
This is the Virasoro algebra. It’s quite famous. The L̃n ’s satisfy exactly the same
algebra, but with c replaced by c̃. Of course, [Ln , L̃m ] = 0. The appearance of c as an
extra term in the Virasoro algebra is the reason it is called the “central charge”. In
general, a central charge is an extra term in an algebra that commutes with everything
else.
– 95 –
Conformal = Diffeo + Weyl
We can build some intuition for the Virasoro algebra. We know that the Ln ’s generate
conformal transformations δz = z n+1 . Let’s consider something closely related: a
coordinate transformation δz = z n+1 . These are generated by the vector fields
ln = z n+1 ∂z
(4.44)
But it’s a simple matter to compute their commutation relations:
[ln , lm ] = (m − n)lm+n
So this is giving us the first part of the Virasoro algebra. But what about the central
term? The key point to remember is that, as we stressed at the beginning of this
chapter, a conformal transformation is not just a reparameterization of the coordinates:
it is a reparameterization, followed by a compensating Weyl rescaling. The central term
in the Virasoro algebra is due to the Weyl rescaling.
4.5.3 Representations of the Virasoro Algebra
With the algebra of conserved charges at hand, we can now start to see how the
conformal symmetry classifies the states into representations.
Suppose that we have some state |ψi that is an eigenstate of L0 and L̃0 .
L0 |ψi = h |ψi
,
L̃0 |ψi = h̃ |ψi
Back on the cylinder, this corresponds to some state with energy
c + c̃
E
= h + h̃ −
2π
24
For this reason, we’ll refer to the eigenvalues h and h̃ as the energy of the state. By
acting with the Ln operators, we can get further states with eigenvalues
L0 Ln |ψi = (Ln L0 − nLn ) |ψi = (h − n)Ln |ψi
This tells us that Ln are raising and lowering operators depending on the sign of n.
When n > 0, Ln lowers the energy of the state and L−n raises the energy of the state. If
the spectrum is to be bounded below, there must be some states which are annihilated
by all Ln and L̃n for n > 0. Such states are called primary. They obey
Ln |ψi = L̃n |ψi = 0
for all n > 0
In the language of representation theory, they are also called highest weight states.
They are the states of lowest energy.
– 96 –
Representations of the Virasoro algebra can now be built by acting on the primary
states with raising operators L−n with n > 0. Obviously this results in an infinite
tower of states. All states obtained in this way are called descendants. From an initial
primary state |ψi, the tower fans out...
|ψi
L−1 |ψi
L2−1 |ψi , L−2 |ψi
L3−1 |ψi , L−1 L−2 |ψi , L−3 |ψi
The whole set of states is called a Verma module. They are the irreducible representations of the Virasoro algebra. This means that if we know the spectrum of primary
states, then we know the spectrum of the whole theory.
Some comments:
• The vacuum state |0i has h = 0. This state obeys
Ln |0i = 0
for all n ≥ 0
(4.45)
Note that this state preserves the maximum number of symmetries: like all primary states, it is annihilated by Ln with n > 0, but it is also annihilated by
L0 . This fits with our intuition that the vacuum state should be invariant under
as many symmetries as possible. You might think that we could go further and
require that the vacuum state obeys Ln | 0i = 0 for all n. But that isn’t consistent
with the central charge term in Virasoro algebra. The requirements (4.45) are
the best we can do.
• This discussion should be ringing bells. We saw something very similar in the
covariant quantization of the string, where we imposed conditions (2.6) as constraints. We will see the connection between the primary states and the spectrum
of the string in Section 5.
• There’s a subtlety that you should be aware of: the states in the Verma module
are not necessarily all independent. It could be that some linear combination
of the states vanishes. This linear combination is known as a null state. The
existence of null states depends on the values of h and c. For example, suppose
that we are in a theory in which the central charge is c = 2h(5 − 8h)/(2h + 1),
where h is the energy of a primary state |ψi. Then it is simple to check that the
following combination has vanishing norm:
3
L2−1 |ψi
(4.46)
L−2 |ψi −
2(2h + 1)
– 97 –
• There is a close relationship between the primary states and the primary operators
defined in Section 4.2.3. In fact, the energies h and h̃ of primary states will turn
out to be exactly the weights of primary operators in the theory. This connection
will be described in Section 4.6.
4.5.4 Consequences of Unitarity
There is one physical requirement that a theory must obey which we have so far neglected to mention: unitarity. This is the statement that probabilities are conserved
when we are in Minkowski signature spacetime. Unitarity follows immediately if we
have a Hermitian Hamiltonian which governs time evolution. But so far our discussion
has been somewhat algebraic and we’ve not enforced this condition. Let’s do so now.
We retrace our footsteps back to the Euclidean cylinder and then back again to
the Minkowski cylinder where we can ask questions about time evolution. Here the
Hamiltonian density takes the form
X
+
−
H = Tww + Tw̄w̄ =
Ln e−inσ + L̃n e−inσ
n
So for the Hamiltonian to be Hermitian, we require
Ln = L†−n
This requirement imposes some strong constraints on the structure of CFTs. Here we
look at a couple of trivial, but important, constraints that arise due to unitarity and
the requirement that the physical Hilbert space does not contain negative norm states.
• h ≥ 0: This fact follows from looking at the norm,
|L−1 |ψi| 2 = hψ| L+1 L−1 |ψi = hψ| [L+1 , L−1 ] |ψi = 2h hψ| ψi ≥ 0
The only state with h = 0 is the vacuum state | 0i.
• c > 0: To see this, we can look at
c
n(n2 − 1) ≥ 0
(4.47)
12
So c ≥ 0. If c = 0, the only state in the vacuum module is the vacuum itself. It
turns out that, in fact, the only state in the whole theory is the vacuum itself.
Any non-trivial CFT has c > 0.
|L−n |0i| 2 = h0| [Ln , L−n ] |0i =
There are many more requirements of this kind that constrain the theory. In fact, it
turns out that for CFTs with c < 1 these requirements are enough to classify and solve
all theories.
– 98 –
4.6 The State-Operator Map
In this section we describe one particularly important aspect of conformal field theories:
a map between states and local operators.
Firstly, let’s get some perspective. In a typical quantum field theory, the states
and local operators are very different objects. While local operators live at a point in
spacetime, the states live over an entire spatial slice. This is most clear if we write
down a Schrödinger-style wavefunction. In field theory, this object is actually a wavefunctional, Ψ[φ(σ)], describing the probability for every field configuration φ(σ) at each
point σ in space (but at a fixed time).
Given that states and local operators are such very different beasts, it’s a little
surprising that in a CFT there is an isomorphism between them: it’s called the stateoperator map. The key point is that the distant past in the cylinder gets mapped to
a single point z = 0 in the complex plane. So specifying a state on the cylinder in the
far past is equivalent to specifying a local disturbance at the origin.
To make this precise, we need to recall how to write down wavefunctions using path
integrals. Different states are computed by putting different boundary conditions on
the functional integral. Let’s start by returning to quantum mechanics and reviewing
a few simple facts. The propagator for a particle to move from position xi at time τi
to position xf at time τf is given by
Z x(τf )=xf
Dx eiS
G(xf , xi ) =
x(τi )=xi
This means that if our system starts off in some state described by the wavefunction
ψi (xi ) at time τi then (ignoring the overall normalization) it evolves to the state
Z
ψf (xf , τf ) = dxi G(xf , xi ) ψi (xi , τi )
There are two lessons to take from this. Firstly, to determine the value of the wavefunction at a given point xf , we evaluate the path integral restricting to paths which
satisfy x(τf ) = xf . Secondly, the initial state ψ(xi ) acts as a weighting factor for the
integral over initial boundary conditions.
Let’s now write down the same formula in a field theory, where we’re dealing with
wavefunctionals. We’ll work with the Euclidean path integral on the cylinder. If we
start with some state Ψi [φi (σ)] at time τi , then it will evolve to the state
Z
Z φ(τf )=φf
Ψf [φf (σ), τf ] = Dφi
Dφ e−S[φ] Ψi [φi (σ), τi ]
φ(τi )=φi
– 99 –
How do we write a similar expression for states after the map
to the complex plane? Now the states are defined on circles
of constant radius, say |z| = r, and evolution is governed by
the dilatation operator. Suppose the initial state is defined
at |z| = ri . In the path integral, we integrate over all fields
with fixed boundary conditions φ(ri ) = φi and φ(rf ) = φf
on the two edges of the annulus shown in the figure,
rf
ri
Figure 26:
Z
Ψf [φf (σ), rf ] =
Z
φ(rf )=φf
Dφi
Dφ e−S[φ] Ψi [φi (σ), ri ]
φ(ri )=φi
This is the traditional way to define a state in field theory,
albeit with a slight twist because we’re working in radial
quantization. We see that the effect of the initial state is to
change the weighting of the path integral over the inner ring
at |z| = ri .
|ψ>
xO
(z=0)
Let’s now see what happens as we take the initial state
back to the far past and, ultimately, to z = 0? We must
Figure 27:
now integrate over the whole disc |z| ≤ rf , rather than the
annulus. The only effect of the initial state is now to change the weighting of the
path integral at the point z = 0. But that’s exactly what we mean by a local operator
inserted at that point. This means that each local operator O(z = 0) defines a different
state in the theory,
Z
Ψ[φf ; r] =
φ(r)=φf
Dφ e−S[φ] O(z = 0)
We’re now integrating over all field configurations within the disc, including all possible
values of the field at z = 0, which is analogous to integrating over the boundary
R
conditions Dφi on the inner circle.
• The state-operator map is only true in conformal field theories where we can
map the cylinder to the plane. It also holds in conformal field theories in higher
dimensions (where R × SD−1 can be mapped to the plane RD ). In non-conformal
field theories, a typical local operator creates many different states.
• The state-operator map does not say that the number of states in the theory is
equal to the number of operators: this is never true. It does say that the states
are in one-to-one correspondence with the local operators.
– 100 –
• You might think that you’ve seen something like this before. In the canonical
quantization of free fields, we create states in a Fock space by acting with creation
operators. That’s not what’s going on here! The creation operators are just about
as far from local operators as you can get. They are the Fourier transforms of
local operators.
• There’s a special state that we can create this way: the vacuum. This arises
by inserting the identity operator 1 into the path integral. Back in the cylinder
picture, this just means that we propagate the state back to time τ = −∞ which
is a standard trick used in the Euclidean path integral to project out all but the
ground state. For this reason the vacuum is sometimes referred to, in operator
notation, as |1i.
4.6.1 Some Simple Consequences
Let’s use the state-operator map to wrap up a few loose ends that have arisen in our
study of conformal field theory.
Firstly, we’ve defined two objects that we’ve called “primary”: states and operators.
The state-operator map relates the two. Consider the state |Oi, built from inserting a
primary operator O into the path integral at z = 0. We can look at,
I
dz n+1
Ln |Oi =
z
T (z) O(z = 0)
2πi
I
dz n+1 hO ∂O
=
z
+
+ ...
(4.48)
2πi
z2
z
R
You may wonder what became of the path integral Dφ e−S[φ] in this expression. The
answer is that it’s still implicitly there. Remember that operator expressions such as
(4.43) are always taken to hold inside correlation functions. But putting an operator in
the correlation function is the same thing as putting it in the path integral, weighted
with e−S[φ] .
From (4.48) we can see the effect of various generators on states
• L−1 |Oi = |∂Oi: In fact, this is true for all operators, not just primary ones. It
is expected since L−1 is the translation generator.
• L0 |Oi = h |Oi: This is true of any operator with well defined transformation
under scaling.
• Ln |Oi = 0 for all n > 0. This is true only of primary operators O. Moreover, it
is our requirement for |Oi to be a primary state.
– 101 –
This has an important consequence. We stated earlier that one of the most important
things to compute in a CFT is the spectrum of weights of primary operators. This
seems like a slightly obscure thing to do. But now we see that it has a much more
direct, physical meaning. It is the spectrum of energy and angular momentum of states
of the theory defined on the cylinder.
Another loose end: when defining operators which carry specific weight, we made
the statement that we could always work in a basis of operators which have specified
eigenvalues under D and L. This follows immediately from the statement that we can
always find a basis of eigenstates of H and L on the cylinder.
X
X
X
X
X
X
=
|ψ>
X
X
X
X
Figure 28:
Finally, we can use this idea of the state-operator map to understand why the OPE
works so well in conformal field theories. Suppose that we’re interested in some correlation function, with operator insertions as shown in the figure. The statement of the
OPE is that we can replace the two inner operators by a sum of operators at z = 0,
independent of what’s going on outside of the dotted line. As an operator statement,
that sounds rather surprising. But this follows by computing the path integral up to
the dotted line, by which point the only effect of the two operators is to determine
what state we have. This provides us a way of understanding why the OPE is exact in
CFTs, with a radius of convergence equal to the next-nearest insertion.
4.6.2 Our Favourite Example: The Free Scalar Field
Let’s illustrate the state-operator map by returning yet again to the free scalar field.
On a Euclidean cylinder, we have the mode expansion
r
α0 X 1
X(w, w̄) = x + α0 p τ + i
αn einw + α̃n einw̄
2 n6=0 n
where we retain the requirement of reality in Minkowski space, which gave us αn? = α−n
and α̃n? = α̃−n . We saw in Section 4.3 that X does not have good conformal properties.
– 102 –
Before transforming to the z = e−iw plane, we should work with the primary field on
the cylinder,
r
r
α0 X
α0
αn einw
with α0 ≡ i
p
∂w X(w, w̄) = −
2 n
2
Since ∂X is a primary field of weight h = 1, its transformation to the plane is given by
(4.18) and reads
r
−1
α0 X αn
∂z
∂w X(w) = −i
∂z X(z) =
∂w
2 n z n+1
¯
and similar for ∂X.
Inverting this gives an equation for αn as a contour integral,
r I
2
dz n
αn = i
z ∂X(z)
(4.49)
α0
2πi
Just as the T T OPE allowed us to determine the [Lm , Ln ] commutation relations in
the previous section, so the ∂X∂X OPE contains the information about the [αm , αn ]
commutation relations. The calculation is straightforward,
I
I
I
I
2
dz
dw
dw
dz
−
[αm , αn ] = − 0
z m wn ∂X(z) ∂X(w)
α
2πi
2πi
2πi
2πi
I
2
dw
−α0 /2
m n
Res z=w z w
+ ...
=− 0
α
2πi
(z − w)2
I
dw m+n−1
=m
w
= mδm+n,0
2πi
where, in going from the second to third line, we have Taylor expanded z around
w. Hearteningly, the final result agrees with the commutation relation (2.2) that we
derived in string theory using canonical quantization.
The State-Operator Map for the Free Scalar Field
Let’s now look at the map between states and local operators. We know from canonical
quantization that the Fock space is defined by acting with creation operators α−m with
m > 0 on the vacuum |0i. The vacuum state itself obeys αm |0i = 0 for m > 0. Finally,
there is also the zero mode α0 ∼ p which provides all states with another quantum
number. A general state is given by
∞
Y
km
|0; pi
α−m
m=1
– 103 –
Let’s try and recover these states by inserting operators into the path integral. Our
first task is to check whether the vacuum state is indeed equivalent to the insertion of
the identity operator. In other words, is the ground state wavefunctional of the theory
on the circle |z| = r really given by
Z
Xf (r)
Ψ0 [Xf ] =
DX e−S[X]
?
(4.50)
We want to check that this satisfies the definition of the vacuum state, namely αm |0i = 0
for m > 0. How do we act on the wavefunctional with an operator? We should still
integrate over all field configurations X(z, z̄), subject to the boundary conditions at
X(|z| = r) = Xf . But now we should insert the contour integral (4.49) at some |w| < r
(because, after all, the state is only going to vanish after we’ve hit it with αm , not
before!). So we look at
Z
αm Ψ0 [Xf ] =
Xf
DX e
−S[X]
I
dw m
w ∂X(w)
2πi
The path integral is weighted by the action (4.19) for a free scalar field. If a given
configuration diverges somewhere inside the disc |z| < r, then the action also diverges.
This ensures that only smooth functions ∂X(z), which have no singularity inside the
disc, contribute. But for such functions we have
I
dw m
w ∂X(w) = 0
for all m ≥ 0
2πi
So the state (4.50) is indeed the vacuum state. In fact, since α0 also annihilates this
state, it is identified as the vacuum state with vanishing momentum.
What about the excited states of the theory?
Claim: α−m |0i = |∂ m Xi. By which we mean that the state α−m |0i can be built
from the path integral,
Z
α−m |0i = DX e−S[X] ∂ m X(z = 0)
(4.51)
Proof: We can check this by acting on |∂ m Xi with the annihilation operators αn .
m
αn |∂ Xi ∼
Z
Xf (r)
DX e
−S[X]
I
– 104 –
dw n
w ∂X(w) ∂ m X(z = 0)
2πi
We can focus on the operator insertions and use the OPE (4.23). We drop the path
integral and just focus on the operator equation (because, after all, operator equations
only make sense in correlation functions which is the same thing as in path integrals).
We have
I
I
dw
1
dw n−m−1
n m−1
w ∂z
w
=0
unless m = n
= m!
2
2πi
(w − z) z=0
2πi
This confirms that the state (4.51) has the right properties.
Finally, we should worry about the zero mode, or momentum α0 ∼ p. It is simple to
show using the techniques above (together with the OPE (4.26)) that the momentum
of a state arises by the insertion of the primary operator eipX . For example,
Z
|0; pi ∼ DX e−S[X] eipX(z=0) .
4.7 Brief Comments on Conformal Field Theories with Boundaries
The open string lives on the infinite strip
with spatial coordinate σ ∈ [0, π]. Here we
make just a few brief comments on the corresponding conformal field theories.
As before, we can define the complex coordinate w = σ + iτ and make the conformal map
z = e−iw
Figure 29:
This time the map takes us to the upper-half plane: Imz ≥ 0. The end points of the
string are mapped to the real axis, Imz = 0.
Much of our previous discussion goes through as before. But now we need to take
care of boundary conditions at Imz = 0. Let’s first look at Tαβ . Recall that the stressenergy tensor exists because of translational invariance. We still have translational
invariance in the direction parallel to the boundary — let’s call the associated tangent
vector tα . But translational invariance is broken perpendicular to the boundary — we
call the normal vector nα . The upshot of this is that Tαβ tβ remains a conserved current.
To implement Neumann boundary conditions, we insist that none of the current flows
out of the boundary. The condition is
Tαβ nα tβ = 0
at Imz = 0
– 105 –
In complex coordinates, this becomes
Tzz = Tz̄z̄
at Imz = 0
There’s a simple way to implement this: we extend the definition of Tzz from the
upper-half plane to the whole complex plane by defining
Tzz (z) = Tz̄z̄ (z̄)
For the closed string we had both functions T and T̄ in the whole plane. But for the
open string, we have just one of these – say, T , — in the whole plane. This contains
the same information as both T and T̄ in the upper-half plane. It’s simpler to work in
the whole plane and focus just on T . Correspondingly, we now have just a single set of
Virasoro generators,
I
dz n+1
z
Tzz (z)
Ln =
2πi
There is no independent L̃n for the open string.
A similar doubling trick works when computing the propagator for the free scalar
field. The scalar field X(z, z̄) is only defined in the upper-half plane. Suppose we want
to implement Neumann boundary conditions. Then the propagator is defined by
hX(z, z̄) X(w, w̄)i = G(z, z̄; w, w̄)
which obeys ∂ 2 G = −2πα0 δ(z − w, z̄ − w̄) subject to the boundary condition
∂σ G(z, z̄; w, w̄)|σ=0 = 0
But we solve problems like this in our electrodynamics courses. A useful way of proceeding is to introduce an “image charge” in the lower-half plane. We now let X(z, z̄)
vary over the whole complex plane with its dynamics governed by the propagator
α0
α0
2
G(z, z̄; w, w̄) = − ln |z − w| −
ln |z − w̄|2
2
2
(4.52)
Much of the remaining discussion of CFTs carries forward with only minor differences.
However, there is one point that is simple but worth stressing because it will be of
importance later. This concerns the state-operator map. Recall the logic that leads
us to this idea: we consider a state at fixed time on the strip and propagate it back
to past infinity τ → −∞. After the map to the half-plane, past infinity is again the
origin. But now the origin lies on the boundary. We learn that the state-operator map
relates states to local operators defined on the boundary.
– 106 –
This fact ensures that theories on a strip have fewer states than those on the cylinder.
¯
For example, for a free scalar field, Neumann boundary conditions require ∂X = ∂X
at Imz = 0. (This follows from the requirement that ∂σ X = 0 at σ = 0, π on the
¯ give rise to different states; on the
strip). On the cylinder, the operators ∂X and ∂X
strip they give rise to the same state. This, of course, mirrors what we’ve seen for the
quantization of the open string where boundary conditions mean that we have only
half the oscillator modes to play with.
– 107 –
5. The Polyakov Path Integral and Ghosts
At the beginning of the last chapter, we stressed that there are two very different
interpretations of conformal symmetry depending on whether we’re thinking of a fixed
2d background or a dynamical 2d background. In applications to statistical physics,
the background is fixed and conformal symmetry is a global symmetry. In contrast, in
string theory the background is dynamical. Conformal symmetry is a gauge symmetry,
a remnant of diffeomorphism invariance and Weyl invariance.
But gauge symmetries are not symmetries at all. They are redundancies in our
description of the system. As such, we can’t afford to lose them and it is imperative
that they don’t suffer an anomaly in the quantum theory. At worst, theories with
gauge anomalies make no sense. (For example, Yang-Mills theory coupled to only lefthanded fundamental fermions is a nonsensical theory for this reason). At best, it may
be possible to recover the quantum theory, but it almost certainly has nothing to do
with the theory that you started with.
Piecing together some results from the previous chapter, it looks like we’re in trouble.
We saw that the Weyl symmetry is anomalous since the expectation value of the stressenergy tensor takes different values on backgrounds related by a Weyl symmetry:
hT αα i = −
c
R
12
On fixed backgrounds, that’s merely interesting. On dynamical backgrounds, it’s fatal.
What can we do? It seems that the only way out is to ensure that our theory has c = 0.
But we’ve already seen that c > 0 for all non-trivial, unitary CFTs. We seem to have
reached an impasse. In this section we will discover the loophole. It turns out that we
do indeed require c = 0, but there’s a way to achieve this that makes sense.
5.1 The Path Integral
In Euclidean space the Polyakov action is given by,
Z
1
√
d2 σ g g αβ ∂α X µ ∂β X ν δµν
SPoly =
0
4πα
From now on, our analysis of the string will be in terms of the path integral5 . We integrate over all embedding coordinates X µ and all worldsheet metrics gαβ . Schematically,
5
The analysis of the string path integral was first performed by Polyakov in “Quantum geometry
of bosonic strings,”, Phys. Lett. B 103, 207 (1981). The paper weighs in at a whopping 4 pages. As
a follow-up, he took another 2.5 pages to analyze the superstring in “Quantum geometry of fermionic
strings,” Phys. Lett. B 103, 211 (1981).
– 108 –
the path integral is given by,
1
Z=
Vol
Z
DgDX e−SPoly [X,g]
The “Vol” term is all-important. It refers to the fact
that we shouldn’t be integrating over all field configurations, but only those physically distinct configurations not
related by diffeomorphisms and Weyl symmetries. Since
the path integral, as written, sums over all fields, the
“Vol” term means that we need to divide out by the volume of the gauge action on field space.
Gauge
Fixing
Gauge Orbits
Figure 30:
To make the situation more explicit, we need to split
the integration over all field configurations into two pieces:
those corresponding to physically distinct configurations — schematically depicted as
the dotted line in the figure — and those corresponding to gauge transformations —
which are shown as solid lines. Dividing by “Vol” simply removes the piece of the
partition function which comes from integrating along the solid-line gauge orbits.
In an ordinary integral, if we change coordinates then we pick up a Jacobian factor for
our troubles. The path integral is no different. We want to decompose our integration
variables into physical fields and gauge orbits. The tricky part is to figure out what
Jacobian we get. Thankfully, there is a standard method to determine the Jacobian,
first introduced by Faddeev and Popov. This method works for all gauge symmetries,
including Yang-Mills and you will also learn about it in the “Advanced Quantum Field
Theory” course.
5.1.1 The Faddeev-Popov Method
We have two gauge symmetries: diffeomorphisms and Weyl transformations. We will
schematically denote both of these by ζ. The change of the metric under a general
gauge transformation is g → g ζ . This is shorthand for,
gαβ (σ) −→
ζ
gαβ
(σ 0 )
2ω(σ)
=e
∂σ γ ∂σ δ
gγδ (σ)
∂σ 0 α ∂σ 0 β
In two dimensions these gauge symmetries allow us to put the metric into any form
that we like — say, ĝ. This is called the fiducial metric and will represent our choice
of gauge fixing. Two caveats:
• Firstly, it’s not true that we can put any 2d metric into the form ĝ of our choosing.
This is only true locally. Globally, it remains true if the worldsheet has the
– 109 –
topology of a cylinder or a sphere, but not for higher genus surfaces. We’ll revisit
this issue in Section 6.
• Secondly, fixing the metric locally to ĝ does not fix all the gauge symmetries. We
still have the conformal symmetries to deal with. We’ll revisit this in the Section
6 as well.
Our goal is to only integrate over physically inequivalent configurations. To achieve
this, first consider the integral over the gauge orbit of ĝ. For some value of the gauge
transformation ζ, the configuration g ζ will coincide with our original metric g. We can
put a delta-function in the integral to get
Z
Dζ δ(g − ĝ ζ ) = ∆−1
(5.1)
F P [g]
This integral isn’t equal to one because we need to take into account the Jacobian
R
factor. This is analogous to the statement that dx δ(f (x)) = 1/|f 0 |, evaluated at
points where f (x) = 0. In the above equation, we have written this Jacobian factor as
∆−1
F P . The inverse of this, namely ∆F P , is called the Faddeev-Popov determinant. We
will evaluate it explicitly shortly. Some comments:
• This whole procedure is rather formal and runs into the usual difficulties with
trying to define the path integral. Just as for Yang-Mills theory, we will find that
it results in sensible answers.
• We will assume that our gauge fixing is good, meaning that the dotted line in the
previous figure cuts through each physically distinct configuration exactly once.
Equivalently, the integral over gauge transformations Dζ clicks exactly once with
the delta-function and we don’t have to worry about discrete ambiguities (known
as Gribov copies in QCD).
• The measure is taken to be the analogue of the Haar measure for Lie groups,
invariant under left and right actions
Dζ = D(ζ 0 ζ) = D(ζζ 0 )
Before proceeding, it will be useful to prove a quick lemma:
Lemma: ∆F P [g] is gauge invariant. This means that
∆F P [g] = ∆F P [g ζ ]
– 110 –
Proof:
ζ
∆−1
F P [g ]
Z
Dζ 0 δ(g ζ − ĝ ζ )
Z
Dζ 0 δ(g − ĝ ζ
=
=
Z
=
0
−1 ζ 0
)
00
Dζ 00 δ(g − ĝ ζ ) = ∆−1
F P [g]
where, in the second line, we have used the fact that the measure is invariant.
Now we can employ the Faddeev-Popov procedure. We start by inserting a factor of
unity into the path integral, in the guise of
Z
1 = ∆F P [g]
Dζ δ(g − ĝ ζ )
We’ll call the resulting path integral expression Z[ĝ] since it depends on the choice of
fiducial metric ĝ. The first thing we do is use the δ(g − ĝ ζ ) delta-function to do the
integral over metrics,
Z
1
DζDXDg ∆F P [g] δ(g − ĝ ζ ) e−SPoly [X,g]
Z[ĝ] =
Vol
Z
1
ζ
DζDX ∆F P [ĝ ζ ] e−SPoly [X,ĝ ]
=
Vol
Z
1
=
DζDX ∆F P [ĝ] e−SPoly [X,ĝ]
Vol
where to go to the last line, we’ve used the fact that the action is gauge invariant,
so S[X, g] = S[X, g ζ ] and, as we proved in the lemma above, the Faddeev-Popov
determinant is also gauge invariant.
But now, nothing depends on the gauge transformation ζ. Indeed, this is precisely
the integration over the gauge orbits that we wanted to isolate and it cancels the “Vol”
factor sitting outside. We’re left with
Z
Z[ĝ] = DX ∆F P [ĝ] e−SPoly [X,ĝ]
(5.2)
This is the integral over physically distinct configurations — the dotted line in the
previous figure. We see that the Faddeev-Popov determinant is precisely the Jacobian
factor that we need.
– 111 –
5.1.2 The Faddeev-Popov Determinant
We still need to compute ∆F P [ĝ]. It’s defined in (5.1). Let’s look at gauge transformations ζ which are close to the identity. In this case, the delta-function δ(g − ĝ ζ )
is going to be non-zero when the metric g is close to the fiducial metric ĝ. In fact, it
will be sufficient to look at the delta-function δ(ĝ − ĝ ζ ), which is only non-zero when
ζ = 0. We take an infinitesimal Weyl transformation parameterized by ω(σ) and an
infinitesimal diffeomorphism δσ α = v α (σ). The change in the metric is
δĝαβ = 2ωĝαβ + ∇α vβ + ∇β vα
Plugging this into the delta-function, the expression for the Faddeev-Popov determinant
becomes
Z
−1
∆F P [ĝ] = DωDv δ(2ωĝαβ + ∇α vβ + ∇β vα )
(5.3)
where we’ve replaced the integral Dζ over the gauge group with the integral DωDv over
the Lie algebra of group since we’re near the identity. (We also suppress the subscript
on vα in the measure factor to keep things looking tidy).
At this stage it’s useful to represent the delta-function in its integral, Fourier form.
R
For a single delta-function, this is δ(x) = dp exp(2πip x). But the delta-function in
(5.3) is actually a delta-functional: it restricts a whole function. Correspondingly, the
integral representation is in terms of a functional integral,
Z
Z
p αβ
−1
2
∆F P [ĝ] = DωDvDβ exp 2πi d σ ĝ β [2ωĝαβ + ∇α vβ + ∇β vα ]
where β αβ is a symmetric 2-tensor on the worldsheet.
R
We now simply do the Dω integral. It doesn’t come with any derivatives, so it
merely acts as a Lagrange multiplier, setting
β αβ ĝαβ = 0
In other words, after performing the ω integral, β αβ is symmetric and traceless. We’ll
take this to be the definition of β αβ from now on. So, finally we have
Z
Z
p αβ
−1
2
∆F P [ĝ] = DvDβ exp 4πi d σ ĝ β ∇α vβ
– 112 –
5.1.3 Ghosts
The previous manipulations give us an expression for ∆−1
F P . But we want to invert it
to get ∆F P . Thankfully, there’s a simple way to achieve this. Because the integrand
is quadratic in v and β, we know that the integral computes the inverse determinant
of the operator ∇α . (Strictly speaking, it computes the inverse determinant of the
projection of ∇α onto symmetric, traceless tensors. This observation is important
because it means the relevant operator is a square matrix which is necessary to talk
about a determinant). But we also know how to write down an expression for the
determinant ∆F P , instead of its inverse, in terms of path integrals: we simply need to
replace the commuting integration variables with anti-commuting fields,
βαβ −→ bαβ
v α −→ cα
where b and c are both Grassmann-valued fields (i.e. anti-commuting). They are known
as ghost fields. This gives us our final expression for the Faddeev-Popov determinant,
Z
∆F P [g] = DbDc exp[iSghost ]
where the ghost action is defined to be
Sghost
1
=
2π
Z
√
d2 σ g bαβ ∇α cβ
(5.4)
and we have chosen to rescale the b and c fields at this last step to get a factor of 1/2π
sitting in front of the action. (This only changes the normalization of the partition
function which doesn’t matter). Rotating back to Euclidean space, the factor of i
disappears. The expression for the full partition function (5.2) is
Z
Z[ĝ] =
DXDbDc exp (−SPoly [X, ĝ] − Sghost [b, c, ĝ])
Something lovely has happened. Although the ghost fields were introduced as some
auxiliary constructs, they now appear on the same footing as the dynamical fields X.
We learn that gauge fixing comes with a price: our theory has extra ghost fields.
The role of these ghost fields is to cancel the unphysical gauge degrees of freedom,
leaving only the D − 2 transverse modes of X µ . Unlike lightcone quantization, they
achieve this in a way which preserves Lorentz invariance.
– 113 –
Simplifying the Ghost Action
The ghost action (5.4) looks fairly simple. But it looks even simpler if we work in
conformal gauge,
ĝαβ = e2ω δαβ
√
The determinant is ĝ = e2ω . Recall that in complex coordinates, the measure is
d2 σ = 12 d2 z, while we can lower the index on the covariant derivative using ∇z =
g zz̄ ∇z̄ = 2e−2ω ∇z̄ . We have
Z
1
Sghost =
d2 z (bzz ∇z̄ cz + bz̄z̄ ∇z cz̄ )
2π
In deriving this, remember that there is no field bzz̄ because bαβ is traceless. Now comes
the nice part: the covariant derivatives are actually just ordinary derivatives. To see
why this is the case, look at
∇z̄ cz = ∂z̄ cz + Γzz̄α cα
But the Christoffel symbols are given by
1
Γzz̄α = g zz̄ (∂z̄ gαz̄ + ∂α gz̄z̄ − ∂z̄ gz̄α ) = 0
2
for α = z, z̄
So in conformal gauge, the ghost action factorizes into two free theories,
Z
1
Sghost =
d2 z bzz ∂z̄ cz + bz̄z̄ ∂z cz̄
2π
The action doesn’t depend on the conformal factor ω. In other words, it is Weyl
invariant without any need to change b and c: these are therefore both neutral under
Weyl transformations.
(It’s worth pointing out that bαβ and cα are neutral under Weyl transformations.
But if we raise or lower these indices, then the fields pick up factors of the metric. So
bαβ and cα would not be neutral under Weyl transformations).
5.2 The Ghost CFT
Fixing the Weyl and diffeomorphism gauge symmetries has left us with two new dynamical ghost fields, b and c. Both are Grassmann (i.e. anti-commuting) variables.
Their dynamics is governed by a CFT. Define
b = bzz
c=c
z
,
b̄ = bz̄z̄
,
c̄ = cz̄
– 114 –
The ghost action is given by
Sghost
1
=
2π
Z
¯ + b̄ ∂c̄
d2 z b ∂c
Which gives the equations of motion
¯ = ∂ b̄ = ∂c
¯ = ∂c̄ = 0
∂b
So we see that b and c are holomorphic fields, while b̄ and c̄ are anti-holomorphic.
Before moving onto quantization, there’s one last bit of information we need from
the classical theory: the stress tensor for the bc ghosts. The calculation is a little
bit fiddly. We use the general definition of the stress tensor (4.4), which requires
us to return to the theory (5.4) on a general background and vary the metric g αβ .
The complications are twofold. Firstly, we pick up a contribution from the Christoffel
symbol that is lurking inside the covariant derivative ∇α . Secondly, we must also
remember that bαβ is traceless. But this is a condition which itself depends on the
metric: bαβ g αβ = 0. To account for this we should add a Lagrange multiplier to the
action imposing tracelessness. After correctly varying the metric, we may safely retreat
back to flat space where the end result is rather simple. We have Tzz̄ = 0, as we must
for any conformal theory. Meanwhile, the holomorphic and anti-holomorphic parts of
the stress tensor are given by,
T = 2(∂c) b + c ∂b
,
¯ b̄ + c̄ ∂¯b̄.
T̄ = 2(∂c̄)
(5.5)
Operator Product Expansions
We can compute the OPEs of these fields using the standard path integral techniques
that we employed in the last chapter. In what follows, we’ll just focus on the holomorphic piece of the CFT. We have, for example,
Z
Z
−S
1 ¯
δ
0
−Sghost
0
0
ghost
e
b(σ ) = DbDc e
− ∂c(σ) b(σ ) + δ(σ − σ )
0 = DbDc
δb(σ)
2π
which tells us that
¯
∂c(σ)
b(σ 0 ) = 2π δ(σ − σ 0 )
Similarly, looking at δ/δc(σ) gives
¯
∂b(σ)
c(σ 0 ) = 2π δ(σ − σ 0 )
– 115 –
¯
We can integrate both of these equations using our favorite formula ∂(1/z)
= 2πδ(z, z̄).
We learn that the OPEs between fields are given by
1
+ ...
z−w
1
+ ...
c(w) b(z) =
w−z
b(z) c(w) =
In fact the second equation follows from the first equation and Fermi statistics. The
OPEs of b(z) b(w) and c(z) c(w) have no singular parts. They vanish as z → w.
Finally, we need the stress tensor of the theory. After normal ordering, it is given by
T (z) = 2 : ∂c(z) b(z) : + : c(z)∂b(z) :
We will shortly see that with this choice, b and c carry appropriate weights for tensor
fields which are neutral under Weyl rescaling.
Primary Fields
We will now show that both b and c are primary fields, with weights h = 2 and h = −1
respectively. Let’s start by looking at c. The OPE with the stress tensor is
T (z) c(w) = 2 : ∂c(z) b(z) : c(w)+ : c(z) ∂b(z) : c(w)
2∂c(z)
c(z)
c(w)
∂c(w)
−
+ ... = −
+
+ ...
=
2
2
z−w
(z − w)
(z − w)
z−w
confirming that c has weight −1. When taking the OPE with b, we need to be a little
more careful with minus signs. We get
T (z) b(w) = 2 : ∂c(z) b(z) : b(w)+ : c(z) ∂b(z) : b(w)
−1
2b(w)
∂b(w)
∂b(z)
= −2b(z)
=
+
+ ...
−
2
2
(z − w)
z−w
(z − w)
z−w
showing that b has weight 2. As we’ve pointed out a number of times, conformal =
diffeo + Weyl. We mentioned earlier that the fields b and c are neutral under Weyl
transformations. This is reflected in their weights, which are due solely to diffeomorphisms as dictated by their index structure: bzz and cz .
The Central Charge
Finally, we can compute the T T OPE to determine the central charge of the bc ghost
system.
T (z) T (w) = 4 : ∂c(z)b(z) : : ∂c(w)b(w) : +2 : ∂c(z)b(z) : : c(w)∂b(w) :
+2 : c(z)∂b(z) : : ∂c(w)b(w) : + : c(z)∂b(z) : : c(w)∂b(w) :
– 116 –
For each of these terms, making two contractions gives a (z − w)−4 contribution to the
OPE. There are also two ways to make a single contraction. These give (z − w)−1 or
(z − w)−2 or (z − w)−3 contributions depending on what the derivatives hit. The end
result is
T (z) T (w) =
−4
4 : ∂c(z)b(w) : 4 : b(z)∂c(w) :
+
−
4
(z − w)
(z − w)2
(z − w)2
4
2 : ∂c(z) ∂b(w) : 4 : b(z)c(w) :
−
−
+
4
(z − w)
z−w
(z − w)3
4
4 : c(z)b(w) : 2 : ∂b(z)∂c(w) :
−
−
+
4
(z − w)
(z − w)3
z−w
1
: c(z)∂b(w) : ∂b(z)c(w) :
−
−
+
+ ...
4
(z − w)
(z − w)2
(z − w)2
After some Taylor expansions to turn f (z) functions into f (w) functions, together with
a little collecting of terms, this can be written as,
T (z) T (w) =
2T (w)
∂T (w)
−13
+
+
+ ...
4
2
(z − w)
(z − w)
z−w
The first thing to notice is that it indeed has the form expected of T T OPE. The second,
and most important, thing to notice is the central charge of the bc ghost system: it is
c = −26
5.3 The Critical “Dimension” of String Theory
Let’s put the pieces together. We’ve learnt that gauge fixing the diffeomorphisms and
Weyl gauge symmetries results in the introduction of ghosts which contribute central
charge c = −26. We’ve also learnt that the Weyl symmetry is anomalous unless c = 0.
Since the Weyl symmetry is a gauge symmetry, it’s crucial that we keep it. We’re forced
to add exactly the right degrees of freedom to the string to cancel the contribution from
the ghosts.
The simplest possibility is to add D free scalar fields. Each of these contributes c = 1
to the central charge, so the whole procedure is only consistent if we pick
D = 26
This agrees with the result we found in Chapter 2: it is the critical dimension of string
theory.
– 117 –
However, there’s no reason that we have to work with free scalar fields. The consistency requirement is merely that the degrees of freedom of the string are described by a
CFT with c = 26. Any CFT will do. Each such CFT describes a different background
in which a string can propagate. If you like, the space of CFTs with c = 26 can be
thought of as the space of classical solutions of string theory.
We learn that the “critical dimension” of string theory is something of a misnomer:
it is really a “critical central charge”. Only for rather special CFTs can this central
charge be thought of as a spacetime dimension.
For example, if we wish to describe strings moving in 4d Minkowski space, we can
take D = 4 free scalars (one of which will be timelike) together with some other c = 22
CFT. This CFT may have a geometrical interpretation, or it may be something more
abstract. The CFT with c = 22 is sometimes called the “internal sector” of the theory.
It is what we really mean when we talk about the “extra hidden dimensions of string
theory”. We’ll see some examples of CFTs describing curved spaces in Section 7.
There’s one final subtlety: we need to be careful with the transition back to Minkowski
space. After all, we want one of the directions of the CFT, X 0 , to have the wrong sign
kinetic term. One safe way to do this is to keep X 0 as a free scalar field, with the
remaining degrees of freedom described by some c = 25 CFT. This doesn’t seem quite
satisfactory though since it doesn’t allow for spacetimes which evolve in time — and, of
course, these are certainly necessary if we wish to understand early universe cosmology.
There are still some technical obstacles to understanding the worldsheet of the string
in time-dependent backgrounds. To make progress, and discuss string cosmology, we
usually bi-pass this issue by working with the low-energy effective action which we will
derive in Section 7.
5.3.1 The Usual Nod to the Superstring
The superstring has another gauge symmetry on the worldsheet: supersymmetry. This
gives rise to more ghosts, the so-called βγ system, which turns out to have central
charge +11. Consistency then requires that the degrees of freedom of the string have
central charge c = 26 − 11 = 15.
However, now the CFTs must themselves be invariant under supersymmetry, which
means that bosons come matched with fermions. If we add D bosons, then we also
need to add D fermions. A free boson has c = 1, while a free fermion has c = 1/2. So,
the total number of free bosons that we should add is D(1 + 1/2) = 15, giving us the
– 118 –
critical dimension of the superstring:
D = 10
5.3.2 An Aside: Non-Critical Strings
Although it’s a slight departure from the our main narrative, it’s worth pausing to
mention what Polyakov actually did in his four page paper. His main focus was not
critical strings, with D = 26, but rather non-critical strings with D 6= 26. From the
discussion above, we know that these suffer from a Weyl anomaly. But it turns out
that there is a way to make sense of the situation.
The starting point is to abandon Weyl invariance from the beginning. We start with
D free scalar fields coupled to a dynamical worldsheet metric gαβ . (More generally, we
could have any CFT). We still want to keep reparameterization invariance, but now
we ignore the constraints of Weyl invariance. Of course, it seems likely that this isn’t
going to have too much to do with the Nambu-Goto string, but let’s proceed anyway.
Without Weyl invariance, there is one extra term that it is natural to add to the 2d
theory: a worldsheet cosmological constant µ,
Z
1
√
Snon−critical =
d2 σ g g αβ ∂α X µ ∂β Xµ + µ
0
4πα
Our goal will be to understand how the partition function changes under a Weyl rescaling. There will be two contributions: one from the explicit µ dependence and one from
the Weyl anomaly. Consider two metrics related by a Weyl transformation
ĝαβ = e2ω gαβ
As we vary ω, the partition function Z[ĝ] changes as
Z
1 ∂Z
1
∂S ∂ĝαβ
−S
=
DX e
−
Z ∂ω
Z
∂ĝαβ ∂ω
Z
1
1 p α
−S
=
DX e
−
ĝ T α
Z
2π
1
c p
=
ĝ R̂ −
µe2ω
24π
2πα0
c √
1
=
g(R − 2∇2 ω) −
µe2ω
24π
2πα0
where, in the last two lines, we used the Weyl anomaly (4.34) and the relationship between Ricci curvatures (1.29). The central charge appearing in these formulae includes
the contribution from the ghosts,
c = D − 26
– 119 –
We can now just treat this as a differential equation for the partition function Z and
solve. This allows us to express the partition function Z[ĝ], defined on one worldsheet
metric, in terms of Z[g], defined on another. The relationship is,
Z
1
cα0 αβ
β
2 √
2ω
g ∂α ω ∂ ω + Rω
Z[ĝ] = Z[g] exp −
d σ g 2µe −
4πα0
6
We see that the scaling mode ω inherits a kinetic term. It now appears as a new
dynamical scalar field in the theory. It is often called the Liouville field on account of
the exponential potential term multiplying µ. Solving this theory is quite hard6 . Notice
also that our new scalar field ω appears in the final term multiplying the Ricci scalar
R. We will describe the significance of this in Section 7.2.1. We’ll also see another
derivation of this kind of Lagrangian in Section 7.4.4.
5.4 States and Vertex Operators
In Chapter 2 we determined the spectrum of the string in flat space. What is the
spectrum for a general string background? The theory consists of the b and c ghosts,
together with a c = 26 CFT. At first glance, it seems that we have a greatly enlarged
Hilbert space since we can act with creation operators from all fields, including the
ghosts. However, as you might expect, not all of these states will be physical. After
correctly accounting for the gauge symmetry, only some subset survives.
The elegant method to determine the physical Hilbert space in a gauge fixed action
with ghosts is known as BRST quantization. You will learn about it in the “Advanced
Quantum Field Theory” course where you will apply it to Yang-Mills theory. Although
a correct construction of the string spectrum employs the BRST method, we won’t
describe it here for lack of time. A very clear description of the general method and its
application to the string can be found in Section 4.2 of Polchinski’s book.
Instead, we will make do with a poor man’s attempt to determine the spectrum of the
string. Our strategy is to simply pretend that the ghosts aren’t there and focus on the
states created by the fields of the matter CFT (i.e. the X µ fields if we’re talking about
flat space). As we’ll explain in the next section, if we’re only interested in tree-level
scattering amplitudes then this will suffice.
To illustrate how to compute the spectrum of the string, let’s go back to flat D = 26
dimensional Minkowski space and the discussion of covariant quantization in Section
6
A good review can be found Seiberg’s article “Notes on Quantum Liouville Theory and Quantum
Gravity”, Prog. Theor. Phys. Supl. 102 (1990) 319.
– 120 –
2.1. We found that physical states |Ψi are subject to the Virasoro constraints (2.6)
and (2.7) which read
Ln |Ψi = 0
for n > 0
L0 |Ψi = a |Ψi
and similar for L̃n ,
L̃n |Ψi = 0
for n > 0
L̃0 |Ψi = ã |Ψi
where we have, just briefly, allowed for the possibility of different normal ordering
coefficients a and ã for the left- and right-moving sectors. But there’s a name for states
in a conformal field theory obeying these requirements: they are primary states of
weight (a, ã).
So how do we fix the normal ordering ambiguities a and ã? A simple way is to
first replace the states with operator insertions on the worldsheet using the stateoperator map: |Ψi → O. But we have a further requirement on the operators O:
gauge invariance. There are two gauge symmetries: reparameterization invariance and
Weyl symmetry. Both restrict the possible states.
Let’s start by considering reparameterization invariance. In the last section, we happily placed operators at specific points on the worldsheet. But in a theory with a
dynamical metric, this doesn’t give rise to a diffeomorphism invariant operator. To
make an object that is invariant under reparameterizations of the worldsheet coordinates, we should integrate over the whole worldsheet. Our operator insertions (in
conformal gauge) are therefore of the form,
Z
V ∼ d2 z O
(5.6)
Here the ∼ sign reflects the fact that we’ve dropped an overall normalization constant
which we’ll return to in the next section.
Integrating over the worldsheet takes care of diffeomorphisms. But what about Weyl
symmetries? The measure d2 z has weight (−1, −1) under rescaling. To compensate,
the operator O must have weight (+1, +1). This is how we fix the normal ordering
ambiguity: we require a = ã = 1. Note that this agrees with the normal ordering
coefficient a = 1 that we derived in lightcone quantization in Chapter 2.
– 121 –
This, then, is the rather rough derivation of the string spectrum. The physical states
are the primary states of the CFT with weight (+1, +1). The operators (5.6) associated
to these states are called vertex operators.
5.4.1 An Example: Closed Strings in Flat Space
Let’s use this new language to rederive the spectrum of the closed string in flat space.
We start with the ground state of the string, which was previously identified as a
tachyon. As we saw in Section 4, the vacuum of a CFT is associated to the identity
operator. But we also have the zero modes. We can give the string momentum pµ by
acting with the operator eip·X . The vertex operator associated to the ground state of
the string is therefore
Z
Vtachyon ∼ d2 z : eip·X :
(5.7)
In Section 4.3.3, we showed that the operator eip·X is primary with weight h = h̃ =
α0 p2 /4. But Weyl invariance requires that the operator has weight (+1, +1). This is
only true if the mass of the state is
M 2 ≡ −p2 = −
4
α0
This is precisely the mass of the tachyon that we saw in Section 2.
Let’s now look at the first excited states. In covariant quantization, these are of
µ
ν
the form ζµν α−1
α̃−1
|0; pi, where ζµν is a constant tensor that determines the type
of state, together with its polarization. (Recall: traceless symmetric ζµν corresponds
to the graviton, anti-symmetric ζµν corresponds to the Bµν field and the trace of ζµν
corresponds to the scalar known as the dilaton). From (4.51), the vertex operator
associated to this state is,
Z
¯ ν : ζµν
Vexcited ∼ d2 z : eip·X ∂X µ ∂X
(5.8)
µ
¯ µ gives a α̃µ excitation. It’s easy to
where ∂X µ gives us a α−1
excitation, while ∂X
−1
0 2
check that the weight of this operator is h = h̃ = 1 + α p /4. Weyl invariance therefore
requires that
p2 = 0
confirming that the first excited states of the string are indeed massless. However, we
still need to check that the operator in (5.8) is actually primary. We know that ∂X is
– 122 –
primary and we know that eip·X is primary, but now we want to consider them both
sitting together inside the normal ordering. This means that there are extra terms in
the Wick contraction which give rise to 1/(z−w)3 terms in the OPE, potentially ruining
the primacy of our operator. One such term arises from a double contraction, one of
which includes the eip·X operator. This gives rise to an offending term proportional to
pµ ζµν . The same kind of contraction with T̄ gives rise to a term proportional to pν ζνµ .
In order for these terms to vanish, the polarization tensor must satisfy
pµ ζµν = pν ζµν = 0
which is precisely the transverse polarization condition expected for a massless particle.
5.4.2 An Example: Open Strings in Flat Space
As explained in Section 4.7, vertex operators for the open-string are inserted on the
boundary ∂M of the worldsheet. We still need to ensure that these operators are diffeomorphism invariant which is achieved by integrating over ∂M. The vertex operator
for the open string tachyon is
Z
Vtachyon ∼
ds : eip·X :
∂M
We need to figure out the dimension of the boundary operator : eip·X :. It’s not the
same as for the closed string. The reason is due to presence of the image charge in
the propagator (4.52) for a free scalar field on a space with boundary. This propagator
appears in the Wick contractions in the OPEs and affects the weights. Let’s see why
this is the case. Firstly, we look at a single scalar field X,
∞
X
α0 1
α0 1
(ip)n
n−1
ipX(w,w̄)
: X(w, w̄)
: −
−
+ ...
∂X(z) : e
:=
(n
−
1)!
2
z
−
w
2
z
−
w̄
n=1
iα0 p
1
1
ipX(w,w̄)
=−
:e
:
+
+ ...
2
z − w z − w̄
With this result, we can now compute the OPE with T ,
2
α 0 p2
1
1
ipX(w,w̄)
ipX
:e
:
+
+ ...
T (z) : e
:=
4
z − w z − w̄
When the operator : eipX(w,w̄) : is placed on the boundary w = w̄, this becomes
T (z) : eipX(w,w̄) :=
α0 p2 : eipX(w,w̄) :
+ ...
(z − w)2
This tells us that the boundary operator : eip·X : is indeed primary, with weight α0 p2 .
– 123 –
For the open string, Weyl invariance requires that operators have weight +1 in order
R
to cancel the scaling dimension of −1 coming from the boundary integral ds. So the
mass of the open string ground state is
M 2 ≡ −p2 = −
1
α0
in agreement with the mass of the open string tachyon computed in Section 3.
The vertex operator for the photon is
Z
Vphoton ∼
ds ζa : ∂X a eip·X :
(5.9)
∂M
where the index a = 0, . . . , p now runs only over those directions with Neumann boundary conditions that lie parallel to the brane worldvolume. The requirement that this is
a primary operator gives pa ζa = 0, while Weyl invariance tells us that p2 = 0. This is
the expected behaviour for the momentum and polarization of a photon.
5.4.3 More General CFTs
Let’s now consider a string propagating in four-dimensional Minkowski space M4 ,
together with some internal CFT with c = 22. Then any primary operator of the
internal CFT with weight (h, h) can be assigned momentum pµ , for µ = 0, 1, 2, 3 by
dressing the operator with eip·X . In order to get a primary operator of weight (+1, +1)
as required, we must have
α 0 p2
=1−h
4
We see that the mass spectrum of closed string states is given by
M2 =
4
(h − 1)
α0
where h runs over the spectrum of primary operators of the internal CFT. Some comments:
• Relevant operators in the internal CFT have h < 1 and give rise to tachyons
in the spectrum. Marginal operators, with h = 1, give massless particles. And
irrelevant operators result in massive states.
• Notice that requiring the vertex operators to be Weyl invariant determines the
mass formula for the state. We say that the vertex operators are “on-shell”, in
the same sense that external legs of Feynman diagrams are on-shell. We will have
more to say about this in the next section.
– 124 –
6. String Interactions
So far, despite considerable effort, we’ve only discussed the free string. We now wish to
consider interactions. If we take the analogy with quantum field theory as our guide,
then we might be led to think that interactions require us to add various non-linear
terms to the action. However, this isn’t the case. Any attempt to add extra non-linear
terms for the string won’t be consistent with our precious gauge symmetries. Instead,
rather remarkably, all the information about interacting strings is already contained in
the free theory described by the Polyakov action. (Actually, this statement is almost
true).
To see that this is at least feasible, try to draw a cartoon
picture of two strings interacting. It looks something like the
worldsheet shown in the figure. The worldsheet is smooth.
In Feynman diagrams in quantum field theory, information
about interactions is inserted at vertices, where different lines
meet. Here there are no such points. Locally, every part
of the diagram looks like a free propagating string. Only
globally do we see that the diagram describes interactions.
6.1 What to Compute?
Figure 31:
If the information about string interactions is already contained in the Polyakov action,
let’s go ahead and compute something! But what should we compute? One obvious
thing to try is the probability for a particular configuration of strings at an early time
to evolve into a new configuration at some later time. For example, we could try to
compute the amplitude associated to the diagram above, stipulating fixed curves for
the string ends.
No one knows how to do this. Moreover, there are words that we can drape around
this failure that suggests this isn’t really a sensible thing to compute. I’ll now try to
explain these words. Let’s start by returning to the familiar framework of quantum
field theory in a fixed background. There the basic objects that we can compute are
correlation functions,
hφ(x1 ) . . . φ(xn )i
(6.1)
After a Fourier transform, these describe Feynman diagrams in which the external legs
carry arbitrary momenta. For this reason, they are referred to as off-shell. To get the
scattering amplitudes, we simply need to put the external legs on-shell (and perform a
few other little tricks captured in the LSZ reduction formula).
– 125 –
The discussion above needs amendment if we turn on gravity. Gravity is a gauge
theory and the gauge symmetries are diffeomorphisms. In a gauge theory, only gauge
invariant observables make sense. But the correlation function (6.1) is not gauge invariant because its value changes under a diffeomorphism which maps the points xi to
another point. This emphasizes an important fact: there are no local off-shell gauge
invariant observables in a theory of gravity.
There is another way to say this. We know, by causality, that space-like separated
operators should commute in a quantum field theory. But in gravity the question of
whether operators are space-like separated becomes a dynamical issue and the causal
structure can fluctuate due to quantum effects. This provides another reason why we
are unable to define local gauge invariant observables in any theory of quantum gravity.
Let’s now return to string theory. Computing the evolution of string configurations
for a finite time is analogous to computing off-shell correlation functions in QFT. But
string theory is a theory of gravity so such things probably don’t make sense. For
this reason, we retreat from attempting to compute correlation functions, back to the
S-matrix.
The String S-Matrix
The object that we can compute in string theory is the
S-matrix. This is obtained by taking the points in the correlation function to infinity: xi → ∞. This is acceptable
because, just like in the case of QED, the redundancy of
the system consists of those gauge transformations which
die off asymptotically. Said another way, points on the
boundary don’t fluctuate in quantum gravity. (Such fluctuations would be over an infinite volume of space and are
suppressed due to their infinite action).
Figure 32:
So what we’re really going to calculate is a diagram of
the type shown in the figure, where all external legs are
taken to infinity. Each of these legs can be placed in a different state of the free string
and assigned some spacetime momentum pi . The resulting expression is the string
S-matrix.
Using the state-operator map, we know that each of these states at infinity is equivalent to the insertion of an appropriate vertex operator on the worldsheet. Therefore,
to compute this S-matrix element we use a conformal transformation to bring each of
these infinite legs to a finite distance. The end result is a worldsheet with the topology
– 126 –
of the sphere, dotted with vertex operators where the legs used to be.
However, we already saw in the previous section that the constraint
of Weyl invariance meant that vertex operators are necessarily onshell. Technically, this is the reason that we can only compute onshell correlation functions in string theory.
6.1.1 Summing Over Topologies
Figure 33:
The Polyakov path integral instructs us to sum over all metrics. But
what about worldsheets of different topologies? In fact, we should also sum over these.
It is this sum that gives the perturbative expansion of string theory. The scattering of
two strings receives contributions from worldsheets of the form
+
+
+
(6.2)
The only thing that we need to know is how to weight these different worldsheets.
Thankfully, there is a very natural coupling on the string that we have yet to consider
and this will do the job. We augment the Polyakov action by
Sstring = SPoly + λχ
(6.3)
Here λ is simply a real number, while χ is given by an integral over the (Euclidean)
worldsheet
Z
1
√
χ=
d2 σ gR
(6.4)
4π
where R is the Ricci scalar of the worldsheet metric. This looks like the EinsteinHilbert term for gravity on the worldsheet. It is simple to check that it is invariant
under reparameterizations and Weyl transformations.
In four-dimensions, the Einstein-Hilbert term makes gravity dynamical. But life is
very different in 2d. Indeed, we’ve already seen that all the components of the metric
can be gauged away so there are no propagating degrees of freedom associated to
gαβ . So, in two-dimensions, the term (6.4) doesn’t make gravity dynamical: in fact,
classically, it doesn’t do anything at all!
– 127 –
The reason for this is that χ is a topological invariant. This means that it doesn’t
actually depend on the metric gαβ at all – it depends only on the topology of the
worldsheet. (More precisely, χ only depends on those global properties of the metric
which themselves depend on the topology of the worldsheet). This is the content of
the Gauss-Bonnet theorem: the integral of the Ricci scalar R over the worldsheet
gives an integer, χ, known as the Euler number of the worldsheet. For a worldsheet
without boundary (i.e. for the closed string) χ counts the number of handles h on the
worldsheet. It is given by,
χ = 2 − 2h = 2(1 − g)
(6.5)
where g is called the genus of the surface. The simplest examples are shown in the
figure. The sphere has g = 0 and χ = 2; the torus has g = 1 and χ = 0. For higher
g > 1, the Euler character χ is negative.
Figure 34: Examples of increasingly poorly drawn Riemann surfaces with χ = 2, 0 and −2.
Now we see that the number λ — or, more precisely, eλ — plays the role of the string
coupling. The integral over worldsheets is weighted by,
Z
X
X
−2λ(1−g)
−Sstring
∼
e
DXDg e−SPoly
e
topologies
metrics
topologies
For eλ 1, we have a good perturbative expansion in which we sum over all topologies.
(In fact, it is an asymptotic expansion, just as in quantum field theory). It is standard
to define the string coupling constant as
gs = e λ
After a conformal map, tree-level scattering corresponds to a worldsheet with the topology of a sphere: the amplitudes are proportional to 1/gs2 . One-loop scattering corresponds to toroidal worldsheets and, with our normalization, have no power of gs . (Although, obviously, these are suppressed by gs2 relative to tree-level processes). The end
– 128 –
result is that the sum over worldsheets in (6.2) becomes a sum over Riemann surfaces
of increasing genus, with vertex operators inserted for the initial and final states,
+
+
+
The Riemann surface of genus g is weighted by
(gs2 )g−1
While it may look like we’ve introduced a new parameter gs into the theory and added
the coupling (6.3) by hand, we will later see why this coupling is a necessary part of
the theory and provide an interpretation for gs .
Scattering Amplitudes
We now have all the information that we need to explain how to compute string scattering amplitudes. Suppose that we want to compute the S-matrix for m states: we
will label them as Λi and assign them spacetime momenta pi . Each has a corresponding vertex operator VΛi (pi ). The S-matrix element is then computed by evaluating
the correlation function in the 2d conformal field theory, with insertions of the vertex
operators.
Z
m
X
Y
−χ 1
(m)
gs
DXDg e−SPoly
A (Λi , pi ) =
VΛi (pi )
Vol
i=1
topologies
This is a rather peculiar equation. We are interpreting the correlation functions of a
two-dimensional theory as the S-matrix for a theory in D = 26 dimensions!
To properly compute the correlation function, we should introduce the b and c ghosts
that we saw in the last chapter and treat them carefully. However, if we’re only interested in tree-level amplitudes, then we can proceed naively and ignore the ghosts. The
reason can be seen in the ghost action (5.4) where we see that the ghosts couple only to
the worldsheet metric, not to the other worldsheet fields. This means that if our gauge
fixing procedure fixes the worldsheet metric completely — which it does for worldsheets
with the topology of a sphere — then we can forget about the ghosts. (At least, we
can forget about them as soon as we’ve made sure that the Weyl anomaly cancels).
However, as we’ll explain in 6.4, for higher genus worldsheets, the gauge fixing does
not fix the metric completely and there are residual dynamical modes of the metric,
known as moduli, which couple the ghosts and matter fields. This is analogous to the
statement in field theory that we only need to worry about ghosts running in loops.
– 129 –
6.2 Closed String Amplitudes at Tree Level
The tree-level scattering amplitude is given by the correlation function of the 2d theory,
evaluated on the sphere,
(m)
A
1 1
= 2
gs Vol
Z
DXDg e
−SPoly
m
Y
VΛi (pi )
i=1
where VΛi (pi ) are the vertex operators associated to the states.
We want to integrate over all metrics on the sphere.
At first glance that sounds rather daunting but, of
course, we have the gauge symmetries of diffeomorphisms and Weyl transformations at our disposal. Any metric on the sphere is conformally
equivalent to the flat metric on the plane. For example, the round metric on the sphere of radius R
can be written as
ds2 =
Figure 35:
4R2
dzdz̄
(1 + |z|2 )2
which is manifestly conformally equivalent to the plane, supplemented by the point at
infinity. The conformal map from the sphere to the plane is the stereographic projection
depicted in the diagram. The south pole of the sphere is mapped to the origin; the
north pole is mapped to the point at infinity. Therefore, instead of integrating over all
metrics, we may gauge fix diffeomorphisms and Weyl transformations to leave ourselves
with the seemingly easier task of computing correlation functions on the plane.
6.2.1 Remnant Gauge Symmetry: SL(2,C)
There’s a subtlety. And it’s a subtlety that we’ve seen before: there is a residual
gauge symmetry. It is the conformal group, arising from diffeomorphisms which can be
undone by Weyl transformations. As we saw in Section 4, there are an infinite number
of such conformal transformations. It looks like we have a whole lot of gauge fixing
still to do.
However, global issues actually mean that there’s less remnant gauge symmetry than
you might think. In Section 4, we only looked at infinitesimal conformal transformations, generated by the Virasoro operators Ln , n ∈ Z. We did not examine whether
these transformations are well-defined and invertible over all of space. Let’s take a
– 130 –
look at this. Recall that the coordinate changes associated to Ln are generated by the
vector fields (4.44),
ln = z n+1 ∂z
which result in the shift δz = z n+1 . This is non-singular at z = 0 only for n ≥ −1. If
we restrict to smooth maps, that gets rid of half the transformations right away. But,
since we’re ultimately interested in the sphere, we now also need to worry about the
point at z = ∞ which, in stereographic projection, is just the north pole of the sphere.
To do this, it’s useful to work with the coordinate
u=
1
z
The generators of coordinate transformations for the u coordinate are
ln = z n+1 ∂z =
1 ∂u
∂u = −u1−n ∂u
∂z
un+1
which is non-singular at u = 0 only for n ≤ 1.
Combining these two results, the only generators of the conformal group that are
non-singular over the whole Riemann sphere are l−1 , l0 and l1 which act infinitesimally
as
l−1 : z → z + l0 : z → (1 + )z
l1 : z → (1 + z)z
The global version of these transformations is
l−1 : z → z + α
l0 : z → λz
l1 : z →
z
1 − βz
which can be combined to give the general transformation
z →
az + b
cz + d
(6.6)
with a, b, c and d ∈ C. We have four complex parameters, but we’ve only got three
transformations. What happened? Well, one transformation is fake because an overall
– 131 –
scaling of the parameters doesn’t change z. By such a rescaling, we can always insist
that the parameters obey
ad − bc = 1
The transformations (6.6) subject to this constraint have the group structure SL(2; C),
which is the group of 2 × 2 complex matrices with unit determinant. In fact, since the
transformation is blind to a flip in sign of all the parameters, the actual group of global
conformal transformations is SL(2; C)/Z2 , which is sometimes written as P SL(2; C).
(This Z2 subtlety won’t be important for us in what follows).
The remnant global transformations on the sphere are known as conformal Killing
vectors and the group SL(2; C)/Z2 is the conformal Killing group. This group allows
us to take any three points on the plane and move them to three other points of our
choosing. We will shortly make use of this fact to gauge fix, but for now we leave the
SL(2; C) symmetry intact.
6.2.2 The Virasoro-Shapiro Amplitude
We will now compute the S-matrix for closed string tachyons. You might think that
this is the least interesting thing to compute: after all, we’re ultimately interested
in the superstring which doesn’t have tachyons. This is true, but it turns out that
tachyon scattering is much simpler than everything else, mainly because we don’t have
a plethora of extra indices on the states to worry about. Moreover, the lessons that we
will learn from tachyon scattering hold for the scattering of other states as well.
The m-point tachyon scattering amplitude is given by the flat space correlation function
Z
m
Y
1
1
(m)
A (p1 , . . . , pm ) = 2
DX e−SPoly
V (pi )
gs Vol(SL(2; C))
i=1
where the tachyon vertex operator is given by,
Z
Z
2
ipi ·X
V (pi ) = gs d z e
≡ gs d2 z V̂ (z, pi )
(6.7)
Note that, in contrast to (5.7), we’ve added an appropriate normalization factor to the
vertex operator. Heuristically, this reflects the fact that the operator is associated to
the addition of a closed string mode. A rigorous derivation of this normalization can
be found in Polchinski.
– 132 –
The amplitude can therefore be written as,
(m)
A
gsm−2
(p1 , . . . , pm ) =
Vol(SL(2; C))
Z Y
m
d2 zi hV̂ (z1 , p1 ) . . . V̂ (zm , pm )i
i=1
where the expectation value h. . .i is computed using the gauge fixed Polyakov action.
But the gauge fixed Polyakov action is simply a free theory and our correlation function
is something eminently computable: a Gaussian integral,
Z
hV̂ (z1 , p1 ) . . . V̂ (zm , pm )i =
!
Z
m
X
1
¯
DX exp −
d2 z ∂X · ∂X
exp i
pi · X(zi , z̄i )
2πα0
i=1
The normalization in front of the Polyakov action is now 1/2πα0 instead of 1/4πα0
because we’re working with complex coordinates and we need to remember that ∂α ∂ α =
4∂ ∂¯ and d2 z = 2d2 σ.
The Gaussian Integral
We certainly know how to compute Gaussian integrals. Let’s go slow. Consider the
following general integral,
0Z
Z
Z
πα
1
1
2
2
0
2
0
0
¯ + iJ · X ∼ exp
X · ∂ ∂X
d zd z J(z, z̄) ¯ J(z , z̄ )
dz
DX exp
2πα0
2
∂∂
Here the ∼ symbol reflects the fact that we’ve dropped a whole lot of irrelevant normal¯ The inverse operator 1/∂ ∂¯ on the right-handization terms, including det−1/2 (−∂ ∂).
side of this equation is shorthand for the propagator G(z, z 0 ) which solves
¯
∂ ∂G(z,
z̄; z 0 , z̄ 0 ) = δ(z − z 0 , z̄ − z̄ 0 )
As we’ve seen several times before, in two dimensions this propagator is given by
G(z, z̄; z 0 , z̄ 0 ) =
1
ln |z − z 0 |2
2π
Back to the Scattering Amplitude
Comparing our scattering amplitude with this general expression, we need to take the
source J to be
J(z, z̄) =
m
X
pi δ(z − zi , z̄ − z̄i )
i=1
– 133 –
Inserting this into the Gaussian integral gives us an expression for the amplitude
!
Z Y
m
α0 X
gsm−2
2
(m)
d zi exp
A ∼
pj · pl ln |zj − zl |
Vol(SL(2; C)) i=1
2 j,l
The terms with j = l seem to be problematic. In fact, they should just be left out.
This follows from correctly implementing normal ordering and leaves us with
Z Y
m
Y
gsm−2
0
(m)
A ∼
d2 zi
(6.8)
|zj − zl | α pj ·pl
Vol(SL(2; C)) i=1
j<l
Actually, there’s something that we missed. (Isn’t there always!). We certainly expect scattering in flat space to obey momentum conservation, so there should be a
P
δ (26) ( m
i=1 pi ) in the amplitude. But where is it? We missed it because we were a little
too quick in computing the Gaussian integral. The operator ∂ ∂¯ annihilates the zero
¯ is not well-defined.
mode, xµ , in the mode expansion. This means that its inverse, 1/∂ ∂,
But it’s easy to deal with this by treating the zero mode separately. The derivatives
∂ 2 don’t see xµ , but the source J does. Integrating over the zero mode in the path
integral gives us our delta function
Z
m
m
X
X
dx exp(i
pi · x) ∼ δ 26 (
pi )
i=1
i=1
So, our final result for the amplitude is
(m)
A
Z Y
m
X
Y
gsm−2
0
26
∼
δ (
pi )
d2 zi
|zj − zl | α pj ·pl
Vol(SL(2; C))
i
i=1
j<l
(6.9)
The Four-Point Amplitude
We will compute only the four-point amplitude for two-to-two scattering of tachyons.
The Vol(SL(2; C)) factor is there to remind us that we still have a remnant gauge
symmetry floating around. Let’s now fix this. As we mentioned before, it provides
enough freedom for us to take any three points on the plane and move them to any
other three points. We will make use of this to set
z1 = ∞ ,
z2 = 0 ,
z3 = z
,
z4 = 1
Inserting this into the amplitude (6.9), we find ourselves with just a single integral to
evaluate,
Z
X
0
0
(4)
2 26
A ∼ gs δ (
pi )
d2 z |z| α p2 ·p3 |1 − z| α p3 ·p4
(6.10)
i
– 134 –
(There is also an overall factor of |z1 |4 , but this just gets absorbed into an overall
normalization constant). We still need to do the integral. It can be evaluated exactly
in terms of gamma functions. We relegate the proof to Appendix 6.5, where we show
that
Z
2πΓ(a)Γ(b)Γ(c)
(6.11)
d2 z |z|2a−2 |1 − z|2b−2 =
Γ(1 − a)Γ(1 − b)Γ(1 − c)
where a + b + c = 1.
Four-point scattering amplitudes are typically expressed in
terms of Mandelstam variables. We choose p1 and p2 to be
incoming momenta and p3 and p4 to be outgoing momenta,
as shown in the figure. We then define
s = −(p1 + p2 )2
,
t = −(p1 + p3 )2
,
u = −(p1 + p4 )2
p3
θ
p1
p2
p4
Figure 36:
These obey
s+t+u=−
X
p2i =
X
i
i
Mi2 = −
16
α0
where, in the last equality, we’ve inserted the value of the tachyon mass (2.27). Writing
the scattering amplitude (6.10) in terms of Mandelstam variables, we have our final
answer
X
Γ(−1 − α0 s/4)Γ(−1 − α0 t/4)Γ(−1 − α0 u/4)
(6.12)
A(4) ∼ gs2 δ 26 (
pi )
0 s/4)Γ(2 + α0 t/4)Γ(2 + α0 u/4)
Γ(2
+
α
i
This is the Virasoro-Shapiro amplitude governing tachyon scattering in the closed
bosonic string.
Remarkably, the Virasoro-Shapiro amplitude was almost the first equation of string
theory! (That honour actually goes to the Veneziano amplitude which is the analogous expression for open string tachyons and will be derived in Section 6.3.1). These
amplitudes were written down long before people knew that they had anything to do
with strings: they simply exhibited some interesting and surprising properties. It took
several years of work to realise that they actually describe the scattering of strings.
We will now start to tease apart the Virasoro-Shapiro amplitude to see some of the
properties that got people hooked many years ago.
6.2.3 Lessons to Learn
So what’s the physics lying behind the scattering amplitude (6.12)? Obviously it is
symmetric in s, t and u. That is already surprising and we’ll return to it shortly. But
we’ll start by fixing t and looking at the properties of the amplitude as we vary s.
– 135 –
The first thing to notice is that A(4) has poles. Lots of poles. They come from the
factor of Γ(−1 − α0 s/4) in the numerator. The first of these poles appears when
−1 −
α0 s
=0
4
⇒
s=−
4
α0
But that’s the mass of the tachyon! It means that, for s close to −4/α0 , the amplitude
has the form of a familiar scattering amplitude in quantum field theory with a cubic
vertex,
∼
1
s − M2
where M is the mass of the exchanged particle, in this case the tachyon.
Other poles in the amplitude occur at s = 4(n − 1)/α0 with n ∈ Z+ . This is precisely
the mass formula for the higher states of the closed string. What we’re learning is
that the string amplitude is summing up an infinite number of tree-level field theory
diagrams,
Mn
=
n
where the exchanged particles are all the different states of the free string.
In fact, there’s more information about the spectrum of states hidden within these
amplitudes. We can look at the residues of the poles at s = 4(n − 1)/α0 , for n =
0, 1, . . .. These residues are rather complicated functions of t, but the highest power of
momentum that appears for each pole is
(4)
A
∼
∞
X
n=0
t2n
s − Mn2
(6.13)
The power of the momentum is telling us the highest spin of the particle states at level
n. To see why this is, consider a field corresponding to a spin J particle. It has a whole
bunch of Lorentz indices, χµ1 ...µJ . In a cubic interaction, each of these must be soaked
up by derivatives. So we have J derivatives at each vertex, contributing powers of
(momentum)2J to the numerator of the Feynman diagram. Comparing with the string
scattering amplitude, we see that the highest spin particle at level n has J = 2n. This
is indeed the result that we saw from the canonical quantization of the string in Section
2.
– 136 –
Finally, the amplitude (6.12) has a property that is very different from amplitudes
in field theory. Above, we framed our discussion by keeping t fixed and expanding in
s. We could just have well done the opposite: fix s and look at poles in t. Now the
string amplitude has the interpretation of an infinite number of t-channel scattering
amplitudes, one for each state of the string
=
Mn
n
Usually in field theory, we sum up both s-channel and t-channel scattering amplitudes.
Not so in string theory. The sum over an infinite number of s-channel amplitudes can
be reinterpreted as an infinite sum of t-channel amplitudes. We don’t include both:
that would be overcounting. (Similar statements hold for u). The fact that the same
amplitude can be written as a sum over s-channel poles or a sum over t-channel poles is
sometimes referred to as “duality”. (A much overused word). In the early days, before
it was known that string theory was a theory of strings, the subject inherited its name
from this duality property of amplitudes: it was called the dual resonance model.
High Energy Scattering
Let’s use this amplitude to see what happens when we collide strings at high energies.
There are different regimes that we could look at. The most illuminating is s, t →
∞, with s/t held fixed. In this limit, all the exchanged momenta become large. It
corresponds to high-energy scattering with the angle θ between incoming and outgoing
particles kept fixed. To see this consider, for example, massless particles (our amplitude
is really for tachyons, but the same considerations hold). We take the incoming and
outgoing momenta to be
√
s
p1 =
(1, 1, 0, . . .)
2
√
s
(1, cos θ, sin θ, . . .)
p3 =
2
√
,
,
s
(1, −1, 0, . . .)
2
√
s
p4 =
(1, − cos θ, − sin θ, . . .)
2
p2 =
Then we see explicitly that s → ∞ and t → ∞ with the ratio s/t fixed also keeps the
scattering angle θ fixed.
– 137 –
We can evaluate the scattering amplitude A(4) in this limit by using Γ(x) ∼ exp(x ln x).
We send s → ∞ avoiding the poles. (We can achieve this by sending s → ∞ in a slightly
imaginary direction. Ultimately this is valid because all the higher string states are actually unstable in the interacting theory which will shift their poles off the real axis once
taken into account). It is simple to check that the amplitude drops off exponentially
quickly at high energies,
X
α0
(4)
2 26
A ∼ gs δ (
pi ) exp − (s ln s + t ln t + u ln u)
as s → ∞
(6.14)
2
i
The exponential fall-off seen in (6.14) is much faster than the amplitude of any field
theory which, at best, fall off with power-law decay at high energies and, at worse,
diverge. For example, consider the individual terms (6.13) corresponding to the amplitude for s-channel processes involving the exchange of particles with spin 2n. We
see that the exchange of a spin 2 particle results in a divergence in this limit. This
is reflecting something you already know about gravity: the dimensionless coupling is
GN E 2 (in four-dimensions) which becomes large for large energies. The exchange of
higher spin particles gives rise to even worse divergences. If we were to truncate the
infinite sum (6.13) at any finite n, the whole thing would diverge. But infinite sums
can do things that finite sums can’t and the final behaviour of the amplitude (6.14)
is much softer than any of the individual terms. The infinite number of particles in
string theory conspire to render finite any divergence arising from an individual particle
species.
Phrased in terms of the s-channel exchange of particles, the high-energy behaviour
of string theory seems somewhat miraculous. But there is another viewpoint where it’s
all very obvious. The power-law behaviour of scattering amplitudes is characteristic of
point-like charges. But, of course, the string
isn’t a point-like object. It is extended and
√
0
fuzzy at length scales comparable to α . This is the reason the amplitude has such
soft high-energy behaviour. Indeed, this idea that smooth extended objects give rise
to scattering amplitudes that decay exponentially at high energies is something that
you’ve seen before in non-relativistic quantum mechanics. Consider, for example, the
scattering of a particle off a Gaussian potential. In the Born approximation, the differential cross-section is just given by the Fourier transform which is again a Gaussian,
now decaying exponentially for large momentum.
It’s often said that theories of quantum gravity should have a “minimum length”,
sometimes taken to be the Planck scale. This is roughly true in string theory, although
not in any crude simple manner. Rather, the minimum length reveals itself in different
– 138 –
ways depending on which question is being asked. The above discussion√highlights
one example of this: strings can’t probe distance scales shorter than ls = α0 simply
because they are themselves fuzzy at this scale. It turns out that D-branes are much
better probes of sub-stringy physics and provide a different view on the short distance
structure of spacetime. We will also see another manifestation of the minimal length
scale of string theory in Section 8.3.
Graviton Scattering
Although we’ve derived the result (6.14) for tachyons, all tree-level amplitudes have this
soft fall-off at high-energies. Most notably, this includes graviton scattering. As we
noted above, this is in sharp contrast to general relativity for which tree-level scattering
amplitudes diverge at high-energies. This is the first place to see that UV problems of
general relativity might have a good chance of being cured in string theory.
Using the techniques described in this section, one can compute m-point tree-level
amplitudes for graviton
scattering. If we restrict attention to low-energies (i.e. much
√
0
smaller than 1/ α ), one can show that these coincide with the amplitudes derived
from the Einstein-Hilbert action in D = 26 dimensions
Z
√
1
S = 2 d26 X −G R
2κ
where R is the D = 26 Ricci scalar (not to be confused with the worldsheet Ricci scalar
which we call R). The gravitational coupling, κ2 is related to Newton’s constant in
26 dimensions. It plays no role for pure gravity, but is important when we couple to
matter. We’ll see shortly that it’s given by
κ2 ≈ gs2 (α0 )12
We won’t explicitly compute graviton scattering amplitudes in this course, partly because they’re fairly messy and partly because building up the Einstein-Hilbert action
from m-particle scattering is hardly the best way to look at general relativity. Instead,
we shall derive the Einstein-Hilbert action in a much better fashion in Section 7.
6.3 Open String Scattering
So far our discussion has been entirely about closed strings. There is a very similar
story for open strings. We again compute S-matrix elements. Conformal symmetry now
maps tree-level scattering to the disc, with vertex operators inserted on the boundary
of the disc.
– 139 –
Figure 37: The conformal map from the open string worldsheet to the disc.
For the open string, the string coupling constant that we add to the Polyakov action
requires the addition of a boundary term to make it well defined,
Z
Z
1
1
2 √
χ=
d σ gR +
ds k
(6.15)
4π M
2π ∂M
where k is the geodesic curvature of the boundary. To define it, we introduce two unit
vectors on the worldsheet: tα is tangential to the boundary, while nα is normal and
points outward from the boundary. The geodesic curvature is defined as
k = −tα nβ ∇α tβ
Boundary terms of the type seen in (6.15) are also needed in general relativity for
manifolds with boundaries: in that context, they are referred to as Gibbons-Hawking
terms.
The Gauss-Bonnet theorem has an extension to surfaces with boundary. For surfaces
with h handles and b boundaries, the Euler character is given by
χ = 2 − 2h − b
Some examples are shown in Figure 38. The expansion for open-string scattering
consists of adding consecutive boundaries to the worldsheet. The disc is weighted by
1/gs ; the annulus has no factor of gs and so on. We see that the open string coupling
is related to the closed string coupling by
2
gopen
= gs
(6.16)
One of the key steps in computing closed string scattering amplitudes was the implementation of the conformal Killing group, which was defined as the surviving gauge
– 140 –
Figure 38: Riemann surfaces with boundary with χ = 1, 0 and −1.
symmetry with a global action on the sphere. For the open string, there is again a
residual gauge symmetry. If we think in terms of the upper-half plane, the boundary
is Imz = 0. The conformal Killing group is composed of transformations
z→
az + b
cz + d
again with the requirement that ad − bc = 1. This time there is one further condition:
the boundary Imz = 0 must be mapped onto itself. This requires a, b, c, d ∈ R. The
resulting conformal Killing group is SL(2; R)/Z2 .
6.3.1 The Veneziano Amplitude
Since vertex operators now live on the boundary, they have a fixed ordering. In computing a scattering amplitude, we must sum over all orderings. Let’s look again at the
4-point amplitude for tachyon scattering. The vertex operator is
Z
√
V (pi ) = gs dx eipi ·X
R
where the integral dx is now over the boundary and p2 = 1/α0 is the on-shell condition
√
for an open-string tachyon. The normalization gs is that appropriate for the insertion
of an open-string mode, reflecting (6.16).
Going through the same steps as for the closed string, we find that the amplitude is
given by
(4)
A
Z Y
4
X
Y
gs
0
26
∼
δ (
pi )
dxi
| xj − xl |2α pj ·pl
Vol(SL(2; R))
i
i=1
j<l
(6.17)
Note that there’s a factor of 2 in the exponent, differing from the closed string expression
(6.8). This comes about because the boundary propagator (4.52) has an extra factor
of 2 due to the image charge.
– 141 –
We now use the SL(2; R) residual gauge symmetry to fix three points on the boundary. We choose a particular ordering and set x1 = 0, x2 = x, x3 = 1 and x4 → ∞. The
only free insertion point is x2 = x but, because of the restriction of operator ordering,
this must lie in the interval x ∈ [0, 1]. The interesting part of the integral is then given
by
Z 1
0
0
(4)
dx |x|2α p1 ·p2 |1 − x|2α p2 ·p3
A ∼ gs
0
This integral is well known: as shown in Appendix 6.5, it is the Euler beta function
Z 1
Γ(a)Γ(b)
dx xa−1 (1 − x)b−1 =
B(a, b) =
Γ(a + b)
0
After summing over the different orderings of vertex operators, the end result for the
amplitude for open string tachyon scattering is,
A(4) ∼ gs [B(−α0 s − 1, −α0 t − 1) + B(−α0 s − 1, −α0 u − 1) + B(−α0 t − 1, −α0 u − 1)]
This is the famous Veneziano Amplitude, first postulated in 1968 to capture some
observed features of the strong interactions. This was before the advent of QCD and
before it was realised that the amplitude arises from a string.
The open string scattering amplitude contains the same features that we saw for the
closed string. For example, it has poles at
s=
n−1
α0
n = 0, 1, 2, . . .
which we recognize as the spectrum of the open string.
6.3.2 The Tension of D-Branes
Recall that we introduced D-branes as surfaces in space on which strings can end. At the
time, I promised that we would eventually discover that these D-branes are dynamical
objects in their own right. We’ll look at this more closely in the next section, but for
now we can do a simple computation to determine the tension of D-branes.
The tension Tp of a Dp-brane is defined as the energy per spatial volume. It has
dimension [Tp ] = p + 1. The tension is telling us the magnitude of the coupling between
the brane and gravity. Or, in our new language, the strength of the interaction between
a closed string state and an open string. The simplest such diagram is shown in the
figure, with a graviton vertex operator inserted. Although we won’t compute this
– 142 –
diagram completely, we can figure out its most important property just by looking at
it: it has the topology of a disc, so is proportional to 1/gs . Adding powers of α0 to get
the dimension right, the tension of a Dp-brane must scale as
Tp ∼
1
lsp+1
1
gs
(6.18)
√
where the string length is defined as ls = α0 . The 1/gs scaling of
the tension is one of the key characteristic features of a D-brane.
I should confess that there’s a lot swept under the carpet in the
above discussion, not least the question of the correct normalization
of the vertex operators and the difference between the string frame
and the Einstein frame (which we will discuss shortly). Nonetheless,
the end result (6.18) is correct. For a fuller discussion, see Section
8.7 of Polchinski.
Figure 39:
6.4 One-Loop Amplitudes
We now return to the closed string to discuss one-loop effects. As we saw above, this
corresponds to a worldsheet with the topology of a torus. We need to integrate over
all metrics on the torus.
For tree-level processes, we used diffeomorphisms and Weyl transformations to map
an arbitrary metric on the sphere to the flat metric on the plane. This time, we use
these transformations to map an arbitrary metric on the torus to the flat metric on
the torus. But there’s a new subtlety that arises: not all flat metrics on the torus are
equivalent.
6.4.1 The Moduli Space of the Torus
Let’s spell out what we mean by this. We can construct a torus by identifying a region
in the complex z-plane as shown in the figure. In general, this identification depends
on a single complex parameter, τ ∈ C.
z ≡ z + 2π
and
z ≡ z + 2πτ
Do not confuse τ with the Minkowski worldsheet time: we left that behind way back
in Section 3. Everything here is Euclidean worldsheet and τ is just a parameter telling
us how skewed the torus is. The flat metric on the torus is now simply
ds2 = dzdz̄
subject to the identifications above.
– 143 –
A general metric on a torus can always be transformed to a flat metric for some value of τ . But the
question that interests us is whether two tori, parameterized by different τ , are conformally equivalent. In
general, the answer is no. The space of conformally inequivalent tori, parameterized by τ , is called the moduli space M.
Im(z)
2πτ
Re(z)
2π
However, there are some values of τ that do correspond to the same torus. In particular, there are
Figure 40:
a couple of obvious ways in which we can change τ
without changing the torus. They go by the names of the S and T transformations:
• T : τ → τ + 1: This clearly gives rise to the same torus, because the identification
is now
z ≡ z + 2π
and
z ≡ z + 2π(τ + 1) ≡ z + 2πτ
• S : τ → −1/τ : This simply flips the sides of the torus. For example, if τ = ia
is purely imaginary, then this transformation maps τ → i/a, which can then be
undone by a scaling.
It turns out that these two changes S and T are the
only ones that keep the torus intact. They are sometimes called modular transformations. A general modular transformations is constructed from combinations of S and T and takes the form,
τ→
aτ + b
cτ + d
with ad − bc = 1
(6.19)
where a, b, c and d ∈ Z. This is the group SL(2, Z). (In fact, we have our usual Z2
identification and the group is actually P SL(2, Z) = SL(2; Z)/Z2 ). The moduli space
M of the torus is given by
M∼
= C/SL(2; Z)
What does this space look like? Using T : τ → τ + 1, we can always shift τ until it lies
within the interval
Re τ ∈ [− 12 , + 12 ]
where the edges of the interval are identified. Meanwhile, S : τ → −1/τ inverts the
– 144 –
Figure 41: The fundamental domain.
modulus |τ |, so we can use this to map a point inside the circle |τ | < 1 to a point outside
|τ | > 1. One can show that by successive combinations of S and T , it is possible to
map any point to lie within the shaded region shown in the figure, defined by
|τ | ≥ 1
and Re τ ∈ [− 12 , + 12 ]
This is referred to as the fundamental domain of SL(2; Z).
We could have just as easily chosen one of the other fundamental domains shown in
the figure. But the shaded region is the standard one.
Integrating over the Moduli Space
In string theory we’re invited to sum over all metrics. After gauge fixing diffeomorphisms and Weyl invariance, we still need to integrate over all inequivalent tori. In other
words, we integrate over the fundamental domain. The SL(2; Z) invariant measure over
the fundamental domain is
Z
d2 τ
(Im τ )2
To see that this is SL(2; Z) invariant, note that under a general transformation of the
form (6.19) we have
d2 τ →
d2 τ
|cτ + d|4
and Im τ →
– 145 –
Im τ
|cτ + d|2
There’s some physics lurking within these rather mathematical statements. The integration over the fundamental domain in string theory is analogous to the loop integral
over momentum in quantum field theory. Consider the square tori defined by Re τ = 0.
The tori with Im τ → ∞ are squashed and chubby. They correspond to the infra-red
region of loop momenta in a Feynman diagram. Those with Im τ → 0 are long and
thin. Those correspond to the ultra-violet limit of loop momenta in a Feynman diagram. Yet, as we have seen, we should not integrate over these UV regions of the loop
since the fundamental domain does not stretch down that far. Or, more precisely, the
thin tori are mapped to chubby tori. This corresponds to the fact that any putative
UV divergence of string theory can always be reinterpreted as an IR divergence. This
is the second manifestation of the well-behaved UV nature of string theory. We will
see this more explicitly in the example of Section 6.4.2.
Finally, when computing a loop amplitude in string theory, we still need to worry
about the residual gauge symmetry that is left unfixed after the map to the flat torus.
In the case of tree-level amplitudes on the sphere, this residual gauge symmetry was
due to the conformal Killing group SL(2; C). For the torus, the conformal Killing
group is generated by the obvious generators ∂z and ∂¯z̄ . It is U (1) × U (1).
Higher Genus Surfaces
The moduli space Mg of the Riemann surface of genus g > 1 can be shown to have
dimension,
dim Mg = 3g − 2
There are no conformal Killing vectors when g > 1. These facts can be demonstrated
as an application of the Riemann-Roch theorem. For more details, see section 5.2 of
Polchinski, or sections 3.3 and 8.2 of Green, Schwarz and Witten.
6.4.2 The One-Loop Partition Function
We won’t compute any one-loop scattering amplitudes in string theory. Instead, we
will look at something a little simpler: the one-loop vacuum to vacuum amplitude.
A Euclidean worldsheet with periodic time has the interpretation of a finite temperature partition function for the theory defined on a cylinder. In D = 26 dimensional
spacetime, it is related to the cosmological constant in bosonic string theory.
Consider firstly the partition function of a theory on a square torus, with Re τ = 0.
Compactifying Euclidean time, with period (Im τ ) is equivalent to putting the theory
at temperature T = 1/(Im τ ),
Z[τ ] = Tr e−2π(Im τ )H
– 146 –
where the Tr is over all states in the theory. For any CFT defined on a cylinder, the
Hamiltonian given by
H = L0 + L̃0 −
c + c̃
24
where the final term is the Casimir energy computed in Section 4.4.1.
What then is the interpretation of the vacuum amplitude
computed on a torus with Re τ 6= 0? From the diagram,
we see that the effect of such a skewed torus is to translate a given point around the cylinder by Re τ . But we
know which operator implements such a translation: it is
exp(2πi(Re τ )P ), where P is the momentum operator on
the cylinder. After the map to the plane, this becomes the
rotation operator
Im(z)
2πτ
Re(z)
2π
Figure 42:
P = L0 − L̃0
So the vacuum amplitude on the torus has the interpretation of the sum over all states
in the theory, weighted by
Z[τ ] = Tr e−2π(Im τ )(L0 +L̃0 ) e−2πi(Re τ )(L0 −L̃0 ) e2π(Im τ )(c+c̃)/24
We define
q = e2πiτ
,
q̄ = e−2πiτ̄
The partition function can then be written in slick notation as
Z[τ ] = Tr q L0 −c/24 q̄ L̃0 −c̃/24
Let’s compute this for the free string. We know that each scalar field X decomposes
into a zero mode and an infinite number harmonic oscillator modes α−n which create
states of energy n. We’ll deal with the zero mode shortly but, for now, we focus on the
oscillators. Acting d times with the operator α−n creates states with energy dn. This
gives a contribution to Trq L0 of the form
∞
X
d=0
q nd =
1
1 − qn
– 147 –
But the Fock space of a single scalar field is built by acting with oscillator modes
n ∈ Z+ . Including the central charge, c = 1, the contribution from the oscillator modes
of a single scalar field is therefore
Tr q L0 −c/24 =
1
q 1/24
∞
Y
1
1 − qn
n=1
There is a similar expression from the q̄ L̃0 −c̃/24 sector. We’re still left with the contribution from the zero mode p of the scalar field. The contribution to the energy H of
the state on the worldsheet is
Z
1
1
dσ (α0 p)2 = α0 p2
0
4πα
2
The trace in the partition function requires us to sum over all states, which gives
Z
dp −πα0 (Im τ )p2
1
e
∼√
0
2π
α Im τ
So, including both the zero mode and oscillators, we get the partition function for a
single free scalar field,
∞
∞
Y
Y
1
1
1
1
Zscalar [τ ] ∼ √
1/24
n
1 − q n=1 1 − q̄ n
α0 Im τ (q q̄)
n=1
(6.20)
where I haven’t been careful to keep track of constant factors.
To build the string partition function, we should really work in covariant quantization
and include the ghost fields. Here we’ll cheat and work in lightcone gauge. This is dodgy
because, if we do it honestly, much of the physics gets pushed to the p+ = 0 limit of
the lightcone momentum where the gauge choice breaks down. So instead we’ll do it
dishonestly.
In lightcone gauge, we have 24 oscillator modes. But we have 26 zero modes. (You
may worry that we still have to impose level matching...this is the dishonest part of
the calculation. We’ll see partly where it comes from shortly). Finally, there’s a couple
of extra steps. We need to divide by the volume of the conformal Killing group. This
is just U (1) × U (1), acting by translations along the cycles of the torus. The volume
is just Vol = 4π 2 Im τ . Finally, we also need to integrate over the moduli space of the
torus. Our final result, neglecting all constant factors, is
!24 ∞
!24
Z
∞
Y
Y
1
1
1
1
1
Zstring = d2 τ
(6.21)
0
13
n
(Im τ ) (α Im τ )
q q̄ n=1 1 − q
1 − q̄ n
n=1
– 148 –
Modular Invariance
The function appearing in the partition function for the scalar field has a name: it is
the inverse of the Dedekind eta function
η(q) = q
1/24
∞
Y
(1 − q n )
n=1
It was studied in the 1800s by mathematicians interested in the properties of functions
under modular transformations T : τ → τ + 1 and S : τ → −1/τ . The eta-function
satisfies the identities
√
η(τ + 1) = e2πi/24 η(τ ) and η(−1/τ ) = −iτ η(τ )
These two statements ensure that the scalar partition function (6.20) is a modular
invariant function. Of course, that kinda had to be true: it follows from the underlying
physics.
Written in terms of η, the string partition function (6.21) takes the form
24
Z
d2 τ
1
1 1
√
Zstring =
(Im τ )2
Im τ η(q) η̄(q̄)
Both the measure and the integrand, are individually modular invariant.
6.4.3 Interpreting the String Partition Function
It’s probably not immediately obvious what the string partition function (6.21) is telling
us. Let’s spend some time trying to understand it in terms of some simpler concepts.
We know that the free string describes an infinite number of particles with mass
= 4(n − 1)/α0 , n = 0, 1, . . .. The string partition function should just be a sum
over vacuum loops of each of these particles. We’ll now show that it almost has this
interpretation.
m2n
Firstly, let’s figure out what the contribution from a single particle would be? We’ll
consider a free massive scalar field φ in D dimensions. The partition function is given
by,
Z
Z
1
D
2
2
Z =
Dφ exp −
d x φ(−∂ + m )φ
2
∼ det−1/2 (−∂ 2 + m2 )
Z
1
dD p
2
2
= exp
ln(p + m )
2
(2π)D
– 149 –
This is the partition function of a field theory. It contains vacuum loops for all numbers
of particles. To compare to the string partition function, we want the vacuum amplitude
for just a single particle. But that’s easy to extract. We write the field theory partition
function as,
∞
X
Zn
1
Z = exp (Z1 ) =
n=0
n!
Each term in the sum corresponds to n particles propagating in a vacuum loop, with
the n! factor taking care of Bosonic statistics. So the vacuum amplitude for a single,
free massive particle is simply
1
Z1 =
2
dD p
ln(p2 + m2 )
(2π)D
Z
Clearly this diverges in the UV range of the integral, p → ∞. There’s a nice way to
rewrite this integral using something known as Schwinger parameterization. We make
use of the identity
Z
∞
dl e
0
∞
Z
1
=
x
−xl
⇒
dl
0
e−xl
= − ln x
l
We then write the single particle partition function as
Z
Z1 =
dD p
(2π)D
∞
Z
0
dl −(p2 +m2 )l
e
2l
(6.22)
It’s worth mentioning that there’s another way to see that this is the single particle
partition function that is a little closer in spirit to the method we used in string theory.
We could start with the einbein form of the relativistic particle action (1.8). After
fixing the gauge to e = 1, the exponent in (6.22) is the energy of the particle traversing
a loop of length l. The integration measure dl/l sums over all possible sizes of loops.
We can happily perform the
have
R
dD p integral in (6.22). Ignoring numerical factors, we
Z
Z1 =
∞
dl
0
1
l1+D/2
e−m
2l
(6.23)
Note that the UV divergence as p → ∞ has metamorphosised into a divergence associated to small loops as l → 0.
– 150 –
Equation (6.23) gives the answer for a single particle of mass m. In string theory,
we expect contributions from an infinite number of species of particles of mass mn .
Specializing to D = 26, we expect the partition function to be
Z ∞
∞
1 X −m2n l
Z=
dl 14
e
l n=0
0
But we know that the mass spectrum of the free string: it is given in terms of the L0
and L̃0 operators by
m2 =
4
4
2
(L0 − 1) = 0 (L̃0 − 1) = 0 (L0 + L̃0 − 2)
0
α
α
α
subject to the constraint of level matching, L0 = L̃0 . It’s easy to impose level matching:
we simply throw in a Kronecker delta in its integral representation,
Z +1/2
1
ds e2πis(L0 −L̃0 ) = δL0 ,L̃0
(6.24)
2π −1/2
Replacing the sum over species, with the trace over the spectrum of states subject to
level matching, the partition function becomes,
Z +1/2
Z ∞
1
0
ds Tr e2πis(L0 −L̃0 ) e−2(L0 +L̃0 −2)l/α
(6.25)
Z=
dl 14
l
−1/2
0
We again use the definition q = exp(2πiτ ), but this time the complex parameter τ is a
combination of the length of the loop l and the auxiliary variable that we introduced
to impose level matching,
τ =s+
2li
α0
The trace over the spectrum of the string once gives the eta-functions, just as it did
before. We’re left with the result for the partition function,
24
Z
d2 τ
1
1 1
√
Zstring =
(Im τ )2
Im τ η(q) η̄(q̄)
But this is exactly the same expression that we saw before. With a difference! In fact,
the difference is hidden in the notation: it is the range of integration for d2 τ which can
be found in the original expressions (6.23) and (6.24). Re τ runs over the same interval
[− 21 , + 12 ] that we saw in string theory. As is clear from this discussion, it is this integral
which implements level matching. The difference comes in the range of Im τ which, in
this naive analysis, runs over [0, ∞). This is in stark contrast to string theory where
we only integrate over the fundamental domain.
– 151 –
This highlights our previous statement: the potential UV divergences in field theory
are encountered in the region Im τ ∼ l → 0. In the above analysis, this corresponds to
particles traversing small loops. But this region is simply absent in the correct string
theory computation. It is mapped, by modular invariance, to the infra-red region of
large loops.
It is often said that in the gs → 0 limit string theory becomes a theory of an infinite
number of free particles. This is true of the spectrum. But this calculation shows that
it’s not really true when we compute loops because the modular invariance means that
we integrate over a different range of momenta in string theory than in a naive field
theory approach.
So what happens in the infra-red region of our partition function? The easiest place
to see it is in the l → ∞ limit of the integral (6.25). We see that the integral is
dominated by the lightest state which, for the bosonic string is the tachyon. This has
m2 = −4/α0 , or (L0 + L̃0 − 2) = −2. This gives a contribution to the partition function
of,
Z ∞
dl +4l/α0
e
l14
which clearly diverges. This IR divergence of the one-loop partition function is another
manifestation of tachyonic trouble. In the superstring, there is no tachyon and the IR
region is well-behaved.
6.4.4 So is String Theory Finite?
The honest answer is that we don’t know. The UV finiteness that we saw above
holds for all one-loop amplitudes. This means, in particular, that we have a one-loop
finite theory of gravity interacting with matter in higher dimensions. This is already
remarkable.
There is more good news: One can show that UV finiteness continues to hold at
the two-loops. And, for the superstring, state-of-the-art techniques using the “purespinor” formalism show that certain objects remain finite up to five-loops. Moreover,
the exponential suppression (6.14) that we saw when all momentum exchanges are large
continues to hold for all amplitudes.
However, no general statement of finiteness has been proven. The danger lurks in
the singular points in the integration over Riemann surfaces of genus 3 and higher.
– 152 –
6.4.5 Beyond Perturbation Theory?
From the discussion in this section, it should be clear that string perturbation theory
is entirely analogous to the Feynman diagram expansion in field theory. Just as in field
theory, one can show that the expansion in gs is asymptotic. This means that the series
does not converge, but we can nonetheless make sense of it.
However, we know that there are many phenomena in quantum field theory that
aren’t captured by Feynman diagrams. These include confinement in the strongly
coupled regime and instantons and solitons in the weakly coupled regime. Does this
mean that we are missing similarly interesting phenomena in string theory? The answer
is almost certainly yes! In this section, I’ll very briefly allude to a couple of more
advanced topics which allow us to go beyond the perturbative expansion in string
theory. The goal is not really to teach you these things, but merely to familiarize you
with some words.
One way to proceed is to keep quantum field theory as our guide and try to build a
non-perturbative definition of string theory in terms of a path integral. We’ve already
seen that the Polyakov path integral over worldsheets is equivalent to Feynman diagrams. So we need to go one step further. What does this mean? Recall that in QFT,
a field creates a particle. In string theory, we are now looking for a field which creates
a loop of string. We should have a different field for each configuration of the string.
In other words, our field should itself be a function of a function: Φ(X µ (σ)). Needless
to say, this is quite a complicated object. If we were brave, we could then consider the
path integral for this field,
Z
Z = DΦ eiS[Φ(X(σ))]
for some suitable action S[Φ]. The idea is that this path integral should reproduce the
perturbative string expansion and, furthermore, defines a non-perturbative completion
of the theory. This line of ideas is known as string field theory. It should be clear
that this is one step further in the development: particles → fields → string fields. Or,
in more historical language, if field theory is “second quantization”, then string field
theory is “third quantization”.
String field theory has been fairly successful for the open string and some interesting
non-perturbative results have been obtained in this manner. However, for the closed
string this approach has been much less useful. It is usually thought that there are
deep reasons behind the failure of closed string field theory, related to issues that we
mentioned at the beginning of this section: there are no off-shell quantities in a theory
– 153 –
of gravity. Moreover, we mentioned in Section 4 that a theory of interacting open
strings necessarily includes closed strings, so somehow the open string field theory
should already contain gravity and closed strings. Quite how this comes about is still
poorly understood.
There are other ways to get a handle on non-perturbative aspects of string theory
using the low-energy effective action (we will describe what the “low-energy effective
action” is in the next section). Typically these techniques rely on supersymmetry to
provide a window into the strongly coupled regime and so work only for the superstring.
These methods have been extremely successful and any course on superstring theory
would be devoted to explaining various aspects of such as dualities and M-theory.
Finally, in asymptotically AdS spacetimes, the AdS/CFT correspondence gives a nonperturbative definition of string theory and quantum gravity in the bulk in terms of
Yang-Mills theory, or something similar, on the boundary. In some sense, the boundary
field theory is a “string field theory”.
6.5 Appendix: Games with Integrals and Gamma Functions
The gamma function is defined by the integral representation
Z ∞
Γ(z) =
dt tz−1 e−t
(6.26)
0
which converges if Rez > 0. It has a unique analytic
expression to the whole z-plane. The absolute value
of the gamma function over the z-plane is shown in
the figure.
Figure 43:
The gamma function has a couple of important properties. Firstly, it can be thought
of as the analytic continuation of the factorial function for positive integers, meaning
Γ(n) = (n − 1)!
n ∈ Z+
Secondly, Γ(z) has poles at non-positive integers. More precisely when z ≈ −n, with
n = 0, 1, . . ., there is the expansion
Γ(z) ≈
1 (−1)n
z + n n!
– 154 –
The Euler Beta Function
The Euler beta function is defined for x, y ∈ C by
B(x, y) =
Γ(x)Γ(y)
Γ(x + y)
It has the integral representation
1
Z
dt tx−1 (1 − t)y−1
B(x, y) =
(6.27)
0
Let’s prove this statement. We start by looking at
Z ∞ Z ∞
dv e−u ux−1 e−v v y−1
du
Γ(x)Γ(y) =
0
0
We write u = a2 and v = b2 so the integral becomes
Z ∞ Z ∞
2
2
db e−(a +b ) a2x−1 b2y−1
da
Γ(x)Γ(y) = 4
0
Z ∞0
Z ∞
2
2
=
da
db e−(a +b ) |a|2x−1 |b|2y−1
−∞
−∞
We now change coordinates once more, this time to polar a = r cos θ and b = r sin θ.
We get
Z ∞
Z 2π
−r2 2x+2y−2
Γ(x)Γ(y) =
rdr e r
dθ | cos θ|2x−1 | sin θ|2y−1
0
0
Z π/2
1
= Γ(x + y) × 4
dθ (cos θ)2x−1 (sin θ)2y−1
2
0
Z 1
dt (1 − t)y−1 tx−1
= Γ(x + y)
0
where, in the final line, we made the substitution t = cos2 θ. This completes the proof.
The Virasoro-Shapiro Amplitude
In the closed string computation, we came across the integral
Z
C(a, b) = d2 z |z|2a−2 |1 − z|2b−2
We will now evaluate this and show that it is given by (6.11). We start by using a
trick. We can write
Z ∞
1
2
2a−2
|z|
=
dt t−a e−|z| t
Γ(1 − a) 0
– 155 –
which follows from the definition (6.26) of the gamma function. Similarly, we can write
Z ∞
1
2
2b−2
du u−b e−|1−z| u
|1 − z|
=
Γ(1 − b) 0
We decompose the complex coordinate z = x + iy, so that the measure of the integral
is d2 z = 2dxdy. We can then write the integral C(a, b) as
Z
d2 z du dt
2
2
C(a, b) =
t−a u−b e−|z| t e−|1−z| u
Γ(1 − a)Γ(1 − b)
Z
dx dy du dt
2
2
t−a u−b e−(t+u)(x +y )+2xu−u
=2
Γ(1 − a)Γ(1 − b)
"
!
#
2
Z
2
dx dy du dt
u
u
=2
t−a u−b exp −(t + u) x −
+ y2 − u +
Γ(1 − a)Γ(1 − b)
t+u
t+u
Now we do the dxdy integral which is simply Gaussian. We find
Z ∞
2π
t−a u−b −tu/(t+u)
C(a, b) =
du dt
e
Γ(1 − a)Γ(1 − b) 0
t+u
Finally, we make a change of variables. We write t = αβ and u = (1 − β)α. In order
for t and u to take values in the range [0, ∞), we require α ∈ [0, ∞) and β ∈ [0, 1].
Taking into account the Jacobian arising from this transformation, which is simply α,
the integral becomes
Z
α1−a−b −a
2π
dα dβ
β (1 − β)−b e−αβ(1−β)
C(a, b) =
Γ(1 − a)Γ(1 − b)
α
But we recognize the integral over dα: it is simply
Z ∞
dα α−a−b e−βα(1−β) = [β(1 − β)]a+b−1 Γ(1 − a − b)
0
We write c = 1 − a − b. Finally, we’re left with
Z 1
2πΓ(c)
C(a, b) =
dβ (1 − β)a−1 β b−1
Γ(1 − a)Γ(1 − b) 0
But the final integral is the Euler beta function (6.27). This gives us our promised
result,
C(a, b) =
2πΓ(a)Γ(b)Γ(c)
Γ(1 − a)Γ(1 − b)Γ(1 − c)
– 156 –
7. Low Energy Effective Actions
So far, we’ve only discussed strings propagating in flat spacetime. In this section we
will consider strings propagating in different backgrounds. This is equivalent to having
different CFTs on the worldsheet of the string.
There is an obvious generalization of the Polyakov action to describe a string moving
in curved spacetime,
Z
1
√
S=
d2 σ g g αβ ∂α X µ ∂β X ν Gµν (X)
(7.1)
0
4πα
Here gαβ is again the worldsheet metric. This action describes a map from the worldsheet of the string into a spacetime with metric Gµν (X). (Despite its name, this metric
is not to be confused with the Einstein tensor which we won’t have need for in this
lecture notes).
Actions of the form (7.1) are known as non-linear sigma models. (This strange name
has its roots in the history of pions). In this context, the D-dimensional spacetime is
sometimes called the target space. Theories of this type are important in many aspects
of physics, from QCD to condensed matter.
Although it’s obvious that (7.1) describes strings moving in curved spacetime, there’s
something a little fishy about just writing it down. The problem is that the quantization
of the closed string already gave us a graviton. If we want to build up some background
metric Gµν (X), it should be constructed from these gravitons, in much the same manner
that a laser beam is made from the underlying photons. How do we see that the metric
in (7.1) has anything to do with the gravitons that arise from the quantization of the
string?
The answer lies in the use of vertex operators. Let’s expand the metric as a small
fluctuation around flat space
Gµν (X) = δµν + hµν (X)
Then the partition function that we build from the action (7.1) is related to the partition
function for a string in flat space by
Z
Z
1
−SPoly −V
Z = DXDg e
= DXDg e−SPoly (1 − V + V 2 + . . .)
2
– 157 –
where SPoly is the action for the string in flat space given in (1.22) and V is the
expression
Z
1
√
V =
d2 σ g g αβ ∂α X µ ∂β X ν hµν (X)
(7.2)
0
4πα
But we’ve seen this before: it’s the vertex operator associated to the graviton state of
the string! For a plane wave, corresponding to a graviton with polarization given by
the symmetric, traceless tensor ζµν and momentum pµ , the fluctuation is given by
hµν (X) = ζµν eip·X
With this choice, the expression (7.2) agrees with the vertex operator (5.8). But in
general, we could take any linear superposition of plane waves to build up a general
fluctuation hµν (X).
We know that inserting a single copy of V in the path integral corresponds to the
introduction of a single graviton state. Inserting eV in the path integral corresponds
to a coherent state of gravitons, changing the metric from δµν to δµν + hµν . In this
way we see that the background curved metric of (7.1) is indeed built of the quantized
gravitons that we first met back in Section 2.
7.1 Einstein’s Equations
In conformal gauge, the Polyakov action in flat space reduces to a free theory. This
fact was extremely useful, allowing us to compute the spectrum of the theory. But on a
curved background, it is no longer the case. In conformal gauge, the worldsheet theory
is described by an interacting two-dimensional field theory,
Z
1
S=
d2 σ Gµν (X) ∂α X µ ∂ α X ν
(7.3)
4πα0
To understand these interactions in more detail, let’s expand around a classical solution
which we take to simply be a string sitting at a point x̄µ .
√
X µ (σ) = x̄µ + α0 Y µ (σ)
Here Y µ are the
√ dynamical fluctuations about the point which we assume to be small.
The factor of α0 is there for dimensional reasons: since [X] = −1, we have [Y ] = 0
and statements like Y 1 make sense. Expanding the Lagrangian gives
0
√
α
µ
ν
0
ω
ω
ρ
Gµν (X) ∂X ∂ X = α Gµν (x̄) + α0 Gµν,ω (x̄) Y + Gµν,ωρ (x̄) Y Y + . . . ∂Y µ ∂Y ν
2
Each of the coefficients Gµν,... in the Taylor expansion are coupling constants for the
interactions of the fluctuations Y µ . The theory has an infinite number of coupling
constants and they are nicely packaged into the function Gµν (X).
– 158 –
We want to know when this field theory is weakly coupled. Obviously this requires
the whole infinite set of coupling constants to be small. Let’s try to characterize this in
a crude manner. Suppose that the target space has characteristic radius of curvature
rc , meaning schematically that
∂G
1
∼
∂X
rc
The radius of curvature is a length scale, so [rc ] = −1. From the expansion of the
metric, we see that the effective dimensionless coupling is given by
√
α0
rc
(7.4)
This means that we can use perturbation theory√
to study the CFT (7.3) if the spacetime
√
metric only varies on scales much greater than α0 . The perturbation series in α0 /rc
is usually called the α0 -expansion to distinguish it from the gs expansion that we saw
in the previous section. Typically a quantity computed in string theory is given by a
double perturbation expansion: one in α0 and one in gs .
If there are regions of spacetime
√ where the radius of curvature becomes comparable
to the string length scale, rc ∼ α0 , then the worldsheet CFT is strongly coupled and
we will need to develop new methods to solve it. Notice that strong coupling in α0 is
hard, but the problem is at least well-defined in terms of the worldsheet path integral.
This is qualitatively different to the question of strong coupling in gs for which, as
discussed in Section 6.4.5, we’re really lacking a good definition of what the problem
even means.
7.1.1 The Beta Function
Classically, the theory defined by (7.3) is conformally invariant. But this is not necessarily true in the quantum theory. To regulate divergences we will have to introduce a
UV cut-off and, typically, after renormalization, physical quantities depend on the scale
of a given process µ. If this is the case, the theory is no longer conformally invariant.
There are plenty of theories which classically possess scale invariance which is broken
quantum mechanically. The most famous of these is Yang-Mills.
As we’ve discussed several times, in string theory conformal invariance is a gauge
symmetry and we can’t afford to lose it. Our goal in this section is to understand the
circumstances under which (7.3) retains conformal invariance at the quantum level.
– 159 –
The object which describes how couplings depend on a scale µ is called the β-function.
Since we have a functions worth of couplings, we should really be talking about a βfunctional, schematically of the form
βµν (G) ∼ µ
∂Gµν (X; µ)
∂µ
The quantum theory will be conformally invariant only if
βµν (G) = 0
We now compute this for the non-linear sigma model at one-loop. Our strategy will be
to isolate the UV divergence of the theory and figure out what kind of counterterm we
should add. The beta-function will vanish if this counterterm vanishes.
The analysis is greatly simplified by a cunning choice of coordinates. Around any
point x̄, we√can always pick Riemann normal coordinates such that the expansion in
X µ = x̄µ + α0 Y µ gives
Gµν (X) = δµν −
α0
Rµλνκ (x̄)Y λ Y κ + O(Y 3 )
3
To quartic order in the fluctuations, the action becomes
Z
α0
1
d2 σ ∂Y µ ∂Y ν δµν − Rµλνκ Y λ Y κ ∂Y µ ∂Y ν
S=
4π
3
We can now treat this as an interacting quantum field theory in two dimensions. The
quartic interaction gives a vertex with the Feynman rule,
∼ Rµλνκ (k µ · k ν )
where kαµ is the 2d momentum (α = 1, 2 is a worldsheet index) for the scalar field Y µ .
It sits in the Feynman rules because we are talking about derivative interactions.
Now we’ve reduced the problem to a simple interacting quantum field theory, we can
compute the β-function using whatever method we like. The divergence in the theory
comes from the one-loop diagram
– 160 –
It’s actually simplest to think about this diagram in position space. The propagator
for a scalar particle is
1
hY λ (σ)Y κ (σ 0 )i = − δ λκ ln |σ − σ 0 |2
2
For the scalar field running in the loop, the beginning and end point coincide. The
propagator diverges as σ → σ 0 , which is simply reflecting the UV divergence that we
would see in the momentum integral around the loop.
To isolate this divergence, we choose to work with dimensional regularization, with
d = 2 + . The propagator then becomes,
Z
0
d2+ k eik·(σ−σ )
λ
κ
0
λκ
hY (σ)Y (σ )i = 2πδ
(2π)2+
k2
δ λκ
−→
as σ → σ 0
The necessary counterterm for this divergence can be determined simply by replacing
Y λ Y κ in the action with hY λ Y κ i. To subtract the 1/ term, we add the counterterm
1
Rµλνκ Y λ Y κ ∂Y µ ∂Y ν → Rµλνκ Y λ Y κ ∂Y µ ∂Y ν − Rµν ∂Y µ ∂Y ν
One can check that this can be absorbed by a wavefunction renormalization Y µ →
Y µ + (α0 /6) Rµν Y ν , together with the renormalization of the coupling constant which,
in our theory, is the metric Gµν . We require,
Gµν → Gµν +
α0
Rµν
(7.5)
From this we learn the beta function of the theory and the condition for conformal
invariance. It is
βµν (G) = α0 Rµν = 0
(7.6)
This is a magical result! The requirement for the sigma-model to be conformally invariant is that the target space must be Ricci flat: Rµν = 0. Or, in other words, the
background spacetime in which the string moves must obey the vacuum Einstein equations! We see that the equations of general relativity also describe the renormalization
group flow of 2d sigma models.
There are several more magical things just around the corner, but it’s worth pausing
to make a few diverse comments.
– 161 –
Beta Functions and Weyl Invariance
The above calculation effectively studies the breakdown of conformal invariance in the
CFT (7.3) on a flat worldsheet. We know that this should be the same thing as the
breakdown of Weyl invariance on a curved worldsheet. Since this is such an important
result, let’s see how it works from this other perspective. We can consider the worldsheet
metric
gαβ = e2φ δαβ
Then, in dimensional regularization, the theory is not Weyl invariant in d = 2 + √
dimensions because the contribution from g does not quite cancel that from the
inverse metric g αβ . The action is
Z
1
S=
d2+ σ eφ ∂α X µ ∂ α X ν Gµν (X)
4πα0
Z
1
d2+ σ (1 + φ) ∂α X µ ∂ α X ν Gµν (X)
≈
0
4πα
where, in this expression, the α = 1, 2 index is now raised and lowered with δαβ . If we
replace Gµν in this expression with the renormalized metric (7.5), we see that there’s
a term involving φ which remains even as → 0,
Z
1
d2 σ ∂α X µ ∂ α X ν [Gµν (X) + α0 φ Rµν (X)]
S=
4πα0
This indicates a breakdown of Weyl invariance. Indeed, we can look at our usual
diagnostic for Weyl invariance, namely the vanishing of T αα . In conformal gauge, this
is given by
4π ∂S
∂S
Tαβ = + √
= −2π δαβ
αβ
g ∂g
∂φ
1
⇒ T αα = − Rµν ∂X µ ∂X ν
2
In this way of looking at things, we define the β-function to be the coefficient in front
of ∂X∂X, namely
T αα = −
1
βµν ∂X µ ∂X ν
2α0
Again, we have the result
βµν = α0 Rµν
– 162 –
7.1.2 Ricci Flow
In string theory we only care about conformal theories with Ricci flat metrics. (And
generalizations of this result that we will discuss shortly). However, in other areas of
physics and mathematics, the RG flow itself is important. It is usually called Ricci
flow,
µ
∂Gµν
= α0 Rµν
∂µ
(7.7)
which dictates how the metric changes with scale µ.
As an illustrative and simple example, consider the target space S2 with radius r.
This is an important model in condensed matter physics where it describes the lowenergy limit of a one-dimensional Heisenberg spin chain. It is sometimes called the
O(3) sigma-model. Because the sphere is a symmetric space, the only effect of the RG
flow is to make the radius scale dependent: r = r(µ). The beta function is given by
µ
∂r2
α0
=
∂µ
2π
Hence r gets large as we go towards the UV and small towards the IR. Since the coupling
is 1/r, this means that the non-linear sigma model with S2 target space is asymptotically free. At low energies, the theory is strongly coupled and perturbative calculations
— such as this one-loop beta function — are no longer trusted. In particular, one can
show that the S2 sigma-model develops a mass gap in the IR.
The idea of Ricci flow (7.7) was recently used by Perelman to prove the Poincaré
conjecture. In fact, Perelman used a slightly generalized version of Ricci flow which we
will see shortly. In the language of string theory, he introduced the dilaton field.
7.2 Other Couplings
We’ve understood how strings couple to a background spacetime metric. But what
about the other modes of the string? In Section 2, we saw that a closed string has
further massless states which are associated to the anti-symmetric tensor Bµν and the
dilaton Φ. We will now see how the string reacts if these fields are turned on in
spacetime.
7.2.1 Charged Strings and the B field
Let’s start by looking at how strings couple to the anti-symmetric field Bµν . We
discussed the vertex operator associated to this state in Section 5.4.1. It is given in
– 163 –
(5.8) and takes the same form as the graviton vertex operator, but with ζµν antisymmetric. It is a simple matter to exponentiate this, to get an expression for how
strings propagate in background Bµν field. We’ll keep the curved metric Gµν as well to
get the general action,
Z
1
2 √
µ
ν αβ
µ
ν αβ
(7.8)
S=
d
σ
g
G
(X)
∂
X
∂
X
g
+
iB
(X)
∂
X
∂
X
µν
α
β
µν
α
β
4πα0
√
Where αβ is the anti-symmetric 2-tensor, normalized such that g12 = +1. (The
factor of i is there in the action because we’re in Euclidean space and this new term
has a single “time” derivative). The action retains invariance under worldsheet reparameterizations and Weyl rescaling.
So what is the interpretation of this new term? We will now show that we should
think of the field Bµν as analogous to the gauge potential Aµ in electromagnetism. The
action (7.8) is telling us that the string is “electrically charged” under Bµν .
Gauge Potentials
We’ll take a short detour to remind ourselves about some pertinent facts in electromagnetism. Let’s start by returning to a point particle. We know that a charged point
particle couples to a background gauge potential Aµ through the addition of a worldline
term to the action,
Z
dτ Aµ (X) Ẋ µ .
(7.9)
If this relativistic form looks a little unfamiliar, we can deconstruct it by working in
static gauge with X 0 ≡ t = τ , where it reads
Z
dt A0 (X) + Ai (X) Ẋ i ,
which should now be recognizable as the Lagrangian that gives rise to the Coulomb
and Lorentz force laws for a charged particle.
So what is the generalization of this kind of coupling for a string? First note that (7.9)
has an interesting geometrical structure. It is the pull-back of the one-form A = Aµ dX µ
in spacetime onto the worldline of the particle. This works because A is a one-form and
the worldline is one-dimensional. Since the worldsheet of the string is two-dimensional,
the analogous coupling should be to a two-form in spacetime. This is an anti-symmetric
– 164 –
tensor field with two indices, Bµν . The pull-back of Bµν onto the worldsheet gives the
interaction,
Z
d2 σ Bµν (X) ∂α X µ ∂β X ν αβ .
(7.10)
This is precisely the form of the interaction we found in (7.8).
The point particle coupling (7.9) is invariant under gauge transformations of the
background field Aµ → Aµ + ∂µ α. This follows because the Lagrangian changes by a
total derivative. There is a similar statement for the two-form Bµν . The spacetime
gauge symmetry is,
Bµν → Bµν + ∂µ Cν − ∂ν Cµ
(7.11)
under which the Lagrangian (7.10) changes by a total derivative.
In electromagnetism, one can construct the gauge invariant electric and magnetic
fields which are packaged in the two-form field strength F = dA. Similarly, for Bµν ,
the gauge invariant field strength H = dB is a three-form,
Hµνρ = ∂µ Bνρ + ∂ν Bρµ + ∂ρ Bµν .
This 3-form H is sometimes known as the torsion. It plays the same role as torsion in
general relativity, providing an anti-symmetric component to the affine connection.
7.2.2 The Dilaton
Let’s now figure out how the string couples to a background dilaton field Φ(X). This is
more subtle. A naive construction of the vertex operator is not primary and one must
work a little harder. The correct derivation of the vertex operators can be found in
Polchinski. Here I will simply give the coupling and explain some important features.
The action of a string moving in a background involving profiles for the massless
fields Gµν , Bµν and Φ(X) is given by
Z
1
√
S=
d2 σ g Gµν (X) ∂α X µ ∂β X ν g αβ + iBµν (X) ∂α X µ ∂β X ν αβ
0
4πα
+ α0 Φ(X) R(2)
(7.12)
where R(2) is the two-dimensional Ricci scalar of the worldsheet. (Up until now, we’ve
always denoted this simply as R but we’ll introduce the superscript from hereon to
distinguish the worldsheet Ricci scalar from the spacetime Ricci scalar).
– 165 –
The coupling to the dilaton is surprising for several reasons. Firstly, we see that the
term in the action vanishes on a flat worldsheet, R(2) = 0. This is one of the reasons
that it’s a little trickier to determine this coupling using vertex operators.
However, the most surprising thing about the coupling to the dilaton is that it
does not respect Weyl invariance! Since a large part of this course has been about
understanding the implications of Weyl invariance, why on earth are we willing to
throw it away now?! The answer, of course, is that we’re not. Although the dilaton
coupling does violate Weyl invariance, there is a way to restore it. We will explain
this shortly. But firstly, let’s discuss one crucially important implication of the dilaton
coupling (7.12).
The Dilaton and the String Coupling
There is an exception to the statement that the classical coupling to the dilaton violates
Weyl invariance. This arises when the dilaton is constant. For example, suppose
Φ(X) = λ , a constant
Then the dilaton coupling reduces to something that we’ve seen before: it is
Sdilaton = λχ
where χ is the Euler character of the worldsheet that we introduced in (6.4). This tells
us something important: the constant mode of the dilaton, hΦi determines the string
coupling constant. This constant mode is usually taken to be the asymptotic value of
the dilaton,
Φ0 = limit Φ(X)
X→∞
(7.13)
The string coupling is then given by
gs = eΦ0
(7.14)
So the string coupling is not an independent parameter of string theory: it is the
expectation value of a field. This means that, just like the spacetime metric Gµν (or,
indeed, like the Higgs vev) it can be determined dynamically.
We’ve already seen that our perturbative expansion around flat space is valid as long
as gs 1. But now we have a stronger requirement: we can only trust perturbation
theory if the string is localized in regions of space where eΦ(X) 1 for all X. If the
string ventures into regions where eΦ(X) is of order 1, then we will need to use techniques
that don’t rely on string perturbation theory as described in Section 6.4.5.
– 166 –
7.2.3 Beta Functions
We now return to understanding how we can get away with the violation of Weyl
invariance in the dilaton coupling (7.12). The key to this is to notice the presence of
α0 in front of the dilaton coupling. It’s there simply on dimensional grounds. (The
other two terms in the action both come with derivatives [∂X] = −1, so don’t need
any powers of α0 ).
However, recall that α0 also plays the role of the loop-expansion parameter (7.4) in
the non-linear sigma model. This means that the classical lack of Weyl invariance in
the dilaton coupling can be compensated by a one-loop contribution arising from the
couplings to Gµν and Bµν .
To see this explicitly, one can compute the beta-functions for the two-dimensional
field theory (7.12). In the presence of the dilaton coupling, it’s best to look at the
breakdown of Weyl invariance as seen by hT αα i. There are three different kinds of
contribution that the stress-tensor can receive, related to the three different spacetime
fields. Correspondingly, we define three different beta functions,
hT αα i = −
i
1
1
βµν (G) g αβ ∂α X µ ∂β X ν − 0 βµν (B) αβ ∂α X µ ∂β X ν − β(Φ)R(2)
0
2α
2α
2
(7.15)
We will not provide the details of the one-loop beta function computations. We merely
state the results7 ,
βµν (G) = α0 Rµν + 2α0 ∇µ ∇ν Φ −
α0
Hµλκ Hν λκ
4
α0 λ
∇ Hλµν + α0 ∇λ Φ Hλµν
2
α0 2
α0
β(Φ) = − ∇ Φ + α0 ∇µ Φ ∇µ Φ − Hµνλ H µνλ
2
24
βµν (B) = −
A consistent background of string theory must preserve Weyl invariance, which now
requires βµν (G) = βµν (B) = β(Φ) = 0.
7.3 The Low-Energy Effective Action
The equations βµν (G) = βµν (B) = β(Φ) = 0 can be viewed as the equations of motion
for the background in which the string propagates. We now change our perspective: we
7
The relationship between the beta function and Einstein’s equations was first shown by Friedan
in his 1980 PhD thesis. A readable account of the full beta functions can be found in the paper by
Callan, Friedan, Martinec and Perry “Strings in Background Fields”, Nucl. Phys. B262 (1985) 593.
The full calculational details can be found in TASI lecture notes by Callan and Thorlacius which can
be downloaded from the course webpage.
– 167 –
look for a D = 26 dimensional spacetime action which reproduces these beta-function
equations as the equations of motion. This is the low-energy effective action of the
bosonic string,
Z
√
1
1
−2Φ
26
µνλ
µ
R − Hµνλ H
S = 2 d X −G e
+ 4∂µ Φ ∂ Φ
(7.16)
2κ0
12
where we have taken the liberty of Wick rotating back to Minkowski space for this
expression. Here the overall constant involving κ0 is not fixed by the field equations
but can be determined by coupling these equations to a suitable source as described,
for example, in 7.4.2. On dimensional grounds alone, it scales as κ20 ∼ ls24 where α0 = ls2 .
Varying the action with respect to the three fields can be shown to yield the beta
functions thus,
Z
√
1
δS = 2 0 d26 X −G e−2Φ (δGµν β µν (G) − δBµν β µν (B)
2κ0 α
1 µν
λ
−(2δΦ + G δGµν )(β λ (G) − 4β(Φ))
2
Equation (7.16) governs the low-energy dynamics of the spacetime fields. The caveat
“low-energy” refers to the fact that we only worked with the one-loop beta functions
which requires large spacetime curvature.
Something rather remarkable has happened here. We started, long ago, by looking
at how a single string moves in flat space. Yet, on grounds of consistency alone, we’re
led to the action (7.16) governing how spacetime and other fields fluctuate in D = 26
dimensions. It feels like the tail just wagged the dog. That tiny string is seriously highmaintenance: its requirements are so stringent that they govern the way the whole
universe moves.
You may also have noticed that we now have two different methods to compute the
scattering of gravitons in string theory. The first is in terms of scattering amplitudes
that we discussed in Section 6. The second is by looking at the dynamics encoded in
the low-energy effective action (7.16). Consistency requires that these two approaches
agree. They do.
7.3.1 String Frame and Einstein Frame
The action (7.16) isn’t quite of the familiar Einstein-Hilbert form because of that
strange factor of e−2Φ that’s sitting out front. This factor simply reflects the fact
that the action has been computed at tree level in string perturbation theory and, as
we saw in Section 6, such terms typically scale as 1/gs2 .
– 168 –
It’s also worth pointing out that the kinetic terms for Φ in (7.16) seem to have
the wrong sign. However, it’s not clear that we should be worried about this because,
again, the factor of e−2Φ sits out front meaning that the kinetic terms are not canonically
normalized anyway.
To put the action in more familiar form, we can make a field redefinition. Firstly,
it’s useful to distinguish between the constant part of the dilaton, Φ0 , and the part
that varies which we call Φ̃. We defined the constant part in (7.13); it is related to the
string coupling constant. The varying part is simply given by
Φ̃ = Φ − Φ0
(7.17)
In D dimensions, we define a new metric G̃µν as a combination of the old metric and
the dilaton,
G̃µν (X) = e−4Φ̃/(D−2) Gµν (X)
(7.18)
Note that this isn’t to be thought of as a coordinate transformation or symmetry of
the action. It’s merely a relabeling, a mixing-up, of the fields in the theory. We could
make such redefinitions in any field theory. Typically, we choose not to because the
fields already have canonical kinetic terms. The point of the transformation (7.18) is
to get the fields in (7.16) to have canonical kinetic terms as well.
The new metric (7.18) is related to the old by a conformal rescaling. One can check
that two metrics related by a general conformal transformation G̃µν = e2ω Gµν , have
Ricci scalars related by
R̃ = e−2ω R − 2(D − 1)∇2 ω − (D − 2)(D − 1)∂µ ω ∂ µ ω
(We used a particular version of this earlier in the course when considering D = 2
conformal transformations). With the choice ω = −2Φ̃/(D−2) in (7.18), and restricting
back to D = 26, the action (7.16) becomes
Z
p
1
1 −Φ̃/3
1
26
µνλ
µ
S = 2 d X −G̃ R̃ − e
Hµνλ H
− ∂µ Φ̃∂ Φ̃
(7.19)
2κ
12
6
The kinetic terms for Φ̃ are now canonical and come with the right sign. Notice that
there is no potential term for the dilaton and therefore nothing that dynamically sets
its expectation value in the bosonic string. However, there do exist backgrounds of
the superstring in which a potential for the dilaton develops, fixing the string coupling
constant.
– 169 –
The gravitational part of the action takes the standard Einstein-Hilbert form. The
gravitational coupling is given by
κ2 = κ20 e2Φ0 ∼ ls24 gs2
(7.20)
The coefficient in front of Einstein-Hilbert term is usually identified with Newton’s
constant
8πGN = κ2
Note, however, that this is Newton’s constant in D = 26 dimensions: it will differ from
Newton’s constant measured in a four-dimensional world. From Newton’s constant, we
define the D = 26 Planck length 8πGN = lp24 and Planck mass Mp = lp−1 . (With the
factor of 8π sitting there, this is usually called the reduced Planck mass). Comparing
to (7.20), we see that weak string coupling, gs 1, provides a parameteric separation
between the Planck scale and the string scale,
gs 1
⇒
lp ls
Often the mysteries of gravitational physics are associated with the length scale lp . We
understand string theory best when gs 1 where much of stringy physics occurs at
ls lp and can be disentangled from strong coupling effects in gravity.
The original metric Gµν is usually called the string metric or sigma-model metric. It
is the metric that strings see, as reflected in the action (7.1). In contrast, G̃µν is called
the Einstein metric. Of course, the two actions (7.16) and (7.19) describe the same
physics: we have simply chosen to package the fields in a different way in each. The
choice of metric — Gµν or G̃µν — is usually referred to as a choice of frame: string
frame, or Einstein frame.
The possibility of defining two metrics really arises because we have a massless scalar
field Φ in the game. Whenever such a field exists, there’s nothing to stop us measuring
distances in different ways by including Φ in our ruler. Said another way, massless
scalar fields give rise to long range attractive forces which can mix with gravitational
forces and violate the principle of equivalence. Ultimately, if we want to connect to
Nature, we need to find a way to make Φ massive. Such mechanisms exist in the context
of the superstring.
7.3.2 Corrections to Einstein’s Equations
Now that we know how Einstein’s equations arise from string theory, we can start to
try to understand new physics. For example, what are the quantum corrections to
Einstein’s equations?
– 170 –
On general grounds, we expect these corrections to kick in√ when the curvature rc
of spacetime becomes comparable to the string length scale α0 . But that dovetails
very nicely with the discussion above where we saw that the perturbative expansion
parameter for the non-linear sigma model is α0 /rc2 . Computing the next loop correction
to the beta function will result in corrections to Einstein’s equations!
If we ignore H and Φ , the 2-loop sigma-model beta function can be easily computed
and results in the α0 correction to Einstein’s equations:
1
βµν = α0 Rµν + α0 2 Rµλρσ Rνλρσ + . . . = 0
2
Such two loop corrections also appear in the heterotic superstring. However, they are
absent for the type II string theories, with the first corrections appearing at 4-loops
from the perspective of the sigma-model.
String Loop Corrections
Perturbative string theory has an α0 expansion and gs expansion. We still have to
discuss the latter. Here an interesting subtlety arises. The sigma-model beta functions
arise from regulating the UV divergences of the worldsheet. Yet the gs expansion cares
only about the topology of the string. How can the UV divergences care about the
global nature of the worldsheet. Or, equivalently, how can the higher-loop corrections
to the beta-functions give anything interesting?
The resolution to this puzzle is to remember that, when computing higher gs corrections, we have to integrate over the moduli space of Riemann surfaces. But this moduli
space will include some tricky points where the Riemann surface degenerates. (For
example, one cycle of the torus may pinch off). At these points, the UV divergences
suddenly do care about global topology and this results in the gs corrections to the
low-energy effective action.
7.3.3 Nodding Once More to the Superstring
In section 2.5, we described the massless bosonic content for the four superstring theories: Heterotic SO(32), Heterotic E8 × E8 , Type IIA and Type IIB. Each of them
contains the fields Gµν , Bµν and Φ that appear in the bosonic string, together with a
collection of further massless fields. For each, the low-energy effective action describes
the dynamics of these fields in D = 10 dimensional spacetime. It naturally splits up
into three pieces,
Ssuperstring = S1 + S2 + Sfermi
– 171 –
Here Sfermi describes the interactions of the spacetime fermions. We won’t describe
these here. But we will briefly describe the low-energy bosonic action S1 + S2 for each
of these four superstring theories.
S1 is essentially the same for all theories and is given by the action we found for the
bosonic string in string frame (7.16). We’ll start to use form notation and denote Hµνλ
simply as H3 , where the subscript tells us the degree of the form. Then the action
reads
Z
√
1
1
10
−2Φ
2
µ
S1 = 2 d X −G e
R − |H̃3 | + 4∂µ Φ ∂ Φ
(7.21)
2κ0
2
There is one small difference, which is that the field H̃3 that appears here for the
heterotic string is not quite the same as the original H3 ; we’ll explain this further
shortly.
The second part of the action, S2 , describes the dynamics of the extra fields which
are specific to each different theory. We’ll now go through the four theories in turn,
explaining S2 in each case.
• Type IIA: For this theory, H̃3 appearing in (7.21) is H3 = dB2 , just as we saw
in the bosonic string. In Section 2.5, we described the extra bosonic fields of the
Type IIA theory: they consist of a 1-form C1 and a 3-form C3 . The dynamics of
these fields is governed by the so-called Ramond-Ramond part of the action and
is written in form notation as,
Z
h√
i
1
10
2
2
S2 = − 2 d X
−G |F2 | + |F̃4 | + B2 ∧ F4 ∧ F4
4κ0
Here the field strengths are given by F2 = dC1 and F4 = dC3 , while the object
that appears in the kinetic terms is F̃4 = F4 − C1 ∧ H3 . Notice that the final term
in the action does not depend on the metric: it is referred to as a Chern-Simons
term.
• Type IIB: Again, H̃3 ≡ H3 . The extra bosonic fields are now a scalar C0 , a
2-form C2 and a 4-form C4 . Their action is given by
Z
√
1
1
10
2
2
2
S2 = − 2 d X
−G |F1 | + |F̃3 | + |F̃5 | + C4 ∧ H3 ∧ F3
4κ0
2
where F1 = dC0 , F3 = dC2 and F5 = dC4 . Once again, the kinetic terms involve
more complicated combinations of the forms: they are F̃3 = F3 − C0 ∧ H3 and
– 172 –
F̃5 = F5 − 12 C2 ∧ H3 + 12 B2 ∧ F3 . However, for type IIB string theory, there is one
extra requirement on these fields that cannot be implemented in any simple way
in terms of a Lagrangian: F̃5 must be self-dual
F̃5 = ? F̃5
Strictly speaking, one should say that the low-energy dynamics of type IIB theory
is governed by the equations of motion that we get from the action, supplemented
with this self-duality requirement.
• Heterotic: Both heterotic theories have just one further massless bosonic ingredient: a non-Abelian gauge field strength F2 , with gauge group SO(32) or E8 ×E8 .
The dynamics of this field is simply the Yang-Mills action in ten dimensions,
Z
√
α0
S2 = 2 d10 X −G Tr |F2 |2
8κ0
The one remaining subtlety is to explain what H̃3 means in (7.21): it is defined
as H̃3 = dB2 − α0 ω3 /4 where ω3 is the Chern-Simons three form constructed from
the non-Abelian gauge field A1
2
ω3 = Tr A1 ∧ dA1 + A1 ∧ A1 ∧ A1
3
The presence of this strange looking combination of forms sitting in the kinetic
terms is tied up with one of the most intricate and interesting aspects of the
heterotic string, known as anomaly cancelation.
The actions that we have written down here probably look a little arbitrary. But they
have very important properties. In particular, the full action Ssuperstring of each of the
Type II theories is invariant under N = 2 spacetime supersymmetry. (That means 32
supercharges). They are the unique actions with this property. Similarly, the heterotic
superstring actions are invariant under N = 1 supersymmetry and, crucially, do not
suffer from anomalies. The second book by Polchinski is a good place to start learning
more about these ideas.
7.4 Some Simple Solutions
The spacetime equations of motion,
βµν (G) = βµν (B) = β(Φ) = 0
have many solutions. This is part of the story of vacuum selection in string theory.
What solution, if any, describes the world we see around us? Do we expect this putative
– 173 –
solution to have other special properties, or is it just a random choice from the many
possibilities? The answer is that we don’t really know, but there is currently no known
principle which uniquely selects a solution which looks like our world — with the gauge
groups, matter content and values of fundamental constants that we observe — from
the many other possibilities. Of course, these questions should really be asked in the
context of the superstring where a greater understanding of various non-perturbative
effects such as D-branes and fluxes leads to an even greater array of possible solutions.
Here we won’t discuss these problems. Instead, we’ll just discuss a few simple solutions that are well known. The first plays a role when trying to make contact with the
real world, while the value of the others lies mostly in trying to better understand the
structure of string theory.
7.4.1 Compactifications
We’ve seen that the bosonic string likes to live in D = 26 dimensions. But we don’t.
Or, more precisely, we only observe three macroscopically large spatial dimensions.
How do we reconcile these statements?
Since string theory is a theory of gravity, there’s nothing to stop extra dimensions of
the universe from curling up. Indeed, under certain circumstances, this may be required
dynamically. Here we exhibit some simple solutions of the low-energy effective action
which have this property. We set Hµνρ = 0 and Φ to a constant. Then we are simply
searching for Ricci flat backgrounds obeying Rµν = 0. There are solutions where the
metric is a direct product of metrics on the space
R1,3 × X
(7.22)
where X is a compact 22-dimensional Ricci-flat manifold.
The simplest such manifold is just X = T22 , the torus endowed with a flat metric. But there are a whole host of other possibilities. Compact, complex manifolds
that admit such Ricci-flat metrics are called Calabi-Yau manifolds. (Strictly speaking,
Calabi-Yau manifolds are complex manifolds with vanishing first Chern class. Yau’s
theorem guarantees the existence of a unique Ricci flat metric on these spaces).
The idea that there may be extra, compact directions in the universe was considered
long before string theory and goes by the name of Kaluza-Klein compactification. If the
characteristic length scale L of the space X is small enough then the presence of these
extra dimensions would not have been observed in experiment. The standard model
of particle physics has been accurately tested to energies of a TeV or so, meaning that
– 174 –
if the standard model particles can roam around X, then the length scale must be
L . (TeV)−1 ∼ 10−16 cm.
However, one can cook up scenarios in which the standard model is stuck somewhere
in these extra dimensions (for example, it may be localized on a D-brane). Under
these circumstances, the constraints become much weaker because we would rely on
gravitational experiments to detect extra dimensions. Present bounds require only
L . 10−5 cm.
Consider the Einstein-Hilbert term in the low-energy effective action. If we are
interested only in the dynamics of the 4d metric on R1,3 , this is given by
Z
Z
p
p
1
Vol(X)
26
4
SEH = 2 d X −G̃ R̃ =
−G4d R4d
d
X
2κ
2κ2
(There are various moduli of the internal manifold X that are being neglected here).
From this equation, we learn that effective 4d Newton constant is given in terms of 26d
Newton constant by,
8πG4d
N =
κ2
Vol(X)
p
(4d)
Rewriting this in terms of the 4d Planck scale, we have lp ∼ gs ls12 / Vol(X). To
trust this whole analysis, we require gs 1 and all length scales of the internal space
(4d)
to be bigger than ls . This ensures that lp < ls . Although the 4d Planck length
(4d)
is ludicrously small, lp ∼ 10−33 cm, it may be that we don’t have to probe to this
distance to uncover UV gravitational physics. The back-of-the-envelope calculation
above shows that the string scale ls could be much larger, enhanced by the volume of
extra dimensions.
7.4.2 The String Itself
We’ve seen that quantizing small loops of string gives rise to the graviton and Bµν field.
Yet, from the sigma model action (7.12), we also know that the string is charged under
the Bµν . Moreover, the string has tension, which ensures that it also acts as a source
for the metric Gµν . So what does the back-reaction of the string look like? Or, said
another way: what is the sigma-model describing a string moving in the background
of another string?
Consider an infinite, static, straight string stretched in the X 1 direction. We can solve
for the background fields by coupling the equations of motion to a delta-function string
– 175 –
source. This is the same kind of calculation that we’re used to in electromagnetism.
The resulting spacetime fields are given by
P
2
ds2 = f (r)−1 (−dt2 + dX12 ) + 25
i=2 dXi
B = (f (r)−1 − 1) dt ∧ dX1
e2Φ = f (r)−1
(7.23)
P
2
The function f (r) depends only on the transverse direction r2 = 25
i=2 Xi and is given
by
,
gs2 N ls22
r22
Here N is some constant which we will shortly demonstrate counts the number√
of strings
which source the background. The string length scale in the solutions is ls = α0 . The
function f (r) has the property that it is harmonic in the space transverse to the string,
meaning that it satisfies ∇2R24 f (r) = 0 except at r = 0.
f (r) = 1 +
Let’s compute the B-field charge of this solution. We do exactly what we do in
electromagnetism: we integrate the total flux through a sphere which surrounds the
object. The string lies along the X 1 direction so the transverse space is R24 . We can
consider a sphere S23 at the boundary of this transverse space. We should be integrating
the flux over this sphere. But what is the expression for the flux?
To see what we should do, let’s look at the action for Hµνρ in the presence of a string
source. We will use form notation since this is much cleaner and refer to Hµνρ simply
as H3 . Schematically, the action takes the form
Z
Z
Z
1
1
?
H3 ∧ H3 +
B2 = 2
H3 ∧ ? H3 + gs2 B2 ∧ δ(ω)
gs2 R26
g
2
26
R
s R
Here δ(ω) is a delta-function source with support on the 2d worldsheet of the string.
The equation of motion is
d? H3 ∼ gs2 δ(ω)
From this we learn that to compute the charge of a single string we need to integrate
Z
1
?
H3 = 1
gs2 S23
After these general comments, we now return to our solution (7.23). The above discussion was schematic and no attention was paid to factors of 2 and π. Keeping in this
spirit, the flux of the solution (7.23) can be checked to be
Z
1
?
H3 = N
2
gs S23
– 176 –
This is telling us that the solution (7.23) describes the background sourced by N
coincident, parallel fundamental strings. Another way to check this is to compute
the ADM mass per unit length of the solution: it is N T ∼ N/α0 as expected.
Note as far as the low-energy effective action is concerned, there is nothing that
insists N ∈ Z. This is analogous to the statement that nothing in classical Maxwell
theory requires e to be quantized. However, in string theory, as in QED, we know the
underlying sources of the microscopic theory and N must indeed take integer values.
Finally, notice that as r → 0, the solution becomes singular. It is not to be trusted
in this regime where higher order α0 corrections become important.
7.4.3 Magnetic Branes
We’ve already seen that string theory is not just a theory of strings; there are also
D-branes, defined as surfaces on which strings can end. We’ll have much more to say
about D-branes in Section 7.5. Here, we will consider a third kind of object that exists
in string theory. It is again a brane – meaning that it is extended in some number
of spacetime directions — but it is not a D-brane because the open string cannot end
there. In these lectures we will call it the magnetic brane.
Electric and Magnetic Charges
You’re probably not used to talking about magnetically charged objects in electromagnetism. Indeed, in undergraduate courses we usually don’t get much further than
pointing out that ∇ · B = 0 does not allow point-like magnetic charges. However, in
the context of quantum field theory, much of the interesting behaviour often boils down
to understanding how magnetic charges behave. And the same is true of string theory.
Because this may be unfamiliar, let’s take a minute to discuss the basics.
In electromagnetism in d = 3 + 1 dimensions, we measure electric charge q by inte~ over a sphere S2 that surrounds the particle,
grating the electric field E
Z
Z
?
~
~
q=
E · dS =
F2
(7.24)
S2
S2
In the second equality we have introduced the notation of differential forms that we
also used in the previous example to discuss the string solutions.
Suppose now that a particle carries magnetic charge g. This can be measured by
~ over the same sphere. This means
integrating the magnetic field B
Z
Z
~
~
g=
B · dS =
F2
(7.25)
S2
S2
– 177 –
In d = 3+1 dimensions, both electrically and magnetically charged objects are particles.
But this is not always true in any dimension! The reason that it holds in 4d is because
both the field strength F2 and the dual field strength ? F2 are 2-forms. Clearly, this is
rather special to four dimensions.
In general, suppose that we have a p-brane that is electrically charged under a suitable
gauge field. As we discussed in Section 7.2.1, a (p + 1)-dimensional object naturally
couples to a (p + 1)-form gauge potential Cp+1 through,
Z
µ
Cp+1
W
where µ is the charge of the object, while W is the worldvolume of the brane. The
(p + 1)-form gauge potential has a (p + 2)-form field strength
Gp+2 = dCp+1
To measure the electric charge of the p-brane, we need to integrate the field strength
over a sphere that completely surrounds the object. A p-brane in D-dimensions has a
transverse space RD−p−1 . We can integrate the flux over the sphere at infinity, which is
SD−p−2 . And, indeed, the counting works out nicely because, in D dimensions, the dual
field strength is a (D − p − 2)-form, ? Gp+2 = G̃D−p−2 , which we can happily integrate
over the sphere to find the charge sitting inside,
Z
?
Gp+2
q=
SD−p−2
This equation is the generalized version of (7.24)
Now let’s think about magnetic charges. The generalized version of (7.25) suggest
that we should compute the magnetic charge by integrating Gp+2 over a sphere Sp+2 .
What kind of object sits inside this sphere to emit the magnetic charge? Doing the
sums backwards, we see that it should be a (D − p − 4)-brane.
We can write down the coupling between the (D −p−4)-brane and the field strength.
To do so, we first need to introduce the magnetic gauge potential defined by
?
Gp+2 = G̃D−p−2 = dC̃D−p−3
(7.26)
We can then add the magnetic coupling to the worldvolume W̃ of a (D − p − 4)-brane
simply by writing
Z
µ̃
C̃D−p−3
W̃
– 178 –
where µ̃ is the magnetic charge. Note that it’s typically not possible to write down
a Lagrangian that includes both magnetically charged object and electrically charged
objects at the same time. This would need us to include both Cp+1 and C̃D−p−3 in
the Lagrangian, but these are not independent fields: they’re related by the rather
complicated differential equations (7.26).
The Magnetic Brane in Bosonic String Theory
After these generalities, let’s see what it means for the bosonic string. The fundamental
string is a 1-brane and, as we saw in Section 7.2.1, carries electric charge under the
2-form B. The appropriate object carrying magnetic charge under B is therefore a
(D − p − 4) = (26 − 1 − 4) = 21-brane.
To stress a point: neither the fundamental string, nor the magnetic 21-brane are
D-branes. They are not surfaces where strings can end. We are calling them branes
only because they are extended objects.
The magnetic 21-brane of the bosonic string can be found as a solution to the lowenergy equations of motion. The solution can be written in terms of the dual potential
B̃22 such that dB̃22 = ? dB2 . It is
!
21
X
2
2
ds2 = −dt2 +
dXi2 + h(r) dX22
+ . . . + dX25
(7.27)
i=1
−2
B̃22 = (1 − h(r) ) dt ∧ dX1 ∧ . . . ∧ dX21
e2Φ = h(r)
The function h(r) depends only on the radial direction in R4 transverse to the brane:
P
2
4
r2 = 25
i=22 Xi . It is a harmonic function in R , given by
h(r) = 1 +
N ls2
r2
The role of this function in the metric (7.27) is to warp the transverse R4 directions.
Distances get larger as you approach the brane and the origin, r = 0, is at infinite
distance.
It can be checked that the solution carried N units of magnetic charge and has
tension
T ∼
N 1
ls22 gs2
– 179 –
Let’s summarize how the tension of different objects scale in string theory. The powers
of α0 = ls2 are entirely fixed on dimensional grounds. (Recall that the tension is mass
per spatial volume, so the tension of a p-brane has [Tp ] = p + 1). More interesting is
the dependence on the string coupling gs . The tension of the fundamental string does
not depend on gs , while the magnetic brane scales as 1/gs2 . This kind of 1/g 2 behaviour
is typical of solitons in field theories. The D-branes sit between the two: their tension
scales as 1/gs . Objects with this behaviour are somewhat rarer (although not unheard
of) in field theory.
In the perturbative limit, gs → 0, both D-branes and magnetic branes are heavy.
The coupling of an object with tension T to gravity is governed by T κ2 where the gravitational coupling scales as κ ∼ gs2 (7.20). This means that in the weak coupling limit,
the gravitational backreaction of the string and D-branes can be neglected. However,
the coupling of the magnetic brane to gravity is always of order one.
The Magnetic Brane in Superstring Theory
Superstring theories also have a brane magnetically charged under B. It is a (D − p −
4) = (10 − 1 − 4) = 5-brane and is usually referred to as the NS5-brane. The solution
in the transverse R4 again takes the form (7.27).
The NS5-brane exists in both type II and heterotic string. In many ways it is
more mysterious than D-branes and its low-energy effective dynamics is still poorly
understood. It is closely related to the 5-brane of M-theory.
7.4.4 Moving Away from the Critical Dimension
The beta function equations provide a new view on the critical dimension D = 26 of
the bosonic string. To see this, let’s look more closely at the dilaton beta function
β(Φ) defined in (7.15): it takes the same form as the Weyl anomaly that we discussed
back in Section 4.4.2. This means that if we consider a string propagating in D 6= 26
then the Weyl anomaly simply arises as the leading order term in the dilaton beta
function. So let’s relax the requirement of the critical dimension. The equations of
motion arising from βµν (G) and βµν (B) are unchanged, while the dilaton beta function
equation becomes
D − 26 α0 2
α0
− ∇ Φ + α0 ∇µ Φ ∇µ Φ − Hµνλ H µνλ = 0
(7.28)
6
2
24
The low-energy effective action in string frame picks up an extra term which looks like
a run-away potential for Φ,
Z
√
1
1
2(D − 26)
D
−2Φ
µνλ
µ
S = 2 d X −G e
R − Hµνλ H
+ 4∂µ Φ ∂ Φ −
2κ0
12
3α0
β(Φ) =
– 180 –
This sounds quite exciting. Can we really get string theory living in D = 4 dimensions
so easily? Well, yes and no. Firstly, with this extra potential term, flat D-dimensional
Minkowski space no longer solves the equations of motion. This is in agreement with
the analysis in Section 2 where we showed that full Lorentz invariance was preserved
only in D = 26.
Another, technical, problem with solving the string equations of motion this way is
that we’re playing tree-level term off against a one-loop term. But if tree-level and
one-loop terms are comparable, then typically all higher loop contributions will be as
well and it is likely that we can’t trust our analysis.
The Linear Dilaton CFT
In fact, there is one simple solution to (7.28) which we can trust. It is the solution to
∂µ Φ ∂ µ Φ =
26 − D
6α0
Recall that we’re working in signature (−, +, +, . . .), meaning that Φ takes a spacelike
profile if D < 26 and a timelike profile if D > 26,
r
26 − D 1
X
D < 26
Φ=
0
r 6α
D − 26 0
X
D > 26
Φ=
6α0
This gives a dilaton which is linear in one direction. This can be compared to the study
of the path integral for non-critical strings that we saw in 5.3.2. There are two ways of
seeing the same physics.
The reason that we can trust this solution is that there is an exact CFT underlying
it which we can analyze to all orders in α0 . It’s called, for obvious reasons, the linear
dilaton CFT. Let’s now look at this in more detail.
Firstly, consider the worldsheet action associated to the dilaton coupling. For now
we’ll consider an arbitrary dilaton profile Φ(X),
Z
1
√
d2 σ g Φ(X)R(2)
Sdilaton =
(7.29)
4π
Although this term vanishes on a flat worldsheet, it nonetheless changes the stressenergy tensor Tαβ because this is defined as
∂S Tαβ = −4π
∂g αβ gαβ =δαβ
– 181 –
The variation of (7.29) is straightforward. Indeed, the term is akin to the EinsteinHilbert term in general relativity but things are simpler in 2d because, for example
Rαβ = 12 gαβ R. We have
√
√
√
δ( gg αβ Rαβ ) = gg αβ δRαβ = g ∇α vα
where
vα = ∇β δgαβ − g γδ ∇α δgγδ
Using this, the variation of the dilaton term in the action is given by
Z
1
√
δSdilaton =
d2 σ g ∇α ∇β Φ − ∇2 Φ g αβ δgαβ
4π
which, restricting to flat space gαβ = δαβ , finally gives us the stress-energy tensor of a
theory with dilaton coupling
dilaton
Tαβ
= −∂α ∂β Φ + ∂ 2 Φ δαβ
Note that this stress tensor is not traceless. This is to be expected because, as we
described above, the dilaton coupling is not Weyl invariant at tree-level. In complex
coordinates, the stress tensor is
T dilaton = −∂ 2 Φ
,
T̄ dilaton = −∂¯2 Φ
Linear Dilaton OPE
The stress tensor above holds for any dilaton profile Φ(X). Let’s now restrict to a
linear dilaton profile for a single scalar field X,
Φ = QX
where Q is some constant. We also include the standard kinetic terms for D scalar
fields, of which X is a chosen one, giving the stress tensor
1
T = − 0 : ∂X ∂X : −Q ∂ 2 X
α
It is a simple matter to compute the T T OPE using the techniques described in Section
4. We find,
2T (w)
∂T (w)
c/2
+
+
+ ...
T (z) T (w) =
4
2
(z − w)
(z − w)
z−w
where the central charge of the theory is given by
c = D + 6α0 Q2
Note that Q2 can be positive or negative depending on the whether we have a timelike
or spacelike linear dilaton. In this way, we see explicitly how a linear dilaton gradient
can absorb central charge.
– 182 –
7.4.5 The Elephant in the Room: The Tachyon
We’ve been waxing lyrical about the details of solutions to the low-energy effective
action, all the while ignoring the most important, relevant field of them all: the tachyon.
Since our vacuum is unstable, this is a little like describing all the beautiful pictures
we could paint if only that damn paintbrush would balance, unaided, on its tip.
Of course, the main reason for discussing these solutions is that they all carry directly
over to the superstring where the tachyon is absent. Nonetheless, it’s interesting to ask
what happens if the tachyon is turned on. Its vertex operator is simply
Z
√
Vtachyon ∼ d2 σ g eip·X
where p2 = 4/α0 . Piecing together a general tachyon profile V (X) from these Fourier
modes and exponentiating, results in a potential on the worldsheet of the string
Z
√
Spotential = d2 σ g α0 V (X)
This is a relevant operator for the worldsheet CFT. Whenever such a relevant operator
turns on, we should follow the RG flow to the infra-red until we land on another CFT.
The c-theorem tells us that cIR < cU V , but in string theory we always require c = 26.
The deficit, at least initially, is soaked up by the dilaton in the manner described above.
The end point of the tachyon RG flow for the bosonic string is not understood. It may
be that there is no end point and the bosonic string simply doesn’t make sense once
the tachyon is turned on. Or perhaps we haven’t yet understood the true ground state
of the bosonic string.
7.5 D-Branes Revisited: Background Gauge Fields
Understanding the constraints of conformal invariance on the closed string backgrounds
led us to Einstein’s equations and the low-energy effective action in spacetime. Now we
would like to do the same for the open string. We want to understand the restrictions
that consistency places on the dynamics of D-branes.
We saw in Section 3 that there are two types of massless modes that arise from the
quantization of an open string: scalars, corresponding to the fluctuation of the D-brane,
and a U (1) gauge field. We will ignore the scalar fluctuations for now, but will return
to them later. We focus initially on the dynamics of a gauge field Aa , a = 0, . . . , p
living on a Dp-brane
– 183 –
The first question that we ask is: how does the end of the string react to a background
gauge field? To answer this, we need to look at the vertex operator associated to the
photon. It was given in (5.9)
Z
Vphoton ∼
dτ ζa ∂ τ X a eip·X
∂M
which is Weyl invariant and primary only if p2 = 0 and pa ζa = 0. Exponentiating this
vertex operator, as described at the beginning of Section 7, gives the coupling of the
open string to a general background gauge field Aa (X),
Z
dX a
Send−point =
dτ Aa (X)
dτ
∂M
But this is a very familiar coupling — we’ve already mentioned it in (7.9). It is telling
us that the end of the string is charged under the background gauge field Aa on the
brane.
7.5.1 The Beta Function
We can now perform the same type of beta function calculation that we saw for the
closed string8 . To do this, it’s useful to first use conformal invariance to map the open
string worldsheet to the Euclidean upper-half plane as we described in Section 4.7. The
action describing an open string propagating in flat space, with its ends subject to a
background gauge field on the D-brane splits up into two pieces
S = SNeumann + SDirichlet
where SNeumann describes the fluctuations parallel to the Dp-brane and is given by
Z
Z
1
2
α a
b
SNeumann =
d σ ∂ X ∂α X δab + i
dτ Aa (X)Ẋ a
(7.30)
4πα0 M
∂M
Here a, b = 0, . . . , p. The extra factor of i arises because we are in Euclidean space.
Meanwhile, the fields transverse to the brane have Dirichlet boundary conditions and
take range I = p + 1, . . . , D − 1. Their dynamics is given by
Z
1
SDirichlet =
d2 σ ∂ α X I ∂α X J δIJ
0
4πα M
8
We’ll be fairly explicit here, but if you want to see more details then the best place to look is the
original paper by Abouelsaood, Callan, Nappi and Yost, “Open Strings in Background Gauge Fields”,
Nucl. Phys. B280 (1987) 599.
– 184 –
The action SDirichlet describes free fields and doesn’t play any role in the computation
of the beta-function. The interesting part is SNeumann which, for non-zero Aa (X), is
an interacting quantum field theory with boundary. Our task is to compute the beta
function associated to the coupling Aa (X). We use the same kind of technique that we
earlier applied to the closed string. We expand the fields X a (σ) as
√
X a (σ) = x̄a (σ) + α0 Y a (σ)
where x̄a (σ) is taken to be some fixed background which obeys the classical equations
of motion,
∂ 2 x̄a = 0
(In the analogous calculation for the closed string we chose the special case of x̄a
constant. Here we are more general). However, we also need to impose boundary
conditions for this classical solution. In the absence of the gauge field Aa , we require
Neumann boundary conditions ∂σ X a = 0 at σ = 0. However, the presence of the gauge
field changes this. Varying the full action (7.30) shows that the relevant boundary
condition is supplemented by an extra term,
∂σ x̄a + 2πα0 i F ab ∂τ x̄b = 0
at σ = 0
(7.31)
where the Fab is the field strength
Fab (X) =
∂Ab
∂Aa
−
≡ ∂a Ab − ∂b Aa
a
∂X
∂X b
The√fields Y a (σ) are the fluctuations which are taken to be small. Again, the presence
of α0 in the expansion ensures that Y a are dimensionless. Expanding the action
SNeumann (which we’ll just call S from now on) to second order in fluctuations gives,
Z
√
1
0
d2 σ ∂Y a ∂Y b δab
S[x̄ + α Y ] = S[x̄] +
4π M
Z
1
0
a b
a b ˙c
+ iα
dτ ∂a Ab Y Ẏ + ∂a ∂b Ac Y Y x̄ + . . .
2
∂M
where all expressions involving the background gauge fields are now evaluated on the
classical solution x̄. We can rearrange the boundary terms by splitting the first term
up into two halves and integrating one of these pieces by parts,
Z
Z
1
c
a b
dτ (∂a Ab )Y Ẏ =
dτ ∂a Ab Y a Ẏ b − ∂a Ab Ẏ a Y b − ∂c ∂a Ab Y a Y b x̄˙
2
– 185 –
Combining this with the second term means that we can write all interactions in terms
of the gauge invariant field strength Fab ,
Z
√
1
0
S[x̄ + α Y ] = S[x̄] +
d2 σ ∂Y a ∂Y b δab
4π M
Z
iα0
c
+
dτ Fab Y a Ẏ b + ∂b Fac Y a Y b x̄˙ + . . . (7.32)
2 ∂M
where the + . . . refer to the higher terms in the expansion which come with higher
derivatives of Fab , accompanied by powers of α0 . We can neglect them for the purposes
of computing the one-loop beta function.
The Propagator
This Lagrangian describes our interacting boundary theory to leading order. We can
now use this to compute the beta function. Firstly, we should determine where possible
divergences arise. The offending term is the last one in (7.32). This will lead to a
divergence when the fluctuation fields Y a are contracted with their propagator
hY a (z, z̄)Y b (w, w̄)i = Gab (z, z̄; w, w̄)
We should be used to these free field Green’s functions by now. The propagator satisfies
∂ ∂¯ Gab (z, z̄) = −2πδ ab δ(z, z̄)
(7.33)
in the upper half plane. But now there’s a subtlety. The Y a fields need to satisfy a
boundary condition at Im z = 0 and this should be reflected in the boundary condition
for the propagator. We discussed this briefly for Neumann boundary conditions in
Section 4.7. But we’ve also seen that the background field strength shifts the Neumann
boundary conditions to (7.31). Correspondingly, the propagator G(z, z̄; w, w̄) must now
satisfy
∂σ Gab (z, z̄; w, w̄) + 2πα0 i F ac ∂τ Gcb (z, z̄; w, w̄) = 0
at σ = 0
(7.34)
In Section 4.7, we showed how Neumann boundary conditions could be imposed by
considering an image charge in the lower half plane. A similar method works here.
We extend Gab ≡ Gab (z, z̄; w, w̄) to the entire complex plane. The solution to (7.33)
subject to (7.34) is given by
1
G = −δ ln |z − w| −
2
ab
ab
1 − 2πα0 F
1 + 2πα0 F
ab
1
ln(z − w̄) −
2
– 186 –
1 + 2πα0 F
1 − 2πα0 F
ab
ln(z̄ − w)
The Counterterm and Beta Function
Let’s now return to the interacting theory (7.32) and see what counterterm is needed
to remove the divergence. Since all interactions take place on the boundary, we should
evaluate our propagator on the boundary, which means z = z̄ and w = w̄. In this case,
all the logarithms become the same and, in the limit that z → w, gives the leading
divergence ln |z − w| → −1 . We learn that the UV divergence takes the form,
"
ab
ab
ab #
1 ab 1 1 − 2πα0 F
1 1 + 2πα0 F
2
1
− δ +
+
=−
2 1 + 2πα0 F
2 1 − 2πα0 F
1 − 4π 2 α0 2 F 2
It’s now easy to determine the necessary counterterm. We simply replace Y a Y b in the
final term with Y a Y b . This yields
i2πα0 2
−
Z
dτ ∂b Fac
∂M
1
1 − 4π 2 α0 2 F 2
ab
c
x̄˙
For the open string theory to retain conformal invariance, we need the associated beta
function to vanish. This gives us the condition on the field strength Fab : it must satisfy
the equation
ab
1
∂b Fac
=0
(7.35)
1 − 4π 2 α0 2 F 2
This is our final equation governing the equations of motion that Fab must satisfy to
provide a consistent background for open string propagation.
7.5.2 The Born-Infeld Action
Equation (7.35) probably doesn’t look too familiar! Following the path we took for the
closed string, we wish to write down an action whose equations of motion coincide with
(7.35). The relevant action was actually constructed many decades ago as a non-linear
alternative to Maxwell theory: it goes by the name of the Born-Infeld action:
Z
p
S = −Tp dp+1 ξ − det (ηab + 2πα0 Fab )
(7.36)
Here ξ are the worldvolume coordinates on the brane and Tp is the tension of the Dpbrane (which, since it multiplies the action, doesn’t affect the equations of motion).
The gauge potential is to be thought of as a function of the worldvolume coordinates:
Aa = Aa (ξ). It actually takes a little work to show that the equations of motion that
we derive from this action coincide with the vanishing of the beta function (7.35). Some
hints on how to proceed are provided on Example Sheet 4.
– 187 –
For small field strengths, Fab 1/α0 , the action (7.36) coincides with Maxwell’s
action. To see this, we need simply expand to get
Z
(2πα0 )2
ab
p+1
Fab F + . . .
S = −Tp d ξ 1 +
4
The leading order term, quadratic in field strengths, is the Maxwell action. Terms with
higher powers of Fab are suppressed by powers of α0 .
So, for small field strengths, the dynamics of the gauge field on a D-brane is governed
by Maxwell’s equations. However, as the electric and magnetic field strengths increase
and become of order 1/α0 , non-linear corrections to the dynamics kick in and are
captured by the Born-Infeld action.
The Born-Infeld action arises from the one-loop beta function. It is the exact result
for constant field strengths. If we want to understand the dynamics of gauge fields with
large gradients, ∂F , then we will have determine the higher loop contributions to the
beta function.
7.6 The DBI Action
We’ve understood that the dynamics of gauge fields on the brane is governed by the
Born-Infeld action. But what about the fluctuations of the brane itself. We looked at
this briefly in Section 3.2 and suggested, on general grounds, that the action should
take the Dirac form (3.6). It would be nice to show this directly by considering the beta
function equations for the scalar fields φI on the brane. Turning these on corresponds
to considering boundary conditions where the brane is bent. It is indeed possible to
compute something along the lines of beta-function equations and to show directly that
the fluctuations of the brane are governed by the Dirac action9 .
More generally, one could consider both the dynamics of the gauge field and the
fluctuation of the brane. This is governed by a mixture of the Dirac action and the
Born-Infeld action which is usually referred to as the DBI action,
Z
p
SDBI = −Tp dp+1 ξ − det(γab + 2πα0 Fab )
As in Section (3.2), γab is the pull-back of the the spacetime metric onto the worldvolume,
γab =
∂X µ ∂X ν
ηµν
∂ξ a ∂ξ b
9
A readable discussion of this calculation can be found in the original paper by Leigh, Dirac-BornInfeld Action from Dirichlet Sigma Model, Mod. Phys. Lett. A4: 2767 (1989).
– 188 –
The new dynamical fields in this action are the embedding coordinates X µ (ξ), with
µ = 0, . . . , D − 1. This appears to be D new degrees of freedom while we expect only
D − p − 1 transverse physical degrees of freedom. The resolution to this should be
familiar by now: the DBI action enjoys a reparameterization invariance which removes
the longitudinal fluctuations of the brane.
We can use this reparameterization invariance to work in static gauge. For an infinite,
flat Dp-brane, it is useful to set
X a = ξa
a = 0, . . . , p
so that the pull-back metric depends only on the transverse fluctuations X I ,
γab = ηab +
∂X I ∂X J
δIJ
∂ξ a ∂ξ b
If we are interested in situations with small field strengths Fab and small derivatives
∂a X, then we can expand the DBI action to leading order. We have
Z
1
1
ab
I a I
0 2
p+1
Fab F + ∂a φ ∂ φ + . . .
S = −(2πα ) Tp d ξ
4
2
where we have rescaled the positions to define the scalar fields φI = X I /2πα0 . We have
also dropped an overall constant term in the action. This is simply free Maxwell theory
coupled to free massless scalar fields φI . The higher order terms that we have dropped
are all suppressed by powers of α0 .
7.6.1 Coupling to Closed String Fields
The DBI action describes the low-energy dynamics of a Dp-brane in flat space. We
could now ask how the motion of the D-brane is affected if it moves in a background
created by closed string modes Gµν , Bµν and Φ. Rather than derive this, we’ll simply
write down the answer and then justify each term in turn. The answer is:
Z
p
SDBI = −Tp dp+1 ξ e−Φ̃ − det(γab + 2πα0 Fab + Bab )
Let’s start with the coupling to the background metric Gµν . It’s actually hidden in the
notation in this expression: it appears in the pull-back metric γab which is now given
by
γab =
∂X µ ∂X ν
Gµν
∂ξ a ∂ξ b
It should be clear that this is indeed the natural place for it to sit.
– 189 –
Next up is the dilaton. As in (7.17), we have decomposed the dilaton into a constant
piece and a varying piece: Φ = Φ0 + Φ̃. The constant piece governs the asymptotic
string coupling, gs = eΦ0 , and is implicitly sitting in front of the action because the
tension of the D-brane scales as
Tp ∼ 1/gs
This, then, explains the factor of e−Φ̃ in front of the action: it simply reunites the
varying part of the dilaton with the constant piece. Physically, it’s telling us that the
tension of the D-brane depends on the local value of the dilaton field, rather than its
asymptotic value. If the dilaton varies, the effective string coupling at a point X in
spacetime is given by gsef f = eΦ(X) = gs eΦ̃(X) . This, in turn, changes the tension of the
D-brane. It can lower its tension by moving to regions with larger gsef f .
Finally, let’s turn to the Bµν field. This is a 2-form in spacetime. The function Bab
appearing in the DBI action is the pull-back to the worldvolume
Bab =
∂X µ ∂X ν
Bµν
∂ξ a ∂ξ b
Its appearance in the DBI action is actually required on grounds of gauge invariance
alone. This can be seen by considering an open string, moving in the presence of both
a background Bµν (X) in spacetime and a background Aa (X) on the worldvolume of a
brane. The relevant terms on the string worldsheet are
Z
Z
1
2
αβ
µ
ν
dτ Aa Ẋ a
d σ ∂α X ∂β X Bµν +
4πα0 M
∂M
Under a spacetime gauge transformation
Bµν → Bµν + ∂µ Cν − ∂ν Cµ
(7.37)
the first term changes by a total derivative. This is fine for a closed string, but it
doesn’t leave the action invariant for an open string because we pick up the boundary
term. Let’s quickly look at what we get in more detail. Under the gauge transformation
(7.37), we have
Z
1
SB =
d2 σ αβ ∂α X µ ∂β X ν Bµν
4πα0 M
Z
1
dσdτ αβ ∂α X µ ∂β X ν ∂µ Cν
−→ SB +
2πα0 M
Z
1
= SB +
dσdτ αβ ∂α (∂β X ν Cν )
2πα0 M
Z
Z
1
1
ν
dτ Ẋ Cν = SB +
dτ Ẋ a Ca
= SB +
0
0
2πα ∂M
2πα ∂M
– 190 –
where, in the last line, we have replaced the sum over all directions X ν with the sum
over those directions obeying Neumann boundary conditions X a , since Ẋ I = 0 at the
end-points for any directions with Dirichlet boundary conditions.
The result of this short calculation is to see that the string action is not invariant
under (7.37). To restore this spacetime gauge invariance, this boundary contribution
must be canceled by an appropriate shift of Aa in the second term,
Aa → Aa −
1
Ca
2πα0
(7.38)
Note that this is not the usual kind of gauge transformation that we consider in electrodynamics. In particular, the field strength Fab is not invariant. Rather, the gauge
invariant combination under (7.37) and (7.38) is
Bab + 2πα0 Fab
This is the reason that this combination must appear in the DBI action. This is
also related to an important physical effect. We have already seen that the string in
spacetime is charged under Bµν . But we’ve also seen that the end of the string is
charged under the gauge field Aa on the D-brane. This means that the open string
deposits B charge on the brane, where it is converted into A charge. The fact that the
gauge invariant field strength involves a combination of both Fab and Bab is related to
this interplay of charges.
7.7 The Yang-Mills Action
Finally, let’s consider the case of N coincident D-branes. We discussed this in Section
3.3 where we showed that the massless fields on the brane could be naturally packaged
as N × N Hermitian matrices, with the element of the matrix telling us which brane
the end points terminate on. The gauge field then takes the form
(Aa )mn
with a = 0, . . . , p and m, n = 1, . . . , N . Written this way, it looks rather like a U (N )
gauge connection. Indeed, this is the correct interpretation. But how do we see this?
2
Why is the gauge field describing a U (N ) gauge symmetry rather than, say, U (1)N ?
The quickest way to see that coincident branes give rise to a U (N ) gauge symmetry
is to recall that the end point of the string is charged under the U (1) gauge field
that inhabits the brane it’s ending on. Let’s illustrate this with the simplest example.
Suppose that we have two branes. The diagonal components (Aa )11 and (Aa )22 arise
– 191 –
from strings which begin and end on the same brane. Each is a U (1) gauge field. What
about the off-diagonal terms (Aa )12 and (Aa )21 ? These come from strings stretched
between the two branes. They are again massless gauge bosons, but they are charged
under the two original U (1) symmetries; they carry charge (+1, −1) and (−1, +1)
respectively. But this is precisely the structure of a U (2) gauge theory, with the offdiagonal terms playing a role similar to W-bosons. In fact, the only way to make sense
of massless, charged spin 1 particles is through non-Abelian gauge symmetry.
So the massless excitations of N coincident branes are a U (N ) gauge field (Aa )mn ,
together with scalars (φI )mn which transform in the adjoint representation of the U (N )
gauge group. We saw in Section 3 that the diagonal components (φI )mm have the
interpretation of the transverse fluctuations of the mth brane. Can we now write down
an action describing the interactions of these fields?
In fact, there are several subtleties in writing down a non-Abelian generalization of
the DBI action and such an action is not known (if, indeed, it makes sense at all).
However, we can make progress by considering the low-energy limit, corresponding to
small field strengths. The field strength in question is now the appropriate non-Abelian
expression which, neglecting the matrix indices, reads
Fab = ∂a Ab − ∂b Aa + i[Aa , Ab ]
The low-energy action describing the dynamics of N coincident Dp-branes can be shown
to be (neglecting an overall constant term),
!
Z
X
1
1
1
S = −(2πα0 )2 Tp dp+1 ξ Tr
Fab F ab + Da φI Da φI −
[φI , φJ ]2
(7.39)
4
2
4 I6=J
We recognize the first term as the U (N ) Yang-Mills action. The coefficient in front of
the Yang-Mills action is the coupling constant 1/gY2 M . For a Dp-brane, this is given by
α0 2 Tp , or
gY2 M ∼ lsp−3 gs
The kinetic term for φI simply reflects the fact that these fields transform in the adjoint
representation of the gauge group,
Da φI = ∂a φI + i[Aa , φI ]
We won’t derive this action in these lectures: the first two terms basically follow from
gauge invariance alone. The potential term is harder to see directly: the quick ways to
derive it use T-duality or, in the case of the superstring, supersymmetry.
– 192 –
A flat, infinite Dp-brane breaks the Lorentz group of spacetime to
S(1, D − 1) → SO(1, p) × SO(D − p − 1)
(7.40)
This unbroken group descends to the worldvolume of the D-brane where it classifies all
low-energy excitations of the D-brane. The SO(1, p) is simply the Lorentz group of the
D-brane worldvolume. The SO(D − p − 1) is a global symmetry of the D-brane theory,
rotating the scalar fields φI .
The potential term in (7.39) is particularly interesting,
V =−
1X
Tr [φI , φJ ]2
4 I6=J
The potential is positive semi-definite. We can look at the fields that can be turned
on at no cost of energy, V = 0. This requires that all φI commute which means that,
after a suitable gauge transformation, they take the diagonal form,


φI = 


φI1
..
.
φIN



(7.41)
The diagonal component φIn describes the position of the nth brane in transverse space
RD−p−1 . We still need to get the dimensions right. The scalar fields have dimension
[φ] = 1. The relationship to the position in space (which we mentioned before in 3.2)
is
~n
~ n = 2πα0 φ
X
(7.42)
where we’ve swapped to vector notation to replace the I index.
The eigenvalues φIn are not quite gauge invariant: there is a residual gauge symmetry
— the Weyl group of U (N ) — which leaves φI in the form (7.41) but permutes the
entries by SN , the permutation group of N elements. But this has a very natural
interpretation: it is simply telling us that the D-branes are indistinguishable objects.
When all branes are separated, the vacuum expectation value (7.41) breaks the gauge
group from U (N ) → U (1)N . The W-bosons gain a mass MW through the Higgs
mechanism. Let’s compute this mass. We’ll consider a U (2) theory and we’ll separate
– 193 –
the two D-branes in the direction X D ≡ X. This means that we turn on a vacuum
expectation value for φD = φ, which we write as
!
φ1 0
φ=
(7.43)
0 φ2
The values of φ1 and φ2 are the positions of the first and second brane. Or, more
precisely, we need to multiply by the conversion factor 2πα0 as in (7.42) to get the
position Xm of the m = 1st , 2nd brane,
Let’s compute the mass of the W-boson from the YangMills action (7.39). It comes from the covariant derivative terms
Dφ. We expand out the gauge field as
!
W
A11
a
a
Aa =
†
22
Wa Aa
with A11 and A22 describing the two U (1) gauge fields and W
the W-boson. The mass of the W-boson comes from the [Aa , φ]
term inside the covariant derivative which, using the expectation
value (7.43), is given by
Figure 44:
1
Tr [Aa , φ]2 = −(φ2 − φ1 )2 |Wa |2
2
This gives us the mass of the W-boson: it is
2
MW
= (φ2 − φ1 )2 = T 2 |X2 − X1 |2
where T = 1/2πα0 is the tension of the string. But this has a very natural interpretation.
It is precisely the mass of a string stretched between the two D-branes as shown in the
figure above. We see that D-branes provide a natural geometric interpretation of the
Higgs mechanism using adjoint scalars.
Notice that when branes are well separated, and the strings that stretch between
them are heavy, their positions are described by the diagonal elements of the matrix
given in (7.41). However, as the branes come closer together, these stretched strings
become light and are important for the dynamics of the branes. Now the positions of
the branes should be described by the full N × N matrices, including the off-diagonal
elements. In this manner, D-branes begin to see space as something non-commutative
at short distances.
– 194 –
~ m , m = 1, . . . , N in
In general, we can consider N D-branes located at positions X
th
th
transverse space. The string stretched between the m and n brane has mass
~n − φ
~ m | = T |X
~n − X
~ m|
MW = |φ
which again coincides with the mass of the appropriate W-boson computed using (7.39).
7.7.1 D-Branes in Type II Superstring Theories
As we mentioned previously, D-branes are ingredients of the Type II superstring theories. Type IIA has Dp-branes with p even, while Type IIB is home to Dp-branes with
p odd. The D-branes have a very important property in these theories: they preserve
half the supersymmetries.
Let’s take a moment to explain what this means. We’ll start by returning to the
Lorentz group SO(1, D − 1) now, of course, with D = 10. We’ve already seen that
an infinite, flat Dp-brane is not invariant under the full Lorentz group, but only the
subgroup (7.40). If we act with either SO(1, p) or SO(D − p − 1) then the D-brane
solution remains invariant. We say that these symmetries are preserved by the solution.
However, the role of the preserved symmetries doesn’t stop there. The next step is
to consider small excitations of the D-brane. These must fit into representations of the
preserved symmetry group (7.40). This ensures that the low-energy dynamics of the Dbrane must be governed by a theory which is invariant under (7.40) and we have indeed
seen that the Lagrangian (7.39) has SO(1, p) as a Lorentz group and SO(D − p − 1)
as a global symmetry group which rotates the scalar fields.
Now let’s return to supersymmetry. The Type II string theories enjoy a lot of supersymmetry: 32 supercharges in total. The infinite, flat D-branes are invariant under half
of these; if we act with one half of the supersymmetry generators, the D-brane solutions
don’t change. Objects that have this property are often referred to as BPS states. Just
as with the Lorentz group, these unbroken symmetries descend to the worldvolume of
the D-brane. This means that the low-energy dynamics of the D-branes is described
by a theory which is itself invariant under 16 supersymmetries.
There is a unique class of theories with 16 supersymmetries and a non-Abelian gauge
field and matter in the adjoint representation. This class is known as maximally supersymmetric Yang-Mills theory and the bosonic part of the action is given by (7.39).
Supersymmetry is realized only after the addition of fermionic fields which also live on
the brane. These theories describe the low-energy dynamics of multiple D-branes.
– 195 –
As an illustrative example, consider D3-branes in the Type IIB theory. The theory
describing N D-branes is U (N ) Yang-Mills with 16 supercharges, usually referred to as
U (N ) N = 4 super-Yang-Mills. The bosonic part of the action is given by (7.39), where
there are D − p − 1 = 6 scalar fields φI in the adjoint representation of the gauge group.
These are augmented with four Weyl fermions, also in the adjoint representation.
– 196 –
8. Compactification and T-Duality
In this section, we will consider the simplest compactification of the bosonic string: a
background spacetime of the form
R1,24 × S1
(8.1)
The circle is taken to have radius R, so that the coordinate on S1 has periodicity
X 25 ≡ X 25 + 2πR
We will initially be interested in the physics at length scales R where motion on the
S1 can be ignored. Our goal is to understand what physics looks like to an observer
living in the non-compact R1,24 Minkowski space. This general idea goes by the name of
Kaluza-Klein compactification. We will view this compactification in two ways: firstly
from the perspective of the spacetime low-energy effective action and secondly from
the perspective of the string worldsheet.
8.1 The View from Spacetime
Let’s start with the low-energy effective action. Looking at length scales R means
that we will take all fields to be independent of X 25 : they are instead functions only
on the non-compact R1,24 .
Consider the metric in Einstein frame. This decomposes into three different fields
on R1,24 : a metric G̃µν , a vector Aµ and a scalar σ which we package into the D = 26
dimensional metric as
ds2 = G̃µν dX µ dX ν + e2σ dX 25 + Aµ dX µ
2
(8.2)
Here all the indices run over the non-compact directions µ, ν = 0, . . . 24 only.
The vector field Aµ is an honest gauge field, with the gauge symmetry descending from diffeomorphisms in D = 26 dimensions. To see this recall that under the
transformation δX µ = V µ (X), the metric transforms as
δGµν = ∇µ Λν + ∇ν Λµ
This means that diffeomorphisms of the compact direction, δX 25 = Λ(X µ ), turn into
gauge transformations of Aµ ,
δAµ = ∂µ Λ
– 197 –
We’d like to know how the fields Gµν , Aµ and σ interact. To determine this, we simply
insert the ansatz (8.2) into the D = 26 Einstein-Hilbert action. The D = 26 Ricci
scalar R(26) is given by
1
R(26) = R − 2e−σ ∇2 eσ − e2σ Fµν F µν
4
where R in this formula now refers to the D = 25 Ricci scalar. The action governing
the dynamics becomes
Z
Z
p
p
1
2πR
1
(26)
σ
2σ
µν
µ
26
25
S = 2 d X −G̃(26) R
=
d X −G̃ e R − e Fµν F + ∂µ σ∂ σ
2κ
2κ2
4
The dimensional reduction of Einstein gravity in D dimensions gives Einstein gravity
in D − 1 dimensions, coupled to a U (1) gauge theory and a single massless scalar. This
illustrates the original idea of Kaluza and Klein, with Maxwell theory arising naturally
from higher-dimensional gravity.
The gravitational action above is not quite of the Einstein-Hilbert form. We need to
again change frames, absorbing the scalar σ in the same manner as we absorbed the
dilaton in Section 7.3.1. Moreover, just as for the dilaton, there is no potential dictating
the vacuum expectation value of σ. Changing the vev of σ corresponds to changing
R, so this is telling us that nothing in the gravitational action fixes the radius R of
the compact circle. This is a problem common to all Kaluza-Klein compactifications10 :
there are always massless scalar fields, corresponding to the volume of the internal
space as well as other deformations. Massless scalar fields, such as the dilaton Φ or the
volume σ, are usually referred to as moduli.
If we want this type of Kaluza-Klein compactification to describe our universe —
where we don’t see massless scalar fields — we need to find a way to “fix the moduli”.
This means that we need a mechanism which gives rise to a potential for the scalar
fields, making them heavy and dynamically fixing their vacuum expectation value. Such
mechanisms exist in the context of the superstring.
Let’s now also look at the Kaluza-Klein reduction of the other fields in the low-energy
effective action. The dilaton is easy: a scalar in D dimensions reduces to a scalar in
D − 1 dimensions. The anti-symmetric 2-form has more structure: it reduces to a
2-form Bµν , together with a vector field õ = Bµ 25 .
10
The description of compactification on more general manifolds is a beautiful story involving aspects
differential geometry and topology. This story is told in the second volume of Green, Schwarz and
Witten.
– 198 –
In summary, the low-energy physics of the bosonic string in D−1 dimensions consists
of a metric Gµν , two U (1) gauge fields Aµ and õ and two massless scalars Φ and σ.
8.1.1 Moving around the Circle
In the above discussion, we assumed that all fields are independent of the periodic
direction X 25 . Let’s now look at what happens if we relax this constraint. It’s simplest
to see the resulting physics if we look at the scalar field Φ where we don’t have to worry
about cluttering equations with indices. In general, we can expand this field in Fourier
modes around the circle
µ
25
Φ(X ; X ) =
∞
X
Φn (X µ )einX
25 /R
n=−∞
where reality requires Φ?n = Φ−n . Ignoring the coupling to gravity for now, the kinetic
terms for this scalar are
Z
Z
∞ X
n2
26
µ
2
25
µ
2
d X ∂µ Φ ∂ Φ + (∂25 Φ) = 2πR d X
∂µ Φn ∂ Φ−n + 2 | Φn |
R
n=−∞
This simple Fourier decomposition is telling us something very important: a single
scalar field on R1,D−1 × S1 splits into an infinite number of scalar fields on R1,D−2 ,
indexed by the integer n. These have mass
Mn2 =
n2
R2
(8.3)
For R small, all particles are heavy except for the massless zero mode n = 0. The
heavy particles are typically called Kaluza-Klein (KK) modes and can be ignored if
we’re probing energies 1/R or, equivalently, distance scales R.
There is one further interesting property of the KK modes Φn with n 6= 0: they
are charged under the gauge field Aµ arising from the metric. The simplest way to
see this is to look at the appropriate gauge transformation which, from the spacetime
perspective, is the diffeomorphism X 25 → X 25 + Λ(X µ ). Clearly, this shifts the KK
modes
inΛ
Φn → exp
Φn
R
This tells us that the nth KK mode has charge n/R. In fact, one usually rescales the
gauge field to A0µ = Aµ /R, under which the charge of the KK mode Φn is simply n ∈ Z.
– 199 –
8.2 The View from the Worldsheet
We now consider the Kaluza-Klein reduction from the perspective of the string. We
want to study a string moving in the background R1,24 × S1 . There are two ways in
which the compact circle changes the string dynamics.
The first effect of the circle is that the spatial momentum, p, of the string in the
circle direction can no longer take any value, but is quantized in integer units
n
n∈Z
p25 =
R
The simplest way to see this is simply to require that the string wavefunction, which
includes the factor eip·X , is single valued.
The second effect is that we can allow more general boundary conditions for the mode
expansion of X. As we move around the string, we no longer need X(σ + 2π) = X(σ),
but can relax this to
X 25 (σ + 2π) = X 25 (σ) + 2πmR
m∈Z
The integer m tells us how many times the string winds around S1 . It is usually simply
called the winding number.
Let’s now follow the familiar path that we described in Section 2 to study the spectrum of the string on the spacetime (8.1). We start by considering only the periodic field
X 25 , highlighting the differences with our previous treatment. The mode expansion of
X 25 is now given by
X 25 (σ, τ ) = x25 +
α0 n
τ + mRσ + oscillator modes
R
which incorporates both the quantized momentum and the possibility of a winding
number. Before splitting X 25 (σ, τ ) into right-moving and left-moving parts, it will be
useful to introduce the quantities
pL =
n mR
+ 0
R
α
,
pR =
n
mR
− 0
R
α
Then we have X 25 (σ, τ ) = XL25 (σ + ) + XR25 (σ − ), where
r
α0 X 1 25 −inσ+
XL25 (σ + ) = 12 x25 + 12 α0 pL σ + + i
α̃ e
,
2 n6=0 n n
r
α0 X 1 25 −inσ−
XR25 (σ − ) = 12 x25 + 12 α0 pR σ − + i
α e
2 n6=0 n n
– 200 –
(8.4)
This differs from the mode expansion (1.36) only in the terms pL and pR . The mode
expansion for all the other scalar fields on flat space R1,24 remains unchanged and we
don’t write them explicitly.
Let’s think about what the spectrum of this theory looks like to an observer living in
D = 25 non-compact directions. Each particle state will be described by a momentum
pµ with µ = 0, . . . , 24. The mass of the particle is
M2 = −
24
X
p µ pµ
µ=0
As before, the mass of these particles is fixed in terms of the oscillator modes of the
string by the L0 and L̃0 equations. These now read
M 2 = p2L +
4
4
2
(
Ñ
−
1)
=
p
(N − 1)
+
R
α0
α0
where N and Ñ are the levels, defined in lightcone quantization by (2.24). (One should
take the lightcone coordinate inside R1,24 rather than along the S1 ). The factors of −1
are the necessary normal ordering coefficients that we’ve seen in several guises in this
course.
These equations differ from (2.25) by the presence of the momentum and winding
terms around S1 on the right-hand side. In particular, level matching no longer tells
us that N = Ñ , but instead
N − Ñ = nm
(8.5)
Expanding out the mass formula, we have
M2 =
n2
m2 R2
2
+
+ 0 (N + Ñ − 2)
2
0
2
R
α
α
(8.6)
The new terms in this formula have a simple interpretation. The first term tells us that
a string with n > 0 units of momentum around the circle gains a contribution to its
mass of n/R. This agrees with the result (8.3) that we found from studying the KK
reduction of the spacetime theory. The second term is even easier to understand: a
string which winds m > 0 times around the circle picks up a contribution 2πmRT to
its mass, where T = 1/2πα0 is the tension of the string.
– 201 –
8.2.1 Massless States
We now restrict attention to the massless states in R1,24 . This can be achieved in the
mass formula (8.6) by looking at states with zero momentum n = 0 and zero winding
m = 0, obeying the level matching condition N = Ñ = 1. The possibilities are
µ
ν
• α−1
α̃−1
|0; pi: Under the SO(1, 24) Lorentz group, these states decompose into a
metric Gµν , an anti-symmetric tensor Bµν and a scalar Φ.
µ
25 µ
25
α̃−1 |0; pi: These are two vector fields. We can identify the
|0; pi and α−1
• α−1
α̃−1
µ
25
25 µ
α̃−1 ) |0; pi with the vector field Aµ coming from the
sum of these (α−1 α̃−1 + α−1
µ
25 µ
25
α̃−1 ) |0; pi with the vector field õ coming
−α−1
metric and the difference (α−1 α̃−1
from the anti-symmetric field.
25 25
α̃−1 |0; pi: This is another scalar. It is identified with the scalar σ associated
• α−1
to the radius of S1 .
We see that the massless spectrum of the string coincides with the massless spectrum
associated with the Kaluza-Klein reduction of the previous section.
8.2.2 Charged Fields
One can also check that the KK modes with n 6= 0 have charge n under the gauge
field Aµ . We can determine the charge of a state under a given U (1) by computing the
3-point function in which two legs correspond to the state of interest, while the third
is the appropriate photon. We have two photons, with vertex operators given by,
Z
V± (p) ∼ d2 z ζµ (∂X µ ∂¯X̄ 25 ± ∂X 25 ∂¯X̄ µ )eip·X
where + corresponds to Aµ and − to õ and we haven’t been careful about the overall
normalization. Meanwhile, any state can be assigned momentum n and winding m by
25
25
dressing the operator with the factor eipL X (z)+ipR X̄ (z̄) . As always, it’s simplest to
work with the momentum and winding modes of the tachyon, whose vertex operators
are of the form
Z
25
25
Vm,n (p) ∼ d2 z eip·X eipL X +ipR X̄
The charge of a state is the coefficient in front of the 3-point coupling of the field and
the photon,
X
hV± (p1 )Vm,n (p2 )V−m,−n (p3 )i ∼ δ 25 (
pi ) ζµ (pµ2 − pµ3 ) (pL ± pR )
i
– 202 –
The first few factors are merely kinematical. The interesting information is in the last
factor. It is telling us that under Aµ , fields have charge pL + pR ∼ n/R. This is in
agreement with the Kaluza-Klein analysis that we saw before. However, it’s also telling
us something new: under õ , fields have charge pL − pR ∼ mR/α0 . In other words,
winding modes are charged under the gauge field that arises from the reduction of Bµν .
This is not surprising: winding modes correspond to strings wrapping the circle and
we saw in Section 7 that strings are electrically charged under Bµν .
8.2.3 Enhanced Gauge Symmetry
With a circle in the game, there are other ways to build massless states that don’t
require us to work at level N = Ñ = 1. For example, we can set N = Ñ = 0 and look
at winding modes m 6= 0. The level matching condition (8.5) requires n = 0 and the
mass of the states is
2
mR
4
2
M =
− 0
0
α
α
and states can be massless whenever the radius takes special values R2 = 4α0 /m2 with
m ∈ Z. Similarly, we can set the winding to zero m = 0 and consider the KK modes
of the tachyon which have mass
4
n2
M = 2− 0
R
α
2
which become massless when R2 = n2 α0 /4.
However, the richest spectrum of massless states occurs when the radius takes a very
special value, namely
√
R = α0
Solutions to the level matching condition (8.5) with M 2 = 0 are now given by
• N = Ñ = 1 with m = n = 0. These give the states described above: a metric,
two U (1) gauge fields and two neutral scalars.
• N = Ñ = 0 with n = ±2 and m = 0. These are KK modes of the tachyon field.
They are scalars in spacetime with charges (±2, 0) under the U (1) × U (1) gauge
symmetry.
• N = Ñ = 0 with n = 0 and m = ±2. This is a winding mode of the tachyon
field. They are scalars in spacetime with charges (0, ±2) under U (1) × U (1).
– 203 –
µ
• N = 1 and Ñ = 0 with n = m = ±1. These are two new spin 1 fields, α−1
|0; pi.
They carry charge (±1, ±1) under the two U (1) × U (1).
• N = 1 and Ñ = 0 with n = −m = ±1. These are a further two spin 1 fields,
µ
|0; pi, with charge (±1, ∓1) under U (1) × U (1).
α̃−1
How do we interpret these new massless states? Let’s firstly look at the spin 1 fields.
These are charged under U (1) × U (1). As we mentioned in Section 7.7, the only way
to make sense of charged massless spin 1 fields is in terms of a non-Abelian
√ gauge
symmetry. Looking at the charges, we see that at the critical radius R = α0 , the
theory develops an enhanced gauge symmetry
U (1) × U (1) → SU (2) × SU (2)
The massless scalars from the N = Ñ = 0 now join with the previous scalars to form
adjoint representations of this new symmetry. We move away from the critical radius
by changing the vacuum expectation value for σ. This breaks the gauge group back to
the Cartan subalgebra by the Higgs mechanism.
From the discussion above, it’s clear that this mechanism for generating non-Abelian
gauge symmetries relies on the existence of the tachyon. For this reason, this mechanism
doesn’t work in Type II superstring theories. However, it turns out that it does work
in the heterotic string, even though it has no tachyon in its spectrum.
8.3 Why Big Circles are the Same as Small Circles
The formula (8.6) has a rather remarkable property: it is invariant under the exchange
R ↔
α0
R
(8.7)
if, at the same time, we swap the quantum numbers
m ↔ n
(8.8)
This means that a string moving on a circle of radius R has the same spectrum as a
string moving on a circle of radius α0 /R. It achieves this feat by exchanging what it
means to wind with that it means to move.
As the radius of the circle becomes large, R → ∞, the winding modes become
very heavy with mass ∼ R/α0 and are irrelevant for the low-energy dynamics. But
the momentum modes become very light, M ∼ 1/R, and, in the strict limit form a
continuum. From the perspective of the energy spectrum, this continuum of energy
states is exactly what we mean by the existence of a non-compact direction in space.
– 204 –
In the other limit, R → 0, the momentum modes become heavy and can be ignored:
it takes way too much energy to get anything to move on the S1 . In contrast, the
winding modes become light and start to form a continuum. The resulting energy
spectrum looks as if another dimension of space is opening up!
The equivalence of the string spectrum on circles of radii R and α0 /R extends to the
full conformal field theory and hence to string interactions. Strings are unable to tell
the difference between circles that are very large and circles that are very small. This
striking statement has a rubbish name: it is called T-duality.
This provides another mechanism in which string theory exhibits
√ a minimum length
scale: as you shrink a circle to smaller and smaller sizes, at R = α0 , the theory acts
as if the circle is growing again, with winding modes playing the role of momentum
modes.
The New Direction in Spacetime
So how do we describe this strange new spatial direction that opens up as R → 0?
Under the exchange (8.7) and (8.8), we see that pL and pR transform as
pL → pL
,
pR → −pR
Motivated by this, we define a new scalar field,
Y 25 = XL25 (σ + ) − XR25 (σ − )
It is simple to check that in the CFT for a free, compact scalar field all OPEs of Y 25
coincide with the OPEs of X 25 . This is sufficient to ensure that all interactions defined
in the CFT are the same.
We can write the new spatial direction Y directly in terms of the old field X, without
first doing the split into left and right-moving pieces. From the definition of Y , one
can check that ∂τ X = ∂σ Y and ∂σ X = ∂τ Y . We can write this in a unified way as
∂α X = αβ ∂ β Y
(8.9)
where αβ is the antisymmetric matrix with τ σ = −στ = +1. (The minus sign from
στ in the above equation is canceled by another from the Minkowski worldsheet metric
when we lower the index on ∂ β ).
– 205 –
The Shift of the Dilaton
The dilaton, or string coupling, also transforms under T-duality. Here we won’t derive
this in detail, but just give a plausible explanation for why it’s the case. The main idea
is that a scientist living in a stringy world shouldn’t be able to do any experiments
that distinguish between a compact circle of radius R and one of radius α0 /R. But the
first place you would look is simply the low-energy effective action which, working in
Einstein frame, contains terms like
Z
p
2πR
25
d X −G̃ eσ R + . . .
24
2
2ls gs
A scientist cannot tell the difference between R and R̃ = α0 /R only if the value of the
dilaton is also ambiguous so that the term in front of the action remains invariant: i.e.
R/gs2 = R̃/g̃s2 . This means that, under T-duality, the dilaton must shift so that the
coupling constant becomes
√
α0 gs
gs → g̃s =
(8.10)
R
8.3.1 A Path Integral Derivation of T-Duality
There’s a simple way to see T-duality of the quantum theory using the path integral.
We’ll consider just a single periodic scalar field X ≡ X + 2πR on the worldsheet. It’s
useful to change normalization and write X = Rϕ, so that the field ϕ has periodicity
2π. The radius R of the circle now sits in front of the action,
Z
R2
S[ϕ] =
d2 σ ∂α ϕ ∂ α ϕ
(8.11)
4πα0
R
The Euclidean partition function for this theory is Z = Dϕ e−S[ϕ] . We will now play
around with this partition function and show that we can rewrite it in terms of new
variables that describe the T-dual circle.
The theory (8.11) has a simple shift symmetry ϕ → ϕ + λ. The first step is to make
this symmetry local by introducing a gauge field Aα on the worldsheet which transforms
as Aα → Aα − ∂α λ. We then replace the ordinary derivatives with covariant derivatives
∂α ϕ → Dα ϕ = ∂α ϕ + Aα
This changes our theory. However, we can return to the original theory by adding a
new field, θ which couples as
Z
Z
R2
i
2
α
S[ϕ, θ, A] =
d σ Dα ϕ D ϕ +
d2 σ θ αβ ∂α Aβ
(8.12)
0
4πα
2π
– 206 –
The new field θ acts as a Lagrange multiplier. Integrating out θ sets αβ ∂α Aβ = 0. If
the worldsheet is topologically R2 , then this condition ensures that Aα is pure gauge
which, in turn, means that we can pick a gauge such that Aα = 0. The quantum theory
described by (8.12) is then equivalent to that given by (8.11).
Of course, if the worldsheet is topologically R2 then we’re missing the interesting
physics associated to strings winding around ϕ. On a non-trivial worldsheet, the condition αβ ∂α Aβ = 0 does not mean that Aα is pure gauge. Instead, the gauge field
can have non-trivial holonomy around the cycles of the worldsheet. One can show that
these holonomies are gauge trivial if θ has periodicity 2π. In this case, the partition
function defined by (8.12),
Z
1
DϕDθDA e−S[ϕ,θ,A]
Z=
Vol
is equivalent to the partition function constructed from (8.11) for worldsheets of any
topology.
At this stage, we make use of a clever and ubiquitous trick: we reverse the order of
integration. We start by integrating out ϕ which we can do by simply fixing the gauge
symmetry so that ϕ = 0. The path integral then becomes
Z
Z
Z
R2
i
2
α
2
αβ
Z = DθDA exp −
d σ Aα A +
d σ (∂α θ)Aβ
4πα0
2π
where we have also taken the opportunity to integrate the last term by parts. We can
now complete the procedure and integrate out Aα . We get
!
Z
Z
R̃2
Z = Dθ exp −
d2 σ ∂α θ ∂ α θ
4πα0
we threw away
with R̃ = α0 /R the radius of the T-dual circle. In the final integration,
p
0
the overall factor in the path integral, which is proportional to α /R. A more careful
treatment shows that this gives rise to the appropriate shift in the dilaton (8.10).
8.3.2 T-Duality for Open Strings
What happens to open strings and D-branes under T-duality? Suppose firstly that we
compactify a circle in direction X transverse to the brane. This means that X has
Dirichlet boundary conditions
X = const ⇒ ∂τ X 25 = 0
– 207 –
at σ = 0, π
But what happens in the T-dual direction Y ? From the definition (8.9) we learn that
the new direction has Neumann boundary conditions,
∂σ Y = 0
at σ = 0, π
We see that T-duality exchanges Neumann and Dirichlet boundary conditions. If we
dualize a circle transverse to a Dp-brane, then it turns into a D(p + 1)-brane.
The same argument also works in reverse. We can start with a Dp-brane wrapped
around the circle direction X, so that the string has Neumann boundary conditions.
After T-duality, (8.9) changes these to Dirichlet boundary conditions and the Dp-brane
turns into a D(p − 1)-brane, localized at some point on the circle Y .
In fact, this was how D-branes were originally discovered: by following the fate of
open strings under T-duality.
8.3.3 T-Duality for Superstrings
To finish, let’s nod one final time towards the superstring. It turns out that the tendimensional superstring theories are not invariant under T-duality. Instead, they map
into each other. More precisely, Type IIA and IIB transform into each other under Tduality. This means that Type IIA string theory on a circle of radius R is equivalent to
Type IIB string theory on a circle of radius α0 /R. This dovetails with the transformation
of D-branes, since type IIA has Dp-branes with p even, while IIB has p odd. Similarly,
the two heterotic strings transform into each other under T-duality.
8.3.4 Mirror Symmetry
The essence of T-duality is that strings get confused. Their extended nature means that
they’re unable to tell the difference between big circles and small circles. We can ask
whether this confusion extends to more complicated manifolds. The answer is yes. The
fact that strings can see different manifolds as the same is known as mirror symmetry.
Mirror symmetry is cleanest to state in the context of the Type II superstring, although similar behaviour also holds for the heterotic strings. The simplest example is
when the worldsheet of the string is governed by a superconformal non-linear sigmamodel with target space given by some Calabi-Yau manifold X. The claim of mirror
symmetry is that this CFT is identical to the CFT describing the string moving on a
different Calabi-Yau manifold Y. The topology of X and Y is not the same. Their
Hodge diamonds are the mirror of each other; hence the name. The subject of mirror
symmetry is an active area of research in geometry and provides a good example of the
impact of string theory on mathematics.
– 208 –
8.4 Epilogue
We are now at the end of this introductory course on string theory. We began by trying
to make sense of the quantum theory of a relativistic string moving in flat space. It is,
admittedly, an odd place to start. But from then on we had no choices to make. The
relativistic string leads us ineluctably to conformal field theory, to higher dimensions
of spacetime, to Einstein’s theory of gravity at low-energies, to good UV behaviour
at high-energies and to Yang-Mills theories living on branes. There are few stories in
theoretical physics where such meagre input gives rise to such a rich structure.
This journey continues. There is one further ingredient that it is necessary to add:
supersymmetry. Even this is in some sense not a choice, but is necessary to remove the
troublesome tachyon that plagued these lectures. From there we may again blindly follow where the string leads, through anomalies (and the lack thereof) in ten dimensions,
to dualities and M-theory in eleven dimensions, to mirror symmetry and moduli stabilization and black hole entropy counting and holography and the miraculous AdS/CFT
correspondence.
However, the journey is far from complete. There is much about string theory that
remains to be understood. This is true both of the mathematical structure of the theory
and of its relationship to the world that we observe. The problems that we alluded to
in Section 6.4.5 are real. Non-perturbative completions of string theory are only known
in spacetimes which are asymptotically anti-de Sitter, but cosmological observations
suggest that our home is not among these. In attempts to make contact with the
standard models of particle physics and cosmology, we typically return to the old idea
of Kaluza-Klein compactifications. Is this the right approach? Or are we missing some
important and subtle conceptual ingredient? Or is the existence of this remarkable
mathematical structure called string theory merely a red-herring that has nothing to
do with the real world?
In the years immediately after its birth, no one knew that string theory was a theory
of strings. It seems very possible that we’re currently in a similar situation. When the
theory is better understood, it may have little to do with strings. We are certainly still
some way from answering the simple question: what is string theory really?
– 209 –
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement