ON FACTS IN

SUPERSTRING THEORY1
Oswaldo Zapata Marı́n

Abstract: Despite the lack of experimental confirmation and of unambiguous theoretical

proof, superstring theory has long been considered by many the only consistent quantized
theory of gravity and the unique viable framework for the unification of all fundamental
forces of nature. In the first part of this essay I explore the type of reasoning used to
support such statements. In order to illustrate the argument, in the second part I focus
on one of the most acclaimed achievements of the theory: the AdS/CFT correspondence.
Finally, I conclude by observing that what constitutes a result in superstring theory involves
more than purely theoretical arguments. Specifically, the acceptance of facts in superstring
theory is inextricably linked to the large group of people that make it possible, whether they
are string practitioners or not.

1 The evolving scientific status of string the-

ory results
Trying to overcome the impasse with what a massless particle with spin two
meant for the dual model of strong nuclear interactions, in 1974 Joël Scherk
and John Schwarz proposed a reinterpretation of this particle as the quantum
carrier of gravitational force: “The possibility of describing particles other than
hadrons (leptons, photons, gauge bosons, gravitons, etc.) by a dual model is
explored. The Virasoro-Shapiro model is studied first, interpreting the mass-
less spin-two state of the model as a graviton.”2 In their seminal paper Scherk
and Schwarz showed that consistency of the dual model entailed a higher di-
mensional version of the Hilbert action. From this it then followed that the
model included gravitational forces as described by Einstein’s equation. These
were the primary motives driving the authors to propose that string theory
quantized gravity: the spectrum showed a massless spin-2 particle, and, more-
over, a ten-dimensional Einstein equation could be derived. In those days
many theoretical physicists found these two results too weak to allege that
a quantized theory of gravity had been achieved. This explains the cautious
reception the theory received in its early years. Unexpectedly, however, this
1
A detailed report on the sociological process here discussed is avail-
able at Spinning the Superweb: Essays on the history of superstring theory
(www.spinningthesuperweb.blogspot.com). See section III and IV of the first essay
“On Facts in Superstring Theory.” These last two sections are not included in this arXiv
version!
2
J. Scherk and J. H. Schwarz, “Dual Models for Nonhadrons,” Nuclear Physics B 81
(1974): 118-144.
1. The evolving scientific status of string theory results 2

once feeble proposal has become widely established within theoretical physics,
even though the mathematical support has remained almost unchanged for
more than thirty years. Let us comment further on this.
In the last chapter of a classic string theory graduate textbook written in
1989, here is how the quantization of gravity is presented:
String theory is claimed to be a unifying framework for the de-
scription of all particles and their interactions, including gravity.
However, up to now our exposition of the subject was rather for-
mal and it is not at all transparent how it can be relevant for low
energy phenomenology. The only hint we got so far was from look-
ing at the spectrum. There especially the occurrence of a spin two
tensor particle indicated that gravity might be contained in string
theory.3
Note that this observation is relegated to the last part of the book, after
some three hundred pages of mathematical details. No doubt this is a queer
situation. Why leave the most important argument in favour of the theory, at
least in popular accounts and undergraduate level materials, to the final pages
of the textbook? The answer to this question is provided by the cautious words
of the authors. Even more surprising is that the same argument has been used
for decades. The only difference is that nowadays the quantization of gravity
is not considered an “elusive task,” as many string theoreticians used to say,
but rather an accomplished one. For example, in the midst of what string
theorists consider the second major revolution of the field, Edward Witten
wrote in Physics Today: “Moreover, these theories have (or this one theory
has) the remarkable property of predicting gravity – that is, of requiring the
existence of a massless spin-2 particle whose coupling at long distances are
those of general relativity.”4 (Italics in the original.) And in an up to date
textbook, aimed at undergraduate physics students, the author says: “The
striking quantum emergence of gravitation in string theory has the full flavor
of a prediction.”5 Understandably, declarations of this kind have given rise to
hot discussions among supporters and detractors of the theory. The question
is: what happened in those intervening years? Why are string theorists so
optimistic now? Did they really find an unquestionable proof, experimental
or theoretical, that their theory quantizes gravity?
In point of fact, to this day nobody has presented an entirely convincing
proof of this. For some theoretical physicists, the presence of a massless spin-2
particle in its spectrum is not enough to a quantized theory of gravity. I is also
argued that the low energy limit analysis of superstring theory does not imply
that that particle is the graviton. The former string theorist Daniel Friedan,
one of the early major contributors, is emphatic about this:
3
D. Lüst and S. Theisen, Lectures on String Theory (Berlin: Springer-Verlag, 1989).
4
E. Witten, “Reflections on the Fate of Spacetime,” Physics Today, April, 1996.
5
B. Zwiebach, A First Course in String Theory (Cambridge: Cambridge University
Press, 2004).
1. The evolving scientific status of string theory results 3

In particular, there is no justification for the claim that string the-

ory explains or predicts gravity. String theory gives perturbative
scattering amplitudes of gravitons. Gravitons have never been ob-
served. Gravity in the real world is accurately described by general
relativity, which is a classical field theory. There is no derivation
of general relativity from string theory. ... String theory does not
produce any mechanical theory of gravity, much less a quantum
mechanical theory.6

Why then do practitioners believe that “string theory is a quantum theory,

and, because it includes gravitation, it is a quantum theory of gravity”? How
have string theorists arrived at the conclusion that “the harmonious union
of general relativity and quantum mechanics is a major success”7 of super-
strings? Moreover, how have they managed to convince other theoreticians of
the validity of their explanations? Let us look at another example.
In the abovementioned paper, Scherk and Schwarz also declared that string
theory could unify all the fundamental interactions: “If it is, a scheme of this
sort might provide a unified theory of weak, electromagnetic, and gravitational
interactions.”8 (These known interactions were then complemented with the
strong nuclear force, the latter successfully described by quantum chromody-
namics.) This was in 1974. Years later, and after intense work, the proposal
had still not been proved. The four-dimensional standard model (it exludes
gravity), with all its details, could not be deduced from string theory. In a
lecture given at the International Centre for Theoretical Physics (ICTP) in
1986, one of the leading string phenomenologists of the time stated: “Be-
ing defined in d=10, some compactification of the six dimensions would be
required to make contact with phenomenology. This process is at the mo-
ment not understood at all; one has to make crude approximations and then
check for consistency a posteriori.”9 The reason why the process of superstring
compactification “was not understood at all” is due to the stringent constrains
that supersymmetry imposes on the four dimensional model. As expected from
the standard model of particle physics, any physical result with supersymme-
try must be renormalizable10 and requires the existence of chiral spinors.11
However, there are some difficulties with this: firstly, supergravity12 with one
6
D. Friedan, “A Tentative Theory of Large Distance Physics,” [arXiv: hep-ph/0204131].
7
B. Greene, The Elegant Universe: Superstrings, Hidden Dimensions, and the Quest for
the Ultimate Theory (New York: W. W. Norton & Co., 1999).
8
J. Scherk and J. H. Schwarz, ”Dual Models for Nonhadrons,” op. cit.
9
H. P. Nilles, “Supergravity And The Low-Energy Limit Of Superstring Theories,”
lectures given at the 1986 ICTP Spring School on Supersymmetry and Supergravity, in B.
de Wit et al. (eds.), Supersymmetry, Supergravity and Superstrings 1986 (Singapore: World
Scientific, 1986).
10
In broad terms, this means that the theory cannot contain quantities which take on an
infinite value.
11
Chirality is an essential property of the particles constituting matter: leptons and
quarks.
12
Supergravity is the theory known to be the low energy limit of superstring theory.
1. The evolving scientific status of string theory results 4

supersymmetry is not renormalizable, and, secondly, models with higher num-

bers of supersymmetries do not include chiral spinors. Satisfying these two
conditions is the difficult mission assigned to string phenomenologists. There
are several approaches to the problem: the simplest model considers com-
pactification on a multi-dimensional torus, other string theorists prefer to use
constructs known as orbifolds or orientifolds; more recently G2 manifolds were
tested. Other Calabi-Yau manifolds are currently under examination. Despite
this confusing situation, there is one thing string theorists know they must
answer: “Why do we live on this particular string vacuum or SSC [superstring
compactification]?”13 This is the most urgent question that needs to be ad-
dressed in order to make contact with physical reality. As Michio Kaku writes
in the introduction to the 2000 edition of his textbook on elementary string
theory: “The search for the true vacuum of string theory is therefore the cen-
tral theme of this book.”14 So, if we could explain why the universe chose this
particular vacuum we would be able to understand how the standard model
arises from superstring theory and why the universe expands as it does. It has
been argued by critics that this reformulation of the problem does not solve
it. On the contrary, it makes it harder and moves it, dangerously, towards the
realm of theology.15
In this case, as in the previous case of the quantization of gravity, su-
perstring theorists have been unable to offer an accurate and comprehensive
explanation of four-dimensional physics. To be sure, string phenomenology,
from the old heterotic string to recent brane models, does not provide the cor-
rect value for the quantities associated to the elementary particles known so
far. In addition to this, critics emphasize, it does not answer crucial questions
that intrigue particle physicists: how is the electroweak symmetry broken?
what fixes the masses of the Higgs boson, quarks, neutrinos and charged lep-
tons? what are the sources of the cold dark matter? what produced the big
bang? why is there matter-antimatter asymmetry? This state of affairs has
lead Sheldom Glashow, an eminent particle physicist, to declare that string
13
F. Quevedo, “Lectures on Superstring Phenomenology,” [arXiv: hep-th/9603074].
14
“Thus, the real problem facing us, in our opinion, is to theoretically settle the following
question as quickly as possible: what is the vacuum (ground state) of superstring theory?
Since the ground state should correspond to our physical universe, if the true vacuum could
be discovered, we might be able to decisively settle whether superstring theory is a theory
of the universe or just the latest in a series of failed efforts to discover the Holy Grail
of physics, the unified field theory.” M. Kaku, Strings, Conformal Fields, and M-Theory
(Berlin: Springer-Verlag, 2000), second edition.
15
By the year 2005 there was a vigorous debate within theoretical physics concerning
the role of anthropic arguments in determining the real vacuum of the universe. Leonard
Susskind’s The Cosmic Landscape triggered this passionate debate. The attacks on string
theory during that period were unprecedented. For example, Lawrence Krauss, a reputed as-
trophysicist, wrote that “it is perhaps not too surprising that when one approaches the limits
of our knowledge, theologians and scientists alike tend to appeal to new hidden universes for,
respectively, either redemption or understanding.” L. Krauss, “Science and Religion Share
Fascination in Things Unseen,” The New York Times, 8 November 2005. In summary, as
Paul Feyerabend would put it, are they scientists or priests?
1. The evolving scientific status of string theory results 5

theory “has failed in its primary goal, which is to incorporate what we already
know into a consistent theory that explains gravity as well. The new theory
must incorporate the old theory and say something more. String theory has
not succeeded in this fashion.”16
From the previous examples we have learnt some important things about
the development of string theory. Firstly, as research progresses in a given
topic, an explicit reference to the unsolved problem tends to disappear from
the literature. For instance, we saw how the quantization of gravity is con-
sidered by string theorists to be an accomplished task that does not deserve
further study, or even a mention. Secondly, while research advances, the initial
problem changes in such a way that it becomes increasingly difficult to un-
ravel the convoluted relationship connecting the final problem to the original
one. This was illustrated by our second example concerning string theory and
the unification of the forces. Originally the idea was to extract the standard
model from superstring theory, an investigation encouraged during the second
half of the eighties by the promising results obtained from the heterotic string.
Then, by the mid-nineties, the goal was to determine the unique vacuum of the
mother of all the theories, the M-Theory. And, more recently, the focus was on
the right “environment” of the anthropic solution. Things have changed, but
the fundamental query remains unsolved: how do we get the standard model
from string theory? With these examples we have learnt something else: this
occurs while an “outward” discourse (from the “inside” to the “outside” of the
professional community) proclaims that the theory has solved such problems.
Indeed, in this movement disadvantages have been transmuted into virtues.
In spite of these fundamental flaws in the theory, enthusiasts proclaim
that “in string theory all forces are truly unified in a deep and significant
way,” or, a bit more prudently, “string theory leads in a remarkably simple
way to a reasonable rough draft of particle physics unified with gravity.”17
The final outcome of this discourse is the same: the stabilization of string
theory as a quantized theory of gravity and unified model. Before concluding
this introduction, I would like to add two more quotations. In Zwiebach’s
undergraduate textbook he asks:

Why is string theory truly a unified theory? The reason is simple

and goes to the heart of the theory. In string theory, each parti-
cle is identified as a particular vibrational mode of an elementary
microscopic string.18

Thus, string theory is a unified theory thanks to its extreme reductionist

approach. Brian Greene, in his best-selling book, backs up this statement.
Years of hard work have shown that the reductionist approach to string theory
is correct.
16
Sheldom Glashow interview; on the website of The Elegant Universe,
http://www.pbs.org/wgbh/nova/elegant/view-glashow.html
17
E. Witten, “Unravelling String Theory,” Nature 438, December, 2005.
18
B. Zwiebach, A First Course in String Theory, op. cit.
2. A case study: the AdS/CFT correspondence 6

These works showed conclusively that numerous features of the

standard model – features that had been painstakingly discovered
over the course of decades of research – emerged naturally and
simply from the grand structure of string theory.19

I think what we have seen in these examples is a characteristic of string

theory research and its elaboration of physical reality. At first, a hypothesis is
made, explaining openly its significance as well as its difficulties. At this stage
no one is sure of the real value of the conjecture, however, it is interesting
enough to drive a significant part of the physics community to devote itself to
its development. Step by step “evidence” accumulates and after a while the
string theory fact emerges. String theorists have created in this way their own
nature: a supersymmetric world, a big bang with all the fundamental forces
combined, a multi-dimensional universe, and so forth. Although I have pro-
vided support to this thesis in the case of the two sanctioned “achievements”
of superstring theory, quantization of gravity and unification of the fundamen-
tal forces, I will now illustrate in full detail this complex process with another
ground-breaking proposal: the AdS/CFT correspondence.

2 A case study: the AdS/CFT correspondence

At the end of 1997, Juan Maldacena, at that time a young researcher at Har-
vard University, proposed what some physicists consider to be one of the main
breakthroughs in the history of string theory and even of theoretical physics.
Approaching the physics of black holes with the powerful mathematical tools of
superstring theory, he conjectured the existence of a deep relationship between
pure non-gravitational theories and superstring theories.20 Even though the
proposal was not well understood by everybody, it was welcomed and enjoyed
rapid acceptance within the community.

This subject has developed with breathtaking speed: Maldacena’s

paper appeared in November 1997, yet by the Strings 98 conference
19
B. Greene, The Elegant Universe, op. cit.
20
J. M. Maldacena, “The Large N Limit of Superconformal Field Theory and Super-
gravity,” [arXiv: hep-th/9711200]. To put it simply, the correspondence says, as it is cur-
rently understood, that string theory defined in a negatively curved anti-de Sitter space
(AdS) is equivalent to a certain conformal field theory (CFT) living on its boundary. One
concrete example is AdS5 /CFT4 . It states that type IIB superstring theory in AdS5 is
equivalently described by an extended N =4 super-CFT in four dimensions. The other five
dimensions of the ten-dimensional space defined by superstring theory are compactified on a
five-dimensional sphere, S5 . The five-sphere with isometry group SO(6) is chosen in order to
match with the SU(4) R-symmetry of the super Yang-Mills theory defined on the boundary.
More complicated models are also possible. Note that there are two different N’s involved
in our discussion: N , in italics, for the number of supersymmetries of the conformal theory,
and N, not italicized, for the dimension of its gauge group. For a short introduction to string
theory and the AdS/CFT correspondence see, for example, O. Zapata, “String Theory: A
Theory of Unification,” [arXiv: hep-th/0612004].
2. A case study: the AdS/CFT correspondence 7

seven months later, more than half the invited speakers chose to
speak on this subject.21

Today, Maldacena’s publication is one of the most well-known papers ever

written in theoretical high energy physics.
The first papers submitted to the electronic preprint library citing Mal-
dacena’s conjecture did not delve deeply into the original proposal; they sim-
ply mentioned it in a superficial way. In these papers we find the following
assertions: “It would be interesting to understand the relation between our
arguments and those of [reference to Maldacena’s paper],” and “Maybe an ar-
gument along the lines of [reference to Maldacena’s paper] can be carried out
here as well.” Things changed dramatically when Edward Witten published a
paper formalizing many of the original ideas put forward by Maldacena.22 He
discovered a precise correspondence (here the term “correspondence” is used
for the first time) between string states on the ten-dimensional spacetime,
dubbed the bulk, and operators of the particle physics-like model. He also
computed some scattering processes. To this end Witten identified the bound-
ary of the ten-dimensional bulk with the space where the non-gravitational
particles reside and interact. After this essential contribution, more and more
people started to work on this correspondence.
Here is how Witten referred to Maldacena’s proposal (first lines of the
abstract):

Recently, it has been proposed by Maldacena that large N limits of

certain conformal field theories in d dimensions can be described
in terms of supergravity (and string theory) on the product of d
+ 1-dimensional AdS space with a compact manifold. Here we
elaborate on this idea and propose a precise correspondence be-
tween conformal field theory observables and those of supergrav-
ity.23 (Italics added.)

Note Witten’s prudence when referring to it: “It has been proposed.” A
year and a half later Maldacena’s “conjecture” was still a conjecture, that is,
nothing exceptional demanded that its scientific status should be upgraded.
This was the state of affairs in 1999 when a group of leading string theorists,
including Maldacena himself, published a review article on the subject. This
report comprises more than two hundred and fifty pages and is still considered
one of the most complete accounts on the subject.
21
J. H. Schwarz, “Introduction to M Theory and AdS/CFT Duality,” [arXiv: hep-
th/9812037].
22
Another important paper deserving special analysis, what I will do in a next version
of this essay, is S.S. Gubser, I. R. Klebanov and A. M. Polyakov, “Gauge theory correlators
from noncritical string theory,” [arXiv: hep-th/9802109]. I thank Igor Klebanov for pointing
me out this omission.
23
E. Witten, “Anti de Sitter Space and Holography,” [arXiv: hep-th/9802150].
2. A case study: the AdS/CFT correspondence 8

So, we conclude that N =4 U(N) Yang-Mills theory could be the

same as ten dimensional superstring theory on AdS5 x S5 [reference
to Maldacena’s paper]. Here we have presented a very heuristic
argument for this equivalence; later we will be more precise and
give more evidence for this correspondence.24 (Italics added.)

In the authors’ opinion the correspondence was still in the phase of gath-
ering evidence. It was not yet established as scientific fact. At this point it is
worth digressing a moment in order to say a few words about the physics of
the correspondence. This will help us to understand its successive evolution
towards a higher degree of truthfulness.
AdS/CFT in its strongest version states that superstring theory in the
bulk corresponds to a full quantum non-gravitational theory on the boundary
of such volume. But, so far support for it has been provided only in the su-
pergravity approximation: the point-like model where the length of the string
is equal to zero (α0 → 0). In addition, since these computations are also very
difficult to carry out, the classical limit is necessary. In this last approximation
quantum corrections are discarded (only tree diagrams are considered). The
theory is said to be weakly coupled.
The consistency of the theory relies on every operation being done on
the gravitational theory having a counterpart on the boundary. (Due to this
relationship between what happens in the bulk and on its boundary, the corre-
spondence has been called holographic.) Thus, the supergravity and classical
limits must have their corresponding procedure in the boundary theory. Ex-
perts have found that this relationship is of the type weak ←→ strong. In short,
this means that the easier the computations in the gravitational theory, as in
the limits above, the harder it is to find the corresponding non-gravitational
results on the conformal field theory side. In turn, when the string calculations
are difficult, the boundary computations are easier to perform. This explains
why the AdS/CFT correspondence is also often called “AdS/CFT duality.”
After these brief observations, we are now ready to evaluate the follow-
ing extract from MAGOO (as the report by Maldacena and collaborators is
sometimes labelled):

One might wonder why the above argument was not a proof rather
than a conjecture. It is not a proof because we did not treat the
string theory non-perturbatively (not even non-perturbatively in
α0 → 0). We could also consider different forms of the conjecture.
In its weakest form the gravity description would be valid for large
g s N [supergravity description with no quantum corrections as ex-
plained above], but the full string theory on AdS might not agree
with the field theory. A not so weak form would say that the con-
jecture is valid even for finite g s N, but only in the N → ∞ limit (so
24
O . Aharony, S. S. Gubser, J. M. Maldacena, H. Ooguri and Y. Oz, “Large N Field
Theories, String Theory and Gravity,” [arXiv: hep-th/9905111].
2. A case study: the AdS/CFT correspondence 9

that the α0 corrections would agree with the field theory, but the
gs corrections may not). The strong form of the conjecture, which
is the most interesting one and which we will assume here, is that
the two theories are exactly the same for all values of gs and N.25
(Italics added.)

This passage from MAGOO suggests that in those days many string theo-
rists were not fully convinced of the validity of the correspondence; although
something like 1500 papers had already been published on the subject (MA-
GOO includes 757 references in its bibliography). Despite such wide interest
and some important contributions to theoretical physics, the general opinion
was that the correspondence was in the process of being proved. In their
“Summary and Discussion” the authors concluded saying:

To summarize, the past 18 months have seen much progress in our

understanding of string/M theory compactifications on AdS and
related spaces, and in our understanding of large N field theories.
However, the correspondence is still far from realizing the hopes
that it initially raised, and much work still remains to be done.26
(Italics added.)

So, by May 1999 string theory experts were convinced that the “simple and
powerful observation”27 made by Maldacena was in its infancy. At the same
time, an optimistic vision was transmitted to young researchers by means of
courses and written materials. In a widely used introductory review written by
Jens Petersen, which appeared three months earlier than MAGOO, we read:

The Maldacena conjecture [reference to Maldacena’s paper] is a

conjecture concerning string theory or M theory on certain back-
grounds of the form AdSd MD−d . ... The conjecture asserts that
the quantum string- or M-theory on this background is mathemat-
ically equivalent – or dual as the word goes – to an ordinary but
conformally invariant quantum field theory in a space-time of di-
mension d-1, which in fact has the interpretation of “the boundary”
of AdSd .28 (Italics in the original.)

Similarly, in the written translation of a couple of lectures delivered during

the spring of 1998 at The Abdus Salam International Centre for Theoretical
Physics, two leading theoretical physicists wrote: “Assuming this conjecture,
one can derive results for the large ’t Hooft coupling limit of gauge theory, by
25
Ibid.
26
Ibid.
27
I. Klebanov, “TASI Lectures: Introduction to the AdS/CFT Correspondence,” [arXiv:
hep-th/0009139].
28
J. L. Petersen, “Introduction to the Maldacena Conjecture on AdS/CFT,” [arXiv:
hep-th: 9902131].
2. A case study: the AdS/CFT correspondence 10

doing computations in AdS supergravity.”29 And concluded, in the last lines,

by saying: “Nevertheless, now that we have a precise and better motivated con-
jecture for the appropriate string in this case, we can hope that progress along
these lines will be made in the near future.” Analysis of written publications
at that time shows that the veracity of the holographic correspondence was
understood by string theoreticians more as a hope rather than as a completed
task.
The lines of development for the following three years, from May 1999 to
February 2002, were foreseen, and in a certain sense dictated, by Maldacena
and collaborators: in chapter 5 they summarized the main results of BTZ black
holes and showed how this was related to the boundary theory; in chapter 6,
the final one, they focused on QCD-like theories.
Thanks to the great amount of available results on the physics of three-
dimensional black holes and a manageable two-dimensional conformal field
theory, the AdS3 /CFT2 model was for many years the favourite setting for
analyzing the correspondence beyond the supergravity limit.

In this paper we study the spectrum of critical bosonic string theory

on AdS3 x M with NS-NS backgrounds, where M is a compact
space. Understanding string theory on AdS3 is interesting from the
point of view of the AdS/CFT correspondence since it enables us
to study the correspondence beyond the gravity approximation.30

Juan Maldacena and Hirosi Ooguri continued working on this framework

for some time, arriving at several remarkable results. Unfortunately, the cor-
respondence was not demonstrated beyond the supergravity approximation
as first expected. The other main line of research within AdS/CFT was the
construction and description of viable QCD-like theories by means of weak
gravitational processes.

A fruitful extension of the basic AdS/CFT correspondence [refer-

ence to Maldacena] stems from studying branes at conical singu-
larities [references]. Consider, for instance, a stack of D3-branes
placed at the apex of a Ricci-flat 6-d cone Y6 whose base is a 5-d
Einstein manifold X5 .31

Even though the two approaches were different, both were trying to provide
evidence for the stronger versions of the correspondence. The AdS3 /CFT2 ef-
fort wanted to prove the exact correspondence in a special case (classical limit
29
M. Douglas and S. Randjbar-Daemi, “Two Lectures on AdS/CFT Correspondence,”
[arXiv: hep-th/9902022].
30
J. Maldacena and H. Ooguri, “Strings on AdS3 and the SL(2,R) WZW Model,” [arXiv:
hep-th/0001053].
31
I. R. Klebanov and M. J. Strassler, “Supergravity and a Confining Gauge Theory:
Duality Cascades and χSB-Resolution of Naked Singularities,” [arXiv: hep-th/0007191].
2. A case study: the AdS/CFT correspondence 11

with α0 6= 0), and the AdS/QCD attempt tried to find plausible phenomeno-
logical results. However, by the end of 2001, after four years of intense work
and more than two thousand citations to Maldacena’s original paper, the cor-
respondence was still waiting for a definitive proof.
The Berenstein-Maldacena-Nastase (BMN) conjecture was proposed in Febru-
ary 2002 and it rapidly seized the attention of string theorists working on
AdS/CFT. This fervent interest on BMN was reflected by the large number
of publications that followed. In the month the paper appeared, a fifth of the
publications on AdS/CFT was on BMN or at least mentioned it in the main
text. The following month, articles on BMN grabbed half the attention of the
research on AdS/CFT. A few months later, up four fifths (June and Septem-
ber 2002) of the citations to Maldacena’s 1997 proposal came from the novel
BMN conjecture. This rough count clearly shows that the new conjecture was
an essential breakthrough within the field. Moreover, as we will see next, it
represented the end of a period and the beginning of a new one. After BMN,
the truthfulness of the AdS/CFT correspondence changed: it was nearing a
scientific fact.
In this new conjecture Maldacena and collaborators envisaged an alterna-
tive setting to verify the correctness of the AdS/CFT correspondence beyond
the supergravity limit. The idea was to concentrate on a very special case
of the original formulation and see how the standard correspondence between
string states and operators matched within this new framework. On the bulk
side of the correspondence the spacetime background was changed to paral-
lel plane waves. The new condition, pp-waves, was obtained by taking the
Penrose limit of the anti-de Sitter space. In the conformal field theory this
corresponded to a truncation of the number of operators. It was believed that
this new model could shed light on the full quantum correspondence.
It is interesting to see how Berenstein, Maldacena and Nastase referred to
the AdS/CFT correspondence, the basis of their new proposal:

The fact that large N gauge theories have a string theory descrip-
tion was believed for a long time. These strings live in more than
four dimensions. One of the surprising aspects of AdS/CFT corre-
spondence is the fact that for N = 4 super Yang Mills these strings
move in ten dimensions and are the usual string of type IIB string
theory.32 (Italics added.)

These are the first lines of the paper. Such a presentation suggests that
the relationship between gravity and particle physics is a matter of “fact.”
They take it for granted. Obviously, there is something paradoxical in all this:
what is expected to be proven is at the same time considered true knowledge.
But, was this an isolated judgement or rather a belief shared by other string
theorists?
32
D. Berenstein, J. M. Maldacena and H. Nastase, “Strings in Flat Space and PP Waves
from N = 4 Super Yang Mills,” [arXiv: hep-th/0202021].
2. A case study: the AdS/CFT correspondence 12

Two months after the BMN proposal, Steven Gubser, Igor Klebanov and
Alexander Polyakov, collaborating then at Princeton, submitted a paper where
it is said:

It was found in [reference to Maldacena, Witten, and a previous

article by GKP], developing some earlier findings of [reference to
Polyakov], that the desired string theory in this case lives in the
space AdS5 x S5 and that there is a unique prescription relating
physical quantities in the string and gauge pictures. Many more
complicated examples have been analyzed since then, confirming
the existence of a dual string picture for various gauge theories.33
(Italics added.)

Though the authors confess in the next lines that the correspondence has
only been “tested” “mostly in the supergravity limit,” as BMN they also pre-
suppose the full validity of the correspondence. Notice the use of the terms “it
was found” and “confirming the existence.” The same predisposition is shown
in another important paper written by a group of researchers from MIT and
Harvard:

More recently the Maldacena conjecture has established a duality

between a conformal gauge theory (with a fixed line of couplings)
and string theories on an AdS background. However these duali-
ties are well understood only at large values of the gauge coupling
[supergravity limit in the bulk].34 (Italics added.)

A widespread trait among publications following the BMN proposal is that

the few lines making explicit reference to the AdS/CFT correspondence are
often in the abstract or in the first paragraphs. For instance, the well known
article by Joseph Minahan and Konstantin Zarembo begins with a very short
discussion on AdS/CFT results and limitations. After the six-line review of
AdS/CFT, they move on to the main subject of the paper: BMN.35 The func-
tion of this brief reference to AdS/CFT in the opening to the paper is simply
to contextualize the article; a context that everybody must be familiar with,
and accept, in order to proceed further. Post-BMN doctoral dissertations show
a similar pattern: the correspondence is assumed and chapters once intended
to explain it are systematically dropped. A short section or even several ci-
tations now replaced the detailed summary. Another confirmation that the
correspondence was entering a new state regarding its factuality is that some
33
S. S. Gubser, I. R. Klebanov and A. M. Polyakov, “A Semi-Classical Limit of the
Gauge/String Correspondence,” [arXiv: hep-th/0204051].
34
N. R. Constable, D. Z. Freedman, M. Headrick, S. Minwalla, L. Motl, A. Postnikov
and W. Skiba, “PP-Wave String Interactions from Perturbative Yang-Mills Theory,” [arXiv:
hep-th/0205089].
35
J. A. Minahan and K. Zarembo, “The Bethe-Ansatz for N =4 super Yang-Mills,” [arXiv:
hep-th/0212208].
2. A case study: the AdS/CFT correspondence 13

authors did not even consider relevant the citation of Maldacena’s original
paper. From the eight most important papers36 on AdS/CFT published after
BMN, only four of them cited it.
What followed in the next years was a confirmation of the previous anal-
ysis. As a sample, let us consider the nine most cited articles on AdS/CFT
during that period.37 Three of the papers deal with phenomenological issues
and concentrate on the implications of the AdS/QCD duality; that is, the pos-
sibility of using the holographic correspondence to obtain precious information
on strongly coupled particle physics processes. As stated in one of these pa-
pers: “Recently, the gravity/gauge, or anti-de Sitter/conformal field theory
(AdS/CFT) correspondence [reference to Maldacena] has revived the hope
that QCD can be reformulated as a solvable string theory.”38 Another three
articles focus on a different spacetime background for the correspondence, the
Lunin-Maldacena background. This include one written by Oleg Lunin and
Juan Maldacena. The other two are by Sergey Frolov and collaborators: “A
relative simplicity of the Lunin-Maldacena supergravity background and the
N =1 superconformal theory makes the conjectured duality a new promising
arena for studying the AdS/CFT correspondence.”39 In contrast to the articles
on AdS/QCD, these last three are not phenomenologically motivated; rather
they try to prove the correspondence beyond a constraining condition called
the BPS limit. It is interesting to notice that Lunin and Maldacena called
the new proposal “conjecture duality,” while the original AdS/CFT proposal
is simply called “correspondence.” This subtlety differentiation suggests that
the latter is in a higher, better consolidated, factual level. Another of the nine
papers on AdS/CFT concerns integrable models, a subject seeking a solution
to superstring theory on non-trivial backgrounds with RR-fluxes. There are
two more papers. One proposes a sort of AdS/CFT correspondence for Sasaki-
Einstein backgrounds, and the other is about flux compactifications. Strictly
speaking, the last article is not about the correspondence; it simply acknowl-
edges the important contribution of the latter to the renewal of the studies on
flux compactifications. And it does it in a single line.
Here concludes our short story of the AdS/CFT correspondence. In it we
saw how string theorists treated the conjecture when it was proposed for the
first time; how they changed their view in the course of time; and how they
communicated it to younger members of the community. We discovered that
the AdS/CFT conjecture became a fact at the same time as most of the talks
36
I am considering papers with more than 150 citations on SPIRES (accessed in December
2006).
37
Since it is impossible for a single person to check the hundreds of papers written in
that period, January 2005 – December 2006, my discussion only contemplates the nine most
cited papers (each with more than 50 citations). This segment of the spectrum will show
the main stream of research.
38
J. Erlich, E. Katz, D. T. Son and M. A. Stephanov, “QCD and a Holographic Model
of Hadrons,” [arXiv: hep-ph/0501128].
39
S. Frolov, “Lax Pair for Strings in Lunin-Maldacena Background,” [arXiv: hep-
th/0503201].
2. A case study: the AdS/CFT correspondence 14

and papers changed the “recently Maldacena conjectured that ...” to “as the
AdS/CFT correspondence teaches us ...” and, finally, to the more impersonal
“as the AdS/CFT establishes.” We saw how the sentence “Maldacena has
recently conjectured that ...” transformed into a single number that pointed
to the original paper. Nonetheless this was not imposed, as some interpreters
would be incline to declare, by a “great leader,” nor by the “will of power” of an
authoritarian group of researchers, nor by mere convention. Instead, it is the
end result of several years of long, hard, and exhausting work. I have sustained
that the breaking point was the new “bold” conjecture of BMN, a hypothesis
that assumed implicitly the correctness of the old AdS/CFT correspondence.
After years spent accumulating “evidence,” but without a definitive proof in
sight, there was the desire and need within the community to surmount the
old correspondence. The research could safely continue only by protecting
Maldacena’s conjecture from profanation, namely, elevating it to the factual
level of the more authentic mathematical demonstrations. In another context,
the historian of science Steven Shapin wrote: “It was necessary to speak con-
fidently of matters of fact because, as the foundations of proper philosophy,
they required protection. And it was proper to speak confidently of matters
of fact, because they were not of one’s own making; they were, in the empiri-
cist model, discovered rather than invented.”40 To shield the correspondence
from attacks was a necessity for the whole community of practitioners. Con-
sequently, more and more discussions on the correspondence were transferred
from research papers and PhD theses to graduate and even undergraduate
courses. This was the final step towards its final entrance into public lectures
and popular science books. Today, the AdS/CFT correspondence pervades
the public debate on superstring theory.

40
S. Shapin, “Pump and Circumstance: Robert Boyle’s Literary Technology,” Social
Studies of Science 14, No.4 (Nov., 1984): 481-520.