arXiv:physics/0603112 v3 28 Mar 2006

String theory and the crisis in particle physics

Bert Schroer
CBPF, Rua Dr. Xavier Sigaud 150
22290-180 Rio de Janeiro, Brazil
and Institut fuer Theoretische Physik der FU Berlin, Germany
December 2005
Abstract
In the first section the history of string theory starting from its Smatrix bootstrap predecessor up to Susskinds recent book is critically
reviewed. The aim is to understand its amazing popularity which starkly
constrasts its fleeting physical content. A partial answer can be obtained
from the hegemonic ideological stance which some of its defenders use
to present and defend it. The second section presents many arguments
showing that the main tenet of string theory which culminated in the
phrase that it represents the only game in town is untenable. It is
based on a wrong view about QFT being a mature theory which (apart
from some missing details) already reached its closure.

An anthology of the crisis in fundamental physics

There can be no doubt that after almost a century of impressive success fundamental physics is in the midst of a deep crisis. Although the epicenter of the
quake which shakes its foundations is located in particle theory, its repercussions are not only beginning to derail experimental particle physics and to affect
adjacent areas of fundamental research, but they are also leading to a weird
philosophical fallout. One does not have to be a physicist in order to notice
that there is something strange going on if reputable people [1] are advocating
the abdication of the kind of physics, which since the time of Galilei, Newton,
Einstein and the protagonists of quantum theory has been a de-mystification
of nature by rational mathematically formulated concepts with experimentally
verifiable consequences. Instead of cogito ergo sum the new maxim is: I exist
and therefore my existence tells me what universe of the multiverses of solutions
of string theory I have to chose to understand my physical world. With other
words I throw my own existence into the theory as a kind of boundary condition,
or to put it more bluntly, I am integrating myself into string theory in order to
make it a work.

One is used to view the exact sciences in particular particle physics as a human activity which follows an intrinsic autonomous path. The resulting knowledge may have an impact on philosophy, the humanities and the development
of society, but their feedback into fundamental research in physics is generally
believed to be negligible. One of the points of this essay is to raise the question
whether such a view can be maintained in a globalized world. Here I am not
thinking primarily of the loss of moral standards in the increasing number of
falsified or plagiarized scientific results. As long as the natural sciences maintain the primate of verifiability this cannot derail its content. The danger of
derailment of particle physics is rather emanating from the hegemonic tendencies of monocultures which are pursued by globalized groups within science with
an ever increasing unwillingness and inability to subject their results to critical
scrutiny. These developments have been looming in string theory for a long
time and finally found their explicit outing in a recent manifesto [1]. Before I
return to the question to what extend string theory and its new outing reflects
the unpleasant side of the present Zeitgeist, it is helpful to review some relevant
episodes in the history of particle physics.
The beginning of relativistic particle theory is inexorably linked to Jordans
Quantelung der Wellenfelder which merged with Diracs multi-particle formalism. QFT according to Jordan draws on a parallelism to classical fields
which amounted to a quantization of the classical canonical structure of the
Lagrangian formalism. Jordans point of view about this was more radical than
Diracs because any structure which fitted into the classical field formalism was
a potential candidate for quantization, it was not necessary that it belonged to
real observed classical physics1 as initially required by Dirac. In the light of
this it is remarkable that two years after his discovery Jordan was dissatisfied
with the use of what he called the classical crutches (klassische Kruecken)
of quantization and pleaded for a more intrinsic access to the physical content
of QFT [3]. It is perhaps not accidental that this happened exactly at a time
when he realized together with Pauli that the pointlike nature of quantum fields
required by quantization are (unlike quantum mechanical observables) necessarily rather unwieldy singular objects which could lead to trouble in building up
interactions by their local couplings. But Jordans dream of an intrinsic and
more autonomous approach remained unfulfilled for a long time to come, the
history of QFT between the early 1930s and the late 40s is characterized by
apparently incurable ultraviolet problems. This ultraviolet crisis led to many
speculative and revolutionary ideas (some of them quite weird, at least in
retrospect). As a result of the Tomonaga-Schwinger-Feynman-Dyson renormalized perturbation theory and its immediate physical success in QED, these ideas
went into the dustbin of history.
Despite a formidable technical and conceptual enrichment, renormalization
was extremely conservative in terms of physical principles; it was not necessary
to introduce a single physical principle beyond the ones which were already
1 In [2] it is mentioned that by the time the Goettinger group learned about Schroedingers
results, Jordan already had a field quantized version.

(indirectly or directly) in place in the pre-renormalization setting. But without renormalization theory and its successful application to QED, quantum field
theory probably would not have been able to survive in the pantheon of physics.
However Jordans plea for intrinsicness was not laid to rest; rather perturbative
renormalization elevated it to a higher level. Instead of the pre-renormalization
ultraviolet problem there was the new problem which consisted in the realization that the perturbative division into non-renormalizable and renormalizable
models was rather technical; it was based on a formal quantization procedure
which required the implementation of interactions by the recipe of coupling
pointlike free fields rather than by letting the principles which underlie QFT
determine an intrinsic way. Hence it is not possible to decide if and how this
formal division into renormalizable and nonrenormalizable models is related to
the genuine intrinsic frontier between well- and ill- defined theories. What complicates the situation is the fact that one could not even decide whether the
reasonably looking renormalizable family really corresponds to models which
exist in any mathematical sense so that one at least would know that the principles allow for nontrivial realizations. The problem that particle physics most
successful theory of QED is also its mathematically most fragile has not gone
away. These problems become even worse in theories as string theory which in
the eyes of string protagonists are supposed to supersede QFT. In this case one
has in addition to the existence problem the conceptual difficulty of not having
been able to extract characterizing principles from ad hoc recipes.
Even though renormalization theory (despite its crucial role in arriving at
a new ultraviolet-controlled computational setting) did not really add a new
physical principle, its covariant formulation was instrumental for a profound
understanding of the particle-field relation. It replaced the relativistic quantum mechanical setting (the basis of the older textbooks of Wenzel and Heitler)
which turned out to be unsuitable for this task2 . In the new setting it was possible to derive a large-time asymptotic relation which illuminated the connection
between particles and fields without invoking perturbation theory; the resulting
LSZ scattering theory was shown to be a direct consequence of the principles
[4]. The upshot of this conceptual progress was that asymptotic particle states
and the scattering matrix had the status of truly intrinsic quantities, whereas
quantum fields (apart from symmetry-generating currents) loose the aspect of
individuality which they enjoyed in classical physics before quantization; in fact
there are infinitely many ways of choosing field coordinatizations and they can
be grouped into local equivalence classes of which every (interpolating) member field is describing the same particles whose charge numbers and spacetime
representation invariants (mass, spin) are related to the class (i.e. the particle
state is uniquely related to the class and not to the field). The deepening of the
field-particle relation led also to the revival of an idea by Heisenberg (which in
its original form was discussed at some conferences shortly before the start of
2 The quantum mechanical view of the particle-field relation became problematic after Furry
and Oppenheimer observed that the presence of interactions made it impossible to obtain a
space-time separation of one field states A(x) |0i into a one particle states and their vacuum
polarization companions.

the second world war) to explore the feasibility of a direct S-matrix construction without passing through a field theory setting; in this way one expected
to obtain a non-singular and more intrinsic way to pursue particle physics. For
lack of a clear implementing strategy letting alone concrete results the interest
waned and only reappeared in the 1950s and 1960s in connection with some nonperturbative conceptual advances in the wake of renormalization theory (which
were mainly driven by the issue to make QFT fit for describing strong interactions). On the one hand the adaptation of the Kramers-Kronig dispersion
relation to relativistic particle scattering and the related more general interest
in analytic properties of scattering amplitudes and (electromagnetic) formfactors nourished the (exaggerated hope, as it turned out later) hope that many
dynamical aspects could perhaps be understood as consistency relations between analytic properties and reasonable assumptions about asymptotic highand low-energy (threshold) behavior. Initially the aims of these investigations
were quite modest in that particle physicists were already satisfied when they
succeeded to derive experimentally verifiable sum rules from current commutation relations and analytic properties (the prototype being the Thomas-Kuhn
sum rules which played an important role in the transition from old to new QT).
The most important post-Heisenberg conceptual enrichment of S-matrix theory (coming originally from Feynmans graphical formulation of renormalized
perturbation theory) was however the crossing property often called (by abuse
of language) crossing symmetry. Its off-shell version is a formal property which
relates scattering graphs with other ones obtained by crossing pairs of external
lines in incoming and outgoing configurations; in certain cases it could be verified outside perturbation theory by studying axiomatically derived analytic
behavior following from locality and positivity of energy. For on-shell quantities as scattering amplitudes or formfactors of currents such derivations (for
those configurations for which they were possible) were quite demanding; often
on-shell crossing properties were simply postulated in order to explore their observable consequences, leaving their derivation from the causality and spectral
principles of QFT to future progress. It is of course standard practice to use
interesting properties as working hypothesis, even if one is not yet in the possession of a conceptual derivation3 from known principles; the only problem with
that is that after the passing of time, even with better conceptual insights and
mathematical tools available, one may have forgotten about their problematic
status.
The emerging so-called S-matrix bootstrap program, which owed its second life (after Heisenbergs first attempt) to the important but incompletely
understood crossing property, was on top of the physical fashions of the late
50s and maintained its leading role through the 60s. Some of its points were
quite reasonable; but the human weakness of making exaggerated promises (a
theory of everything except gravity), and finally the strong return of QFT in
the form of nonabelian gauge theories, led to its demise. Among the few valu3 The conjecture that on-shell crossing is a relic left by spacelike (anti)commutativity is
plausible but there exists up this date no mathematical derivation (see remarks in next section).

able points it contributed is the attention it drew to the fact that an on-shell
approach (if it succeeds) is not only free of ultraviolet problems but it is also
manifestly independent of the infinite number of possibilities for pointlike field
coordinatizations (leading all to the same physics); in this sense it addressed,
together with the independently emerging algebraic QFT during the 60s, for
the first time Jordans plea for intrinsicness of the description (abandonment of
classical crutches).
The emphasis on an intrinsic description is well known in mathematics; it
underlies the invariant description of modern differential geometry which not
only liberated geometry from the use of singular coordinates, but from the use
of any coordinatization altogether. This should be understood as nothing more
than an analogy, i.e. the use of modern differential geometry in the formulation
of QFT does not make the latter more intrinsic; in fact the implementation of
intrinsicness in QFT turns out to be a much more radical step than in differential geometry. But the antagonistic bellicose tone in which the underlying
bootstrap philosophy was set against QFT (QFT is like a mortally wounded
soldier dying on the battlefield...) undermined its credibility in the eyes of
many physicists, and in my opinion, looking at this situation in retrospect, this
delayed an interesting chance to enrich QFT by adding a new viewpoint for several decades (see next section); only now, as a result of the crisis, the situation
forces young physicists who are out to re-capture the lost path of innovations
to revisit the relics on the wayside of the great caravan of particle physics (for
more details see [5]).
Since there was no operator formalism in which the underlying ideas (invariance, unitarity, crossing, maximal analyticity) could be implemented, the problem of constructing a crossing symmetric unitary maximally analytic S-matrix
was ill-formulated. Tinkering with properties of Beta functions and their representation in terms of Gamma functions, Veneziano was able to construct the first
model for a crossing symmetric elastic scattering amplitude. His proposal did
not satisfy unitarity, but the realization that his on-shell prescription allowed
an auxiliary field theoretic description in terms of a (off-shell) two-dimensional
conformal field operators theory and that it also admitted an auxiliary presentation in terms of the canonical quantization of a classical relativistic Nambu-Goto
string Lagrangian contributed significantly to its theoretical attraction. It also
nourished the hope that the model can be unitarized in a later stage. Its main
popularity it however enjoyed among strong interaction phenomenologists who
actually (for reasons which nowadays hardly anybody remembers) liked the idea
of satisfying crossing already with infinite towers of intermediate particle states
(duality = one-particle crossing) without the participation of the multi-particle
scattering continua as would be required for an S-matrix coming from QFT. In
conjunction with the phenomenological use of ideas of Regge poles the emerging trajectory pictures (mass versus spin) had a certain phenomenological charm
(and probably were Venezianos physical motivation), and although infinite particle towers cause some field theoretic headache4 , there was no reason to reject
4 Unless

only a finite number of particles remain as stable objects (and infinitely many are

it as a phenomenological proposal which captures some aspects of strong interactions. and whose possible relation with the known principles could be clarified
after establishing some phenomenological success.
The subject became somewhat problematic when the phenomenological interpretation in terms of Regge trajectories for strongly interacting scattering
amplitudes was abandoned (partially because the description of high transverse
momentum scattering data worsened) and a new gravitational interpretation on
the level of the Planck scale was proposed (the Bartholomew day massacre
of the old dual model string theory by J. Scherk et.al. 1975 in Paris). The
new message was the suggestion that string theory (as a result of the presence
of spin two and the apparent absence of perturbative ultraviolet divergencies)
should be given the status of a fundamental theory at an energy scale of the
gravitational Planck mass 1019 GeV i.e. as a true theory of everything (TOE),
this time including gravity. Keeping in mind that the frontiers of fundamental
theoretical physics (and in particular of particle physics) are by their very nature a quite speculative subject, one should not be surprised about the highly
speculative radical aspects of this proposals; we know from history that some
of our most successful theories originated in a similar manner. What is worrisome about this episode is not so much its speculative nature but rather its
uncritical reception. After all there is no precedent in the history of physics of
a phenomenologically conceived idea for laboratory energies to became miraculously transmuted into a theory of everything by just sliding the energy scale
upward through 15 orders of magnitudes and changing the terminology without
a change in its mathematical-conceptual setting5 .
One would have expected that a speculative idea which has lost its phenomenological support basis and instead jumped into a conceptual blue yonder beyond the range of QFT would receive at least that amount of critical
attention with which particle theorists analyzed previous situations of apparent conceptual conflicts from new proposals with established principles. After
all antinomies and paradoxical situations were the source of great conceptual
advances in the last century; often the results obtained from their resolution
were more important than the original idea by itself. Unfortunately this is not
what happened in the case of string theory. The times of critical spirits of
the caliber of Pauli had long passed and the new love affair between physicists
and mathematicians which left its mark in the great Atiyah-Witten dialogue
was more directed towards finding beauty and harmony (between differential
geometry and topology on the one side with the artistry of the Euclidean functional integral on the QFT side) than to the task of carrying antinomies between
known structures of local quantum physics and new speculative ideas to their
breaking point. The enthusiasm about a new golden age between mathematics and theoretical particle physics was mutual, but I think it was based on a
deep misunderstanding. Physicists were impressed by depth of modern coordiconverted to poles in the second Riemann sheet) one gets into trouble with locality (this can
be seen in the context of d=1+1 factorizing theories).
5 Some geometrical enrichments came much later and did not really modify the core of the
situation.

natization free differential geometry and topology and started to write articles
in which they explained to their fellow physicists what they have learned and
how it can be used for the quasiclassical exploration of functional integrals6 .
On the other hand mathematicians were impressed by the geometrical power
of some computational tools of physicists. They begun to look in awe at the
Euclidean functional integrals7; fortunately for mathematics they often found
rigorous ways to confirm conjectures suggested by these formal tools of the
physicists. Only very few mathematicians (including Vaughn Jones) were actually aware that some of their colleagues were erecting their beautiful edifices
on some of the ruins of particle physics. The interesting geometrical content of
string theory contributed to the rather uncritical support it received even from
reputable people. This in my view was the historical beginning of a mode of
free-floating discursive thinking which finally after almost 3 decades led to the
abdication of what used to be fundamental physics. As already mentioned this
erosion process had its recent outing in a semipopular anthropic manifesto by
L. Susskind.
Before looking more closely to the sociological changes which came with the
increasing popularity of string theory, it is quite instructive to analyse the origin of its somewhat confusing terminology. The historical reason why the name
dual model (after its interpretative paradigmatic change which allegedly incorporates gravity) was converted to string theory had nothing to do with its
quantum localization in space time but refers to the infinite particle tower
mass spectrum which the canonical quantization of a free classical relativistic
string leads to; the intrinsic quantum localization of e.g. a canonical quantized Nambu-Goto string is despite its classical origin not string-like but rather
pointlike (more on this see below and next section), a first indication that the
cherished quantization parallelism between classical and quantum localization
is limited to pointlike objects. In particular those Euclidean tube pictures cannot be interpreted as the time-development of physically string-like localized
objects. One may justifiably fear that this could create a conceptual salad in
the minds of newcomers unless they are especially warned that the terminology
should not be confused with the physical content. Its more recent descendant
cannot create this kind of confusion since it receives its name M-theory just
from the alphabet of big Latin letters, but far from this being a precautionary
measure for not prejudicing its frail content, one is invited (according to Wit6 The euclidean functional integral representation as well as the closely related canonical
quantization are formal artistic devices which acquire a mathematical meaning in the quasiclassical interpretation; renormalized correlation functions in strictly renormalizable theories
are however not representable in this manner (only their meaningless unrenormalized analogs
are formally consistent with functional integral representations).
7 This deepened already existing misunderstandings between the concept of localization
(the living space) of a theory and its Bargmann-Wightman-Hall analyticity region and led
to such confusing terminology as chiral QFT on Riemann surfaces. It also led to statements
that their symmetry group is the group of analytic maps (only in [6] a mathematician points
out that this does not define a group and wonders what physicists could possibly mean by
this; in order to resign with the statement that in view of the interesting results they must be
in the pocession of a correct meaning) instead of the group of diffeomorphisms of the circle.

ten) to interpret M among other things as standing for mystery (also matrix,
mother, magic, in short a true theory of everything). Interpretation like this
facilitates my attempts in this essay to describe string theory a la Susskind
as a return to scholastic pre-Galilei-Newton-Einstein re-mystification of nature,
which attaches weight to the question of how many vacua fit onto a universal
string (alias how many angels fit onto the pin of a needle).
In this essay I will argue that this crisis cannot really be understood in terms
of physics without taking into account sociological changes of the increasingly
globalized community of physicists who is pursuing this kind ideas. My experience with string theoreticians is admittedly marked by a certain distance since
my underlying philosophy has basically remained a bit more Wilsonian in the
sense that reality in my view consists of an infinitely many shells of asymptotically inclusive theories (like the sheets of a possibly infinite onion). Perhaps
somewhat different from Ken Wilson (I do not know his opinion on this point),
I believe that every shell can be described in conceptually complete and mathematically rigorous way, i.e. mathematical consistency alone does not require
knowledge of the next layer (the unreasonable efficiency of mathematics in the
sense of Wigner). In fact the attempt to dump the details of the next layer into
a cutoff parameter could mathematically be even counterproductive8 . In other
words I believe that the 70 year old weakness of not knowing whether those
objects we talk about in QFT really exist still will be overcome, in fact I think
(see remarks in the next section) that the day when QFT will enjoy the same
conceptual status as other areas of physics is not very far; but for the time being
QFT did not reach closure.
I do not subscribe to the underlying idea of string theory that a TOE is at all
possible. In fact the idea that Einsteins Dear Lord (the one who does not throw
dice) permits some string theorist to find a closure to fundamental physics (so
that for the rest of all days the curiosity of humanity about its material world will
end in intellectual boredom or at best re-directed into a plumbing job for some of
the still missing details or perhaps subsumed into a branch of the entertainment
industry) appears to me outright ridiculous. Ideas like this probably will be
cited by historians in a distant future as representing the hubris and intellectual
(not necessarily personal) arrogance of a past Zeitgeist. The best one can do to
illustrate these points is to quote statements from string theorists. One citation
by which an apparently reputable young (but apparently very immature) string
theorist9 used to sign off his contributions to discussion groups reads as follows
[7]:
Superstring/M-theory is the language in which God wrote the world.
Whenever I looked at that line (I stopped looking 3 years ago, hence I do
not know whether dictums like this still enjoys popularity) an old limerick came
to my mind. It originates from pre-war multi-cultural Prague where, after a
performance of Wagners Tristan and Isolde by a maestro named Motl, an art
8 Note that string theory, if anything, has a much worse existence problem; it is not even
possible to give an intrinsic characterization, the only thing one has are cooking recipes.
9 As in the religious domain there is of course the risk of violating somebodies feeling. But
without using the freedom of expression this essay would lack substance and credibility.

critic wrote instead of the usual critical review in next days newspaper the
following limerick10 (unfortunately this cannot be translated from German to
English without a complete loss of charm):
Gehe nicht zu Motls Tristan
schau Dir nicht dieses Trottels Mist an,
schaff dir lieber nen drittel Most an
und trink dir mit diesem Mittel Trost an
What a difference of this God who has to follow the logic of string theory
(perhaps because there is no other game in town, see next section) to Einsteins Dear Lord! Whom God allowed to have a copy of his blueprint of the
world does of course not have to bother about earthly worries of experimental
verifications; those remaining earthlings which are not in its possession, and
want to verify results of this divine theory in the old fashioned way by experiments and observations will have no chance since they never can get to the
required energy scale. Even a theoretical comparison with existing principles
of quantum field theory is not possible, because string theory is only known in
terms of recipes which have resisted an intrinsic characterization by principles.
This situation is without precedent in the history of physics.
String theorists offer their well known scale-sliding arguments to show that
Lagrangian field theory is a low energy footnote of string theory. But those
arguments are no substitute for a structural comparison. To illustrate what is
meant by structural consider the somewhat analogous relation between QM
and QFT. The structural reason is that there exist relativistic field operators
which in their one-time application to the vacuum create one particle states
without any admixture of vacuum polarization. These are the free Bose/Fermi
fields. If we would live in a (d=1+2) world of anyons and plektons (particles with
braid-group statistics) the nonrelativistic version would not be QM but rather
a quantum field theory11 with a nonrelativistic dispersion law. The relevance
of QM to the physical world is related to the fact that particle number conservation is compatible with the relativistic spin-statistics connections of Bosons
and Fermions. It has not been possible to make such structural arguments in
string theory because recipes are no substitute for principles and the fact that
one knows a lot about the principles which characterize QFT is of no help here.
For readers who want to see string-theoretic adulation in its purest form, I
recommend the article in [8] where one finds the following passage:
In the 1990s, Edward Witten and others found strong evidence that the different superstring theories were different limits of an unknown 11-dimensional
10 I am indepted to a reader for pointing out to me that the viertel in my original version
should be replaced by drittel. I appologize for having caused this temporary distortion
of the most perfect example of a German Schuettelreim (spoonerism), but the old measure
(Schoppen=1/3) has not been used any longer during my wining years; nowadays it is 1/4.
11 Wilszeks quantum mechanical anyons (Aharonov-Bohm like dyons) violate the spin statistic relation for an anomalous value of the spin; one consequence of this violation is that that
the resulting multi-dyon QM does not permit a description in the second quantization Fock
space setting. True anyons however remain field theoretical for all energies and never loose
their vaccum polarization clouds (even in the absence of genuine interactions). A formal
scale-sliding argument could easily overlook this subtle structural point.

theory called M-theory. These discoveries sparked the second superstring revolution. When Witten named M-theory, he didnt specify what the M stood
for, presumably because he didnt feel he had the right to name a theory which
he hadnt been able to fully describe. Guessing what the M stands for has become a kind of game among theoretical physicists. The M sometimes is said
to stand for Mystery, or Magic, or Mother. More serious suggestions include
Matrix or Membrane. Sheldon Glashow has noted that the M might be an
upside down W, standing for Witten. Others have suggested that the M
in M-theory should stand for Missing, Monstrous or even Murky. According
to Witten himself, as quoted in the PBS documentary based on Brian Greenes
The Elegant Universe, the M in M-theory stands for magic, mystery, or
matrix according to taste.
With such statements of scholastic adulation and a dash of talk-show entertainment12 , one is almost inclined to consider the first phrase of this article:
String theory is a model of fundamental physics whose building blocks are onedimensional extended objects (strings) rather than the zero-dimensional points
(particles) i.e. the standard opening mantra used by string theorists, as a lighthearted lapsus linguae rather than a potential source of misconception. It was
already mentioned before that whereas it is true that the extension of the classical Nambu-Goto string (whose canonically quantized version is the mother
of the string-theoretic reformulation of Venezianos dual model) is really a relativistic spacetime string, this is not the case for its canonically quantized counterparts. The only intrinsic notion of a quantum relativistic localization is that
obtained by translating this object into other positions and checking whether
its commutator with the original object is nonvanishing once the second would
be string shape enters the causal influence region of the first. Quantum localization is an autonomous property and the case where it coincides with the
classical meaning is one of the greatest lucky coincidences and Jordan with his
field quantization was very lucky indeed. This quantum localization test was
made (even by string theorists) and the result was negative [9][10]; the commutator is just that of two pointlike objects localized at the center of mass points
of the imagined strings. But the commutator test is lakmus test which decides
whether the use of the word string has an intrinsic quantum meaning. The
fact that string theory fails this test is not totally unexpected since the string
description is strictly auxiliary, it belongs to the classical side of the cooking
recipe for the construction of a particular crossing symmetric (and at the end
hopefully unitary) S-matrix and it certainly does not have a similar conceptual
physical status as a generalization of the interpolating pointlike fields of the LSZ
scattering theory to a string-like Heisenberg field. Without the use of scattering theory in terms of large timelike limits (of either point-like or semi-infinite
12 The author of the Wikipedia article informed me that I misunderstood the intention of
his remarks concerning the interpretation of M, the main purpose was to expose to his readers
the vagueness of the subject. But I am also a reader and a bit of irony or tongue in cheek
in the phrasing would have helped me. If I am the only one who incorrectly interpreted this
as adulation through boastful mystification (that is at least the intention of Brian Green, a
humble admission of its speculative content would look different), I will retract my remarks.

10

string-like localized interacting fields the relation of the S-matrix with locality
is extremely indirect (and has been noticed only recently with the help of the
concept of modular localization for wedge regions [5], see also remarks in next
section). Relativistic string-localized quantum objects which could substitute
the interpolating pointlike fields and those auxiliary constructs of string theorists are very different concepts. Our presently best knowledge about them may
be condensed into the statement: quantum string-localized objects (example the
quantum fields behind the zero mass infinite spin Wigner representations) do
not allow a representation in terms of Lagrangian quantization and on the other
hand the quantization of classical (Nambu-Goto type) strings does not lead to
quantum string localization. This may at first sight come as a shock even to ordinary field theorists who expected the quantization approach (i.e. the invocation
of a parallelism with classical structures) to be of general validity transcending
the boundaries of pointlike localization. To contemplate a less singular more
intrinsic description without classical crutches, as Jordan did, is one thing,
but to be confronted with a conceptual barrier which prevents to view certain
a quantum object as a quantized version of a classical analog is a different and
more surprising issue. String theory is build on the supposition that geometric
aspects and localization properties are maintained by quantization even if one
goes beyond pointlike fields.
There is another result, this time not referring to localization, which highlights the strictly auxiliary nature of the meaning of the word string in string
theory. It has been known that the Nambu-Goto string is an integrable system. The infinite family of conserved charges has been identified, and their
Poisson relations are known. One of the important messages coming in particular from the early work of Faddeev on integrable systems that it is easier
to arrive at the (nonperturbative) solution if one directly quantizes the closed
system of algebraic Poisson-bracket solutions between the complete system of
conserved charges. But what should be done in a situation where the intrinsic quantization of charges leads to results which disagree13 with the standard
canonical quantization as it is the case for N-G strings? [11]. The answer is:
the dual model S-matrix construction requires the canonical quantization since
the notion of string is just a short hand computational device for re-producing
the mass tower spectrum of the dual model, the question of what is the correct
quantization of a physical N-G string may be interesting but is irrelevant in the
present context since the string of string theory is only a computational trick
to find a Lagrangian packing for the dual model with its infinite mass tower.
With other words the second canonically quantized Nambu-Goto field theory is
a special kind of infinite component theory
The above remarks should not be misunderstood as a plea for a moratorium on speculative ideas. Particle physics needs in certain situations jumps
into the blue yonder and the risk that they turn out to have no experimental support and that their conceptual relation to established principles remains
13 Among other things the invariant quantization of the N-G string does not distinguish any
particular spacetime dimension.

11

obscure has to be taken. The real worrisome aspect of string theory (and to a
lesser degree with supersymmetry) is not that it turned into a fashionable subject without an observational basis or that it is unable to clarify its conceptual
status with respect to the principles of QFT (which remains despite its incompleteness the best reference point); there were fashions before (the S-matrix
bootstrap, Regge poles,...) which have attracted a lot of attentions and also
suffered from the same shortcomings. But in the past there was enough critical
potential to prevent such a condensation along a fashionable monoculture14 .
The situation has radically changed in that reputable and influential physicists
whose critical judgement could have a sobering impact on a confusing situation
give their uncritical support to these tendencies and in this way become part
of the problem. Particle physics was at its best when physicists like Pauli who
were strongly committed to the truth were at the helm of particle physics. This
led to a situation where contributions in this area are submitted to so called
high impact journals whose editorial board already consists of people who are
highly specialized on hot subjects (if not directly related to string theory, then
at least to its sociological fall-out) but are incapable (lack of background of a
general physical culture) to evaluate anything else which does not follow these
globalized lines15 .
The main difference is that string theory (together and in union with supersymmetry) is the first theory which despite of these mentioned faults managed
to survive for more than 3 decades. After such a long time the solid conceptual
grip on QFT which would be necessary to press ahead and carry string theory
to the breaking point in order to arrive at an enigmatic solutions of the points
of structural tensions with QFT (or bury it, as it was done with several other
failed theories before) is lost or (in case of younger string theorist) never existed.
On the other hand those quantum field theorist who have this strength are not
motivated to look into the details of prescriptions for what they believe could
be another failed theory; in addition they know that as a result of the sectarian
ideological attitude their critical remarks would not be welcome anyhow. Particle physics in the past has successfully maintained its unity despite the use
of widely different concepts and mathematical tools, but this unity is rapidly
dissipating.
There is a related sociological problem. If an idea which promoted the
careers of many physicists is kept alive for such a long time it becomes immune
against criticism. I think nobody at this late point would seriously expect that
somebody who invested more than 3 decades into a theory which led to tens
of thousands of publications but failed to make contact with real physics will
come up and say sorry folks this was it!
14 An beautiful example of a critical essay against some excrescences of the S-matrix bootstrap program and brilliant illustration of highly informative scientific polemicat its best one
finds in [12].
15 But

contrary to a normal referee who recognizes his limitations and sends it back to the
editor, this new breed rejects a paper without ever criticising its content, since in order to to
merit their attention they believe that you should belong to what they consider as their class
(use their conceptual framework etc.).

12

Theories which occupy minds for such a long time escape the critical radar
screen and develop a life of their own. Unfortunately there is no prior experience
with similar problem in the past from which we could learn how to protect the
content of fundamental physics against such new market tendencies. But even if
one does not know what to do about it, there is at least no problem to describe
the situation in terms of the following 5 thesis:
1. Fundamental knowledge about particle theory and in particular about
QFT has and is being lost.
2. The research in string theory is done and marketed by globalized groups
without encountering an independent critical evaluation.
3. Newcomers to particle physics are prematurely exposed to highly speculative and controversial ideas before their critical immune system had a
chance to develop through a study of concepts of QFT.
4. String theorists dismiss the enigmatic power of antinomies and structural
clashes, papers which contain theorems which do not fit their prejudices
are simply ignored. The string theoretic literature consists mainly of calculations based on certain recipes and speculations, there is hardly any
attempt to provide a conceptual framework.
5. A large number of positions in particle theory have gone to string theorists.
There are good reasons to worry that this could have negative long-lasting
ramifications for particle physics.
The case in favor of these statements is based on the following arguments.
In talking to members of the younger generation of physicists one is often
surprised that they know very little about the great conceptual achievments of
the 1950/60 post renormalization era. Although the recipes of the renormalization technology are usually part of their working knowledge (perhaps because
it makes the string recipes more palatable), the conceptual achievments as the
derivation of time-dependent scattering theory as a consequence of causality
(LSZ, Haag-Ruelle) or the subtleties in the particle-field relation remain outside
their stock of knowledge. It seems that they have been replaced by differential
geometry (often only in a Euclidean setting), and such special ideas as the possible use of Calabi-Yau manifolds in particle physics. One does not have to be
old-fashioned in order to be concerned about such trends.
Unfortunately those QFT textbooks which have been written by veterans of
string theory emphasize this cooking book recipe approach to QFT. Why should
somebody to whom the S-matrix is nothing but the result of a string theory
cooking recipe mention the (profoundly beautiful) result of scattering theory
being a consequence of spacetime causality and ensuing cluster properties? This
is time consumimg and could be counterproductive for the marketing of string
theory since it could undermine confidence in its recipes.
In order to illustrate the second point in the above list it is helpful to recall
that in earlier times there was a clear separation between the act of creation by
13

the author and the critical evaluation of its content by his colleagues. Nowadays
authors are often embedded in a globalized group and they communicate an
innovative idea among members of this group already before it could reach a
certain amount of maturity. This has two consequences, on the one hand the
idea becomes modified and its originally intended content may get diluted, and
on the other hand the chance that the work receives a critical evaluation from
an independent viewpoint is significantly reduced. The likely kind of criticism
in such a setting is that of an ideological battle between two globalized groups
rather than a critical analysis of its content. Just imagine that Einsteins general
relativity would have come into being within such a sociological environment.
Imagine in particular that instead of struggling for years with his famous hole
problem and thinking it through by himself, he would have explained it to the
other members of such a globalized group. The topic would have entered emails
and discussion groups and the result may have well been a total confusion for
many years to come.
Another danger emanating from big groups whose interests is focussed on
the promotion of a particle physics monoculture is that they tend to control
editorial boards of so-called high impact journals and in this way prejudice the
research topics of particle physics. The easiest way to obtain a list of who is
who in administrating the crisis of particle physics is to look at the editorial
board of a new electronic journal as JHEP (Journal of High Energy Physics).
The third point of the list is perhaps the most serious one. A theory which
has in more than 30 years been unable to get its relation to physics straightened
out is presented to physics students as a theory of everything which supersedes QFT before their critical faculties have been strengthened by learning
about the concepts and open problems of the most successful theory of particle physics. To get an impression about the extend to which the borders
between education and indoctrination have been blurred already, it is instructive to take a look an undergraduate course on string theory offered at MIT
(http://ocw.mit.edu/OcwWeb/Physics/8-251Spring-2005/CourseHome/index.htm).
Let us explain the fourth thesis in the context of a concrete example. QFT
in its present state is in most cases not able to produce a convincing proof
of a conjecture about a concrete Lagrangian model. But if this model is part
of a larger class of models which can be characterized by certain structural
properties, one is in a much better situation. An example is the Maldacena
conjecture about a possible connection between a special conformal theory (a
supersymmetric gauge model) and a certain gravitational anti de Sitter (AdS)
model. This conjecture has not been proven and according to our first remark
it is questionable that this will ever be possible. But since the generic relation between AdS-QFT and a conformal theory on its boundary is an easily
established structural (model independent) fact (which is supported by AQFT
in a exceptionally clear way), there is the interesting chance to come up with
a mathematical theorem about which structural property on the conformal side
implies/excludes a corresponding property on the AdS side. The rigorous result
of a structural investigation is that a standard conformal quantum field theory
described in terms of pointlike fields (e.g. a Lagrangian QFT) corresponds on
14

the AdS side to a QFT which does not have sufficiently many degrees of freedom
in order to be presentable in terms of Lagrangian pointlike fields. Vice versa, if
on starts with a standard theory on the AdS side there are too many degrees
of freedom on the conformal side so that it is again not possible to relate this
structure with any kind of Lagrangian QFT. Hence if the Maldacena conjecture
was meant to relate two standard Lagrangian field theories it is contradicted
by a theorem. If on the other hand the conjecture involves a string theory
on the AdS side then nothing can be said because the structural properties
of string theory which one would need for making structural comparisons are
not known. So instead of proving the conjecture one could perhaps turn the
Maldecena conjecture into a working hypothesis and see whether the rigorous
structural properties which AQFT leads to on the AdS side are compatible with
what one expects of a string theory; I am not aware of any such attempt. But
after 30 years it is probably more realistic to believe that structural properties
(of the kind one has been able to establish from the principles of QFT) may
never be available in string theory because it always may remain a collection
of prescriptions; after all there is no reason why any collection of not outright
contradictory prescriptions can be backed up by a consistent theory. In other
words the lack of structural understanding may not be the result of limitations
on the part of string theorists, but rather may indicate that there is no coherent
theory in the standard sense of these words behind the string prescriptions.
Among the thousands of AdS-CFT papers I only know of three which address this problem, but the string community has ignored them. I think that
this is not the result of a conscious act (a kind of intellectual boycott) but a
genuine loss of understanding of what structural theorems are about. In this
situation it makes almost no difference whether this is plain ignorance or results
from the kind of hubris about having a TOE at ones disposal; in short these
are the visible marks of a profound self-created crisis in particle physics. With
acclamation (instead of critical cutting edge arguments) it is much easier to
enhance impact parameters for ones career. The present system of assigning
high impact parameter to publication in certain journals which support fashinable subject rewards acclamation and punishes attempts to look for antinomies
and breaking points (and thus doing research in the best of our traditions). As
mentioned before it is not difficult to find out about who is who in the administration of the crisis in particle theory and verify that they enjoy a high social
status in the community of string theorists. No theory before string theory led
to such a deep schism in particle physics; not only have string theorist problems
to communicate with their more modest competitors from loop gravity, but also
those part of QFT which are of a more conceptual nature and less suited for
immediate computational use for specific models (but very efficient to unravel
structures of families of models) are beyond string theoretic radar screens. In
fact the unity of particle physics may already have been lost.
The last point in the list expresses concern with the future of particle physics.
Given the human nature, it is very improbable that high-flying string theorists
who enjoyed social success and recognition in their globalized group and in the
media are able to muster the intellectual modesty which one needs in order to
15

clean up ones desk and start something else; it is more realistic to expect that
the present situation will continue or even aggravate.
In the past fundamental physics has proven remarkably resistent against
outside pressure, neither the Nazis nor the Stalinists were able to make its
underlying philosophy subservient to their ideologies. But its future state of
health depends crucially on whether the acting physicists are able to protect it
against certain tendencies which already dominate our life through the economic
hegemomy of globalized capitalism, i.e. a kind of Enron-Worldcom-Parmetat
analog of physicists acquiring personal influence, wealth and fame at the expense
of the search for truth in physics.
Is science loosing its autonomy and are we reaching the end of the path which
begun with Planck and Einstein? The recent book by Susskind [1] has brought
questions like this into the forefront. In order to spare string theory the fate of
becoming a kind of gigantic 20 century phlogiston (the obsolete pre-Lavoisier
theory of combustion) he proposes to change the meaning of what constitutes
science and to redefine of what makes an argument scientific.
Attempts to look for sociological causes behind paradigmatic changes in
physics are not new. After the discovery of quantum mechanics, P. Foreman
[13] offered a sociological explanation for why these new ideas originated in warshattered Germany. He made the argument that the unusual new indeterminism
which originated from the probabilistic interpretation of quantum theory could
only have been thought through in the necessary radical manner by individuals
who saw their world destroyed. According to Foreman, those who were deeply
affected by the gloom and doom Zeitgeist expressed in Spenglers influential
book the demise of the west (which was widely red in post-war Germany) had
the best precondition for being able to contemplate a new kind of mechanics in
which acausality, indeterminism and chance played an important role.
This does not sound very palatable. Physics in war-shattered Germany
remained a strong autonomous science and QM has nothing to do with the
Spengler kind of gloom; actually the influence went occasionally in the reverse
direction e. i. misunderstandings of new developments in physics led to misleading philosophical and sociological interpretations16 . If one wants to find
marks of the Zeitgeist in physics one should look less at the subject of research
and more to the manner in which physicist present their arguments and communicate their findings; in the actual sociological setting a relevant question
would be: do physicist uphold their traditional auto-critical manner or do they
succumb to the forces of a neo-liberal globalized market economy where the
exploration of the entertainment value by the media promises instant reward
and recognition. String theory serves as an excellent illustration of the power
which public relation has acquired in particle physics. Its negative influence is
not just the result of its highly speculative content and its lack of observability
(which it shares with other previous fashions). More dangerous for the future
of particle physics are the attempts by its protagonists to elevate these flaws
16 It is well-known that Einstein disliked the terminology relativity because his main
achievement was to identify the invariants which are independent of the observer.

16

into a virtue and try to convince newcomers that there is no other game in
town. That such attempts have a chance of success is very much related to
the conceptual confusions which already started at the cradle of this theory and
became solidified as a result of its several revolutions during its more than 30
years of existence. The conceptual dust it left on each of its crossroads of metamorphoses prevented a return to the many interesting open problems which it
left behind; at no time before were so many computations done on such a muddy
conceptual ground.
There is of course in addition the human element of what one may call a
pitbull syndrome17: after having invested almost 30 year of their scientific
life in making string theory work, its protagonists cannot let loose anymore.
As a result we have to be prepared to coexist for a long time to come with
string theory, its supersymmetry support, its noncommutative spin-off and with
whatever big Latin letters stand for.
What might be the thoughts of a physicist 100 years from now who enters
a physics library (assuming that nonelectronic publications will continue) and
walks along the corridors looking at those long shelves filled with hundred of
thousands of publications on supersymmetry, string theory and related subjects?
Will he have similar thoughts as we have when we look at the great achievements
100 years ago? Or will they be more similar to our thoughts when we look at
publications from the time of the pre-Lavoisier phlogiston theory of combustion
(which was wrong despite the fact that it explained some observations)? If there
will be bad luck with the new generation of particle accelerators, and among our
descendents there would be somebody with Paulis sarcasm, he might actually
point out that string theory was perfect: it predicted nothing and that was
exactly what was found.
Since hubris, prepotency and ignorance appear to be the hallmark of a hegemonic new Zeitgeist and physics does not seem to be left unaffected, it may be
interesting to know whether there exists a contemporary analog of Spenglers
critical socio-philosophical studies. A fundamental criticism of modernity is extremely hard because of the problem to understand how in the midst of a highly
civilized culture with impressive contributions to music, arts, literature and science a monstrous crime of genocide can take place. This is precisely the starting
point of Theodor Adornos critical analysis of modernity [14]. The part which
in my view is relevant for arts and sciences is his Dialectics of Enlightenment
(with Max Horkheimer) his later essays on Kants criticism of pure reason as
well as on Negative Dialectics18 . His focal point is the mechanism by which
rationality and enlightenment can turn into irrationality. Adorno illustrates his
ideas mainly in arts and philosophy, but I think that an adaptation to science
and in particular to the crisis in particle physics is possible and could shed more
17 The pittbull is a combat dog which was bred in the UK during the second worldwar.
Below a certain threshold of incitement it has the temper of a normal dog, beyond this it is
incapable to let loose.
18 These writings had an enormous impact on the postwar philosophy and sociology in Germany but (probably because they use all the resources of the German language and philosophy)
they have not played a comperable role in the anglo-saxon cultural sphere.

17

light on the points raised in the present essay.

Is it really the only game in town?

A guy with the gambling sickness loses his shirt every night in a
poker game. Somebody tells him that the game is crooked, rigged
to send him to the poorhouse. And he says, haggardly, I know, I
know. But its the only game in town.
Kurt Vonnegut, The Only Game in Town [15]
This is a quotation from a short story by Kurt Vonnegut which Peter Woit
[16] recently used in one of the chapters in his forthcoming book entitled Not
Even Wrong : The Failure of String Theory & the Continuing Challenge to
Unify the Laws of Physics (using a famous phrase by which Wolfgang Pauli
characterized ideas which either had not even the quality of being wrong in
an interesting way or simply lacked the scientific criterion of being falsifiable).
The most prominent string theorist who used the no other game in town
phrase in interviews when asked about alternatives by science journalists was
David Gross, and of course we do not know whether (despite his different social
conditions) he sometimes thinks along similar lines as the Vonnegut character.
In this section I would like to convince the reader why this statement is
wrong; the sociological question why so many physicists seem to accept this
dictum was already addressed in the previous section. In order to avoid sidetracking the reader from my main aim which is just to answer the question in
title of this section (and not to use a criticism of string theory as a pretence for
proselyting for approaches which have attracted my attention during the last
decade) I will be scarce on references; the reader wants to have more information, it should not be too difficult to track down more relevant references.
My illustrations of alternative games are different from those by Peter
Woit; in particular I have less faith in differential and algebraic geometry methods in particle physics. As a result I also see the past Atiyah-Witten collaboration in a somewhat different light. On a superficial level, taking Wittens early
work (e.g. on the screening effects of infrared clouds and the Kosterlitz-Thouless
phase transition) into account, I am inclined to see a negative influence more
in the opposite direction from how Woit sees it. But on a more profound level
I find it very difficult to comment on a complex relation between two such eminent scientist and prefer to leave the answer to historians of science, especially
since it does not contribute anything to the theme at hand.
The way out which I advocate is to revisit some old unsolved deep conceptual and mathematical problems which got lost in the dust on the wayside left
by several so-called revolutions. I believe that the deployment of new powerful
conceptual and mathematical tools of advanced QFT to such problems offer the
best hope for superseding the present crisis. This third way19 becomes particu19 For the present discussion the first way denotes the speculative jump into a blue yonder
and the a posteriori justification via conceptional and mathematical safeguarding; the second
way refers to cataclystic experimental discoveries.

18

larly important if the first two ways failed (e.g. because a long-time speculative
journey into the blue yonder has led to a loss of conceptual orientation as it is
the case with string theory, or because the unreasonable efficiency of a high parameter description with many conceptual deficiencies has led into a labyrinth
i.e. the reality of the standard model) and the second path of experimental
discoveries gives no new clues (since high energy experiments are not any more
autonomous endeavors, but have an increasing dependence on the quality of the
leading theory). This third way requires significant theoretical resources and is
certainly more difficult than that based on free-floating speculations. Its main
tool are cutting edge arguments (including gedankenexperiments) by which apparent conceptual antinomies and gaps in a given setting are brought to the
breaking point so that the new structures can emerge in contrast to the existing
ones. The transition from the old quantum theory (with its strong experimental
basis) to the new QM illustrates the power of this method. The first step is
to become aware that quantum field theory, apart from the highly developed
technology of perturbative renormalization, has remained an largely unfinished
business; never mind its string theoretic caricature. Apart from some general
structural consequences of the underlying principles (TCP, spin-statistics, the
DHR theory of superselection sectors) I cannot think of any problem in QFT
which was successfully solved and really brought to a closure in the sense as
it happened a long time ago in other areas of fundamental physics e.g. QM.
The post renormalization research in QFT has left us with an enormous amount
of half-baked results even in those areas in which the theory was very successful on the observational side. A third way in QFT would primarily consist in
finding an intrinsic formulation and computational methods which are not just
prescriptions but follow directly from the underlying principles.
I have already touched upon some of these questions in the previous section
when I referred to Jordans plea for a more intrinsic formulation of QFT without
the use of classical crutches and to the aborted S-matrix bootstrap approach
as a potential start for an on-shell reformulation of QFT. Let us recall here
also the Wigner 1939 characterization of particles in terms of positive energy
representations of the Poincare group as the first completely intrinsic way to do
(interaction-free) quantum physics without invoking any classical parallelism.
Although it is of enormous conceptual value, it was certainly too limited for
generating widespread attention since it said nothing about how to incorporate
interactions. There was however one conceptual spike in an otherwise perfect
setting. Whereas the relation of most finite energy representations with the
Lagrangian quantization approach for free fields was understood during the
subsequent two decades in terms of intertwiners between the Wigner canonical
representations and their covariant counterparts, there were rather big positive
energy families which resisted attempts to incorporate them into a setting of
Lagrangian quantization. For the large infinite spin family20 there were many
failed attempts by several generation of physicists to encode this representation
20 For this family Wigners little group is the two-dimensional euclidean group whose
irreducible unitary representation lead to a continuous range of Casimir values corresponding
to an infinite (half)integer helicity tower.

19

into the setting of pointlike fields which finally found their explanation in a
No-Go theorem [17]. The remaining question of what is the best (sharpest)
localization consistent with the infinite helicity tower led to the new concept of
string-localization in the general setting of modular localization [18][19]. Stringlocalized fields are operator-valued distributions A(x, e) which fluctuate in two
spatial variables where x is a d-dimensional Minkowski spacetime point event at
which the semiinfinite spacelike string starts and e is its the spacelike directional
(unit) vector which (as far as far as covariance and fluctuations are concerned)
can also be viewed as a point in a d-1 dimensional de Sitter space. For the
(anti)commutator of two such fields to vanish it is not sufficient that the x s
are spacelike, one string contour must also be outside the causal influence zone
of the other i.e. the string is visible in the sense of quantum localization[19]
(whereas that of a canonically quantized N-G string is fictitious apart from the
localization of its centre of mass). This illustrates that it is not always the
grand design but little observations on the wayside which lead to interesting
conceptual progress.
At this point it is helpful for the reader to be reminded that there are two
important concepts of localizations in relativistic quantum theory: the NewtonWigner localization and modular localization. Whereas the N-W localization
results from the adaptation of the Born x-space localization probability to relativistic wave functions and connects to a position operator and associated
localization projectors and probabilities, modular localization results from the
attempt to liberate the causal localization inherent in pointlike quantum fields
from the non-intrinsic aspects of field-coordinatization. The lack of covariance
and causality of the N-W localization21 does not have any unpleasant consequences as long as one uses it in scattering theory; the fact that N-W localization regains covariance for asymptotically large separation makes it in fact
indispensable for the derivation of the Poincare-invariant S-matrix where the
projectors and their probability interpretation is crucial. The use of N-W localization becomes however deadly wrong (superluminal acausalities) if used
for propagation over finite distances. Modular localization [19] on the other
hand does not lead to projectors and probabilities but is the correct concept
for the covariant causal localization of states and operator algebras. Unlike
the Lagrangian quantization, it is capable to control situation where a coordinatization by pointlike fields is impossible (as in the mentioned case of the
Wigner massless helicity tower representation). In such cases the issue of local
functions, Wick polynomials, vacuum properties, an energy-momentum tensor
and statistical mechanics properties are sufficiently different from the standard
pointlike case and require new conceptual attention before deciding if such a
matter is unnatural or has a role to play (e.g. in the astrophysical discussion
of what constitutes black matter/energy22 ).
21 More generally a theory with positivity of energy does not admit a system of covariant
localization projectors.
22 Despite the existence of KMS states on the infinite helicity tower string-localized fields,
their thermal properties are quite unusual (as already nocticed by Wigner himself [18]) and
remain essentially ununderstood.

20

There are massive Wigner representations in d=1+2 dimensions with anomalous (non-halfinteger) spin whose associated fields have plektonic (braid group)
statistics which is known to be inconsistent with a pointlike localization as well
with an on-shell structure. More precisely even in the freest version (vanishing
scattering cross section) the realization of braid group statistics requires that
any operator whose one-time application to the vacuum is a state with a oneparticle component has necessarily a nonvanishing vacuum polarization cloud23
[23], in other words there is no on-shell free anyon field. Many properties of
anyons (anyon=abelian plekton) can be seen by applying modular localization
to the Wigner representation. Using similar modular intersection ideas as in
the case of factorizing models [24], there are good reasons to expect that in due
time one will have an explicit construction of integrable anyons.
It is somewhat surprising that modular localization within the Wigner setting also leads to new viewpoint about gauge theory. It was well known that the
covariant field description of the Wigner theory for massless helicity one representations relates to the electromagnetic field strength but does not support
the introduction of pointlike vector potentials (without additional unphysical
extensions). This is very different from the classical situation where the positivity of the quantum theoretical state space is of no relevance; in that case the
idea of a gauge principle is useful in order to select among all possible covariant
couplings of a classical vector potential the Maxwells classical electrodynamics, This selection principle has a straightforward extension to the quasiclassical
realm of quantum particles/fields in an external vector potential. The situation
changes radically if vector potentials become quantum objects. The standard
way of handling this problem is to temporarily forget the positivity requirement
and to uphold the pointlike structure so that the usual perturbative Lagrangian
quantization approach could be applied. This is achieved by artificially extending the quantum theory by adding in unphysical ghosts which at the end of
the calculation have to be removed. From the outset it is not clear that after
having done the perturbation theory in this unphysical setting one can remove
the ghosts from quantities which in classical sense would be gauge invariant and
the best formulation which makes such a descend manifest (by formulating the
physical descend as a cohomological problem) is the well known BRST formalism of gauge theory. There is nothing to say against doing this and getting
good results for gauge invariant quantities. The problem starts if such a BRST
catalyzer (Ghosts are neither in the original problem of spin one particle
representations nor in the final physical answers) is not considered as a temporary computational trick, but becomes elevated to the status of a fundamental
physical tool.
There are two arguments against accepting such a catalyzer.
One is that in the quantum context one does not need a gauge principle
for the purpose of selecting between different possibility because (different from
23 Only if one does not insist in the best localization which is stringlike but extends the
localization space to a (Rindler) wedge one can even in the presence of interactions create
polarization-free states.

21

interactions involving lower spin) unitarity and renormalizability 24 already fix

the standard theory without the necessity of an additional choice. In different
words the other non-gauge classical possibilities which are eliminated by the
gauge principle do not occur in local quantum physics and hence there is no
need for a selecting gauge principle; to the contrary, accepting renormalizability
as a principle of local quantum physics25 , the classical gauge principle is the
quasiclassical relic of a property of local quantum physics.
The second point is of a more philosophical nature in that the assignment of a
fundamental role to the BRST formalism would go against Heisenbergs dictum
that concepts (and if possible also computations) must be based on observables.
In this context it is deeply satisfying that the Wigner theory, extended by the
concept of modular localization, leads to a physical string-localized potential
A (x, e) in the Fock space of photons [19]. The existence of this string-like
object also explains an old observation in algebraic QFT stating that the operator algebra generated by the field strength violates Haag duality for toroidal
spacetime regions. All these statements have extensions to higher helicity representations and the case of helicity two, where the Wigner covariant field
strength is a 4-tensor with the algebraic properties of the Riemann tensor and
where the string-localized potential is the metric 2-tensor of general relativity,
is for obvious reasons particular interesting. Contrary to the field strength, the
short distance properties of the potentials do not grow with increasing helicity. But it is presently not known how to formulate interactions and set up a
perturbation theory for string-like localized fields; as expected, the formal imitation of the standard pointlike formalism leads to infrared divergencies. So
an alternative formulation of the content of gauge theory becomes synonymous
with the problem whether one is able to implement interactions starting with
string-localized physical vectorpotentials. It would be very interesting to make
some progress on this issue since the whole massless finite helicity family admits
such physical string-localized ultraviolet-good-mannered potentials, even (as for
helicity=2) a BRST-like gauge formulation does not seem to be in sight [19].
Behind all these remarks is a theory, which after the Hilbert space operator formulation of quantum mechanics is the most impressive examples of a
perfect matching of physical and mathematical concepts: the modular (TomitaTakesaki) theory of operator algebras and its unifying awe-abiding power to
relate statistical mechanics, quantum field theoretical localization and the local quantum physical reason detre for the emergence of internal and external
symmetries from general properties of operator algebras. Whereas the Hilbert
space theory was completely in place at the time of the discovery of QM and the
physicists only had to learn it26 , the development of the modular theory in the
24 In this argument there is an underlying tacit assumption that any quantum requirement
(even if it is presently not completely understood) is more fundamental than a classical argument i.e. quantum renormalizability explains the classical gauge principle in the quasiclassical
limit.
25 This is reasonable because any quantum selection principle, even if presently insufficiently
understood, is more basic than a classical principle (even if completely understood)
26 It is often overlooked that it was Fritz London [26] (and not John von Neumann) who

22

middle 60s was a genuine give and take between physics and mathematics. In the
physical side it was Haag, Hugenholz and Winnink in their pursuit of statistical
mechanics for open systems (i.e. directly in the infinite thermodynamic infinite
volume limit, bypassing quantization boxes) where the true equilibrium thermodynamics becomes valid without finite volume correction terms, who came
up with this new mathematical structure in a very spacial setting (physicists
usually do not aim at the generality of their arguments). One of their findings
was that in the thermal state the generator of the time translation has a twosided symmetric spectrum and that this phenomenon is related to a symmetry
between the observable algebra and its commutant. They also identified the
so called KMS property as the characteristic thermal equilibrium property, thus
giving a fundamental significance to an observation which in the hands of Kubo,
Martin and Schwinger was merely an analytic technical trick to avoid computing
traces for Gibbs states. On the mathematical side Tomita succeeded to push
the modular property of group algebras (associated with the relation between
right and left Haar measures) right into the heart of general operator algebras;
Takesaki extended the theory and together with Winnink (shortly after the protagonists from mathematicians and physics met at a conference 1967 in Baton
Rouge) made the relation with the physicists KMS structure whereas Connes
used this new concepts (including the physicists terminology) to significantly extend the classification of factor algebras started by Murray and von Neumann.
It took another decade before the physicist Wichmann together with his young
PhD student and collaborator Bisognano discovered a geometrical aspect of that
theory in the setting of localized subalgebras of QFT; in this case they proved
that if the local algebras associated with spacetime regions are generated by
pointlike fields, then the modular data of a wedge-localized algebra consists of
the modular group being identical to the wedge-preserving Lorentz boost and
the Tomita involution being related to the reflection on the edge of the wedge
which (up to a -rotation around the spatial axis perpendicular to the edge)
is implemented by the famous TCP operator of QFT. The thermodynamical
KMS aspect attributed to the Lorentz boost (upon restriction of the vacuum
state to the wedge algebra) appeared at first sight a bit unusual. But after
Sewell was able to connect this observation with another unexpected property
namely Hawkings thermal radiation of black holes, and after Unruh analysed
the gedankenexperiment involving an uniformly accelerated observer in the field
theoretic vacuum, both situations became considerably de-mystified by finding
their common root. The following years were characterized by a deepening of
structural insight about operator algebraic modularity and its connections with
geometrical aspects of symmetry and localization (modular localization) as well
as thermal manifestation (localization thermality).
Within recent years the new framework of modular localization has entered
a new phase. Whereas until recently it mainly served to study rather general
for the first time introduced this concepts of Hilbert space and rotations in Hilbert space
(unitary operators) into quantum physics. Since at that time he was an assistant at the TU
Stuttgart, he was outside the quantum dialog and his paper went into oblivion despite an
enthusiastic citation in Jordans paper on transformation theory [27].

23

structural properties and prove mathematical theorems for families of QFTs

which shared certain properties, it became clear that it containes the long
sought-after concepts for a new constructive approach. This approach for the
first time in the history of QFT does not use the (quasi)-classical crutches of
Lagrangian quantization. One of its first successful tests was to fill the above
mentioned loopholes left by certain positive energy Wigner representation whose
localization aspects cannot be described in terms of pointlike fields and which
escape Lagrangian quantization attempts.
Another successful application of modular localization led to a profound conceptual understanding of the bootstrap-formfactor construction for 2-dimensional
factorizing models. Several years after the dreams of a unique S-matrix bootstrap came to an abrupt halt and its remainders became recycled into string
theory of everything, some physicists made the modest observation that the
S-matrix bootstrap can be made to work for special families of two-dimensional
integrable models, the so-called factorizing models [25]. Far from being a
TOE, the S-matrix bootstrap for factorizing S-matrices turned out to be the
start of something really interesting. The simpler structure of unitarity, crossing and covariance in the context of factorization permits not only to classify
factorizing S-matrices, but also to start a formfactor formalism which leads to
a unique determination of formfactors of would-be QFT models whose largetime asymptotic limits reproduce the bootstrap S-matrix. The implementation
of crossing followed the field theoretic expectation (crossing as a collective effect
of all intermediate states) and not the string theoretic duality (crossing solely
in terms of infinite one-particle towers).
Most of these results were obtained by adding additional recipes to the LSZ
framework of scattering theory which then led to unique multiparticle formfactors whose correctness was supported by consistency checks. These recipes
complicated however the recognition of the conceptual positioning of factorizing models within the general setting of QFT, which is an important issue if
one wants to use them as a theoretical laboratory for a nonperturbative access to QFT and not just for their own sake. It turns out that these models are characterized by the existence of vacuum-polarization-f ree generators
(PFG) for wedge-localized algebras27 whose Fourier transforms are related to
Zamolodchikov-Faddeev algebras. Modular localization and interaction-caused
vacuum polarization are inexorably related to causal locality i.e. to those properties which are at the root of the sharp separation between QFT and any
form of relativistic QM. It is profoundly astonishing that the interaction-caused
vacuum polarization, whose inexorable structural presence was first noted in
the standard setting of QFT way back by Furry and Oppenheimer (and which
has been the cause of conceptional complexity in the particle-field relation), is
now returning as a central quantitative concept in the nonperturbative intrinsic
construction of models.
Although in more realistic QFTs there are no PFGs which fulfill the phys27 The wedge region is the smallest region for which the existence of PFGs is compatible
with the presence of interactions.

24

ically required operator domain properties [22], the idea that there may be
a classification of (non-PFG, but nevertheless constructively accessible) generators of wedge algebras even in the general case is very suggestive. Such
generators are non-local in the old sense, but their purpose is not the construction of a nonlocal QFT28 but rather to use the structural simplicity gained by
relaxing localization29 in intermediate steps of the construction. Modular localization theory assigns a preferred role to operator algebras associated with
spatial wedge regions; in some sense which can be made precise wedge algebras
implement the best compromise between particles and fields. As in the Lagrangian quantization approach the perturbative construction of a model is in
principle determined once one specified the Lagrangian, the QFT in the modular localization setting is uniquely determined in terms of the structure of its
wedge algebras (the position of the wedge algebra within the algebra of all operators, or the algebraic structure of generators). The algebras for smaller regions
(spacelike cones, double cones) are computed in terms of algebraic intersections
of wedge algebras. Unlike the Lagrangian approach there is no place where
ultraviolet divergences can occur; the worst which can happen is that the intersection of wedge algebras associated with double cone algebras is trivial (a
multiple of the identity). In that case the expected local QFT associated with
the wedge data simply does not exist. In the Lagrangian approach there are
no credible criteria for existence, but nobody really expects that Lagrangian
models with nonrenormalizable short distance behavior can be associated with
mathematically existing QFTs.
It is extremely interesting to note that modular localization unites three
important unsolved ideas from the past and places them into a new perspective.
The idea that a construction which avoids correlation functions in favor of
the S-matrix and formfactors is necessarily free of ultraviolet divergencies;
this is the extension of the old S-matrix bootstrap.
The idea of using instead of the rather singular quantum fields the local
net of spacetime-indexed operator algebras which they generate. This step
is analogous to the coordinate-independent setting achieved in modern
differential geometry. It has been advocated by Haag since the end of the
50s and as a result of a collective effort over several decades it grew into
the body of what is now called algebraic QFT or Local Quantum Physics
[4].
The spirit underlying the Wigner representation theoretical approach to
interaction-free theories. This provides an illustration of intrinsicness (no
28 The

attempts to construct physically interpretable (macro-causal) nonlocal QFTs has

been marred by serious causality obstacles through all its history starting from the early
Moeller-Kristensen work to the present attempts of noncommutative (and therefore necessarily
nonlocal) theories [21].
29 The Zamolodchikov-Faddeev algebra operators furnish the simplest illustration of such
wedge-algebra generators. The use of simple non-local operators for the construction of localized algebras is not limited to two dimensions [20].

25

classical crutches) in a very special context, but in contrast to algebraic

QFT, whose main use has been structural side, the Wigner classification
of positive energy representation and the associated algebraic net in Fock
space is totally constructive. One would like to be able to carry these
ideas into the realm of interactions without being forced to introduce nonintrinsic field coordinatizations
In some way this could be considered as a renovation of some ideas of the
old bootstrap S-matrix approach on a superior conceptual and mathematical
level, but now not in antagonism to the off-shell localization concepts of QFT
but rather as a new conceptional enrichment. For two-dimensional models a
rigorous criterion for the solution of the age old problem of mathematical existence (which perturbative Lagrangian quantization theory for known reasons
cannot deliver) in form of a checkable degree of freedom property (analogous
to modular nuclearity) for wedge algebras has been formulated and successfully
tested for factorizing models [29]. Since the knowledge of wedge algebras (i.e.
generating operators) is known to guaranty uniqueness of the possibly associated local QFT, it is reasonable to expect that there exists an algebraic property
related to the cardinality of degrees of freedom which also in the general case
secures the nontriviality of local intersections. As in the old days where the
existence of a theory was threatened by the bad short distance behavior in the
integration over fluctuations of singular pointlike field coordinates, in the new
completely coordinatization- and singularity-free intrinsic setting the quantum
phase space degree of properties decide whether a particular starting wedge
algebra structure leads to a QFT or not.
Perhaps it is helpful at this point to remind the reader that some of the great
theoretical discoveries of last century were made by re-thinking and re-working
already existing theoretical elements. In Einsteins discovery of special relativity, the formula for Lorentz transformation and its relation to Maxwells equation
was already noticed before but its revolutionary message for a new concept of
spacetime was not seen30 . The great achievment of perturbative renormalization did not consist in discovering new principles underlying QFT, but rather
in implementing the known principles by new concepts and more appropriate
computational technics which allowed to control certain aspects of ultraviolet
divergences in terms of a more physical parametrization. Technically speaking
the old formulation of QFT which was heavily burdened by the quantum mechanics formalism (Wenzel, Heitler) was traded with a manifestly relativistic one
which placed the need to distinguish between Lagrangian (unrenormalized) parameters and physical ones into a clearer perspective. Although both, QM and
QFT are quantum theories, the inexorable presence of vacuum polarization and
the resulting significant structural changes make it impossible to picture QFT
as some kind of relativistic quantum mechanics, a fact which even nowadays is
30 There is a controversy among historians if Einstein who was on top of the ideas around
the turn of the 19 century may have known about these formulas perhaps unconciously. In
any case his derivation in his new spacetime interpretation is totally novel and extremely
beautiful.

26

often overlooked [35].

Although QFT is an amazingly successful framework when it comes to
agreement with experiments, the lack of any mathematically controllable model
(apart from a special low-dimensional cases) and the absence of a comprehensive
constructive strategy which explores the underlying principles without classical crutches sets it apart from QM and most other area in theoretical physics.
The crucial question, which arises from the aforementioned historical remarks,
is whether a theory which has been around for more than 70 years and undergone already one revolutionary change of its formalism allows yet another postrenormalization revolutionary step which would convert it into a autonomous
theory like previous theories. The existence of a radically different formulation which de-emphasizes field-coordinatizations in favor of spacetime indexed
algebras [4] is a very strong hint in this direction31 .
A particularly radical result in comparison with the standard Lagrangian
setting (which shows that good old QFT still good for amazing surprises) is the
possibility to describe a full-fledged QFT with all its structural richness in terms
of a finite number of monades. Monades are copies of one unique object in a
common environment such that all physics is encoded in the relative positions
of these copies. If one interprets the word monade in this physical realization
of Leibnizs philosophy as the unique (up to isomorphism) hyperfinite type III1
Murray von Neumann factor32 , the environment as a joint Hilbert space in which
this operator algebra sits in different positions and if furthermore these relative
positions are defined in an appropriate way in terms of modular operator algebra
concepts (modular inclusions and intersections with a joint vacuum), then the
existence of a Poincare (or conformal) spacetime symmetry group and of a net
of local algebras (generated from the action of these symmetries on the monads) are consequences33 . However such new structural insight cannot directly
be converted into new ways of classifying and constructing models, they are
presently in a too abstract setting. Constructing (extended) generators of local
algebras by modular theory treads new ground and (as mentioned before); in
the special context of factorizing models where the generators of wedge algebras
are of a particular simple kind (closely related to the Zamolodchikov-Faddeev
algebra) one already has a nice understanding about their construction in the
spacetime setting of modular localization. A particular rewarding case to test
the new algebraic concepts is the setting of chiral conformal theories. Here the
aim of present investigations is to classify and construct such theories in terms
of specialization of general principles and concepts of QFT rather than bas31 Many of the concepts (modular inclusions, modular intersections) have been developed
by H.-W. Wiesbrock [28] who, despite his amazing original work, had no chance to continue
his academic career at a time when most university positions went to string theorist.
32 This algebra, which appears naturally in the setting of quantum field theoretic localization
and the ensuing vacuum polarization, does not appear in QM and hence underlines the big
structural differences between QM and QFT.
33 The protagonist of this deep work on modular inclusions and modular intersections had
no chance to develop an academic carreer, whereas his contemporaries who also started in
QFT, but then realized that a career in particle physics is only possible through string theory,
obtained high academic positions.

27

ing constructions on mathematical structures (e.g. loop groups) which do not

appear to admit a higher dimensional counterpart (see however [30]).
A recently solved interesting problem of QFT which required a conceptual
insight beyond the standard setting is the quantum adaptation of Einsteins local covariance principle to QFT in curved spacetime. The reason why it took
such a long time to understand this issue is that the local (patch-wise) isometric
diffeomorphisms of the classical theory have no straightforward implementation
on the level of quantum states (as compared to the unitarily implemented standard global spacetime symmetries as Poincare invariance of the vacuum state
in Minkowski spacetime). The standard formalism for expectation values based
on Lagrangian action functional (or any other quantization formalism) does not
separate states from operators. Only after the algebraic approach led to such a
separation one learned how to do formulate quantum local covariance [31]. From
its algebraic functorial formulation in terms of a functor which relates a category
of causal manifolds with a category of certain algebras, the old problem one had
with states became clear: states are dual to algebras. When one dualizes the
algebraic statement one finds that only folei of states are invariant, the quantum local covariance does not leave them individually unchanged. The upshot
of these investigations34 is a new way of looking at QFT: instead of considering
quantum fields on prescibed Lorentzian causally complete (globally hyperbolic)
manifolds, a field theory model in the new setting is a functor between all
causally complete manifolds and an operator algebraic category (e.g. the Weyl
algebra, the CAR algebra,....). A resulting interesting question is whether each
QFT in the old sense is extendible to one in this new sense, i.e. whether the
algebraic substrate of a QFT on Minkowski spacetime can be resettled (as a net
with a new spacetime indexing) onto any globally hyperbolic manifold. If one
takes free field algebras in their operator algebraic setting (CCR, CAR) this can
be shown but in a more general context this is a very difficult open problem.
In the simplest case of chiral algebras this amounts to the question whether a
Moebius covariant net of algebras allows an extension to arbitrary Diff(S). This
is of course always possible if one assumes that the chiral net has an affiliated
stress energy tensor, so in order to learn something which may be helpful for
higher dimensional QFT one should like to ask for physically plausible properties of the defining system of operator algebras which insure the existence of
this extension. This leads to a rather special version of this question is whether
each chiral algebra in the sense of Moebius invariance can be extended to a
theory with Diff(S1 ) symmetry (without assuming the existence of an energymomentum tensor). It is evident that if one cannot solve this simple question in
this special case there will be no chance for the more general case. There seem
to be reasonable assumptions on modular positioning which indeed permit to
generalize the derivation of Moebius covariance to Diff(S) covariance.
One interesting side result of these remarkable conceptual advances (which
remained largely unnoticed on the wayside of the great string-dominated cara34 A recently observed fascinating possibility is that the perturbative version of local covariance may actually lead to a perturbative background independence [32].

28

van) is that the presently much discussed estimates of the field vacuum contributions to the cosmological constant (more than 50 orders of magnitude too large)
are in contradiction with the aforementioned quantum version of Einsteins local
covariance. The latter does not support the naive zero point energy counting arguments as in [33] which treat the vacuum as a relativistic quantum mechanical
level system. These arguments have been uncritical used by many particle physicists (example [34]). In a very interesting paper Hollands and Wald [35] show
that the local covariance setting of QFT contradicts such relativistic quantum
mechanical picture of filling momentum space levels (in agreement with similar
remarks made above) which is in harmony with the idea the momentum space
(Fourier transform) only acquires its physical interpretation through covariant
localization and not the other way around. Unfortunately the incorrect idea that
QFT is some sort of relativistic quantum mechanics is extremely widespread,
so that their arguments probably will not get the attention which they deserve.
Needless to mention that generically curved spacetime reference states which
replace the Minkowski vacuum do lead to nonvanishing expectation of the correctly (in agreement with the local covariance principle) defined stress-energy
tensor. There is however a new coupling parameter involving the curvature and
within a curved spacetime setting one has to make assumptions about its numerical strength. Hollands and Wald show that the problem of a theory based
on a energy-stress tensor quantized according to the requirement of local covariance does not lead to such gigantic values for the cosmological constant.
In fact their problem is the opposite, under reasonable assumptions about that
unknown coupling the vacuum polarization contribution comes out too small
in order to account for the astrophysical data. More reliable estimates which
could clarify whether there is a genuine problem would need a better conceptual
understanding of the local and global properties of the involved quantum states.
One can only hope that observational astrophysics maintains its autonomy and
does not go the way of some parts of experimental high energy physics in whose
evaluation boards are influenced by string theorists (who told them to measure
the manifestations of their little curled up dimensions and to look for a breaking
of Lorentz invariance).
The string-theorists only game in town claim is based on the belief that
the main content of QFT is already known. But if a theory allows for such a
radically different conceptual setting as I have indicated in this section, it is
quite far from having reached its closure. It rather seems to call for another
post-renormalization revolutionary step before it can reach its final form. There
are of course many other topics not mentioned here which equally well show that
the no other game in town claim is media hype for personal promotion and
to the detriment of fundamental physics.
Foremans ill-fated attempt [13] to relate revolutionary changes in the quantum mechanics of the 1920s to the post world war I sociological Zeitgeist and
the Weimar republic political situation was perhaps premature. A more suitable
subject for such a comparison of physics and the Zeitgeist should be provided
by the string revolution at the end of the past century. Whereas quantum
mechanics is one of physics most viable description of reality, whose unusual as29

pect of probability and chance on a fundamental level has nothing to do which

the Spengler type gloom and doom of the Weimar republic, the strong connection between characteristic features of string theory with the Zeitgeist behind
present political events is too evident to be overlooked. They both share the
same strong hegemonic spirit of insatiable ideological domination over the more
rational traditions. The only game in town on one side corresponds to the
lore about weapons of mass destruction on the other side35 and those mentioned string theoretic redefinitions of words which originally had a well-defined
intrinsic physical meaning correspond to the new definitions of words like torture or rendition, since this makes the break with the past appear less drastic
and increases acceptance by the democratic majority.
As long as the internal democratic traditions are still intact there is some
hope that the present bellicose political confrontation may convert back to compromise, diplomacy and a genuine interest in other peoples cultures instead of
the hegemonic compulsion to impose ones own way of life and political system
on others.
The crisis in particle physics, which finds its most visible outing in the hegemony of string theory, is of a more enduring kind. As long as some leading
physicists, including Nobel prize winners, are failing to play their natural role
as critical observers (in contrast to their more critical predecessors as Pauli who
kept particle physics in a healthy rational state), the present situation will continue and may even deteriorate. Without the active support fringe ideas could
not find such big communities of followers. This phenomenon is not any more
limited to string theory. Nobel prize winner Murray Gell-Man is duing his best
in order help create a new fad in statistical mechanics [36]. Boltzmann and
Gibbs would rotate in their graves. Current practitioners, like Lebowitz and
Lieb, have not made a secret of their unhappiness with this development.
The difference to Einstein and Heisenberg, who also in the later parts of
their life worked many decades on failed theories, is that it was clearer to their
colleagues and to the general public at that time that a Nobel prize is given for
having right ideas at the right time (maybe also at the right place) and does not
come with a guaranty of life-long brilliance. Hence they were protected from
harming physics after their pivotal contributions; this situation has changed as a
result of increased media hype. Pauli worked initially together with Heisenberg
on the failed nonlinear spinor theory but when (on the occasion of a lecture tour
through the US) he was criticised by Feynman, he immediatly saw the untenable
points; the reason why upon his return there was a lot of friction with Heisenberg
was that he could not convince his former collaborator that they were on the
wrong track. During Paulis unfortunately short lifetime his cutting sarcasm (he
did not spare himself36 ) and polemic style kept particle physics in an excellent
35 The illustrated parallelity is of course not meant to imply that string theorist are necessarily followers of the Bush administrations policies.
36 In October 1958 he came to Hamburg to give a colloquium lecture on the neutrino. After
his talk Harry Lehmann (at that time my advisor) took him to his office which contained a
turning chair, called the Pauli chair (the one he used at the time of his exclusion principle
epiphany). He was totally exhausted and looked very ill. he started to sway his head like

30

healthy state even though he sometimes took the wrong side. This healthy
state lasted almost three decades because people as Jost, Lehmann, Landau,
and Coleman (at least during his years of good health) and others kept its
critical spirit alive. It begun to wane with the death of these great people as
well as with the demise of the Soviet Union and the end of the era of the great
schools of innovative ideas. The co-option of the physicists from these places
into the US-led globalized market economy considerably weakened the search
of truth as one of the primary motives for doing science. With woefulness one
reads the famous 1918 address Principles of research of Einstein in honour of
Planck:
In the temple of science there are many mansions, and various indeed are
they that dwell therein and the motives that have led them thither. Many take
to science out of a joyful sense of superior intellectual power; science is their
own special sport to which they look for vivid experience and the satisfaction of
ambition; many others are found in the temple who have offered the products
of their brains on this altar for purely utilitarian purposes. Were an angel of
the Lord to come and drive all the people belonging to this two categories out
of the temple, the assembledge would be seriously depleted, but there would still
be some man, both of present and past times, left inside. Our Planck is one of
them, and that is why we love him.
With a large number of chairs at theoretical physics departments worldwide
being occupied by string theorists I do not see much hope. There will be more
big Latin Letters and string protegees (of the kind mentioned in the first section)
to come, and the chance for a radical change of direction through newcomers
entering particle physics will remain extremely dim for several reasons . On the
one hand the situation is conceptually and mathematically much more demanding than it was e.g. a century ago when Einstein wrote his epoch creating 1905
papers; at that time the starting situation was not influenced by the long reign
of a confusing theory. There were many schools of thought, but no hegemonic
control of research directions and institutions via impact indices and globalized
interest groups. Nowadays somebody who has the capabilities and the guts to
resist the lure of the string hegemon in pursuit of his own original ideas will run
a high risk to see the end of his academic career with no old-fashioned patent
office around which could serve as a temporary shelter.
Note added : I think that the use of a polemic style in physics should be
limited to situations of utmost emergency. Polemics, if it is good, sometimes has
the power of opening frozen minds and catalyses a scientific light bulb moment,
but it is of course totally useless with people who allowed fundamentalism to
an orthodox Jew (some people called that his ground state frequency) and said Ja, Ja, ich
glaube der Heisenberg liegt mir noch schwer im Magen. He died 6 weeks later of pancreatic
cancer.
He very much liked Harry Lehmann and considered him the leader of the Feldverein.
Harry Lehmann had a very similar temper. Once in a Festvortrag by Carl-Friedrich von
Weiszaecker about the Ure (two-valued spin alternatives), when von Weiszaecker estimated
the number of Ure in the universe and came up with a number 1060 (or alike), Lehmann
said in a voice which was well audible in his surroundings: minus one, please count me out
because I made my decision right now.

31

take over. A perfect example of scientific polemics at its best is Res Josts
article [12]. This was written at the hight of the S-matrix bootstrap fashion and
I think it did at least something to clear the air. The situation is much harder
now, and if my efford is in vain, it is probably not only because it lacks the
elegance and coherence of Josts. But there is no question that this emergency
situation has come; we are at a crossroad where the unity of particle physics
(including the universally binding conceptual basis for rational communications
between different approaches) is at stake. In most previous fashions (I am in
particular thinking about the S-matrix bootstrap) top-notch places as MIT and
some ivy league universities were left undented. It is an indication of the depth
of the present crisis that this does not seem to be the case any more. Perhaps
one needs a kind of academic Noahs ark which could play for nonconformist
innovators an analog role as the Bern patent office did for Einsteins career.

References
[1] L. Susskind, The Cosmic Landscape: String Theory and the Illusion of
Intelligent Design
[2] O. Darrigol, The origin of quantized matter fields, Hist. Stud. Phys. Sci.
16/2, 198
[3] P. Jordan, The Present State of Quantum Electrodynamics, in Talks and
Discussions of the Theoretical-Physical Conference in Kharkov (May 19.25., 1929) Physik.Zeitschr.XXX, (1929) 700
[4] R. Haag, Local Quantum Physics, Springer 1996
[5] B. Schroer, Ann. of Phys. 319, (2005) 48, and references therein
[6] M. Schottenloher, A mathematical introduction to conformal field theory,
Springer 1997
[7] http://www.lns.cornell.edu/spr/2003-12/msg0057032.html
[8] www.Wikipedia\String theory-Wikipedia, the free encyclopedia
[9] E. Martinec, Class. Quant. Grav. 10, (1993) L187
[10] J. Dimock, J. Math. Phys. 41, (2000) 40
[11] D. Bahns, J.Math.Phys. 45 (2004) 4640
[12] Res Jost, TCP-Invarianz der Streumatrix und interpolierennde Felder, Helvetica Physica Acta 36, (1963) 77
[13] P. Foreman, Weimar culture, causality and quantum theory1918-1927:
Adaptation by German physicists and mathematicians to a hostile intellectual environment. Hist. Stud. Sci, 3:1-115

32

[14] click Theodor Adorno in Wikipedia: http://www.wikipedia.org/

[15] K. Vonnegut, The only Game in Town, Natural History, Winter 2001
[16] see http://www.math.columbia.edu/woit/
[17] J. Yngvason, Commun. Math. Phys. 18, (1970) 195
[18] J. Mund, B. Schroer and J. Yngvason, Phys. Lett. B596, (2004) 156
[19] J. Mund, B. Schroer and J. Yngvason, String-localized quantum fields and
modular localization, math-ph/0511042 , to be published in CMP.
[20] D. Buchholz and S. J. Summers, StringLocalized Fields in a Strongly Nonlocal Model, math-ph/0512060
[21] B. Schroer, Ann. of Phys. 319 (2005) 92
[22] H.J. Borchers, D. Buchholz and B. Schroer, Commun. Math. Phys. 219,
(2001) 125, hep-th/0003243
[23] J. Mund, Lett. Math. Phys. 43, (1998) 319
[24] B. Schroer, Ann. Phys. 319, (2005) 48 and previous contributions cited
therein
[25] M. Karowski and P. Weisz, Phys. Rev. B 139, (1978) 445
[26] F. London, Zeitschr. fuer Physik 40 (1926) 193
[27] P. Jordan, Zeitschr. f
ur Physik 40, (1926) 809
[28] R. Kaehler and H.-W.. Wiesbrock, JMP 42, (2000) 74
[29] G. Lechner, An Existence Proof for Interacting Quantum Field Theories
with a Factorizing S-Matrix, math-ph/0601022
[30] Nikolay M. Nikolov, Ivan T. Todorov, Lectures on Elliptic Functions and
Modular Forms in Conformal Field Theory, math-ph/0412039
[31] R. Brunetti and K. Fredenhagen, The generally covariant locality principle A new paradigm for local quantum physics,Commun.Math.Phys. 237
(2003) 31
[32] R. Brunetti and K. Fredenhagen, Towards a background independent formulation of perturbative quantum gravity, gr-qc/0603079
[33] S. Weinberg, Rev. Mod. Phys. 61, (1989) 1
[34] N. Straumann, On the Cosmological Constant Problems and the Astronomical Evidence for a Homogeneous Energy Density with Negative
Pressure,astro-ph/0203330

33

[35] Hollands and R. E. Wald, Quantum Field Theory Is Not Merely Quantum Mechanics Applied to Low Energy Effective Degrees of Freedom,
Gen.Rel.Grav. 36 (2004) 2595
[36] M. Gell-Mann and C. Tsallis, cond-mat/0509229 and special issue of Europhysics news 36, Nr.6, Nov.-Dec. 2005

34