Beruflich Dokumente
Kultur Dokumente
Jean-Pierre Francoise
Universite P.-M. Curie, Paris VI
Paris, France
Gregory L. Naber
Drexel University
Philadelphia, PA, USA
Sergio Albeverio
Rheinische Friedrich-Wilhelms-Universitat Bonn
Bonn, Germany
Lisa Jeffrey
University of Toronto
Toronto, Canada
Huzihiro Araki
Kyoto University
Kyoto, Japan
T.W.B. Kibble
Imperial College of Science, Technology and Medicine
London, UK
Abhay Ashtekar
Pennsylvania State University
University Park, PA, USA
Antti Kupiainen
University of Helsinki
Helsinki, Finland
Andrea Braides
Universita` di Roma Tor Vergata
Roma, Italy
Shahn Majid
Queen Mary, University of London
London, UK
Francesco Calogero
Universita` di Roma La Sapienza
Roma, Italy
Barry M. McCoy
State University of New York Stony Brook
Stony Brook, NY, USA
Cecile DeWitt-Morette
The University of Texas at Austin
Austin, TX, USA
Hirosi Ooguri
California Institute of Technology
Pasadena, CA, USA
Artur Ekert
University of Cambridge
Cambridge, UK
Roger Penrose
University of Oxford
Oxford, UK
Giovanni Gallavotti
Universita` di Roma La Sapienza
Roma, Italy
Pierre Ramond
University of Florida
Gainesville, FL, USA
Simon Gindikin
Rutgers University
Piscataway, NJ, USA
Tudor Ratiu
Ecole Polytechnique Federale de Lausanne
Lausanne, Switzerland
Gennadi Henkin
Universite P.-M. Curie, Paris VI
Paris, France
Rudolf Schmid
Emory University
Atlanta, GA, USA
Allen C. Hirshfeld
Universitat Dortmund
Dortmund, Germany
Albert Schwarz
University of California
Davis, CA, USA
Yakov Sinai
Princeton University
Princeton, NJ, USA
Vladimir Turaev
Institut de Recherche Mathematique Avancee,
Strasbourg, France
Herbert Spohn
Technische Universitat Munchen
Munchen, Germany
Gabriele Veneziano
CERN, Gene`ve, Switzerland
Stephen J. Summers
University of Florida
Gainesville, FL, USA
Roger Temam
Indiana University
Bloomington, IN, USA
Reinhard F. Werner
Technische Universitat Braunschweig
Braunschweig, Germany
C.N. Yang
Tsinghua University
Beijing, China
Craig A. Tracy
University of California
Davis, CA, USA
Eberhard Zeidler
Max-Planck Institut fur Mathematik in
den Naturwissenschaften
Leipzig, Germany
Andrzej Trautman
Warsaw University
Warsaw, Poland
Steve Zelditch
Johns Hopkins University
Baltimore, MD, USA
FOREWORD
n bygone centuries, our physical world appeared to be filled to the brim with mysteries. Divine powers
could provide for genuine miracles; water and sunlight could turn arid land into fertile pastures, but the
same powers could lead to miseries and disasters. The force of life, the vis vitalis, was assumed to be the
special agent responsible for all living things. The heavens, whatever they were for, contained stars and other
heavenly bodies that were the exclusive domain of the Gods.
Mathematics did exist, of course. Indeed, there was one aspect of our physical world that was recognised to
be controlled by precise, mathematical logic: the geometric structure of space, elaborated to become a genuine
form of art by the ancient Greeks. From my perspective, the Greeks were the first practitioners of mathematical
physics, when they discovered that all geometric features of space could be reduced to a small number of
axioms. Today, these would be called fundamental laws of physics. The fact that the flow of time could be
addressed with similar exactitude, and that it could be handled geometrically together with space, was only
recognised much later. And, yes, there were a few crazy people who were interested in the magic of numbers,
but the real world around us seemed to contain so much more that was way beyond our capacities of analysis.
Gradually, all this changed. The Moon and the planets appeared to follow geometrical laws. Galilei and
Newton managed to identify their logical rules of motion, and by noting that the concept of mass could be
applied to things in the sky just like apples and cannon balls on Earth, they made the sky a little bit more
accessible to us. Electricity, magnetism, light and sound were also found to behave in complete accordance
with mathematical equations.
Yet all of this was just a beginning. The real changes came with the twentieth century. A completely new
way of thinking, by emphasizing mathematical, logical analysis rather than empirical evidence, was pioneered
by Albert Einstein. Applying advanced mathematical concepts, only known to a few pure mathematicians, to
notions as mundane as space and time, was new to the physicists of his time. Einstein himself had a hard
time struggling through the logic of connections and curvatures, notions that were totally new to him, but are
only too familiar to students of mathematical physics today. Indeed, there is no better testimony of Einsteins
deep insights at that time, than the fact that we now teach these things regularly in our university classrooms.
Special and general relativity are only small corners of the realm of modern physics that is presently being
studied using advanced mathematical methods. We have notoriously complex subjects such as phase transitions in
condensed matter physics, superconductivity, BoseEinstein condensation, the quantum Hall effect, particularly
the fractional quantum Hall effect, and numerous topics from elementary particle physics, ranging from fibre
bundles and renormalization groups to supergravity, algebraic topology, superstring theory, CalabiYau spaces
and what not, all of which require the utmost of our mental skills to comprehend them.
The most bewildering observation that we make today is that it seems that our entire physical world
appears to be controlled by mathematical equations, and these are not just sloppy and debatable models, but
precisely documented properties of materials, of systems, and of phenomena in all echelons of our universe.
Does this really apply to our entire world, or only to parts of it? Do features, notions, entities exist that are
emphatically not mathematical? What about intuition, or dreams, and what about consciousness? What
about religion? Here, most of us would say, one should not even try to apply mathematical analysis, although
even here, some brave social scientists are making attempts at coordinating rational approaches.
No, there are clear and important differences between the physical world and the mathematical world.
Where the physical world stands out is the fact that it refers to reality, whatever reality is. Mathematics is
the world of pure logic and pure reasoning. In physics, it is the experimental evidence that ultimately decides
whether a theory is acceptable or not. Also, the methodology in physics is different.
A beautiful example is the serendipitous discovery of superconductivity. In 1911, the Dutch physicist Heike
Kamerlingh Onnes was the first to achieve the liquefaction of helium, for which a temperature below 4.25 K
had to be realized. Heike decided to measure the specific conductivity of mercury, a metal that is frozen solid
at such low temperatures. But something appeared to go wrong during the measurements, since the volt
meter did not show any voltage at all. All experienced physicists in the team assumed that they were dealing
with a malfunction. It would not have been the first time for a short circuit to occur in the electrical
equipment, but, this time, in spite of several efforts, they failed to locate it. One of the assistants was
responsible for keeping the temperature of the sample well within that of liquid helium, a dull job, requiring
nothing else than continuously watching some dials. During one of the many tests, however, he dozed off.
The temperature rose, and suddenly the measurements showed the normal values again. It then occurred to
the investigators that the effect and its temperature dependence were completely reproducible. Below 4.19
degrees Kelvin the conductivity of mercury appeared to be strictly infinite. Above that temperature, it is
finite, and the transition is a very sudden one. Superconductivity was discovered (D. van Delft, Heike
Kamerling Onnes, Uitgeverij Bert Bakker, Amsterdam, 2005 (in Dutch)).
This is not the way mathematical discoveries are made. Theorems are not produced by assistants falling
asleep, even if examples do exist of incidents involving some miraculous fortune.
The hybrid science of mathematical physics is a very curious one. Some of the topics in this Encyclopedia
are undoubtedly physical. High Tc superconductivity, breaking water waves, and magneto-hydrodynamics,
are definitely topics of physics where experimental data are considered more decisive than any high-brow
theory. Cohomology theory, DonaldsonWitten theory, and AdS/CFT correspondence, however, are examples
of purely mathematical exercises, even if these subjects, like all of the others in this compilation, are strongly
inspired by, and related to, questions posed in physics.
It is inevitable, in a compilation of a large number of short articles with many different authors, to see quite a
bit of variation in style and level. In this Encyclopedia, theoretical physicists as well as mathematicians together
made a huge effort to present in a concise and understandable manner their vision on numerous important
issues in advanced mathematical physics. All include references for further reading. We hope and expect that
these efforts will serve a good purpose.
Gerard t Hooft,
Spinoza Institute,
Utrecht University,
The Netherlands.
PREFACE
foundation of advanced undergraduate courses, as we know that some authors have already made such use of them.
In addition to the printed version, an on-line version of the Encyclopedia is planned, which will allow both
the contents and the articles themselves to be updated if and when the occasion arises. This is probably a
necessary provision in such a rapidly advancing field.
This project was some four years in the making. Our foremost thanks at its completion go to the members
of our Editorial Advisory Board, who have advised, helped and encouraged us all along, and to all our
authors who have so generously devoted so much of their time to writing these articles and given us much
useful advice as well. We ourselves have learnt a lot from these colleagues, and made some wonderful
contacts with some among them. Special thanks are due also to Arthur Greenspoon whose technical expertise
was indispensable.
The project was started with Academic Press, which was later taken over by Elsevier. We thank warmly
members of their staff who have made this transition admirably seamless and gone on to assist us greatly in
our task: both Carey Chapman and Anne Guillaume, who were in charge of the whole project and have been
with us since the beginning, and Edward Taylor responsible for the copy-editing. And Martin Ruck, who
manages to keep an overwhelming amount of details constantly at his fingertips, and who is never known to
have lost a single email, deserves a very special mention.
As a postscript, we would like to express our gratitude to the very large number of authors who generously
agreed to donate their honorariums to support the Committee for Developing Countries of the European
Mathematical Society in their work to help our less fortunate colleagues in the developing world.
Jean-Pierre Francoise
Gregory L. Naber
Tsou Sheung Tsun
PERMISSION ACKNOWLEDGMENTS
The following material is reproduced with kind permission of Nature Publishing Group
Figures 11 and 12 of Point-vortex Dynamics
http://www.nature.com/nature
The following material is reproduced with kind permission of Oxford University Press
Figure 1 of Random Walks in Random Environments
http://www.oup.co.uk
Path Integral Methods see Functional Integration in Quantum Physics; Feynman Path Integrals
3. Cross-References
All of the articles in the Encyclopedia have been extensively cross-referenced. The cross-references, which appear at the
end of an entry, serve three different functions:
i. To indicate if a topic is discussed in greater detail elsewhere.
ii. To draw the readers attention to parallel discussions in other entries.
iii. To indicate material that broadens the discussion.
Example
The following list of cross-references appears at the end of the entry STOCHASTIC HYDRODYNAMICS
See also: Cauchy Problem for Burgers-Type Equations; Hamiltonian
Fluid Dynamics; Incompressible Euler Equations: Mathematical Theory;
Malliavin Calculus; Non-Newtonian Fluids; Partial Differential Equations:
Some Examples; Stochastic Differential Equations; Turbulence Theories;
Viscous Incompressible Fluids: Mathematical Theory; Vortex Dynamics
Here you will find examples of all three functions of the cross-reference list: a topic discussed in greater detail elsewhere
(e.g. Incompressible Euler Equations: Mathematical Theory), parallel discussion in other entries (e.g. Stochastic Differential Equations) and reference to entries that broaden the discussion (e.g. Turbulence Theories).
The eight Introductory Articles are not cross-referenced from any of the main entries, as it is expected that introductory
articles will be of general interest. As mentioned above, the Introductory Articles may be found at the start of Volume 1.
4. Index
The index will provide you with the volume and page number where the material is located. The index entries
differentiate between material that is a whole entry, is part of an entry, or is data presented in a figure or table. Detailed
notes are provided on the opening page of the index.
5. Contributors
A full list of contributors appears at the beginning of each volume.
CONTRIBUTORS
A Abbondandolo
Universita` di Pisa
Pisa, Italy
M J Ablowitz
University of Colorado
Boulder, CO, USA
S L Adler
Institute for Advanced Study
Princeton, NJ, USA
H Airault
Universite de Picardie
Amiens, France
G Alberti
Universita` di Pisa
Pisa, Italy
S Albeverio
Rheinische FriedrichWilhelms-Universitat Bonn
Bonn, Germany
L Andersson
University of Miami
Coral Gables, FL, USA and Albert Einstein Institute
Potsdam, Germany
B Andreas
Humboldt-Universitat zu Berlin
Berlin, Germany
V Araujo
Universidade do Porto
Porto, Portugal
A Ashtekar
Pennsylvania State University
University Park, PA, USA
W Van Assche
Katholieke Universiteit Leuven
Leuven, Belgium
G Aubert
Universite de Nice Sophia Antipolis
Nice, France
S T Ali
Concordia University
Montreal, QC, Canada
H Au-Yang
Oklahoma State University
Stillwater, OK, USA
R Alicki
University of Gdansk
Gdansk, Poland
M A Aziz-Alaoui
Universite du Havre
Le Havre, France
G Altarelli
CERN
Geneva, Switzerland
V Bach
Johannes Gutenberg-Universitat
Mainz, Germany
C Amrouche
Universite de Pau et des Pays de lAdour
Pau, France
C Bachas
Ecole Normale Superieure
Paris, France
M Anderson
State University of New York at Stony Brook
Stony Brook, NY, USA
V Baladi
Institut Mathematique de Jussieu
Paris, France
xiv
CONTRIBUTORS
D Bambusi
Universita` di Milano
Milan, Italy
M Blasone
Universita` degli Studi di Salerno
Baronissi (SA), Italy
C Bardos
Universite de Paris 7
Paris, France
M Blau
Universite de Neuchatel
Neuchatel, Switzerland
D Bar-Natan
University of Toronto
Toronto, ON, Canada
S Boatto
IMPA
Rio de Janeiro, Brazil
E L Basor
California Polytechnic State University
San Luis Obispo, CA, USA
M T Batchelor
Australian National University
Canberra, ACT, Australia
S Bauer
Universitat Bielefeld
Bielefeld, Germany
V Beffara
Ecole Nomale Superieure de Lyon
Lyon, France
R Beig
Universitat Wien
Vienna, Austria
M I Belishev
Petersburg Department of Steklov Institute
of Mathematics
St. Petersburg, Russia
P Bernard
Universite de Paris Dauphine
Paris, France
D Birmingham
University of the Pacific
Stockton, CA, USA
L V Bogachev
University of Leeds
Leeds, UK
L Boi
EHESS and LUTH
Paris, France
M Bojowald
The Pennsylvania State University
University Park, PA, USA
C Bonatti
Universite de Bourgogne
Dijon, France
P Bonckaert
Universiteit Hasselt
Diepenbeek, Belgium
F Bonetto
Georgia Institute of Technology
Atlanta, GA, USA
G Bouchitte
Universite de Toulon et du Var
La Garde, France
A Bovier
Weierstrass Institute for Applied Analysis and Stochastics
Berlin, Germany
H W Braden
University of Edinburgh
Edinburgh, UK
Jir Bicak
Charles University, Prague, Czech Republic
and Albert Einstein Institute
Potsdam, Germany
H Bray
Duke University
Durham, NC, USA
C Blanchet
Universite de Bretagne-Sud
Vannes, France
Y Brenier
Universite de Nice Sophia Antipolis
Nice, France
CONTRIBUTORS
J Bros
CEA/DSM/SPhT, CEA/Saclay
Gif-sur-Yvette, France
J Cardy
Rudolf Peierls Centre for Theoretical Physics
Oxford, UK
R Brunetti
Universitat Hamburg
Hamburg, Germany
R Caseiro
Universidade de Coimbra
Coimbra, Portugal
M Bruschi
Universita` di Roma La Sapienza
Rome, Italy
A S Cattaneo
Universitat Zurich
Zurich, Switzerland
T Brzezinski
University of Wales Swansea
Swansea, UK
A Celletti
Universita` di Roma Tor Vergata
Rome, Italy
D Buchholz
Universitat Gottingen
Gottingen, Germany
D Chae
Sungkyunkwan University
Suwon, South Korea
N Burq
Universite Paris-Sud
Orsay, France
G-Q Chen
Northwestern University
Evanston, IL, USA
F H Busse
Universitat Bayreuth
Bayreuth, Germany
L Chierchia
Universita` degli Studi Roma Tre
Rome, Italy
G Buttazzo
Universita` di Pisa
Pisa, Italy
S Chmutov
Petersburg Department of Steklov
Institute of Mathematics
St. Petersburg, Russia
P Butta`
Universita` di Roma La Sapienza
Rome, Italy
S L Cacciatori
Universita` di Milano
Milan, Italy
P T Callaghan
Victoria University of Wellington
Wellington, New Zealand
M W Choptuik
University of British Columbia
Vancouver, Canada
Y Choquet-Bruhat
Universite P.-M. Curie, Paris VI
Paris, France
P T Chrusciel
Universite de Tours
Tours, France
Francesco Calogero
University of Rome, Rome, Italy and Institute
Nazionale di Fisica Nucleare
Rome, Italy
Chong-Sun Chu
University of Durham
Durham, UK
A Carati
Universita` di Milano
Milan, Italy
F Cipriani
Politecnico di Milano
Milan, Italy
xv
xvi
CONTRIBUTORS
R L Cohen
Stanford University
Stanford, CA, USA
G W Delius
University of York
York, UK
T H Colding
University of New York
New York, NY, USA
G F dellAntonio
Universita` di Roma La Sapienza
Rome, Italy
J C Collins
Penn State University
University Park, PA, USA
G Comte
Universite de Nice Sophia Antipolis
Nice, France
A Constantin
Trinity College
Dublin, Republic of Ireland
D Crowdy
Imperial College
London, UK
A B Cruzeiro
University of Lisbon
Lisbon, Portugal
G Dal Maso
SISSA
Trieste, Italy
F Dalfovo
Universita` di Trento
Povo, Italy
A S Dancer
University of Oxford
Oxford, UK
P DAncona
Universita` di Roma La Sapienza
Rome, Italy
S R Das
University of Kentucky
Lexington, KY, USA
E Date
Osaka University
Osaka, Japan
N Datta
University of Cambridge
Cambridge, UK
C DeWitt-Morette
The University of Texas at Austin
Austin, TX, USA
L Diosi
Research Institute for Particle and Nuclear Physics
Budapest, Hungary
A Doliwa
University of Warmia and Mazury in Olsztyn
Olsztyn, Poland
G Dolzmann
University of Maryland
College Park, MD, USA
S K Donaldson
Imperial College
London, UK
T C Dorlas
Dublin Institute for Advanced Studies
Dublin, Republic of Ireland
M R Douglas
Rutgers, The State University of New Jersey
Piscataway, NJ, USA
M Dutsch
Universitat Zurich
Zurich, Switzerland
B Dubrovin
SISSA-ISAS
Trieste, Italy
J J Duistermaat
Universiteit Utrecht
Utrecht, The Netherlands
S Duzhin
Petersburg Department of Steklov Institute of
Mathematics
St. Petersburg, Russia
CONTRIBUTORS
G Ecker
Universitat Wien
Vienna, Austria
B Ferrario
Universita` di Pavia
Pavia, Italy
M Efendiev
Universitat Stuttgart
Stuttgart, Germany
R Finn
Stanford University
Stanford, CA, USA
T Eguchi
University of Tokyo
Tokyo, Japan
D Fiorenza
Universita` di Roma La Sapienza
Rome, Italy
J Ehlers
Max Planck Institut fur Gravitationsphysik
(Albert-Einstein Institut)
Golm, Germany
A E Fischer
University of California
Santa Cruz, CA, USA
P E Ehrlich
University of Florida
Gainesville, FL, USA
D Einzel
Bayerische Akademie der Wissenschaften
Garching, Germany
G A Elliott
University of Toronto
Toronto, Canada
G F R Ellis
University of Cape Town
Cape Town, South Africa
C L Epstein
University of Pennsylvania
Philadelphia, PA, USA
J Escher
Universitat Hannover
Hannover, Germany
J B Etnyre
University of Pennsylvania
Philadelphia, PA, USA
A S Fokas
University of Cambridge
Cambridge, UK
J-P Franc oise
Universite P.-M. Curie, Paris VI
Paris, France
S Franz
The Abdus Salam ICTP
Trieste, Italy
L Frappat
Universite de Savoie
Chambery-Annecy, France
J Frauendiener
Universitat Tubingen
Tubingen, Germany
K Fredenhagen
Universitat Hamburg
Hamburg, Germany
S Friedlander
University of Illinois-Chicago
Chicago, IL, USA
G Falkovich
Weizmann Institute of Science
Rehovot, Israel
M R Gaberdiel
ETH Zurich
Zurich, Switzerland
M Farge
Ecole Normale Superieure
Paris, France
G Gaeta
Universita` di Milano
Milan, Italy
xvii
xviii
CONTRIBUTORS
L Galgani
Universita` di Milano
Milan, Italy
H Gottschalk
Rheinische Friedrich-Wilhelms-Universitat Bonn
Bonn, Germany
G Gallavotti
Universita` di Roma La Sapienza
Rome, Italy
O Goubet
Universite de Picardie Jules Verne
Amiens, France
R Gambini
Universidad de la Republica
Montevideo, Uruguay
T R Govindarajan
The Institute of Mathematical Sciences
Chennai, India
G Gentile
Universita` degli Studi Roma Tre
Rome, Italy
A Grassi
University of Pennsylvania
Philadelphia, PA, USA
A Di Giacomo
Universita` di Pisa
Pisa, Italy
P G Grinevich
L D Landau Institute for
Theoretical Physics
Moscow, Russia
P B Gilkey
University of Oregon
Eugene, OR, USA
Ch Gruber
Ecole Polytechnique Federale de Lausanne
Lausanne, Switzerland
R Gilmore
Drexel University
Philadelphia, PA, USA
J-L Guermond
Universite de Paris Sud
Orsay, France
S Gindikin
Rutgers University
Piscataway, NJ, USA
F Guerra
Universita` di Roma La Sapienza
Rome, Italy
A Giorgilli
Universita` di Milano
Milan, Italy
T Guhr
Lunds Universitet
Lund, Sweden
G A Goldin
Rutgers University
Piscataway, NJ, USA
C Guillope
Universite Paris XII Val de Marne
Creteil, France
G Gonzalez
Louisiana State University
Baton Rouge, LA, USA
C Gundlach
University of Southampton
Southampton, UK
R Gopakumar
Harish-Chandra Research Institute
Allahabad, India
S Gutt
Universite Libre de Bruxelles
Brussels, Belgium
D Gottesman
Perimeter Institute
Waterloo, ON, Canada
K Hannabuss
University of Oxford
Oxford, UK
CONTRIBUTORS
M Haragus
Universite de Franche-Comte
Besanc on, France
D D Holm
Imperial College
London, UK
S G Harris
St. Louis University
St. Louis, MO, USA
B Hasselblatt
Tufts University
Medford, MA, USA
A Huckleberry
Ruhr-Universitat Bochum
Bochum, Germany
P Hayden
McGill University
Montreal, QC, Canada
K Hulek
Universitat Hannover
Hannover, Germany
D C Heggie
The University of Edinburgh
Edinburgh, UK
D Iagolnitzer
CEA/DSM/SPhT, CEA/Saclay
Gif-sur-Yvette, France
B Helffer
Universite Paris-Sud
Orsay, France
R Illge
Friedrich-Schiller-Universitat Jena
Jena, Germany
G M Henkin
Universite P.-M. Curie, Paris VI
Paris, France
P Imkeller
Humboldt Universitat zu Berlin
Berlin, Germany
M Henneaux
Universite Libre de Bruxelles
Bruxelles, Belgium
G Iooss
Institut Non Lineaire de Nice
Valbonne, France
S Herrmann
Universite Henri Poincare, Nancy 1
Vandoeuvre-le`s-Nancy, France
M Irigoyen
Universite P.-M. Curie, Paris VI
Paris, France
C P Herzog
University of California at Santa Barbara
Santa Barbara, CA, USA
J Isenberg
University of Oregon
Eugene, OR, USA
J G Heywood
University of British Columbia
Vancouver, BC, Canada
R Ivanova
University of Hawaii Hilo
Hilo, HI, USA
A C Hirshfeld
Universitat Dortmund
Dortmund, Germany
E M Izhikevich
The Neurosciences Institute
San Diego, CA, USA
A S Holevo
Steklov Mathematical Institute
Moscow, Russia
R W Jackiw
Massachusetts Institute of Technology
Cambridge, MA, USA
T J Hollowood
University of Wales Swansea
Swansea, UK
J K Jain
The Pennsylvania State University
University Park, PA, USA
xix
xx CONTRIBUTORS
M Jardim
IMECCUNICAMP
Campinas, Brazil
L H Kauffman
University of Illinois at Chicago
Chicago, IL, USA
L C Jeffrey
University of Toronto
Toronto, ON, Canada
R K Kaul
The Institute of Mathematical Sciences
Chennai, India
J Jimenez
Universidad Politecnica de Madrid
Madrid, Spain
Y Kawahigashi
University of Tokyo
Tokyo, Japan
S Jitomirskaya
University of California at Irvine
Irvine, CA, USA
B S Kay
University of York
York, UK
P Jizba
Czech Technical University
Prague, Czech Republic
R Kenyon
University of British Columbia
Vancouver, BC, Canada
A Joets
Universite Paris-Sud
Orsay, France
M Keyl
Universita` di Pavia
Pavia, Italy
K Johansson
Kungl Tekniska Hogskolan
Stockholm, Sweden
T W B Kibble
Imperial College
London, UK
G Jona-Lasinio
Universita` di Roma La Sapienza
Rome, Italy
S Kichenassamy
Universite de Reims Champagne-Ardenne
Reims, France
V F R Jones
University of California at Berkeley
Berkeley, CA, USA
J Kim
University of California at Irvine
Irvine, USA
N Joshi
University of Sydney
Sydney, NSW, Australia
S B Kim
Chonnam National University
Gwangju, South Korea
D D Joyce
University of Oxford
Oxford, UK
A Kirillov
University of Pennsylvania
Philadelphia, PA, USA
C D Jakel
Ludwig-Maximilians-Universitat Munchen
Munchen, Germany
A Kirillov, Jr.
Stony Brook University
Stony Brook, NY, USA
G Kasperski
Universite Paris-Sud XI
Orsay, France
K Kirsten
Baylor University
Waco, TX, USA
CONTRIBUTORS
F Kirwan
University of Oxford
Oxford, UK
M Krbec
Academy of Sciences
Prague, Czech Republic
S Klainerman
Princeton University
Princeton, NJ, USA
D Kreimer
IHES
Bures-sur-Yvette, France
I R Klebanov
Princeton University
Princeton, NJ, USA
A Kresch
University of Warwick
Coventry, UK
Y Kondratiev
Universitat Bielefeld
Bielefeld, Germany
D Kretschmann
Technische Universitat Braunschweig
Braunschweig, Germany
A Konechny
Rutgers, The State University of New Jersey
Piscataway, NJ, USA
P B Kronheimer
Harvard University
Cambridge, MA, USA
K Konishi
Universita` di Pisa
Pisa, Italy
B Kuckert
Universitat Hamburg
Hamburg, Germany
T H Koornwinder
University of Amsterdam
Amsterdam, The Netherlands
Y Kuramoto
Hokkaido University
Sapporo, Japan
P Kornprobst
INRIA
Sophia Antipolis, France
J M F Labastida
CSIC
Madrid, Spain
V P Kostov
Universite de Nice Sophia Antipolis
Nice, France
G Labrosse
Universite Paris-Sud XI
Orsay, France
R Kotecky
Charles University
Prague, Czech Republic and the
University of Warwick, UK
C Landim
IMPA, Rio de Janeiro, Brazil and UMR 6085
and Universite de Rouen
France
Y Kozitsky
Uniwersytet Marii Curie-Sklodowskiej
Lublin, Poland
E Langmann
KTH Physics
Stockholm, Sweden
P Kramer
Universitat Tubingen
Tubingen, Germany
S Laporta
Universita` di Parma
Parma, Italy
C Krattenthaler
Universitat Wien
Vienna, Austria
O D Lavrentovich
Kent State University
Kent, OH, USA
xxi
xxii CONTRIBUTORS
G F Lawler
Cornell University
Ithaca, NY, USA
C Le Bris
CERMICS ENPC
Champs Sur Marne, France
A Lesne
Universite P.-M. Curie, Paris VI
Paris, France
D Levi
Universita` Roma Tre
Rome, Italy
J Lewandowski
Uniwersyte Warszawski
Warsaw, Poland
R G Littlejohn
University of California at Berkeley
Berkeley, CA, USA
R Livi
Universita` di Firenze
Sesto Fiorentino, Italy
R Longoni
Universita` di Roma La Sapienza
Rome, Italy
M Lyubich
University of Toronto
Toronto, ON, Canada and Stony Brook University
NY, USA
R Leandre
Universite de Bourgogne
Dijon, France
P Levay
Budapest University of Technology and Economics
Budapest, Hungary
R Maartens
Portsmouth University
Portsmouth, UK
N MacKay
University of York
York, UK
J Magnen
Ecole Polytechnique
France
F Magri
Universita` di Milano Bicocca
Milan, Italy
J Maharana
Institute of Physics
Bhubaneswar, India
J Lowengrub
University of California at Irvine
Irvine, USA
S Majid
Queen Mary, University of London
London, UK
C Lozano
INTA
Torrejon de Ardoz, Spain
C Marchioro
Universita` di Roma La Sapienza
Rome, Italy
T T Q Le
Georgia Institute of Technology
Atlanta, GA, USA
K Marciniak
Linkoping University
Norrkoping, Sweden
B Lucquin-Desreux
Universite P.-M. Curie, Paris VI
Paris, France
M Marcolli
Max-Planck-Institut fur Mathematik
Bonn, Germany
V Lyubashenko
Institute of Mathematics
Kyiv, Ukraine
M Marino
CERN
Geneva, Switzerland
CONTRIBUTORS
J Marklof
University of Bristol
Bristol, UK
P K Mitter
Universite de Montpellier 2
Montpellier, France
C-M Marle
Universite P.-M. Curie, Paris VI
Paris, France
V Moncrief
Yale University
New Haven, CT, USA
L Mason
University of Oxford
Oxford, UK
V Mastropietro
Universita` di Roma Tor Vergata
Rome, Italy
V Mathai
University of Adelaide
Adelaide, SA, Australia
J Mawhin
Universite Catholique de Louvain
Louvain-la-Neuve, Belgium
S Mazzucchi
Universita` di Trento
Povo, Italy
B M McCoy
State University of New York at Stony Brook
Stony Brook, NY, USA
E Meinrenken
University of Toronto
Toronto, ON, Canada
I Melbourne
University of Surrey
Guildford, UK
J Mickelsson
KTH Physics
Stockholm, Sweden
W P Minicozzi II
University of New York
New York, NY, USA
Y Morita
Ryukoku University
Otsu, Japan
P J Morrison
University of Texas at Austin
Austin, TX, USA
J Mund
Universidade de Sao Paulo
Sao Paulo, Brazil
F Musso
Universita` Roma Tre
Rome, Italy
G L Naber
Drexel University
Philadelphia, PA, USA
B Nachtergaele
University of California at Davis
Davis, CA, USA
C Nash
National University of Ireland
Maynooth, Ireland
S Necasova
Academy of Sciences
Prague, Czech Republic
A I Neishtadt
Russian Academy of Sciences
Moscow, Russia
S Miracle-Sole
Centre de Physique Theorique, CNRS
Marseille, France
N Neumaier
Albert-Ludwigs-University in Freiburg
Freiburg, Germany
A Miranville
Universite de Poitiers
Chasseneuil, France
S E Newhouse
Michigan State University
E. Lansing, MI, USA
xxiii
xxiv
CONTRIBUTORS
C M Newman
New York University
New York, NY, USA
P E Parker
Wichita State University
Wichita KS, USA
S Nikcevic
SANU
Belgrade, Serbia and Montenegro
S Paycha
Universite Blaise Pascal
Aubie`re, France
M Nitsche
University of New Mexico
Albuquerque, NM, USA
R G Novikov
Universite de Nantes
Nantes, France
J M Nunes da Costa
Universidade de Coimbra
Coimbra, Portugal
S OBrien
Tyndall National Institute
Cork, Republic of Ireland
A Okounkov
Princeton University
Princeton, NJ, USA
A Onuki
Kyoto University
Kyoto, Japan
J-P Ortega
Universite de Franche-Comte
Besanc on, France
H Osborn
University of Cambridge
Cambridge, UK
Maciej P Wojtkowski
University of Arizona
Tucson, AZ, USA and Institute of Mathematics PAN
Warsaw, Poland
P A Pearce
University of Melbourne
Parkville VIC, Australia
P Pearle
Hamilton College
Clinton, NY, USA
M Pedroni
Universita` di Bergamo
Dalmine (BG), Italy
B Pelloni
University of Reading
UK
R Penrose
University of Oxford
Oxford, UK
A Perez
Penn State University,
University Park, PA, USA
J H H Perk
Oklahoma State University
Stillwater, OK, USA
T Peternell
Universitat Bayreuth
Bayreuth, Germany
D Petz
Budapest University of Technology and Economics
Budapest, Hungary
M J Pflaum
Johann Wolfgang Goethe-Universitat
Frankfurt, Germany
J Palmer
University of Arizona
Tucson, AZ, USA
B Piccoli
Istituto per le Applicazioni del Calcolo
Rome, Italy
J H Park
Sungkyunkwan University
Suwon, South Korea
C Piquet
Universite P.-M. Curie, Paris VI
Paris, France
CONTRIBUTORS
L P Pitaevskii
Universita` di Trento
Povo, Italy
K-H Rehren
Universitat Gottingen
Gottingen, Germany
S Pokorski
Warsaw University
Warsaw, Poland
E Remiddi
Universita` di Bologna
Bologna, Italy
E Presutti
Universita` di Roma Tor Vergata
Rome, Italy
E Previato
Boston University
Boston, MA, USA
B Prinari
Universita` degli Studi di Lecce
Lecce, Italy
J Pullin
Louisiana State University
Baton Rouge, LA, USA
M Pulvirenti
Universita` di Roma La Sapienza
Rome, Italy
O Ragnisco
Universita` Roma Tre
Rome, Italy
P Ramadevi
Indian Institute of Technology Bombay
Mumbai, India
S A Ramakrishna
Indian Institute of Technology
Kanpur, India
J Rasmussen
Princeton University
Princeton, NJ, USA
L Rastelli
Princeton University
Princeton, NJ, USA
J E Roberts
Universita` di Roma Tor Vergata
Rome, Italy
L Rey-Bellet
University of Massachusetts
Amherst, MA, USA
R Robert
Universite Joseph Fourier
Saint Martin DHe`res, France
F A Rogers
Kings College London
London, UK
R M S Rosa
Universidade Federal do Rio de Janeiro
Rio de Janeiro, Brazil
C Rovelli
Universite de la Mediterranee et Centre
de Physique Theorique
Marseilles, France
S N M Ruijsenaars
Centre for Mathematics and Computer Science
Amsterdam, The Netherlands
F Russo
Universite Paris 13
Villetaneuse, France
L H Ryder
University of Kent
Canterbury, UK
T S Ratiu
Ecole Polytechnique Federale de Lausanne
Lausanne, Switzerland
S Sachdev
Yale University
New Haven, CT, USA
S Rauch-Wojciechowski
Linkoping University
Linkoping, Sweden
H Sahlmann
Universiteit Utrecht
Utrecht, The Netherlands
xxv
xxvi
CONTRIBUTORS
M Salmhofer
Universitat Leipzig
Leipzig, Germany
P M Santini
Universita` di Roma La Sapienza
Rome, Italy
A Sarmiento
Universidade Federal de Minas Gerais
Belo Horizonte, Brazil
R Sasaki
Kyoto University
Kyoto, Japan
A Savage
University of Toronto
Toronto, ON, Canada
M Schechter
University of California at Irvine
Irvine, CA, USA
D-M Schlingemann
Technical University of Braunschweig
Braunschweig, Germany
R Schmid
Emory University
Atlanta, GA, USA
G Schneider
Universitat Karlsruhe
Karlsruhe, Germany
K Schneider
Universite de Provence
Marseille, France
B Schroer
Freie Universitat Berlin
Berlin, Germany
T Schucker
Universite de Marseille
Marseille, France
M A Semenov-Tian-Shansky
Steklov Institute of Mathematics
St. Petersburg, Russia and and Universite de Bourgogne
Dijon, France
A N Sengupta
Louisiana State University
Baton Rouge LA, USA
S Serfaty
New York University
New York, NY, USA
E R Sharpe
University of Utah
Salt Lake City, UT, USA
D Shepelsky
Institute for Low Temperature Physics and Engineering
Kharkov, Ukraine
S Shlosman
Universite de Marseille
Marseille, France
A Siconolfi
Universita` di Roma La Sapienza
Rome, Italy
V Sidoravicius
IMPA
Rio de Janeiro, Brazil
J A Smoller
University of Michigan
Ann Arbor MI, USA
M Socolovsky
Universidad Nacional Autonoma de Mexico
Mexico DF, Mexico
J P Solovej
University of Copenhagen
Copenhagen, Denmark
S Scott
Kings College London
London, UK
A Soshnikov
University of California at Davis
Davis, CA, USA
P Selick
University of Toronto
Toronto, ON, Canada
J M Speight
University of Leeds
Leeds, UK
CONTRIBUTORS
H Spohn
Technische Universitat Munchen
Garching, Germany
B Temple
University of California at Davis
Davis, CA, USA
J Stasheff
Lansdale, PA, USA
R P Thomas
Imperial College
London, UK
D L Stein
University of Arizona
Tucson, AZ, USA
K S Stelle
Imperial College
London, UK
G Sterman
Stony Brook University
Stony Brook, NY, USA
S Stringari
Universita` di Trento
Povo, Italy
S J Summers
University of Florida
Gainesville, FL, USA
V S Sunder
The Institute of Mathematical Sciences
Chennai, India
U Tillmann
University of Oxford
Oxford, UK
K P Tod
University of Oxford
Oxford, UK
J A Toth
McGill University
Montreal, QC, Canada
C A Tracy
University of California at Davis
Davis, CA, USA
A Trautman
Warsaw University
Warsaw, Poland
D Treschev
Moscow State University
Moscow, Russia
Y B Suris
Technische Universitat Munchen
Munchen, Germany
L Triolo
Universita` di Roma Tor Vergata
Rome, Italy
R J Szabo
Heriot-Watt University
Edinburgh, UK
J Troost
Ecole Normale Superieure
Paris, France
S Tabachnikov
Pennsylvania State University
University Park, PA, USA
H Tasaki
Gakushuin University
Tokyo, Japan
V Turaev
IRMA
Strasbourg, France
M E Taylor
University of North Carolina
Chapel Hill, NC, USA
D Ueltschi
University of Arizona
Tucson, AZ, USA
R Temam
Indiana University
Bloomington, IN, USA
A M Uranga
Consejo Superior de Investigaciones Cientificas
Madrid, Spain
xxvii
xxviii
CONTRIBUTORS
A Valentini
Perimeter Institute for Theoretical Physics
Waterloo, ON, Canada
R F Werner
Technische Universitat Braunschweig
Braunschweig, Germany
M Vaugon
Universite P.-M. Curie, Paris VI
Paris, France
H Widom
University of California at Santa Cruz
Santa Cruz, CA, USA
P Di Vecchia
Nordita
Copenhagen, Denmark
C M Will
Washington University
St. Louis, MO, USA
A F Verbeure
Institute for Theoretical Physics
KU Leuven, Belgium
N M J Woodhouse
University of Oxford
Oxford, UK
Y Colin de Verdie`re
Universite de Grenoble 1
Saint-Martin dHe`res, France
Siye Wu
University of Colorado
Boulder, CO, USA
M Viana
IMPA
Rio de Janeiro, Brazil
V Wunsch
Friedrich-Schiller-Universitat Jena
Jena, Germany
G Vitiello
Universita` degli Studi di Salerno
Baronissi (SA), Italy
D R Yafaev
Universite de Rennes
Rennes, France
D-V Voiculescu
University of California at Berkeley
Berkeley, CA, USA
M Yamada
Kyoto University
Kyoto, Japan
S Waldmann
Albert-Ludwigs-Universitat Freiburg
Freiburg, Germany
M Yuri
Hokkaido University
Sapporo, Japan
J Wambsganss
Universitat Heidelberg
Heidelberg, Germany
D Zubrinic
University of Zagreb
Zagreb, Croatia
R S Ward
University of Durham
Durham, UK
V Zupanovic
University of Zagreb
Zagreb, Croatia
E Wayne
Boston University
Boston, MA, USA
R Zecchina
International Centre for Theoretical Physics (ICTP)
Trieste, Italy
F W Wehrli
University of Pennsylvania
Philadelphia, PA, USA
S Zelditch
Johns Hopkins University
Baltimore, MD, USA
CONTRIBUTORS
S Zelik
Universitat Stuttgart
Stuttgart, Germany
M R Zirnbauer
Universitat Koln
Koln, Germany
S-C Zhang
Stanford University
Stanford, CA, USA
A Zumpano
Universidade Federal de Minas Gerais
Belo Horizonte, Brazil
M B Ziane
University of Southern California
Los Angeles, CA, USA
xxix
INTRODUCTORY ARTICLES
Classical Mechanics 1:1
Differential Geometry 1:33
Electromagnetism 1:40
Equilibrium Statistical Mechanics 1:51
Functional Analysis 1:88
Minkowski Spacetime and Special Relativity
Quantum Mechanics 1:109
Topology 1:131
1:96
PHYSICS SUBJECTS
Classical Mechanics
Boundary Control Method and Inverse Problems of
Wave Propagation 1:340
Constrained Systems 1:611
Cotangent Bundle Reduction 1:658
Gravitational N-body Problem (Classical) 2:575
Hamiltonian Fluid Dynamics 2:593
Hamiltonian Systems: Obstructions to
Integrability 2:624
Infinite-Dimensional Hamiltonian Systems 3:37
Inverse Problem in Classical Mechanics 3:156
KAM Theory and Celestial Mechanics 3:189
Peakons 4:12
Poisson Reduction 4:79
Stability Problems in Celestial Mechanics 5:20
Symmetry and Symplectic Reduction 5:190
Disordered Systems
Cellular Automata 1:455
Lagrangian Dispersion (Passive Scalar) 3:255
Mean Field Spin Glasses and Neural
Networks 3:407
Percolation Theory 4:21
Random Matrix Theory in Physics 4:338
Random Walks in Random Environments 4:353
Short-Range Spin Glasses: The Metastate
Approach 4:570
Spin Glasses 4:655
Stochastic Loewner Evolutions 5:80
Two-Dimensional Ising Model 5:322
Wulff Droplets 5:462
Dynamical Systems
Averaging Methods 1:226
Bifurcations of Periodic Orbits 1:285
Billiards in Bounded Convex Domains 1:296
Central Manifolds, Normal Forms 1:467
Cellular Automata 1:455
Chaos and Attractors 1:477
Cotangent Bundle Reduction 1:658
Diagrammatic Techniques in Perturbation
Theory 2:54
Dissipative Dynamical Systems of Infinite
Dimension 2:101
Dynamical Systems and Thermodynamics 2:125
Dynamical Systems in Mathematical Physics:
An Illustration from Water Waves 2:133
Entropy and Quantitative Transversality 2:237
Ergodic Theory 2:250
Fractal Dimensions in Dynamics 2:394
Generic Properties of Dynamical Systems 2:494
Gravitational N-Body Problem (Classical) 2:575
Hamiltonian Fluid Dynamics 2:593
Hamiltonian Systems: Stability and Instability
Theory 2:631
Holomorphic Dynamics 2:652
Homeomorphisms and Diffeomorphisms of the
Circle 2:665
Homoclinic Phenomena 2:672
h-Pseudodifferential Operators and
Applications 2:701
Hyperbolic Billiards 2:716
Hyperbolic Dynamical Systems 2:721
Isomonodromic Deformations 3:173
Fluid Dynamics
Bifurcations in Fluid Dynamics
Breaking Water Waves 1:383
1:281
Gauge Theory
Abelian and Nonabelian Gauge Theories Using
Differential Forms 1:141
Abelian Higgs Vortices 1:151
AdS/CFT Correspondence 1:174
AharonovBohm Effect 1:191
Anomalies 1:205
BRST Quantization 1:386
ChernSimons Models: Rigorous Results 1:496
Dirac Fields in Gravitation and Nonabelian Gauge
Theory 2:67
DonaldsonWitten Theory 2:110
Effective Field Theories 2:139
ElectricMagnetic Duality 2:201
Electroweak Theory 2:209
Exact Renormalization Group 2:272
Gauge Theories from Strings 2:463
Gauge Theory: Mathematical Applications 2:468
Instantons: Topological Aspects 3:44
Large-N and Topological Strings 3:263
Lattice Gauge Theory 3:275
Measure on Loop Spaces 3:413
Noncommutative Geometry and the Standard
Model 3:509
Nonperturbative and Topological Aspects of Gauge
Theory 3:568
Perturbative Renormalization Theory and
BRST 4:41
Quantum Chromodynamics 4:144
Quantum Electrodynamics and Its Precision
Tests 4:168
xxxiii
General Relativity
General Relativity: Overview 2:487
Asymptotic Structure and Conformal
Infinity 1:221
Black Hole Mechanics 1:300
Boundaries for Spacetimes 1:326
Brane Worlds 1:367
Canonical General Relativity 1:412
Critical Phenomena in Gravitational
Collapse 1:668
Computational Methods in General Relativity:
The Theory 1:604
Cosmology: Mathematical Aspects 1:653
Dirac Fields in Gravitation and Nonabelian Gauge
Theory 2:67
EinsteinCartan Theory 2:189
Einsteins Equations with Matter 2:195
Einstein Equations: Exact Solutions 2:165
Einstein Equations: Initial Value
Formulation 2:173
General Relativity: Experimental Tests 2:481
Geometric Analysis and General Relativity 2:502
Geometric Flows and the Penrose
Inequality 2:510
Gravitational Lensing 2:567
Gravitational Waves 2:582
Hamiltonian Reduction of Einsteins
Equations 2:607
Minimal Submanifolds 3:420
Newtonian Limit of General Relativity 3:503
Quantum Field Theory in Curved
Spacetime 4:202
Relativistic Wave Equations Including Higher Spin
Fields 4:391
Shock Wave Refinement of the Friedman
RobertsonWalker Metric 4:559
Spacetime Topology, Causal Structure and
Singularities 4:617
Spinors and Spin Coefficients 4:667
Stability of Minkowski Space 5:14
Stationary Black Holes 5:38
Twistors 5:311
Integrable Systems
Integrable Systems: Overview 3:106
Abelian Higgs Vortices 1:151
Affine Quantum Groups 1:183
Backlund Transformations 1:241
xxxiv
Quantum Gravity
Knot Invariants and Quantum Gravity 3:215
Knot Theory and Physics 3:220
Loop Quantum Gravity 3:339
Quantum Cosmology 4:153
Quantum Dynamics in Loop Quantum
Gravity 4:165
Quantum Field Theory in Curved
Spacetime 4:202
Quantum Geometry and Its Applications 4:230
Spin Foams 4:645
WheelerDe Witt Theory 5:453
Quantum Mechanics
AharonovBohm Effect 1:191
Arithmetic Quantum Chaos 1:212
Coherent States 1:537
Geometric Phases 2:528
h-Pseudodifferential Operators and
Applications 2:701
N-particle Quantum Scattering 3:585
Normal Forms and Semiclassical
Approximation 3:578
Quantum Entropy 4:177
Quantum Ergodicity and Mixing of
Eigenfunctions 4:183
Quantum Mechanical Scattering
Theory 4:251
Quantum Mechanics: Foundations 4:260
Quantum Mechanics: Generalizations 4:265
xxxv
RELATED MATHEMATICS
SUBJECTS
Algebraic Techniques
Affine Quantum Groups 1:183
Braided and Modular Tensor Categories 1:351
Clifford Algebras and Their
Representations 1:518
Derived Categories 2:41
Finite-Dimensional Algebras and Quivers 2:313
Finite Group Symmetry Breaking 2:322
Hopf Algebras and Q-Deformation Quantum
Groups 2:687
Operads 3:609
xxxvi
Algebraic Topology
Characteristic Classes 1:488
Cohomology Theories 1:545
Derived Categories 2:41
Equivariant Cohomology and the Cartan
Model 2:242
FourierMukai Transform in String
Theory 2:379
Index Theorems 3:23
Intersection Theory 3:151
K-theory 3:246
MathaiQuillen Formalism 3:390
Operads 3:609
Spectral Sequences 4:623
String Topology: Homotopy and Geometric
Perspectives 5:111
Complex Geometry
Derived Categories 2:41
Gauge Theory: Mathematical Applications 2:468
FourierMukai Transform in String Theory 2:379
Knot Homologies 3:208
Mirror Symmetry: A Geometric Survey 3:439
Moduli Spaces: An Introduction 3:449
Quillen Determinant 4:315
Riemann Surfaces 4:419
RiemannHilbert Problem 4:436
Several Complex Variables: Basic Geometric
Theory 4:540
Several Complex Variables: Compact
Manifolds 4:551
Twistor Theory: Some Applications 5:303
Discrete Mathematics
Arithmetic Quantum Chaos 1:212
Combinatorics: Overview 1:553
Number Theory in Physics 3:600
Quasiperiodic Systems 4:308
Noncommutative Geometry
Hopf Algebra Structure of Renormalizable
Quantum Field Theory 2:678
Noncommutative Geometry and the Standard
Model 3:509
Noncommutative Geometry from Strings 3:515
Noncommutative Tori, Yang-Mills, and String
Theory 3:524
Path Integrals in Noncommutative Geometry 4:8
Quantum Group Differentials, Bundles and Gauge
Theory 4:236
Quantum Hall Effect 4:244
RiemannHilbert Problem 4:436
xxxvii
Quantum Groups
Affine Quantum Groups 1:183
Bicrossproduct Hopf Algebras and
Noncommutative Spacetime 1:265
Braided and Modular Tensor Categories 1:351
Classical r-Matrices, Lie Bialgebras, and Poisson Lie
Groups 1:511
Hopf Algebras and q-Deformation Quantum
Groups 2:687
Hopf Algebra Structure of Renormalizable
Quantum Field Theory 2:678
q-Special Functions 4:105
Quantum Group Differentials, Bundles and Gauge
Theory 4:236
YangBaxter Equations 5:465
Stochastic Methods
Determinantal Random Fields 2:47
Free Probability Theory 2:417
Growth Processes in Random Matrix
Theory 2:586
Integrable Systems in Random Matrix
Theory 3:102
Lagrangian Dispersion (Passive Scalar) 3:255
Malliavin Calculus 3:383
Measure on Loop Spaces 3:413
Random Matrix Theory in Physics 4:338
Random Partitions 4:347
Random Walks in Random Environments 4:353
Stochastic Differential Equations 5:63
Stochastic Hydrodynamics 5:71
Stochastic Loewner Evolutions 5:80
Supersymmetry Methods in Random Matrix
Theory 5:151
Symmetry Classes in Random Matrix
Theory 5:204
xxxviii
Variational Techniques
Capillary Surfaces 1:431
Control Problems in Mathematical Physics 1:636
Convex Analysis and Duality Methods 1:642
CONTENTS
VOLUME 1
Introductory Article: Classical Mechanics
G Gallavotti
S Paycha
33
N M J Woodhouse
40
G Gallavotti
S Paycha
88
51
G L Naber
G F dellAntonio
96
109
131
A
Abelian and Nonabelian Gauge Theories Using Differential Forms
Abelian Higgs Vortices
Adiabatic Piston
A C Hirshfeld
141
J M Speight
151
160
AdS/CFT Correspondence
174
183
AharonovBohm Effect
M Socolovsky
191
198
S L Adler
205
J Marklof
212
J Frauendiener
221
A I Neishtadt
226
B Kuckert
232
234
B
Backlund Transformations
D Levi
BatalinVilkovisky Quantization
Bethe Ansatz
BF Theories
241
A C Hirshfeld
247
M T Batchelor
253
M Blau
257
S Majid
265
xl
CONTENTS
Bifurcation Theory
275
G Schneider
281
285
M Pedroni
290
S Tabachnikov
296
A Ashtekar
300
M Pulvirenti
306
318
S G Harris
326
J Cardy
333
M I Belishev
B Pelloni
351
S L Cacciatori
360
R Maartens
367
340
346
V Lyubashenko
312
E Langmann
S R Das
373
A Constantin
383
M Henneaux
386
C
C -Algebras and their Classification
G A Elliott
393
D D Joyce
S N M Ruijsenaars
C Rovelli
398
403
412
P Hayden
418
D Kretschmann
424
R Finn
431
G M Henkin
455
P Bonckaert
467
M Keyl
472
R Gilmore
477
Characteristic Classes
488
A N Sengupta
496
S Gindikin
500
M A Semenov-Tian-Shansky
A Trautman
531
S T Ali
Cohomology Theories
Combinatorics: Overview
511
518
R Kotecky
Cluster Expansion
Coherent States
446
537
U Tillmann
545
C Krattenthaler
553
A Kirillov and A Kirillov, Jr.
M R Douglas
586
G-Q Chen
576
595
M W Choptuik
604
CONTENTS
xli
M Henneaux
611
G Gallavotti
617
J B Etnyre
631
B Piccoli
636
G Bouchitte
642
G F R Ellis
653
658
C Gundlach
668
G A Goldin
674
VOLUME 2
D
Deformation Quantization
A C Hirshfeld
S Waldmann
16
41
A Soshnikov
47
G Gentile
54
R Kenyon
61
24
34
E R Sharpe
M J Pflaum
J A Smoller
67
S N M Ruijsenaars
74
J Bros
87
101
DonaldsonWitten Theory
110
118
125
O Goubet
133
E
Effective Field Theories
G Ecker
139
P A Pearce
Jir Bicak
J Isenberg
A S Dancer
EinsteinCartan Theory
ElectricMagnetic Duality
Electroweak Theory
155
165
173
189
Y Choquet-Bruhat
K Konishi
148
182
A Trautman
Entanglement
J A Toth
195
201
209
216
228
xlii
CONTENTS
Entanglement Measures
R F Werner
233
G Comte
237
E Meinrenken
242
M Yuri
250
F Guerra
256
J Escher
265
P K Mitter
272
F
FalicovKimball Model
283
Fedosov Quantization
N Neumaier
291
V Mastropietro
300
S Mazzucchi
307
322
328
R F Werner
334
D Bar-Natan
340
T T Q Le
348
P B Kronheimer
356
313
D-M Schlingemann
Floer Homology
A Savage
G Gaeta
J-L Guermond
365
374
B Andreas
C Nash
V Zupanovic and D Zubrinic
386
394
J K Jain
402
379
G Buttazzo
D-V Voiculescu
411
417
H W Braden
425
C DeWitt-Morette
434
G
-Convergence and Homogenization
G Dal Maso
S Bauer
P Di Vecchia
449
457
463
S K Donaldson
C M Will
R Penrose
468
481
487
C Bonatti
494
L Andersson
502
H Bray
510
G Alberti
P Levay
528
M B Ziane
520
J Mickelsson
534
539
CONTENTS
GinzburgLandau Equation
Y Morita
547
S Franz
553
560
J Wambsganss
567
D C Heggie
575
Gravitational Waves
xliii
582
K Johansson
586
H
Hamiltonian Fluid Dynamics
P J Morrison
593
L C Jeffrey
600
607
M Irigoyen
624
P Bernard
631
A Siconolfi
636
Hard Hexagon Model see Eight Vertex and Hard Hexagon Models
High Tc Superconductor Theory
Holomorphic Dynamics
S-C Zhang
645
M Lyubich
652
J Palmer
660
S E Newhouse
672
665
D Kreimer
678
S Majid
687
B Helffer
701
H Tasaki
712
M P Wojtkowski
716
B Hasselblatt
721
VOLUME 3
I
Image Processing: Mathematics
D Chae
Indefinite Metric
H Gottschalk
Index Theorems
32
R Schmid
37
M Jardim
44
T J Hollowood
50
O Ragnisco
59
E Previato
65
78
R Caseiro and
87
23
M Vaugon
10
17
A S Fokas
Francesco Calogero
93
102
106
xliv CONTENTS
Interacting Particle Systems and Hydrodynamic Equations
Interacting Stochastic Particle Systems
H Spohn
123
130
135
J Jimenez
Intermittency in Turbulence
Intersection Theory
C Landim
144
A Kresch
151
R G Novikov
156
Inverse Problems in Wave Propagation see Boundary Control Method and Inverse Problems of Wave
Propagation
Inviscid Flows
R Robert
160
Francesco Calogero
Isomonodromic Deformations
166
V P Kostov
173
J
The Jones Polynomial
V F R Jones
179
K
KacMoody Lie Algebras see Solitons and KacMoody Lie Algebras
KAM Theory and Celestial Mechanics
L Chierchia
189
Kinetic Equations
C Bardos
200
Knot Homologies
J Rasmussen
208
220
231
215
L H Kauffman
G Schneider and E Wayne
V Mathai
239
246
L
Lagrangian Dispersion (Passive Scalar)
G Falkovich
255
S Shlosman
261
R Gopakumar
263
A Grassi
269
A Di Giacomo
275
J Mawhin
281
Lie Bialgebras see Classical r-Matrices, Lie Bialgebras, and Poisson Lie Groups
Lie Groups: General Theory
R Gilmore
286
L Frappat
305
C-M Marle
O D Lavrentovich
LjusternikSchnirelman Theory
Loop Quantum Gravity
Lorentzian Geometry
320
J Mawhin
328
S Jitomirskaya
333
C Rovelli
339
312
343
M Viana
349
M
Macroscopic Fluctuations and Thermodynamic Functionals
Magnetic Resonance Imaging
Magnetohydrodynamics
C Le Bris
G Jona-Lasinio
357
367
375
CONTENTS
Malliavin Calculus
A B Cruzeiro
xlv
383
MarsdenWeinstein Reduction see Cotangent Bundle Reduction: Poisson Reduction: Symmetry and
Symplectic Reduction
Maslov Index see Optical Caustics: Semiclassical Spectra and Closed Orbits: Stationary Phase
Approximation
MathaiQuillen Formalism
S Wu
390
L Boi
399
A Bovier
407
H Airault
413
S Shlosman
Minimal Submanifolds
417
420
A Abbondandolo
432
R P Thomas
439
F Kirwan
449
459
A Lesne
465
N
Negative Refraction and Subdiffraction Imaging
Newtonian Fluids and Thermohydraulics
J Ehlers
492
503
T Schucker
483
509
Chong-Sun Chu
515
A Konechny
524
G Gallavotti
530
540
R Livi
544
C Guillope
560
552
R W Jackiw
D Bambusi
568
578
D R Yafaev
585
P T Callaghan
592
M Marcolli
600
O
Operads
J Stasheff
609
A Joets
616
620
H Osborn
M Keyl
Y Brenier
W Van Assche
628
632
637
xlvi CONTENTS
VOLUME 4
P
Painleve Equations
N Joshi
R Temam
Path Integral Methods see Functional Integration in Quantum Physics; Feynman Path Integrals
R Leandre
D D Holm
12
R J Szabo
28
21
A Onuki
47
E Presutti
53
R Kotecky
PirogovSinai Theory
Point-Vortex Dynamics
41
60
66
Poisson Lie Groups see Classical r-Matrices, Lie Bialgebras, and Poisson Lie Groups
Poisson Reduction
79
Polygonal Billiards
S Tabachnikov
84
F Cipriani
88
P E Parker
94
Q
q-Special Functions
T H Koornwinder
105
123
A F Verbeure
130
A S Holevo
144
M Bojowald
153
R Alicki
159
H Sahlmann
D Petz
168
S Zelditch
183
D Gottesman
196
B S Kay
L H Ryder
202
212
S Albeverio and H Gottschalk
T Brzezinski
K Hannabuss
216
221
230
236
244
D R Yafaev
R Penrose
P Pearle and A Valentini
165
177
142
G Sterman
Quantum Entropy
117
R Sasaki
R G Littlejohn
L Diosi
251
260
265
276
283
CONTENTS
Quantum Phase Transitions
S Sachdev
Quantum Spin Systems
B Nachtergaele
Quantum Statistical Mechanics: Overview
xlvii
289
295
302
L Triolo
Quasiperiodic Systems
P Kramer
Quillen Determinant
S Scott
Quivers see Finite-Dimensional Algebras and Quivers
308
315
R
Random Algebraic Geometry, Attractors and Flux Vacua
Random Dynamical Systems
V Araujo
M R Douglas
323
330
338
347
353
371
376
386
391
399
407
415
Riemann Surfaces
K Hulek
RiemannHilbert Methods in Integrable Systems
M Salmhofer
419
429
D Shepelsky
RiemannHilbert Problem
V P Kostov
Riemannian Holonomy Groups and Exceptional Holonomy
436
441
D D Joyce
S
Saddle Point Problems
M Schechter
Scattering in Relativistic Quantum Field Theory: Fundamental Concepts and Tools
S J Summers
Scattering in Relativistic Quantum Field Theory: The Analytic Program
447
D Buchholz and
J Bros
456
465
475
487
494
503
Semiclassical Approximation
Semiclassical Spectra and Closed Orbits
Y Colin de Verdie`re
Semilinear Wave Equations
P DAncona
512
518
526
Separatrix Splitting
D Treschev
535
540
551
559
S N M Ruijsenaars
M Anderson
J-P Franc oise and C Piquet
570
576
584
588
xlviii
CONTENTS
E Date
594
R S Ward
602
R Penrose
609
617
Special Lagrangian Submanifolds see Calibrated Geometry and Special Lagrangian Submanifolds
Spectral Sequences
P Selick
623
M Schechter
633
A Perez
Spin Glasses
645
F Guerra
655
K P Tod
667
VOLUME 5
Stability of Flows
S Friedlander
Stability of Matter
J P Solovej
S Klainerman
14
A Celletti
G Gentile
26
G Altarelli
32
44
R Zecchina
50
S Miracle-Sole
38
J J Duistermaat
55
F Russo
63
B Ferrario
20
71
G F Lawler
80
86
94
103
111
115
122
128
Superstring Theories
R L Cohen
133
S Pokorski
140
145
M R Zirnbauer
151
S Kichenassamy
160
L H Ryder
166
I Melbourne
J E Roberts
M A Aziz-Alaoui
M R Zirnbauer
172
179
184
190
198
204
213
CONTENTS
xlix
T
t HooftPolyakov Monopoles see Solitons and Other Extended Field Configurations
C D Jakel
227
Y B Suris
235
E L Basor
251
257
264
L Boi
271
278
D Birmingham
290
R M S Rosa
T W B Kibble
T Eguchi
244
S J Summers
295
L Mason
303
K P Tod
311
M R Gaberdiel
B M McCoy
317
322
B Schroer
328
U
Universality and Renormalization
M Lyubich
343
V
Variational Methods in Turbulence
F H Busse
351
S Serfaty
355
G Dolzmann
363
Vertex Operator Algebras see Two-Dimensional Conformal Field Theory and Vertex
Operator Algebras
Viscous Incompressible Fluids: Mathematical Theory
J G Heywood
Y Kawahigashi
M Nitsche
369
V S Sunder
379
385
390
W
Wave Equations and Diffraction
M E Taylor
M Yamada
408
420
401
B Dubrovin
426
438
448
453
S Shlosman
462
Y
YangBaxter Equations
INDEX
465
475
Introductory Articles
Introductory Article: Classical Mechanics
G Gallavotti, Universita` di Roma La Sapienza,
Rome, Italy
2006 G Gallavotti. Published by Elsevier Ltd.
All rights reserved.
General Principles
Classical mechanics is a theory of motions of point
particles. If X = (x1 , . . . , xn ) are the particle positions
in a Cartesian inertial system of coordinates, the
equations of motion are determined by their masses
(m1 , . . . , mn ), mj > 0, and by the potential energy of
interaction, V(x1 , . . . , xn ), as
i @xi Vx1 ; . . . ; xn ;
mi x
i 1; . . . ; n
1
n
1X
def
mi y2i VX KY VX,
2 i1
Y y1 , . . . , yn
is called the Lagrangian function and the action can
be written as
Z
t2
_
LXt;
Xt dt
t1
_
The quantity K(X(t))
is called kinetic energy and
motions satisfying [1] conserve energy as time
t varies, that is,
_
KXt
VXt E const:
3
Z q
E Vxs ds
4
Constraints
Often particles are subject to constraints which force
the motion to take place on a surface M Rnd , i.e.,
X(t) is forced to be a point on the manifold
M. A typical example is provided by rigid systems
in which motions are subject to forces which keep
the mutual distances of the particles constant:
jxi xj j = ij , with ij time-independent positive quantities. In essentially all cases, the forces that imply
constraints, called constraint reactions, are velocity
dependent and, therefore, are not in the class of
conservative forces considered here, cf. [1]. Hence,
from a fundamental viewpoint admitting only conservative forces, constrained systems should be regarded
as idealizations of systems subject to conservative
forces which approximately imply the constraints.
5
Z
0
stationary.
t1
_
KXt
Va Xt dt
6
7
def
Lh; q
1;
1X
gij qi j V a q
2 i;j
1
gqh h V a q
2
8
i 1; . . . ;
10
pi gqij q_ j ;
i 1; . . . ;
11
Since g(q) > 0, the motions t ! q(t) and the corre_ can be described equivasponding velocities t ! q(t)
lently by t ! (q(t), p(t)): and the equations of motion
[10] become the first-order equations
q_ i @pi Hp; q;
p_ i @qi Hp; q
12
Hp; q 12gq1 p; p V a q
13
where i, j = 1, . . . , . Let
0
E
1
1
0
17
fF; Ggp; k
X
k1
18
20
p @k p 0 ; k
k 0 @p 0 p 0 ; k
Quadratures
The simplest mechanical systems are integrable by
quadratures. For instance, the Hamiltonian on R2 ,
Hp; q
1 2
p Vq
2m
21
Qt
dx
p
2=mE Vx
q E
23b
24
generates
(locally) the
correspondence between
p
2
TEA
a_ wA
28
p dq
30
Ii @j i SA; j
31
X
1 2 1X
pi
cij qi qj
2mi
2 i; j
i1
X v
; i qi C
1 B X v
; i pi
q A
A
p !2
@
2!
mi
i1
i1
m1
i
for all eigenvectors v
,
= 1, . . . , , of the above
matrix, the vectors A have all components >0.
Even though this system is isochronous, it nevertheless admits a system of canonical actionangle
coordinates in which the Hamiltonian takes the
simplest form
hA
! A w A
33
with
0
P
v
; i pi
p
m
B
C
i
B
C
arctanB i1
C
@ P p
A
mi !
v
; i qi
wA
1
Ai
Ji
d
m2
_ 0
dt
i.e., m2
= G is a constant of motion (it is the
angular momentum), and
m
@ V @ 2
_2
2
2
G
@ V
m3
@ VG
as conjugate angles.
An example of anisochronous system is the free
rotators or free wheels: i.e., noninteracting points
on a circle of radius R or noninteracting homogeneous coaxial wheels of radius R. If Ji = mi R2 or,
respectively, Ji = (1=2)mi R2 are the inertia moments
and if the positions are determined by angles a =
(1 , . . . , ), the angular velocities are constants
related to the angular momenta A = (A1 , . . . , A ) by
!i = Ai =Ji . The Hamiltonian and the spectrum are
def
i1
X
1 2
hA
Ai ;
2J
i
i1
34
p
V
2m
2m 2
i1;...;
Multidimensional Quadratures:
Central Motion
Several important mechanical systems with more
than one degree of freedom are integrable by
canonical quadratures in vast regions of phase
space. This is checked by showing that there is a
foliation into invariant tori T (I) of dimension equal
to the number of degrees of freedom () parametrized by constants of motion I in involution, i.e.,
such that {Ii , Ij } = 0. One then performs, if possible,
the construction of the actionangle variables by
the quadratures discussed in the previous section.
The above procedure is well illustrated by the
theory of the planar motion of a unit mass attracted
by a coplanar center of force: the Lagrangian is, in
polar coordinates (,
),
m 2
_ 2
_2 V
2
35
VG
1 G2
V
2 m2
TE; G 2
E; G
E; G
dx
p
2=mE VG x
Rt
dx
p
2=mE VG x
E; G
36a
or
t
TE; G
2
Rt
dx
p
2=mE VG x
E; G
36b
10
t
0
dt0
2
0 mRt0 t0
a second quadrature. Therefore, we can use as
coordinates for the motion E, G, t0 , which determine
_ 0 , 0 ,
_0 and a fourth coordinate that determines
0
which could be
0 itself but which is conveniently
determined, via the second quadrature, as follows.
The function Gm1 R(t)2 is periodic with period
T(E, G); hence it can be expressed in a Fourier series
X
2
itk
k E; G exp
0 E; G
TE; G
k60
the quadrature for
(t) can be performed by
integrating the series terms. Setting
X k E; G
2
def TE; G
exp
it0 k
t0
2 k60
k
TE; G
0 ), the expression
and 1 (0) =
0
(t
Z t
G
t
0
dt0
2
0 mRt0 t0
37
!0
2
TE; G
and
def
!1 0 E; G
0 !0 t0 ;
2
A1
2
E; G p
2mE VG xdx;
E; G
38
A2 G
According to the general theory (cf. the previous
section) a generating function for the canonical
change of coordinates from (p , , p
,
) to action
angle variables is (if, to fix ideas, p > 0)
Z p
SA1 ; A2 ; ;
G
2mE VG xdx 39
E;
1 @A1 S !0 t;
def
2 @A2 S
t
40
becomes
1 t 1 0 0 E; G t
def
1
0
t0
At E, G fixed, the motion takes place on a twodimensional torus T (E, G) with 0 , 1 as angles.
In the anisochronous cases, i.e., when
det @E, G w(E, G) 6 0, canonical actionangle variables conjugated to (p , , p
,
) can be constructed
via [29], [30] by using two cycles 1 , 2 on the torus
T (E, G). It is convenient to choose
1. 1 as the cycle consisting of the points = x,
= 0
whose first half (where p 0)pconsists
in the
Hp; q
1 2 km
p
2m
jqj
c E
S
e = 0.75
O S
e = 0.3
e = 0.75
Figure 2 Eccentric and true anomalies of P, which moves on a small circle E centered at a point c moving on the circle D located
half-way between the two concentric circles containing the Keplerian ellipse: the anomaly of c with respect to the axis OS is . The
circle D is eccentric with respect to S and therefore is, even today, called eccentric anomaly, whereas the circle D is, in ancient
terminology, the deferent circle (eccentric circles were introduced in astronomy by Ptolemy). The small circle E on which the point P
moves is, in ancient terminology, an epicycle. The deferent and the epicyclical motions are synchronous (i.e., they have the same
period); Kepler discovered that his key a priori hypothesis of inverse proportionality between angular velocity on the deferent and
distance between P and S (i.e., _ = constant) implied both synchrony and elliptical shape of the orbit, with focus in S. The latter law is
_ Small eccentricity ellipses can hardly be distinguished from circles.
equivalent to 2
_ = constant (because of the identity a _ =
).
Rigid Body
Another fundamental integrable system is the rigid
body in the absence of gravity and with a fixed point
O. It can be naturally described in terms of the Euler
angles
0 , 0 , 0 (see Figure 3) and their derivatives
_0 , _ 0 , _ 0 .
Let I1 , I2 , I3 be the three principal inertia moments
of the body along the three principal axes with unit
vectors i1 , i2 , i3 . The inertia moments and the
principal axes are the eigenvalues and the associated
unit eigenvectors of the 3
P 3 inertia matrix I ,
which is defined by I hk = ni= 1 mi (xi )h (xi )k , where
h, k = 1, 2, 3 and xi is the position of the ith particle
in a reference frame with origin at O and in which
e sin
1 e cos 1 e cos
1 e2
Z
d
0
1 e2 3=2
2
0 1 e cos
0
41
O
g
1 e2
a 1 e cos f
i2
1 e2
a 1 e cos
f )
i3
42
i1
12
43
44
1
I1
_0 cos 0 _ 0 sin
0 sin 0 2
2
1
I2
_0 sin 0 _ 0 sin
0 cos
2
1
I3 _ 0 cos
0 _ 0 2
2
0 , 0 , 0 ;
3. the Euler angles of (O; 1, 2, 3) with respect to
(O; x, y, z), which are denoted
, , ; P
4. G, the total angular momentum: G2 = j Ij2 !2j ;
5. M3 , the angular momentum along the z0 axis;
M3 = G cos ; and
6. L, the projection of M on the axis 3, L = G cos
.
The quantities G, M3 , L, , , determine
0 , 0 ,
_ _ 0 , _ 0 , or the p
, p , p
variables
0 and
0 ,
0
0
0
conjugated to
0 , 0 , 0 as shown by the following
comment.
Considering Figure 4, the angles , determine
location, in the fixed frame (O; x0 , y0 , z0 ) of the
direction of M and the node line m, which are,
respectively, the z-axis and the x-axis of the fixed
frame associated with the angular momentum; the
angles
, , then determine the position of the
comoving frame with respect to the fixed frame
(O; x, y, z), hence its position with respect to
(O; x0 , y0 , z0 ), that is, (
0 , 0 , 0 ). From this and
G, it is possible to determine w because
I3 !3
I 2 !2
;
tan
G
I 1 !1
!22 I22 G2 I12 !21 I32 !23
cos
45
Since angular momentum is conserved, it is convenient to introduce the laboratory frame (O; x0 ,
y0 , z0 ) with fixed axes x0 , y0 , z0 and (see Figure 4):
M ||z
z0
j 1; 2; 3
46
47
2
y
x0
0
x =m
y0
0
n
n0
p0 M z0 ;
M i3
48
p
0 d
0
49
50
2 I3
I1
I2
This again shows that G, M3 are constants of
motion, and the L, variables are determined by a
quadrature, because the Hamilton equation for
combined with the energy conservation yields
!
2
2
1
sin
cos
_
I3
I1
I2
v
u
u2E G2 sin2 cos2
u
I1
I2
51
t
sin2
cos2
1
I3 I1 I2
In the integrability region, this motion is periodic
with some period TL (E, G). Once (t) is determined,
the Hamilton equation for leads to the further
quadrature
!
sin2 t cos2 t
_
G
52
I1
I2
which determines a second periodic motion with
period TG (E, G). The , M3 are constants and,
therefore, the motion takes place on threedimensional invariant tori T E, G, M3 in phase space,
each of which is always foliated into twodimensional invariant tori parametrized by the
angle which is constant (by [50], because K is
M3 -independent): the latter are in turn foliated by
one-dimensional invariant tori, that is, by periodic
orbits, with E, G such that the value of
TL (E, G)=TG (E, G) is rational.
1
H
G2
2I3
2I
G2
!
1=2
L2
cos
53
1 2
G
so that, again, the system is integrable by quadratures
(with the roles of and interchanged with respect
to the previous case) in suitable regions of phase space.
This is called the Lagranges gyroscope.
A less elementary integrable case is when the
inertia moments are related as I1 = I2 = 2I3 and the
center of mass is in the i1 i2 plane (rather than on
the i3 -axis) and only gravity acts, besides the
constraint force on the pivot point O; this is called
Kowalevskaias gyroscope.
For more details, see Gallavotti (1983).
Other Quadratures
An interesting classical integrable motion is that of a
point mass attracted by two equal-mass centers of
gravitational attraction, or a point ideally constrained
to move on the surface of a general ellipsoid.
New integrable systems have been discovered
quite recently and have generated a wealth of new
developments ranging from group theory (as integrable systems are closely related to symmetries) to
partial differential equations.
It is convenient to extend the notion of integrability by stating that a system is integrable in a
region W of phase space if
1. there is a change of coordinates (p, q) 2
W ! {A, a, Y, y} 2 (U T ) (V Rm ) where
U R , V Rm , with m 1, are open sets; and
2. the A, Y are constants of motion while the other
coordinates vary linearly:
a; y ! a wA; Yt; y vA; Yt
54
14
forming sets of zero volume). The notion is convenient also because it allows us to say that even the
systems of free particles are integrable.
Two very remarkable systems integrable in the
new sense are the Hamiltonian systems, respectively
called Toda lattice (KRUSKAL, ZABUSKY), and
Calogero lattice (CALOGERO, MOSER); if (pi , qi ) 2 R 2 ,
they are
HT p; q
n
n1
X
1 X
p2i
g eqi1 qi
2m i1
i1
HC p; q
n
n
X
1 X
g
p2i
2m i1
q
qj 2
i
i<j
55
n
1X
m!2 q2i
2 i1
p
1
56
Nhh 0
57
i
1
Mhk
;
Nhk
h 6 k
2
qh qk
qh qk
while for the Toda lattice (with m = g = 12 = 1) the
nonzero matrix elements of M, N are
Mhh ph ;
58
n
n
X
1 X
g
p2k
2
2m ik
h<k sinh qh qk
59
wA0 a 1 A0 ; a f A0 ; a f A0
61
Generic Nonintegrability
H" A; a hA "f A; a
if wA0 n 0;
n 6 0
62
B"; A; a B0 A; a "B1 A; a
"2 B2 A; a
wA @a B0 A; a 0
@A f A; a @a B0 A; a @a f A; a @A B0 A; a
wA @a B1 A; a 0
63
64
16
65
Perturbing Functions
To check, in a given problem, the nonexistence of
nontrivial constants of motion along the lines
indicated in the previous section, it is necessary to
express the potential, usually given in Cartesian
66
1
X
X
1
i0 c0 i0 p
p! k1 kp k1
p1
0 1 p
hkj j ;
k>1
67
4
1
2
5
6
7
8
9
10
and then bifurcate arbitrarily (such trees are sometimes called rooted trees).
Imagine the tree oriented from the endpoints
towards the root r (not to be considered a node)
and given a node v call v0 the node immediately
following it. If v is the first node before the root r,
let v0 = r and v 0 = 1. For each such decorated tree
define its numerical value
Val
Y
i Y
v 0 v
cv
k! lines lv0 v
nodes
68
69
;
order
k
1 X
X
k1
"k ei Val
70
;
order
k
where the
means that the trees are subject to the
further restriction of not containing any simple
node. It should be noted that the above graphical
representation of the solution of the Kepler equation
is strongly reminiscent of the representations of
quantities in terms of graphs that occur often in
quantum field theory. Here the trees correspond to
Feynman graphs, the factors associated with the
nodes are the couplings, the factors associated with
the lines are the propagators, and the resummations
are analogous to the self-energy resummations,
while the cancellations mentioned above can be
related to the class of identities called Ward
identities. Not only the analogy can be shown not
to be superficial, but it also turns out to be very
helpful in key mechanical problems: see Appendix 1.
The existence of a vast number of identities
relating the tree values is shown already by the
simple form of the Lagrange series and by the
even more remarkable resummation (LEVI-CIVITA)
leading to
h
k
1
X
" sin k
1
@
1 " cos
k!
k1
71
18
A A0 H " y ; a y h" y ;
y 2 T 73
def
1
;
Cjnj
for all 0 6 n 2 Z
74
A_ " a f A; a
75
u" A0 "A2
F" A0 ; a
1
X
X
"k
fn eian
76
k1
06n2Z2
i 2 k
i!01 1 !02 2 k1
77
uA0 "A02
F" A0 ; a "
fn eian
i!01 1 !20 "2
2
78
06n2Z
zk
1
1z
for z 6 1
79
20
80
In this simple case (called Thirring model: representing particles on a circle interacting via a potential
"f (a)) the equations for the maximal tori [75]
reduce to equations for the only functions h" :
w y 2 h" y " a f y h" y ;
y 2 T
81
nl
nw
82
wv
and we call n(
) the current flowing in the root
branch.
Here the value Val(
) of a tree has to be defined
differently because the equation to be solved ([81])
contains the differential operator (w 0 y )2 which,
when Fourier transformed, becomes multiplication
of the Fourier component with harmonic n by
(iw n)2 .
The variation due to the presence of the operator
(w 0 y )2 and the necessity of its inversion in the
evaluation of u h(k)
n , that is, of the component of
h(k)
n along an arbitrary unit vector u, is nevertheless
quite simple: the value of a tree graph
of order k
Val
84
; n
n
order
k
where the
means that the sum is only over trees in
which a nonzero current n(l) flows on the lines l 2
.
The quantity u h(k)
0 will be defined to be 0 (see the
previous section).
In the case of [66] zero-current lines could appear:
but the contributions from tree graphs containing at
least one zero current line would cancel. In the
present case, the statement that the above algorithm
actually gives h(k)
n by simply ignoring trees with lines
with zero current is nontrivial. It was Poincares
contribution to the theory of Lindstedt series to show
that even in the general case (cf. [75]) the equations
for the invariant tori can be solved by a formal power
series. Equation [84] is proved by induction on k after
checking it for the first few orders.
The algorithm just described leading to [83] can
be extended to the case of the general Hamiltonian
considered in the KAM theorem.
The convergence proof is more delicate than the
(elementary) one for eqn [66]. In fact, the values of
trees of order k can give large contributions to h(k)
n :
because the new factors (w 0 n(l))2 , although not
zero, can be quite small and their small size can
overwhelm the smallness of the factors fn and ". In
fact, even if f is a trigonometric polynomial (so that fn
vanishes identically for jnj large enough) the currents
flowing in the branches can be very large, of the
order of the number k of nodes in the tree; see [82].
This is called the small-divisors problem. The key
to its solution goes back to a related work (SIEGEL)
which shows that
Theorem 4 Consider the contribution to the sum
in [82] from graphs
in which no pairs of lines
F C2 maxjnjN jfn j
the corresponding Val(
) can be bounded by
1
1 k 2k Y
n= def 1
F N
22n4Nk2 Bk
k!
k!
n0
X
B FN 2 2
8n2n=
85
22
s1
X
1
i1
s2
X
1
j1
86
i ; j > 0
where
(A, a) 2 U Tr , U 2 R r , (p, q) 2 V R 2s1 ,
0
(p, k )2V R2s2 with V,V 0 neighborhoods of the
origin
= r s1 s2 , si 0, s1 s2 > 0 and
p and
p
j , j are called Lyapunov coefficients of
the resonance. The perturbations considered are
supposed to have the form "f (A, a, p, q, p, k ). The
denomination of a priori stable or unstable refers to
the properties of the a priori given unperturbed
Hamiltonian. The label a priori unstable is
certainly appropriate if s1 > 0: here also s1 = 0 is
allowed for notational convenience implying that the
Lyapunov coefficients in a priori unstable cases are all
p
of order 1 (whether real j or imaginary i j ). In
a 0 y h" y ;
B B0 K " y ;
b b 0 k" y
y 2 Tr
87
H" A; a 12 A0 12 B2 "f a 0 ; b
and the motions on it are y ! y wt. The above
property, when satisfied, is summarized by saying
that
the
unperturbed
resonant
motions
A = (A00 , B0 ), a = (a 00 w 0 t, b 0 ) can be continued in
presence of perturbation "f , for small ", to quasiperiodic motions with the same spectrum and on a
slightly deformed torus T A00 , b 0 , " .
A priori unstable resonances: here the question is
whether the special invariant tori continue to exist
in presence of small enough perturbations, of
course slightly deformed. This means asking
whether, given A0 such that w(A0 ) = @A H0 (A0 ) has
rationally independent components, there are functions (H " (y ), h" (y )), (P " (y ), Q" (y )) and (P " (y ),
K " (y )) smooth in " near " = 0, vanishing for " = 0,
analytic in y 2 Tr and such that the r-dimensional
surface
A A0 H " y ;
p P" y ;
a y h" y
q Q" y
p P " y ;
k K " y
y 2 Tr
88
89
24
A21
p2
A2 gcos q 1
2
2
"cos 1 sin 2 cos q 1
90
This is a system describing a motion of a pendulum ((p, q) coordinates) interacting with a rotating wheel ((A1 , 1 ) coordinates) and a clock
((A2 , 2 ) coordinates) a priori unstable near the
p
points
p = 0, q = 0, 2
(s1 = 1, s2 = 0, 1 = g,
cf. [86]). It can be proved that on the energy surface
of energy E and for each " 6 0 small enough (no
matter how small) there are initial data with action
i
coordinates close to Ai = (Ai1 , Ai2 ) with (1=2)Ai2
1 A2
close to E eventually evolving to a datum
A0 = (A01 , A02 ) with A01 at a distance from Af1 smaller
than an arbitrarily prefixed distance (of course with
energy E). Furthermore, during the whole process
the pendulum energy stays close to zero within o(")
(i.e., the pendulum swings following closely the
unperturbed separatrices).
In other words, [90] describes a machine (the
pendulum) which, working approximately in a
cycle, extracts energy from a reservoir (the clock)
to transfer it to a mechanical device (the wheel). The
statement that diffusion is possible means that the
machine can work as soon as " 6 0, if the initial
actions and the initial phases (i.e., 1 , 2 , p, q) are
suitably tuned (as functions of ").
The peculiarity of the system [90] is that the fixed
points P of the unperturbed pendulum (i.e., the
equilibria p = 0, q = 0, 2) remain unstable equilibria
even when " 6 0 and this is an important simplifying feature.
It is a peculiarity that permits bypassing the
obstacle, arising in the analysis of more general
cases, represented by the resonance surfaces consisting of the As with A1 1 2 = 0: the latter
correspond to harmonics (1 , 2 ) present in the
perturbing function, i.e., the harmonics which
would lead to division by zero in an attempt to
construct (as necessary in studying [90] by Arnolds
method) the parametric equations of the perturbed
invariant tori with action close to such As. In the
case of [90] the problem arises only on the
resonance marked in Figure 6 by a heavy line, i.e.,
A1 = 0, corresponding to cos 1 in [90].
If " = 0, the points P with p = 0, q = 0 and the
point P with p = 0, q = 2 are both unstable
equilibria (and they are, of course, the same point,
if q is an angular variable). The unstable manifold
(it is a curve) of P coincides with the stable
manifold of P and vice versa. So that the
unperturbed system admits nontrivial motions leading from P to P and from P to P , both in a biinfinite time interval (1, 1): the p, q variables
describe a pendulum and P are its unstable
equilibria which are connected by the separatrices
(which constitute the zero-energy surfaces for the
pendulum).
Af
Ai
Af
(a)
Ai
(b)
Figure 6 (a) The " = 0 geometry: the partial energy lines are
parabolas, (1=2)A21 A2 = const: The vertical lines are the
resonances A1 = rational (i.e., 1 A1 2 = 0). The disks are
neighborhoods of the points Ai and Af (the dots at their centers).
(b) " 6 0; an artists rendering of a trajectory in A space, driven
by the pendulum swings to accelerate the wheel from Ai1 to Af1 at
the expenses of the clock energy, sneaking through invariant tori
not represented and (approximately) located away from the
intersections between resonances and partial energy lines (a
dense set, however). The pendulum coordinates are not shown:
its energy stays close to zero, within a power of ". Hence the
pendulum swings, staying close to the separatrix. The oscillations symbolize the wiggly behavior of the partial energy
(1=2)A21 A2 in the process of sneaking between invariant tori
which, because of their invariance, would be impossible without
the pendulum. The energy (1=2)A21 of the wheel increases
slightly at each pendulum swing: accurate estimates yield an
increase of the wheel speed A1 of the order of "=( log "1 ) at
each swing of the pendulum implying a transition time of the
order of g 1=2 "1 log "1 .
91
26
called the Arnold diffusion. Simple sufficient conditions for a transition from near Ai to near Af are
expressed by the following result:
Theorem 7 Given the Hamiltonian [91] with Hu
admitting two hyperbolic fixed points P with
heteroclinic connections, t ! (pa (t), qa (t)), a = 1, 2,
suppose that:
(i) On the unperturbed energy surface of energy
E = H(Ai ) Hu (P ) there is a regular curve
: s ! A(s) joining Ai to Af such that the
unperturbed tori {A(s)} T can be continued
at " 6 0 into invariant tori T A(s), " for a set of
values of s which fills the curve leaving only
gaps of size of order o(").
(ii) The matrix Dij of the second derivatives of
the integral of f over the heteroclinic motions is
not degenerate, that is,
j det Dj
Z
det
dt @i j f A; a wAt;
pa t; qa t
> c > 0
1
92
!1 A1 !2 A2
def
1=2
93
1=2
e
with w " = (!1 , !2 ), where !1 = "
!, !2 = " !
e > 0 constants. The three scales are
and p
!,
!
!1
g1 , !1
1 ,
2 . In this case, there are many
(although by no means all) pairs A1 , A2 which can
be connected within a time that can be estimated to
be of order O("1 log "1 ).
This is a rapid-diffusion case in an a priori
unstable system in which condition [92] is not
satisfied: because the "-dependence of w(A) implies
that the lower bound c in [92] must depend on "
(and be exponentially small with an inverse power
of " as " ! 0).
The unperturbed system in [93] is nonresonant in
the H0 part for " > 0 outside a set of zero measure
(i.e., where the vector w " satisfies a suitable
Diophantine property) and, furthermore, it is
a priori unstable: cases met in applications can be
a priori stable and resonant (and often not anisochronous) in the H0 part. In such a system, not only
the speed of diffusion is not understood but
proposals to prove its existence, if present (as
expected), have so far not given really satisfactory
results.
For more details, the reader in referred
to Nekhorossev (1977).
1 2
1
R_ W !0 R? R_ !20 R2
2
2
2 "R
Ri
!0
1"
94
with
def
kmS
kmS "
jR Rij
jRj
1 R
g
"
i
i
R R
R
95
1 R
g2
g
96
H 2 !0 G0 "
i
i
R R
R
2L0
28
0 0 f 0 ;
1=2
G20
e 1 2
;
L0
0 0 0 0 f 0
jRj G20
1
R gR 1 ecos0 f0
97
g2
g
!G0 " FG0 ; L0 ;0 ;0 0
R
2L20
98
@ 2 H00
2
6g2 L4
0 !0 h 6 0
@G0 ; L0
if h g2 L2
0 2!G0 6 0
Therefore, the KAM theorem applies to H0" and
the key observation is that the orbits generated by
the Hamiltonian (H" )2 are geometrically the same as
those generated by the Hamiltonian H" : they are
only run at a different speed because of the need of a
time rescaling by the constant factor 2H" .
This shows that, given an unperturbed ellipse of
parameters (L0 , G0 ) such that w = (g2 =L30 , !),
G0 > 0, with !1 =!2 Diophantine, then the perturbed
system admits a motion which is quasiperiodic with
spectrum proportional to w and takes place on an orbit
which wraps around a torus remaining forever close to
the unperturbed torus (which can be visualized as
described by a point moving, according to the area law
"g G40
2R g2 R2
G L 0 G0
101
0 0 ;
0
pq
so that e = 2GL1 1 G(2L)1 and 0 =
and
0 = 0 f0 , where f0 is defined in [42] (see
also [97]). Hence,
0 f ;
0 0 f
s
p 1
G
1
e 2G
L
2L
102
j%j L2 1 e2
1
R
gR
1 e cos f
and the Hamiltonian [100] takes the form
g2
!L !G
2L2
g
" FL G; L; ;
R
H"
103
H
106
30
n v1 M; nl0 n v2
2
w 0 nl0
1
w 0 nl0 2
107
where M is an matrix
Mrs ; nl0
Y
Y n v n v0
"jj
out; r in; s fn v
2
k!
v2
l2 w 0 nl
n v1
k0
M0 nl0
w 0 nl0
n v1
!k
n v2
1
w 0 nl0 2
1
w 0 nl0 2 M0 nl0
n v2
!
108
32
y Gy hy h
y y h
y a
Z
dy
def
y y h
y
a hy h
2
Z
dy
h
y y h
y
2
109
where = (w y ) and the invariant torus corresponds to A0 = w in the map a = y A F(A, y ) and
A0 = A y (A, y ). In fact, by [109] the latter
becomes A0 = A h and, from the second of [75]
written for f depending only on the angles a, it is
A = w h when A, a are on the invariant torus.
Note that if a exists it is necessarily determined by the
third relation in [109] but the check that the second
equation in [109] is soluble (i.e., that the RHS is an exact
gradient up to a constant) is nontrivial. The canonical
map generated by A y F(A, y ) is also defined for A0
close to w and foliates the neighborhood of the invariant
torus with other tori: of course, for A0 6 w the tori
defined in this way are, in general, not invariant.
The reader is referred to Gallavotti et al. (2004)
for more details.
1
e 2
p A WR
2m
c
110
Further Reading
Arnold VI (1989) Mathematical Methods of Classical Mechanics.
Berlin: Springer.
Calogero F and Degasperis A (1982) Spectral Transform and
Solitons. Amsterdam: North-Holland.
33
Differential geometry appeared later in the eighteenth century with the works of Euler Recherches
sur la courbure des surfaces (1760) (Investigations
on the curvature of surfaces) and Monge Une
application de lanalyse a` la geometrie (1795) (An
application of analysis to geometry). Until Gauss
fundamental article Disquisitiones generales circa
superficies curvas (General investigations of curved
surfaces) published in Latin in 1827 (of which one
can find a partial translation to English in Spivak
(1979)), surfaces embedded in R3 were either
described by an equation, W(x, y, z) = 0, or by
expressing one variable in terms of the others.
Although Euler had already noticed that the
coordinates of a point on a surface could be
expressed as functions of two independent variables,
it was Gauss who first made a systematic use of such
a parametric representation, thereby initiating the
concept of local chart which underlies differential
geometry.
Differentiable Manifolds
The actual notion of n-manifold independent of a
particular embedding in a Euclidean space goes back
ber die Hypothesen, welche der
to a lecture U
Geometrie zu Grunde liegen (On the hypotheses
which lie at the foundations of geometry) (of which
one can find a translation to English and comments
in Spivak (1979)) delivered by Riemann at Gottingen
University in 1854, in which he makes clear the
fact that n-manifolds are locally like n-dimensional
Euclidean space. In his work, Riemann mentions
the existence of infinite-dimensional manifolds,
such as function spaces, which today play an
important role since they naturally arise as configuration spaces in quantum field theories.
In modern language a differentiable manifold
modeled on a topological space V (which can be
34
critical point theory which extends to infinitedimensional manifolds and, among other consequences, leads to conclusions on extremals or closed
extremals of variational problems. Rather than
privileging points on a manifold, one can study
instead the geometry of manifolds from the point of
view of spaces of functions, which leads to an
algebraic approach to differential geometry. The
initial concept there is a commutative ring (which
becomes a possibly noncommutative algebra in the
framework of noncommutative geometry), namely
the ring of smooth functions on the manifold, while
the manifold itself is defined in terms of the ring as the
space of maximal ideals. In particular, this point of
view proves to be fruitful to understand supermanifolds, a generalization of manifolds which is important for supersymmetric field theories.
One can further consider the sheaf of smooth
functions on an open subset of the manifold; this
point of view leads to sheaf theory which provides a
unified approach to establishing connections between
local and global properties of topological spaces.
35
Lobatchevsky in 1829 and Bolyai in 1832. NonEuclidean geometries actually played a major role in
the development of differential geometry and Lobachevskys work inspired Riemann and later Klein.
Dropping the positivity assumption for the
bilinear forms gm on Tm M leads to Lorentzian
manifolds which are (n 1)-dimensional smooth
manifolds equipped with bilinear forms on the
tangent spaces with signature (1, n). These occur in
general relativity and tangent vectors with negative,
positive, or vanishing squared length are called
timelike, spacelike, and lightlike, respectively.
Just as complex vector spaces can be equipped with
positive-definite Hermitian products, a complex
manifold M can come equipped with a Hermitian
metric, namely a positive-definite Hermitian product
hm on Tm M for every point m 2 M depending
smoothly on the point m; every Hermitian metric
induces a Riemannian one given by its real part. The
complex projective space CPn comes naturally
equipped with the FubiniStudy Hermitian metric.
Transformation Groups
Metric Properties
Riemann focused on the metric properties of manifolds
but the first clear formulation of the concept of a
manifold equipped with a metric was given by Weyl in
Die Idee der Riemannsche Flache. A Riemannian
metric on a differentiable manifold M is a positivedefinite scalar product gm on Tm M for every point
m 2 M depending smoothly on the point m. A manifold
equipped with a Riemannian metric is called a
Riemannian manifold. A Weyl transformation, which
is multiplying the metric by a smooth positive function,
yields a new Riemannian metric with the same angle
measurement as the original one, and hence leaves the
conformal structure on M unchanged.
Riemann also suggested considering metrics on
the tangent spaces that are not induced from scalar
products; metrics on the manifold built this way
were first systematically investigated by Finsler and
are therefore called Finsler metrics. Geodesics on a
Riemannian manifold M which correspond to
smooth curves : [a, b] ! M that minimize the
length functional
Z s
1 b
d d
L :
;
gt
dt
2 a
dt dt
then generalize to curves which realize the shortest
distance between two points chosen sufficiently close.
Euclids axioms which naturally lead to Riemannian geometry are also satisfied up to the axiom
of parallelism by a geometry developed by
36
Fiber Bundles
Transformation groups give rise to principal fiber
bundles which play a major role in YangMills
theory. The notion of fiber bundle first arose out of
questions posed in the 1930s on the topology and the
geometry of manifolds, and by 1950 the definition of
fiber bundle had been clearly formulated by Steenrod.
A smooth fiber bundle with typical fiber a
manifold F is a triple (E, , B), where E and B are
smooth manifolds called the total space and the base
space, and : E ! B is a smooth surjective map
called the projection of the bundle such that the
preimage 1 (b) of a point b 2 B called the fiber of
the bundle over b is isomorphic to F and any base
point b has a neighborhood U B with preimage
1 (U) diffeomorphic to U F, where the diffeomophisms commute with the projection on the base
d
RexptX p
dtjt0
37
Connections
On a manifold there is no canonical method to
identify tangent spaces at different points. Such an
identification, which is needed in order to differentiate vector fields, can be achieved on a Riemannian
manifold via parallel transport of the vector fields.
The basic concepts of the theory of covariant
differentiation on a Riemannian manifold were given
at the end of the nineteenth century by Ricci and, in a
more complete form, in 1901 in collaboration with
Levi-Civita in Methodes de calcul differentiel absolu et
leurs applications; on a Riemannian manifold, it is
possible to define in a canonical manner a parallel
displacement of tangent vectors and thereby to
differentiate vector field covariantly using the since
then called Levi-Civita connection.
More generally, a (linear) connection (or equivalently a covariant derivation) on a vector bundle E
over a manifold M provides a way to identify fibers
of the vector bundle at different points; it is a map r
taking sections of E to E-valued 1-forms on M
which satisfies a Leibniz rule, r(f ) = df f r,
for any smooth function f on M. When E is the
tangent bundle over M, curves on the manifold
_ = 0 give rise
with covariantly constant velocity r(t)
_ =X 2
to geodesics. Given an initial velocity (0)
Tm M and provided X has small enough norm, X (1)
defines a point on the corresponding geodesic and
the map exp : X 7! X (1) a diffeomorphism from a
neighborhood of 0 in Tm M to a neighborhood of
m 2 M called the exponential map of r.
The concept of connection extends to principal
bundles where it was developed by Ehresmann
building on the work of Cartan. A connection on a
principal bundle (P, , B) with structure group G,
which is a smooth equivariant (under the action of
8 i 6 j
38
Curvature
The concept of curvature, which is now understood in terms of connections (the curvature of a
connection r is defined by = r2 ), historically
arose prior to that of connection. In its modern
form, the concept of curvature dates back to Gauss.
Using a spherical representation of surfaces the
Gauss map , which sends a point m of an oriented
surface R3 to the outward pointing unit normal
vector m Gauss defined what is since then called
the Gaussian curvature Km at point m 2 U as
the limit when the area of U tends to zero of the
ratio area( (U))=area(U). It measures the obstruction to finding a distance-preserving map from a
piece of the surface around m to a region in the
standard plane. Gauss Teorema Egregium says that
the Gaussian curvature of a smooth surface in R3 is
defined in terms of the metric on the surface so that
it agrees for two isometric surfaces.
From the curvature of a connection on a
Riemannian manifold (M, g), one builds the
@ @ @ @
: g
;
;
@i @j @k @l
Cohomology
Differentiation of functions f 7! df on a differentiable manifold M generalizes to exterior differentiation
7! d
of differential forms. A form
is closed
whenever it is in the kernel of d and it is exact
whenever it lies in the range of d. Since d2 = 0, exact
forms are closed.
Cartans structure equations d! = (1=2)[!, !]
relate the exterior differential of the connection 1-form
! on a principal bundle to its curvature given by
the exterior covariant derivative D! := d! h, where
h : Tp P ! Hp P is the projection onto the horizontal
space.
On a complex manifold, forms split into sums
of (p, q)-forms, those with p-holomorphic and
q-antiholomorphic components, and exterior differentiation splits as d = @ @ into holomorphic and
antiholomorphic derivatives, with @ 2 = @ 2 = 0.
Geometric data are often expressed in terms of
closedness conditions on certain differential forms.
For example, a symplectic manifold is a manifold
M equipped with a closed nondegenerate differential
2-form called the symplectic form. The theory of
J-holomorphic curves on a manifold equipped with
an almost-complex structure J has proved fruitful in
building invariants on symplectic manifolds. A
Kahler manifold is a complex manifold equipped
with a Hermitian metric h whose imaginary part
Im h yields a closed (1, 1)-form. The complex
projective space CPn is Kahler.
The exterior differentation d gives rise to de Rham
cohomology as Ker d=Im d, and de Rhams theorem
establishes an isomorphism between de Rham cohomology and the real singular cohomology of a
manifold. Chern (or characteristic) classes are topological invariants associated to fiber bundles and play
a crucial role in index theory. ChernWeil theory
builds representatives of these de Rham cohomology
classes from a connection r of the form tr(f (r2 )),
where f is some analytic function.
When the manifold is Riemannian, the Laplace
Beltrami operator on functions generalizes to differential forms in two different ways, namely to the
Bochner Laplacian T M on forms (i.e., sections of
T M), where the contangent bundle T M is
equipped with a connection induced by the Levi-Civita
connection and to the LaplaceBeltrami operator on
forms (d d )2 = d d d d , where d is the (formal)
adjoint of the exterior differential d. These are related
via Weitzenbocks formula which in the particular case
of 1-forms states that the difference of those two
operators is measured by the Ricci curvature.
When the manifold is compact, Hodges theorem
asserts that the de Rham cohomology groups are
39
isomorphic to the space of harmonic (i.e., annihilated by the LaplaceBeltrami operator) differential
forms. Thus, the dimension of the set of harmonic
k-forms equals the kth Betti numbers from which
one can define the Euler characteristic (M) of the
manifold M taking their alternate sum. Hodge
theory plays an important role in mirror symmetry
which posits a duality between different manifolds
on the geometric side and between different field
theories via their correlation functions on the
physics side. CalabiYau manifolds, which are
Ricci-flat Kahler manifolds, are studied extensively
in the context of duality.
Index Theory
While the Gaussian curvature is the solution to a
local problem, it has strong influence on the global
topology of a surface. The GaussBonnet formula
(1850) relates the Euler characteristic on a closed
surface to the Gaussian curvature by
Z
1
M
Km dAm
2 M
where dAm is the volume element on M. This is the
first result relating curvature to global properties
and can be seen as one of the starting points for
index theory. It generalizes to the ChernGauss
Bonnet theorem (1944) on an even-dimensional
closed manifold and can be interpreted as an
example of the AtiyahSinger index theorem (1963)
Z
^ g etrW
indDW
A
g
M
40
Further Reading
Bishop R and Crittenden R (2001) Geometry of Manifolds.
Providence, RI: AMS Chelsea Publishing.
Chern SS, Chen WH, and Lam KS (2000) Lectures on Differential
Geometry, Series on University Mathematics. Singapore: World
Scientific.
Choquet-Bruhat Y, de Witt-Morette C, and Dillard-Bleick M
(1982) Analysis, Manifolds and Physics, 2nd edn. Amsterdam
New York: North Holland.
Gallot S, Hulin D, and Lafontaine J (1993) Riemannian Geometry,
Universitext. Berlin: Springer.
Helgason S (2001) Differential, Lie Groups and Symmetric Spaces.
Graduate Studies in Mathematics 36. AMS, Providence, RI.
Introduction
The modern theory of electromagnetism is built on
the foundations of Maxwells equations:
0
div B 0
div E
curl B
1 @E
0 J
c2 @t
@B
0
@t
curl E
41
42
ker
r3
7
kee0
r2
5
6
8
k0 ev ^ r
r3
9
1
;
40
k0
0
4
10
0 1:3 106
11
43
er
40 r3
0
0 r3
40 r3
r5
everywhere except at r = 0. Therefore, by the
divergence theorem,
Z
E dS 0
13
@V
SR
@V
14
15
44
0 ev ^ r
4r3
17
f eE v ^ B
A second observer O0 moving relative to the first with
velocity v will see the same force, but now acting on a
particle at rest. He will therefore measure the electric
field to be E0 = f =e. We conclude that an observer
moving with velocity v through a magnetic field B and
an electric field E should see an electric field
E0 E v ^ B
18
19
0 ev ^ r
4r3
0 ev ^ r
Ov2 =c2
4r3
20
div B 0
21
45
22
So we must have
0 curl E0
curl E curlv ^ B
23
46
@B
0
@t
B
24
0 ev ^ r
Ov2 =c2
4r3
er
40 r3
cB
c
c
c
By taking the curl of both sides, and dropping terms
of order v2 =c2 ,
v^E
curlcB curl
c
1
v div E v grad E
c
But
div E =0 ;
v grad E
@E
@t
by [22]. Therefore,
curlcB
1 @E
1
J c0 J
c @t c0
Maxwells Equations
curl B
1 @E
0 J
c2 @t
div B 0
curl B
27
1 @E
0 J
c2 @t
28
@B
0
@t
29
curl E
Conservation of Charge
To see the connection between the continuity
equation and charge conservation, let us look at
the total charge within a fixed V bounded by a
surface S. If charge is conserved, then any increase
or decrease in a short period of time must be
exactly balanced by an inflow or outflow of charge
across S.
Consider a small element dS of S with outward
unit normal and consider all the particles that have a
particular charge e and a particular velocity v at
time t. Suppose that there are of these per unit
volume ( is a function of position). Those that cross
the surface element between t and t
t are those
that at time t lie in the region of volume
jv n dS
tj
shown in Figure 1. They contribute e v dS
t to the
outflow of charge through the surface element. But
the value of J at the surface element is the sum of
e v over all possible values of v and e. By summing
over v, e, and the elements of the surface, therefore,
and by passing to the limit of a continuous
distribution, the total rate of outflow is
Z
J dS
S
47
30
dS
dt
dt
48
@E
@t
34
div B 0
35
curl B
1 @2
1 @2
@2
@2
@2
2
r
c2 @t2
c2 @t2 @x2 @y2 @z2
39
40
!
@B
e ^ a sin e ^ b cos
42
c
@t
A possible choice is
1 @E
0
c2 @t
36
@B
0
@t
37
curl E
p
where c = 1= 0 0 . By taking the curl of eqn [36]
and by substituting from eqns [35] and [37], we
obtain
1
@E
2
0 grad div B r B 2 curl
c
@t
1 @
r2 B 2 curl E
c @t
1 @2B
r2 B 2 2
38
c @t
e^E 1
e ^ a cos e ^ b sin
c
c
43
Potentials
For every solution of Maxwells equations in vacuo,
the components of E and B satisfy the threedimensional wave equation; but the converse is not
true. That is, it is not true in general that if
&B 0;
&E 0
45
B curl A;
E grad
a a grad u
for any scalar function u of position and time. It
should be kept in mind that the existence statements
are local. If v is defined on a region U with
47
@A
@t
48
curl E
@B
0
@t
@
div A
@t
0
1 @E
r2 A grad div A
c2 @t
1 @
@A
grad
2
c @t
@t
0 J
46
@A
@t
49
If we put
1 @
div A
c2 @t
@t 0
&A grad 0 J
&
50
Gauge Transformations
& =0 ;
51
&A 0
52
1 @
0
c2 @t
@A0
grad 0 E
@t
@A
@u
grad
E
@t
@t
@u
grad
@t
We can, for example, choose three arbitrary solutions of the scalar wave equation for the components of the vector potential, and then define by
Z
c2 div Adt
Historical Note
A 7! A0 A grad u;
7! 0
@u
@t
49
1 @
divA0 &u
c2 @t
&u
1 @0
0
c2 @t
Further Reading
52
N
X
@qi qi qj
Wwall qi
j1
def
@qi Q
1
KP Q
HP; Q
X 1
def
p2i Q K
2m
i
du pdv
is exact
T
4
x U;V
p
U xdx
x U;V
) dS
p
dU @V xdVdx= K
p
R
dx= KK
5
p p
Noting that dx= K = 2=m dt, [3] follows because
time averages
are given by integrating with p
respect
p
3
54
bounce back. One particle in the layer will contribute to the average of @L (x) the amount
Z t1
1
2 0 L j dt
6
total time t0
if t0 is the first instant when the point j enters the
layer and t1 is the instant when the -component of
the velocity vanishes against the wall. Since
0 (L j ) is the -component of the force, the
integral is 2mjj j (by Newtons law), provided, of
course, j > 0:
Suppose that no collisions between particles occur
while the particles travel within the range of the
potential of the wall, i.e., the mean free path is much
greater than the range of the potential defining the
wall. The contribution of collisions to the average
momentum transfer to the wall per unit time is
therefore given by, see [2],
Z
2mv f vwall Av dv
v>0
with
C
HP; Q UdP dQ
def
7
Z
1 T
FSt P; Qdt
T!1 T 0
R
FP 0 ; Q0 HP0 ; Q0 U dP 0 dQ0
R
HP0 ; Q0 U dP 0 dQ0
lim
8
Ensembles
Eventually Boltzmann in 1884 realized that the
validity of the heat theorem for averages computed
via the right-hand side (rhs) of [8] held independently of the ergodic hypothesis, that is, [8] was not
necessary because the heat theorem (i.e., [3]) could
also be derived under the only assumption that the
averages involved in its formulation were computed
1
dP dQ
HP; Q U
Zmc U; N; V
N!hdN
9
1
dPdQ
eHP;Q
Zc ; N; V
N!hdN
10
1
dPdQ
e NHP;Q
Zgc ; ; V
N!hdN
11
Hence, there are several different models of thermodynamics. The key tests for accepting them as real
microscopic descriptions of macroscopic thermodynamics are as follows.
1. A correspondence between the macroscopic
states of thermodynamic equilibrium and the
elements of a collection of probability distributions on phase space can be established by
identifying, on the one hand, macroscopic
thermodynamic states with given values of the
thermodynamic functions and, on the other,
Z
2KP
; @ q1 2mv1 n2 dq1
3N
12
56
Equivalence of Ensembles
BOLTZMANN proved that, computing averages via the
microcanonical or canonical distributions, the essential property [3] was satisfied when changes in their
parameters (i.e., u, v or , v, respectively) induced
changes du and dv on energy and volume, respectively. He also proved that the function s, whose
existence is implied by [3], was the same function
once expressed as a function of u, v (or of any pair
of thermodynamic parameters, e.g., of T, v or p, u).
A close examination of Boltzmanns proof shows
that the [3] holds exactly in the canonical ensemble
and up to corrections tending to 0 as ! 1 in the
microcanonical ensemble. Identity of thermodynamic functions evaluated in the two ensembles
holds, as a consequence, up to corrections of this
order. In addition, Gibbs added that the same held
for the grand canonical ensemble.
Of course, not every collection of stationary
probability distributions on phase space would
provide a model for thermodynamics: Boltzmann
called orthodic the collections of stationary
distributions which generated models of thermodynamics through the above-mentioned identification of its elements with macroscopic equilibrium
states. The microcanonical, canonical, and the later
grand canonical ensembles are the chief examples
of orthodic ensembles. Boltzmann and Gibbs
proved these ensembles to be not only orthodic
but to generate the same thermodynamic functions,
that is to generate the same thermodynamics.
This meant freedom from the analysis of the truth
of the doubtful ergodic hypothesis (still unproved in
any generality) or of the monocyclicity (manifestly
false if understood literally rather than regarding the
phase space as consisting of finitely many small,
discrete cells), and allowed Gibbs to formulate the
problem of statistical mechanics of equilibrium as
follows.
Problem Study the properties of the collection of
probability distributions constituting (any) one of
the above ensembles.
However, by no means the three ensembles just
introduced exhaust the class of orthodic ensembles
producing the same models of thermodynamics in
the limit of infinitely large systems. The wealth of
Thermodynamic Limit
Adopting Gibbs axiomatic point of view, it is
interesting to see the path to be followed to achieve
an equivalence proof of three ensembles introduced
in the section Heat theorem and ergodic
hypothesis.
A preliminary step is to consider, given a cubic
box of volume V = Ld , the normalization factors
Zgc (, , V), Zc (, N, V), and Zmc (U, N, V) in [9],
[10], and [11], respectively, and to check that the
following thermodynamic limits exist:
1
log Zgc ; ; V
V!1 V
1
def
log Zc ; N; V
fc ; lim
N
V!1;N
V
def
pgc ; lim
13
k1
B smc u;
def
lim
V!1;N=V; U=Nu N
def
log Zmc U; N; V
14
58
Z ; N; V
1
X
e N Zc ; N; V
N0
Z 1
15
eU Zmc U; N; V dU
B
1
fc ; v1 maxu k1
B smc u; v
16
kB @u smc uc ; v1
17
Physical Interpretation of
Thermodynamic Functions
The existence of the limits [13] implies several
properties of interest. The first is the possibility of
finding the physical meaning of the functions
pgc , fc , smc and of the parameters ,
Note first that, for all V the grand canonical average
hKi, is (d=2)1 hNi, so that 1 is proportional to
the temperature Tgc = T(, ) in the grand canonical
distribution: 1 = kB T(, ). Proceeding heuristically,
the physical meaning of p(, ) and can be found
through the following remarks.
ConsiderR the microcanonical distribution mc
u, v and
denote by
the integral over (P, Q) extended to the
domain of the (P, Q) such that H(P, Q) = U and, at
the same time, q1 2 dV, where dV is an infinitesimal
volume surrounding the region . Then, by the
microscopic definition of the pressure p (see the
introductory section), it is
Z
N
2 p21 dP dQ
pdV
ZU; N; V
3 2m N!hdN
Z
2
dP dQ
KP
18
3ZU; N; V
N!hdN
where (H(P, Q) U). The RHS of [18] can be
compared with
Z
@V ZU; N; VdV
N
dP dQ
ZU; N; V
ZU; N; V
N!hdN
to give
@V Z dV
p dV
N
p dV
Z
2=3hKi
R R
K= 1,
because hKi , which denotes the average
should be essentially the same as the microcanonical
average hKimc (i.e., insensitive to the fact that one
particle is constrained to the volume dV) if N is
large. In the limit V ! 1, V=N = v, the latter
remark together with the second of [17] yields
1
k1
B @v smc u; v pu; v;
k1
B @u smc u; v
19
20
a independent; a gc; c; mc
21
60
3.
4.
5.
6.
an idealization void of physical reality, it is nevertheless useful to define such states because certain
notions (e.g., that of pure state) can be sharply
defined, with few words and avoiding wide circumvolutions, in terms of them. Therefore, let:
Definition An infinite-volume state with parameters
(, v), (u, v) or (, ) is a collection of average values
F ! hFi obtained, respectively, as limits of finitevolume averages hFin defined from canonical, microcanonical, or grand canonical distributions in n with
fixed parameters (, v), (u, v) or (, ) and with general
boundary condition of fixed external particles, on
sequences n ! 1 for which such limits exist simultaneously for all local observables F.
Having set the definition of infinite-volume
state consider a local observable G(X) and let
22
62
j1
x1 ; x2 ; . . . ; xp lim
n
n
p
Nn Y
X
+
xj qij
i1 ;...;ip j1
n
q; q
R
def
lim
n
i = f i , by
both sides of the equations of motion, mq
(1=2)qi and summing over i, it follows that
N
N
1X
1X
def 1
i
mqi q
qi f i Cq
2 i1
2 i1
2
and
hCext i 3pV
hKi 32 pV 12hCint i
Nn 2!Zc0 ;; Vn
q; q0 x qq0 x ! 0
x!1
23
or
1
hCint i
pv
3N
where
def
Zc0
eUQ dQ
24
N X
X
f j!i qi
i1 i6j
@ qi qi qj qi qj
i<j
25
where
I
1
2
eq 1d3 q
I
Ov2 1
v
O
p 2 v 1
v
v 1 b=v
v2
or
a
p 2 v b 1 Ov2
v
26
27
vi
vg
64
28
a
2 P0 v
v
Maxwell0 s rule
29
Continuous Symmetries: No d = 2
Crystal Theorem
A second case in which it is possible to rule out
existence of phase transitions or at least of certain
kinds of transitions arises when the system under
analysis enjoys large symmetry. By symmetry is
meant a group of transformations acting on the
configurations and transforming each of them into a
XY
p
qpn a n
dQ
N!
j1
66
"!0 N!1
30
q1 ; .. .; qn
1
X
1
znm
Zgc ; ; V m0
Z
dy dym
eq1 ;...;qn ;y1 ;...;ym 1
m!
q1 ; . . . ; qn zn;1 q1 K q1 ; . . . ; qn 33a
with
def
s!
s1
k1
q2 ;. ..;qn ; y1 ; ...; ys
1
X
ck k
k2
32
33b
r3p
35
68
0
0 z
36
Lattice Models
The problem of proving the existence of phase
transitions in models of homogeneous gases with
pair interactions is still open. Therefore, it makes
sense to study the problem of phase transitions
in simpler models, tractable to some extent but
nontrivial, and which are of practical interest in
their own right.
The simplest models are the so-called lattice
models in which particles are constrained to points
of a lattice: they cannot move in the ordinary sense
of the word (but, of course, they could jump) and
eHX
jXj;N
Zcp ; N;
38a
e jXj eHX
Zgc
p ; ;
38b
or
p; X
P
P
exp h x x;y x yx y
40
Zgc
s ; h;
!1
V=Nv
1
log Zcp ; N;
N
1
gc
log Zp ; ;
!1 V
1
g; m lim log Zcs ; M;
!1; V
M=V!m
1
log Zgc
s ; ;
!1 V
42
p; lim
f ; h lim
41
More generally,
let H(s) in [39] be replaced by
P
H(s) = X JX X with JX 0 and X can be any
finite set; then, if Y = (y1 , . . . , yn ), X = (x1 , . . . , xn ),
the following Griffiths inequalities hold:
hX i 0;
70
The inequalities can be used to check, in ferromagnetic Ising models, [39], existence of infinite-volume
states (cf. the sections Phase transitions and boundary
conditions and Lattice models) obtained by fixing
the boundary condition B to be either all external
spins or all external sites empty. If hFiB,
denotes the grand canonical average with boundary
condition B and any fixed , h > 0, this means that
for all local observables F(s ) (i.e., for all F depending
on the spin configuration in any fixed region ) all the
following limits exist:
lim hFiB; hFiB
!1
44
46
1
X
p2 3p e2Jp ! 0
p4
!1
Finite-Volume Effects
The description in the last section of the phase
transition in the nearest-neighbor Ising model can be
made more precise both from physical and mathematical points of view giving insights into the nature
of the phase transitions. Assume that the boundary
condition is the ()-boundary condition and
describe a spin configuration s by means of the
associated closed disjoint polygons (
1 , . . . ,
n ).
Attribute to s = (
1 , . . . ,
n ) a probability proportional to [45]. Then the following MinlosSinais
theorem holds:
Theorem If is large enough there exist C > 0,
(
) > 0 with (
) e2Jj
j and such that a spin
configuration s randomly chosen out of the grand
canonical distribution with boundary conditions
and h = 0 will contain, with probability approaching
1 as ! 1, a number K(
) (s) of contours congruent to
such that
p
47
jK
s
jjj C jj eJj
j
and this relation holds simultaneously for all
s.
72
m(, h)
1
m*()
O(||1/2)
O(||1/2)
m*()
(||1/2)
O (||1/2)
h
h
m*()
m*()
(a)
(b)
Figure 3 The h vs m (, h) graphs for finite and (a) and (b) conditions.
m(, h)
m(, h)
1
m*()
O(||1/2)
m*()
O(||1/2)
h
h
m*()
m*()
(a)
(b)
Figure 4 (a) The h vs m (, h) graph for periodic or empty boundary conditions. (b) The discontinuity (at h = 0) of the thermodynamic limit.
h!0
and if
V 1 number of zeros of the form
zj eij ; j d !
!1
d
2
it is
f ; h 2J
1
2
48
49
X
0
h!0
50
74
51
2J
j i j
j j j
52
53
54
!
where ()
! 1 exponentially fast; the exponent 3/4, P
here and below, is not
P optimal.
(iii) If M
=
x and M
x2
x2 x , then
3=4
jM
m Vj < V
3=4
M
1 m Vj < V
55
(iv) If K
(s) denotes the number of contours congruent to a given
and lying in
then,
simultaneously for all the shapes of
:
jK
s
V j CeJj
j V 1=2 ; C > 0
56
76
lim
Critical Points
Correlation functions for a system with short-range
interactions and in an equilibrium state (which is
a pure phase) have cluster properties (see [22]):
their physical meaning is that in a pure phase there
is independence between fluctuations occurring in
widely separated regions. The simplest cluster
property concerns the pair correlation function,
that is, the probability density (q1 , q2 ) of finding
particles at points q1 , q2 independently of where
the other particles may happen to be (see [23]).
In the case of spin systems, the pair correlation
(q1 , q2 ) = hq1 q2 i will be considered. The pair
correlation of a translation-invariant equilibrium
state has a cluster property ([22], [42]), if
jq1 ; q2 2 j ! 0
jq1 q2 j!1
57
jxyj
e
A p ;
jx yj
1
Ac
;
jx yj1=4
A
jxyj
jx yj2
58
59
Fluctuations
As it appears from the discussion in the last section,
fluctuations of observables around their averages
have interesting properties particularly at critical
points. Of particular interest are observables that
78
Z
p
jj 2 a; b !
!1
dz
2
p ez =2D
2D
60
61
jj
!
p mh2
o dp
const:exp jj
2Dh
62
Renormalization Group
The theory of fluctuations just discussed concerns
only fluctuations of a single quantity. The problem
of joint fluctuations of several quantities is also
interesting and in fact led to really new developments in the 1970s. It is necessary to restrict
attention to rather special cases in order to illustrate
some ideas and the philosophy behind the approach.
Consider, therefore, the equilibrium distribution 0
associated with one of the classical equilibrium
ensembles. To fix the ideas we consider the
equilibrium distribution of an Ising energy function
H0 , having included the temperature factor in the
energy: the inclusion is done because the discussion
will deal with the properties of 0 as a function of .
It will also be assumed that the average of each spin
is zero (no magnetic field, see [39] with h = 0).
Keeping in mind a concrete case, imagine that H0
is the energy function of the nearest-neighbor Ising
ferromagnet in zero field.
Imagine that the volume of the container has
periodic boundary conditions and is very large,
ideally infinite. Define the family of blocks kx,
parametrized by x 2 Zd and with k an integer,
consisting of the lattice sites x = {ki xi < (k 1)
i }. This is a lattice of cubic blocks with side size k
that will be called the k-rescaled lattice.
P
Given , the quantities mx = kd x2kx x are
called the block spins and define the map
R,k 0 = k transforming the initial distribution on
the original spins into the distribution of the block
spins. Note that if the initial spins have only two
values x = 1, the block spins take values between
kd=kd and kd=kd at steps of size 2=kd . Furthermore, the map R, k makes sense independently of
how many values the initial spins can assume, and
even if they assume a continuum of values Sx 2 R.
Taking = 1 means, for k large, looking at the
probability distribution of the joint large fluctuations
in the blocks kx. Taking = 1=2 corresponds to
studying a joint central limit property for the block
variables.
Considering a one-parameter family of initial
distributions 0 parametrized by a parameter
80
fax g
Rn1=2 H0 !
n!1
63
x2
64
X
x2Z
1
S2x
2D
d
65
82
Quantum Statistics
Statistical mechanics is extended to assemblies of
quantum particles rather straightforwardly. In the
case of N identical particles, the observables are
operators O on the Hilbert space
HN L2 N
or
HN L2 C2 N
N
X
X
2 X
h
qj
qj qj0
wqj
2m j1
j<j0
j
def
KF
66
hOi
tr eHN O
tr eHN
67
0
n q ; q1 ; . . . ; qN1 dq1
. . . dqN1
68
V!1
69
xx !1
70
Cooper Pairs
The superconductivity theory has been phenomenologically related to the existence of Cooper pairs.
Consider the Hamiltonian [66] and define (cf. [69])
x y; ; x0 y0 ; 0 ; x x0
def
2 x; ; y; ; x0 ; 0 ; y0 ; 0
84
N2
2
2
so that
Pcollapse
Z
dpdq Kp q
e
h3N N!
ZC
dpdq Kp q
e
h3N N!
p3 !N
4 2m 1
3N b1=2NN 1
e
3
3h
N!
Z
dq q
e
h3N N!
71
as
q!1
pd
1
log Zgc ; ; V e eB 2m 1 73
V
86
pn
Ldn
log Zn
N p2
X
j
j1
UQ "WQ
75
76
jhfA ; Cgij2
hfC; fC ; Hggi
77
AQ
N
X
def
gqj ;
CP; Q
j1
Then H =
P1
2
2 pj
N
X
pj hqj 78
j1
F(q1 , . . . , qN ), if
Fq1 ; . . . ; qN
X
1X
jqj qj0 j "
Wqj
2 j6j0
j
Choose
g(q) ei(k K) q , h(q) = cos q k
and
bound (hj hj0 )2 by k 2 (qj qj0 )2 , (@ j hj )2 by k 2 and
hA ; Ai j
N
X
iq k K 2
j1
and
D X
E2
hj @gj
jhfA ; Cgij2
j
jK k j2 N2 " K " K 2k 2
hence [77] becomes, after multiplying both sides
by the auxiliary function
(k ) (assumed even and
vanishing for jk j > =a) and summing over k ,
*
+
N
X
1 X
def 1
iKk qj 2
j
k
e
j
D1
N k
N j1
1X
k
N k
81
Further Reading
Aizenman M (1980) Translation invariance and instability of phase
coexistence in the two dimensional Ising system. Communications in Mathematical Physics 73: 8394.
Aizenman M (1982) Geometric analysis of 4 fields and Ising
models. 86: 148.
Aizenman M, Lieb EH, Seiringer R, Solovej JP, and Yngvason J
(2004) BoseEinstein condensation as a quantum phase
transition in a optical lattice, Physical Review A 70: 023612.
Baxter R (1982) Exactly Solved Models. London: Academic Press.
Benfatto G and Gallavotti G (1995) Renormalization group.
Princeton: Princeton University Press.
Bleher P and Sinai Y (1975) Critical indices for Dysons asymptotically hierarchical models. Communications in Mathematical
Physics 45: 247278.
ber die mechanische Bedeutung des zweiten
Boltzmann L (1968a) U
Haupsatzes der Warmetheorie. In: Hasenohrl F (ed.) Wissenschaftliche Abhandlungen, vol. I, pp. 933. New York: Chelsea.
Boltzmann L (1968b) Uber die Eigenshaften monzyklischer und
anderer damit verwandter Systeme. In: Hasenohrl FP (ed.)
Wissenshafltliche Abhandlungen, vol. III, pp. 122152.
New York: Chelsea.
Dobrushin RL (1968) Gibbsian random fields for lattice systems
with pairwise interactions. Functional Analysis and Applications 2: 3143.
Domb C and Green MS (1972) Phase Transitions and Critical
Points. New York: Wiley.
Dyson F (1969) Existence of a phase transition in a onedimensional
Ising ferromagnet. Communications in Mathematical Physics 12:
91107.
Dyson F and Lenard A (1967, 1968) Stability of matter. Journal
of Mathematical Physics 8: 423434, 9: 698711.
Friedli S and Pfister C (2004) On the singularity of the free energy at
a first order phase transition. Communications in Mathematical
Physics 245: 69103.
Gallavotti G (1999) Statistical Mechanics. Berlin: Springer.
Gallavotti G, Bonetto F and Gentile G (2004) Aspects of the
Ergodic, Qualitative and Statistical Properties of Motion.
Berlin: Springer.
Gawedzky K and Kupiainen A (1983) Block spin renormalization
group for dipole gas and (@)4 . Annals of Physics 147:
198243.
Gawedzky K and Kupiainen A (1985) Massless lattice 44 theory:
rigorous control of a renormalizable asymptotically free model.
Communications in Mathematical Physics 99: 197252.
Gibbs JW (1981) Elementary Principles in Statistical Mechanics.
Woodbridge (Connecticut): Ox Bow Press (reprint of the 1902
edition).
Higuchi Y (1981) On the absence of non translationally invariant
Gibbs states for the two dimensional Ising system. In: Fritz J,
Lebowitz JL, and Szaz D (eds.) Random Folds. Amsterdam:
North-Holland.
Hohenberg PC (1967) Existence of long range order in one and
two dimensions. Physical Review 158: 383386.
88
Introduction
Functional analysis is concerned with the study of
functions and function spaces, combining techniques
borrowed from classical analysis with algebraic
techniques. Modern functional analysis developed
around the problem of solving equations with
solutions given by functions. After the differential
and partial differential equations, which were
studied in the eighteenth century, came the integral
equations and other types of functional equations
investigated in the nineteenth century, at the end of
which arose the need to develop a new analysis,
with functions of an infinite number of variables
instead of the usual functions. In 1887, Volterra,
inspired by the calculus of variations, suggested a
new infinitesimal calculus where usual functions are
replaced by functionals, that is, by maps from a
function space to R or C, but he and his followers
were still missing some algebraic and topological
tools to be developed later. Modern analysis was
born with the development of an algebra of the
infinite closely related to classical linear algebra
which by 1890 had (up to the concept of duality,
analysis, global analysis, the theory of pseudodifferential operators, differential geometry, operator
algebras, noncommutative geometry, etc.
89
1
X
2k
k1
kf gkk
1 kf gkk
2 N n0
90
P
k
as the fermionic Fock space F (H) = 1
k=0 H
(where k stands for the antisymmetrized (closed)
tensor product).
A prototype of Hilbert space is the space l2 (Z) of
complex-valued
sequences (un )n2Z such that
P
2
ju
j
is
finite,
which is already implicit in
n2Z n
Hilberts Grundzugen. Shortly afterwords, Riesz and
Fischer, with the help of the integration tool
introduced by Lebesgue, showed that the space
L2 (]0, 1[) (first introduced by Riesz) of squaresummable functions on the interval ]0, 1[, that is,
functions f such that
Z 1
1=2
2
kf kL2
jf xj dx
0
Duality
The concept of duality (in a topological sense) was
initiated at the beginning of the twentieth century by
Hadamard, who was looking for continuous linear
functionals on the Banach space C(I) of continuous
functions on a compact interval I equipped with a
uniform topology. It is implicit in Hilberts theory
and plays a central part in Riesz work, who
managed to express such continuous functionals as
Stieltjes integrals, one of the starting points for the
modern theory of integration.
The topological dual of a topological vector space
E is the space E of continuous linear forms on E
which, when E is a normed space, can be equipped
with the dual norm kLkE = supu2E, kuk1 jL(u)j.
Dual spaces often provide a receptacle for singular
objects; any of the functions f 2 Lp (Rn )(p 1) and
the delta-function at point x 2 Rn, x : f 7! f (x), all lie
in the space S 0 (R n ) dual to S(Rn ) of tempered
distributions on Rn , which is itself contained in the
space D0 (Rn ) of distributions dual to D(Rn ).
Furthermore, the topological dual E of a nuclear
space E contains the support of a probability
91
Fourier Transform
In the middle of the eighteenth century, oscillations
of a vibrating string were interpreted by Bernouilli
as a limit case for the oscillation of n-point masses
when n tends the infinity, and Bernouilli introduced
the novel idea of the superposition principle by
which the general oscillation of the string should
decompose in a superposition of proper oscillations. This point of view triggered off a discussion
as to whether or not an arbitrary function can be
expanded as a trigonometric series. Other examples
of expansions in orthogonal functions (this terminology actually only appears with Hilbert) had been
found in the mean time in relation to oscillation
problems and investigations on heat theory, but it
was only in the nineteenth century, with the works
of Fourier and Dirichlet, that the superposition
problem was solved.
Separable Hilbert spaces can be equipped with a
countable orthonormal system {en }n2Z (hen , em iH =
mn with h , iH the scalar product on H) which is
92
n2Z
2
Z
Rn
eix A x; ^f d
^ : eihx;iH
dx
H
Fredholm operators
A complex-valued continuous function K on [0, 1]
[0, 1] gives rise to an integral operator
Z 1
A:f !
Kx; yf y dy
0
8u 2 E; 8v 2 F
is also finite dimensional, so that it has a welldefined index ind(B) = dim(Ker B) dim(Ker B ), a
starting point for index theory. Toplitz operators
T , where is a continuous function on the unit
circle S1 , provide first examples of Fredholm
operators; they act on the Hardy space H2 (S1 ) by
!
X
X
am e m
amn em
Ten
m0
m0
93
Laplace equation
@ t
u 0
@t2
@u
u 0
@t
wave equation
heat equation
@u
u
@t
8u; v 2 DA DA
94
Ut u u
;
t
u2H
Tt u u
;
t
u2H
Spectral Theory
Spectral theory is the study of the distribution of the
values of the complex parameter for which, given
a linear operator A on a normed space E, the
operator A I has an inverse and of the properties
of this inverse when it exists, the resolvent
R(A, ) = (A I)1 of A. The resolvent (A) of A
is the set of complex numbers for which A I is
invertible with densely defined bounded inverse. The
spectrum Sp(A) of A is the complement in C of the
resolvent; it consists of a union of three disjoint sets:
the set of all complex numbers for which A I is
not injective, called the point spectrum such a is
an eigenvalue of A with associated eigenfunction
any u 2 D(A) such that Au = u; the set of points
for which A I has a densely defined unbounded
inverse R(A, ) called the continuous spectrum; and
the set of points for which A I has a welldefined unbounded but not densely defined inverse
R(A, ) called the residual spectrum.
A bounded operator has bounded spectrum and a
self-adjoint operator A acting on a Hilbert space has
real spectrum and no residual spectrum since the
range of A I is dense. As a consequence of the
p
while the map 7! gives rise to a positive
p
p2
operator A such that A = A.
The resolvent can also be used to define new
operators
Z
1
f A
f RA; d
2i C
from a linear operator via a Cauchy-type integral
along a countour C around the spectrum; this way
one defines complex powers Az of (essentially selfadjoint) positive elliptic pseudodiffferential operators which enter the definition of the zeta-function,
z 7! (A, z), of the operator A. The -function is a
useful tool to extend the ordinary determinant to
-determinants of self-adjoint elliptic operators,
thereby providing an ansatz to give a meaning to
partition functions in the path integral approach to
quantum field theory.
Operator Algebras
Bounded linear operators on a Hilbert space H
form an algebra L(H) closed for the operator norm
95
96
Further Reading
Adams R (1975) Sobolev Spaces. London: Academic Press.
Dunford N and Schwartz J (1971) Linear Operators. Part I.
General Theory. Part II. Spectral Theory. Part III. Spectral
Operators. New York: Wiley.
Hille E (1972) Methods in Classical and Functional Analysis.
London: Academic Press and Addison-Wesley.
Kato T (1982) A Short Introduction to Perturbation Theory for
Linear Operators. New YorkBerlin: Springer.
Reed M and Simon B (1980) Methods of Modern Mathematical
Physics vols. IIV, 2nd edn. New York: Academic Press.
Riesz F and SZ-Nagy B (1968) Lecons danalyse fonctionnelle.
Paris: GauthierVillars: Budapest Akademiai Kiado.
Rudin W (1994) Functional Analysis, 2nd edn. New York:
International Series in Pure and Applied Mathematics.
Yosida K (1980) Functional Analysis, 6th edn. Die Grundlehren
der Mathematischen Wissenschaften in Einzeldarstellungen
Band vol. 132. BerlinNew York: Springer.
Introduction
Minkowski spacetime is generally regarded as the
appropriate mathematical context within which to
formulate those laws of physics that do not refer
specifically to gravitational phenomena. Here we
shall describe this context in rigorous terms,
postulate what experience has shown to be its
correct physical interpretation, and illustrate by
means of examples its appropriateness for the
formulation of physical laws.
Minkowski Spacetime
and the Lorentz Group
Minkowski spacetime M is a four-dimensional real
vector space on which is defined a bilinear form
g : M M ! R that is symmetric (g(v, w) = g(w, v)
for all v, w 2 M) and nondegenerate (g(v, w) = 0
1. (orthogonality) T = ,
where T means transpose
0
1
B0
ab B
@0
0
Timelike
CN
Null
Spacelike
97
and
0
1
0
0
1
0 0
0 0C
C
1 0A
0 1
1
then
the
matrix
= (a b )
(a = row index,
b = column index) can be shown to satisfy the
following three conditions (Naber 1992, section 1.3):
Physical Interpretation
For the purpose of describing how one is to think of
Minkowski spacetime and the Lorentz group physically it will be convenient to distinguish (intuitively
and terminologically, if not mathematically) between a
vector in M and a point in M (the tip of a
vector). The points in M are called events and are to be
thought of as actual physical occurrences, albeit
idealized as point events which have no spatial
extension and no duration. One might picture, for
example, an instantaneous collision, or explosion, or
an instant in the history of some point material
particle or photon (particle of light).
Events are observed and identified by the assignment of coordinates. We will be interested in
coordinates assigned in a very particular way by a
98
a 1; 2; 3; 4
4
^3 x3
x
^4 sinh x1 cosh x4
x
6
t v=c2 x1
^t p
1 v2 =c2
x 2
x2
( > 0)
2
5
x 1, x 1
x3
x 3
99
p
jg(x x0 , x x0 )j, usually written (x x0 ), or
simply , is the proper time separation of x0 and x.
One can think of (x x0 ) as a sort of length for
x x0 measured, however, by a clock carried along by
a free material particle that experiences both x0 and x.
It is an odd sort of length, however, since it satisfies
not the usual triangle inequality, but the following
reversed version.
Reversed triangle inequality (Naber 1992, theorem
1.4.2) Let x0 , x and y be events in M for which y x
and x x0 are timelike with the same time orientation.
Then y x0 = (y x) (x x0 ) is timelike and
y x0 y x x x0
7
neous (^
x4 = 0), the quantity g(x x0 , x x0 ) is
the distance between them and for this reason this
quantity is called the proper spatial separation of x0
and x (whenever x x0 is spacelike).
Kinematic Effects
All of the well-known kinematic effects of special
relativity (the addition of velocities formula, the
relativity of simultaneity, time dilation, and length
contraction) follow easily from what we have done.
Because it eases visualization and because, as we
mentioned earlier, it suffices to do so, we will limit our
discussion to the special Lorentz transformations.
Let 1 and 2 be two real numbers and consider
the corresponding elements L(1 ) and L(2 ) of
L defined by [3]. Sum formulas for sinh and
cosh imply that L(1 )L(2 ) = L(1 2 ). Defining
i = tanh i , i = 1, 2, and = tanh (1 2 ), the sum
formula for tanh then gives
1 2
1 1 2
8
p
1
1
2
(= 1 )x = ^
x , which will not be
zero unless = 0 and so there is no relative motion
(^
x1 cannot be zero since then ^
xa = 0 for
a = 1, 2, 3, 4 and x = x0 ). This phenomenon is
called the relativity of simultaneity and we now
construct a simple geometrical representation of it.
Select two perpendicular lines in the plane to
represent the x1 - and x4 -axes (the Euclidean orthogonality of the lines has no physical significance and
x4
x 4
(x 1)2 (x 4)2 = 1
Hatted line of simultaneity
x 1
(x 1, x 4) = (1, 0)
(x 1, x 4) = (1, 0)
Figure 3 Relativity of simultaneity.
x1
x4
x 1 = 1
x 4
101
(x 1)2 (x 4)2 = 1
1
^
x4 p x4 > x4
1 2
This effect is entirely symmetrical since, if ^
x1 = 0,
then [5] implies
1
x4
x4 p ^
x4 > ^
1 2
(x 1, x 4) = (1, 0)
x4
x 4
x 4 = 1
(x 1, x 4) = (0, 1)
(x 1, x 4) = (0, 1)
x 1
x1
x 1
(x 1, x 4) = (1, 0)
x1
complex and subtle and some of which are commonplace (a passenger in a smooth, quiet airplane
traveling at constant groundspeed cannot feel
his motion relative to the earth). It is a powerful
guide for constructing the laws of relativistic
physics, but even more fundamentally it prohibits
us from regarding any particular admissible observer
as having a privileged view of the universe. In
particular, we are forbidden from attaching any
objective significance to such questions as, were the
two supernovae simultaneous?, How long did the
meson survive?, and What is the distance between
the Crab Nebula and Alpha Centauri? This is
severe, but one must deal with it.
dxa
ea
d
dxa dxb
d d
1
0
s
dxa dxb
d
ab
d d
9
0
dxa
ea
d
10
where
s
1 2 2 2 3 2
dx
dx
dx
kVk
4
4
dx
dx
dx4
is the usual magnitude of the particles velocity
vector
V V x4
dx1
dx2
dx3
e
e3
1
2
dx4
dx4
dx4
V i ei
12
13
m
P q V
1 kVk2
which, for kVk 1, is approximately mV. Identifying m with the inertial mass of Newtonian
mechanics (measured by an observer for whom the
particles speed is small), this is simply the classical
momentum of the particle.
Somewhat
more expli
q
2
citly, if one expands 1= 1 kVk by the Binomial
Theorem one finds that
m
Pi q V i
1 kVk2
i 1; 2; 3
14
which gives the components of the classical momentum plus relativistic corrections. In order
to preserve a formal similarity
with Newtonian
q
mechanics one often sees m=
P4 P e4
m
1
q m mkVk2
2
2
1 kVk
1 kVk2 referred
15
The appearance of the term (1=2)mkVk2 corresponding to the Newtonian kinetic energy suggests
that P4 be denoted E and called the total relativistic
energy measured by the given admissible observer
for the particle:
E P e4
16
1
mV i mV i kVk2 ;
2
103
17
18
19
h
d N e1 2 N e2 2
i1=2 h
N e 3 2
N e1 e1
i
N e2 e2 N e3 e3
20
where
21
Then, by [19],
N E d e4
22
1 2
1
1 cos 2 1 cos
2
23
23
24
105
q d
25
1 0
2q d
Since any future-directed timelike unit vector u is
the 4-velocity of some charged particle, we find
~ u = 0 for any such vector. Linearity then
that F(u)
~ v = 0 for any timelike vector. Now,
implies F(v)
if u and v are timelike and future directed, then u v
~ v) (u v) = F(u)
~ v
is timelike so 0 = F(u
~
~
~
u F(v) and therefore F(u) v = u F(v). But M
has a basis of future-directed timelike vectors so
~
~
Fx
y x Fy
26
27
28
and
d FJ
29
: p M ! 4p M;
33
Thus,
F E1 dx1 ^ dx4 E2 dx2 ^ dx4
E3 dx3 ^ dx4 B3 dx1 ^ dx2
B2 dx3 ^ dx1 B1 dx2 ^ dx3
(a more thorough discussion is available in ChoquetBruhat et al. (1977, chapter V A3)). In particular,
[29] is equivalent to
31
35
and
curl E
@B
0
@t
37
while d F = J is equivalent to
div E
d F J
d F 0
div B 0
1p1
32
dF 0
and
p 0; 1; 2; 3; 4
38
and
curl B
@E
J
@t
39
; 1; 2; 3; 4
40
Now, suppose that we wish to describe the electromagnetic field of a uniformly moving charge.
According to the relativity principle, it does not
matter at all whether we view the charge as moving
p
E
2
1 ^r3
3
^
q
x
E3 p 3
2
1 ^r
42
B1 0
3
^
q
x
B p 3
2
1 ^r
2
^
q
x
3
B p 3
2
1 ^r
107
and
!
q
1
B p 3 0e1 x3 e2 x2 e3
1 2 r
!
q
1
p 3 e1 r
1 2 r
44
dU1 bq 2
U ;
m
d
dU3
0;
d
^2 x2 ;
x
^3 x3
x
and so
^r2
1
x1 2 x2 2 x3 2
1 2
dU2
bq
U1
m
d
dU4
0
d
c
e4 45
m2
where x0 = xa0 ea 2 M is constant and a, , and c are
real constants with a > 0 (we have used U U = 1
to eliminate one other arbitrary real constant). Note
that, at each point on , (x1 x10 )2 (x2 x20 )2 = a2 .
Thus, if c 6 0 the spatial trajectory in this coordinate system is a helix along the e3 -direction
(i.e., along the magnetic field lines). If c = 0, the
trajectory is a circle in the x1 x2 plane. This case
is of some practical significance since one can
m2
2
a b2 q2
1
m ajbj
1 kVk2
Since a, b, and kVk are measurable, one obtains the
desired charge-to-mass ratio.
To conclude we wish to briefly consider the
existence and use of potentials for electromagnetic
fields. Suppose F is an electromagnetic field defined
on some connected, open region X in M. Then F is
a 2-form on X which, by [28], is closed. Suppose
also that the second de Rham cohomology H 2 (X ; R)
of X is trivial (since M is topologically R4 this will
be the case, for example, when X is all of M, or an
open ball in M, or, more generally, an open starshaped region in M). Then, by definition, every
closed 2-form on X is exact so, in particular, there
exists a 1-form A on X satisfying
F dA
46
and
d Ag
47
48
49
50
@2
@x1
@2
@x2
@2
@x3
2
@2
@x4 2
a 1; 2; 3; 4
51
Further Reading
Choquet-Bruhat Y, De Witt-Morette C, and Dillard-Bleick M
(1977) Analysis, Manifolds and Physics. Amsterdam: NorthHolland.
Einstein A et al. (1958) The Principle of Relativity. New York:
Dover.
109
Historical Background
In this section we shall briefly recall the basic
empirical facts and the first theoretical attempts
from which the theory and the formalism of presentday quantum mechanics (QM) has grown. In the
next sections we shall give the mathematical and
computational structure of QM, mention the physical problems that QM has solved with much
success, and describe the serious conceptual consistency problems which are posed by QM (and which
remain unsolved up to now).
Empirical rules of discretization were observed
already, starting from the 1850s, in the absorption
and in the emission of light. Fraunhofer noticed
that the dark lines in the absorption spectrum of
the light of the sun coincide with the bright lines in
the emission lines of all elements. G Kirchhoff and
R Bunsen reached the conclusion that the relative
intensities of the emission and absorption of light
implied that the ratio between energy emitted and
absorbed is independent of the atom considered.
This was the starting point of the analysis by
Planck.
On the other hand, by the end of the eighteenth
century, the spatial structure of the atom had been
investigated; the most successful model was that of
Rutherford, in which the atom appeared as a small
nucleus of charge Z surrounded by Z electrons
attracted by the nucleus according to Coulombs
law. This model represents, for distances of the
order of the size of an atom, a complete departure
from Newtons laws combined with the laws of
classical electrodynamics; indeed, according to these
laws, the atom would be unstable against collapse,
and would certainly not exhibit a discrete energy
spectrum. We must conclude that the classical laws
a 30 W
Notice that, for wavelength of 8103 A,
20
1
lamp emits roughly 10 photons s ; for macroscopic objects the discrete nature of light has no
appreciable consequence.
Plancks postulate and energy conservation imply
that in emitting and absorbing light the atoms of the
various elements can lose or gain energy only by
discrete amounts. Therefore, atoms as producers or
absorbers of radiation are better described by a
theory that assigns to each atom a (possible infinite)
discrete set of states which have a definite energy.
The old quantum theory of matter addresses
precisely this question. Its main proponent is
N Bohr (Bohr 1913, 1918). The new theory is
entirely phenomenological (as is Plancks theory)
and based on Rutherfords model and on three
more postulates (Born 1924):
(i) The states of the atom are stable periodic
orbits, as given by Newtons laws, of energy
En , n 2 Z , given by En = hn f (n), where h is
Planks constant, n is the frequency of the
electron on that orbit, and f(n) is for each atom
a function approximately linear in Z at least for
small values of Z.
(ii) When radiation is emitted or absorbed, the
atom makes a transition to a different state.
The frequency of the radiation emitted or
absorbed when making a transition is
n, m = h1 jEn Em j.
(iii) For large values of n and m and small values of
(n m)=(n m) the prediction of the theory
should agree with those of the classical theory
of the interaction of matter with radiation.
Later, A Sommerfeld gave a different version of the
first postulate, by requiring that the allowed orbits
be those for which the classical action is an integer
multiple of Plancks constant.
The old quantum theory met success when
applied to simple systems (atoms with Z < 5) but
it soon appeared evident that a new, radically
different point of view was needed and a fresh
start; the new theory was to contain few free
parameters, and the role of postulate (iii) was now
to fix the value of these parameters.
There were two (successful) attempts to construct
a consistent theory; both required a more sharply
defined mathematical formalism. The first one was
sparked by W Heisenberg, and further important
ideas and mathematical support came from M Born,
P Jordan, W Pauli, P Dirac and, on the mathematical side, also by J von Neumann and A Weyl. This
formulation maintains that one should only consider
relations between observable quantities, described
by elements that depend only on the initial and final
states of the system; each state has an internal
energy. By energy conservation, the difference
between the energies must be proportional (with a
universal constant) to the frequency of the radiation
absorbed or emitted. This is enough to define the
energy of the state of a single atom modulo an
additive constant. The theory must also take into
account the probability of transitions under the
influence of an external electromagnetic field.
We shall give some details later on, which will
help to follow the basis of this approach.
The other attempt was originated by L de Broglie
following early remarks by HW Bragg and
M Brillouin. Instead of emphasizing the discrete
nature of light, he stressed the possible wave nature
of particles, using as a guide the HamiltonJacobi
formulation of classical mechanics. This attempt
was soon supported by the experiments of Davisson
and Germer (1927) of scattering of a beam of ions
from a crystal. These experiments showed that,
while electrons are recorded as point particles,
their distribution follows the law of the intensity for
the diffraction of a (dispersive) wave. Moreover, the
relation between momentum and frequency was,
within experimental errors, the same as that
obtained by Einstein for photons.
The theory started by de Broglie was soon placed
in almost definitive form by E Schrodinger. In this
approach one is naturally led to formulate and solve
partial differential equations and the full development of the theory requires regularity results from
the theory of functions.
Schrodinger soon realized that the relations which
were found in the approach of Heisenberg could be
easily (modulo technical details which we shall
discuss later) obtained within the formalism he was
advocating and indeed he gave a proof that the two
formalisms were equivalent. This proof was later
refined, from the mathematical point of view, by
J von Neumann and G Mackey.
In fact, Schrodingers approach has proved much
more useful in the solution of most physical
problems in the nonrelativistic domain, because it
can rely on the developments and practical use of
the theory of functions and of partial differential
equations. Heisenbergs algebraic approach has
therefore a lesser role in solving concrete problems
in (nonrelativistic) QM.
If one considers processes in which the number of
particles may change in time, one is forced to
h
h
p2
; E
2
2m
111
^(x, !) is an
neighborhood of !0 ), one finds that u
approximate solution of the equation
^
ux; !
!20 2
n x; !^
ux; !
c2
1
@ x; t
h2
x x; t Vx x; t
@t
2m
2
h
x; tr x; t x; tr x; t 3
2m
4
These mathematical properties led to the statistical interpretation given by Max Born: in those
experiments in which the position of the particles is
measured, the integral of j (x, t)j2 over a region of
space gives the probability that at time t the particle
is localized in the region . Moreover, the current
associated with a charged particle is given locally by
j(x, t) defined above.
Let us now briefly review Heisenbergs approach.
At the heart of this approach are: empirical formulas
for the intensities of emission and absorption of
radiation (dispersion relations), Sommerfelds quantum condition for the action and the vague
statement the analogue of the derivative for the
discrete action variable is the corresponding finite
difference quotient. And, most important, the
remark that the correct description of atomic
physics was through quantities associated with
pairs of states, that is, (infinite) matrices and the
X
m>0
2m rJ
jAJj2 m
m 2
6
e2 X
fm;n
fn;m
2
2
2
4m n;m n;m
n;m 2
Em > En
7
113
Formal Structure
In this section we describe the formal mathematical
structure that is commonly associated with QM. It
constitutes a coherent mathematical theory, but the
interpretation axiom it contains leads to conceptual
difficulties.
We state the axioms in the form in which they
were codified by J von Neumann (1966); they
constitute a mathematically precise rendering of the
formalism of Born, Heisenberg, and Jordan. The
formalism of Schrodinger per se does not require
general statements about the category of
observables.
Axiom I
(i) Observables are represented by self-adjoint operators in a complex separable Hilbert space H.
(ii) Every such operator represents an observable.
Remark Axiom I (ii) is introduced only for mathematical simplicity. There is no physical justification
for part (ii). In principle, an observable must be
connected to a procedure of measurement (observation) and for most of the self-adjoint operators on H
(e.g., in the Schrodinger representation for
ixk (@=@xh )xk ) such procedure has not yet been given).
Axiom II
(i) Pure states of the systems are represented by
normalized vectors in H.
(ii) If a measurement of the observable A is made on
a system in the state represented by the element
2 H, the average of the numerical values one
obtains is < , A >, a real number because A is
self-adjoint (we have denoted by < , > the
scalar product in H).
Remark Notice that Axiom II makes no statement
about the outcome of a single measurement.
Using the natural complex structure of B(H), pure
states can be extended as linear real functionals on
B(H).
b, 2 R then immediately after the measurement the system can be in any (not necessarily
pure) state which lies in the convex hull of the
pure states which are in the spectral subspace of
the operator A in the interval b;
(b , b ).
Note Statements (ii) and (iii) can be extended
without modification to the case in which the initial
state is not a pure state, and is represented by a
density matrix
.
Remark 1 Axiom III makes sure that if one
performs, immediately after the first, a further
measurement of the same observable A the outcome
will still lie in the interval b; . This is needed to
give some objectivity to the statement made about
the outcome; notice that one must place the
condition immediately after because the evolution
may not leave invariant the spectral subspaces of A.
If the operator A has, in the interval b; , only
discrete (pure point) spectrum, one can express
Axiom III in the following way: the outcome can
be any state that can be represented by a convex
affine superposition of the eigenstates of A with
eigenvalues contained in b; .
In the very special case when A has only one
eigenvalue in b; and this eigenvalue is not
degenerate, one can state Axiom III in the following
form (commonly referred to as reduction of the
wave packet): the system after the measurement is
pure and is represented by an eigenstate of the
operator A.
Remark 2 Notice that the third axiom makes a
statement about the state of the system after the
measurement is completed.
It follows from Axiom III that one can measure
simultaneously only observables which are represented by self-adjoint operators that commute with
each other (i.e., their spectral projections mutually
commute). It follows from the spectral representation of the self-adjoint operators that a family {Ak }
of commuting operators can be considered (i.e.,
there is a representation in which they are) functions
over a common measure space.
Axioms IIII give a mathematically consistent
formulation of QM and allow a statistical description (and statistical prediction) of the outcome of
the measurement of any observable. It is worth
remarking that while the predictions will have only
a statistical nature, the dynamical evolution of the
observables (and by duality of the states) will be
described by deterministic laws. The intrinsically
statistical aspect of the predictions comes only from
the third postulate, which connects the mathematical content of the theory with the measurement
process.
The third axiom, while crucial for the connection
of the mathematical formalism with the experimental data, contains the seed of the conceptual
difficulties which plague QM and have not been
cured so far.
Indeed, the third axiom indicates that the process
of measurement is described by laws that are
intrinsically different from the laws that rule the
evolution without measurement. This privileged role
of the changing by effect of a measurement leads to
serious conceptual difficulties since the changing is
independent of whether or not the result is recorded
by some observer; one should therefore have a way
to distinguish between measurements and generic
interactions with the environment.
A related problem that is originated by Axiom III
is that the formulation of this axiom refers implicitly
to the presence of a classical observer that certifies
the outcomes of measurements and is allowed to
make use of classical probability theory. This
observer is not subjected therefore to the laws
of QM.
These two aspects of the conceptual difficulties
have their common origin in the separation of the
measuring device and of the measured systems into
disjoint entities satisfying different laws. The difficulties in the theory of measurement have not yet
received a satisfactory answer, but various attempts
have been made, with various degree of success, and
some of them are described briefly in the section
Interpretation problems. It appears therefore that
QM in its present formulation is a refined and
successful instrument for the description of the
nonrelativistic phenomena at the Planck scale, but
its internal consistency is still standing on shaky
ground.
Returning to the axioms, it is worth remarking
explicitly that according to Axiom II a state is a
linear functional over the observables, but it is
represented by a sesquilinear function on the
complex Hilbert space H. Since Axiom II states
that any normalized element of H represents a state
(and elements that differ only by a phase represent
the same state) together with , also
a
b , jaj2 jbj2 = 1 represent a state superposition of
and (superposition principle).
But for an observable A, one has in general
(A) 6 jaj2 (A) jbj2 (A), due to the cross-terms
in the scalar product. The superposition principle is
one of the characteristic features of QM. The
superposition of the two pure states and
has
properties completely different from those of a
115
n
The map 2 :
!
1 is called reduction or also
conditioning) with respect to H2 ; it is also called
partial trace with respect to H2 . The first notation
reflects the analogy with conditioning in classical
probability theory.
The map 2 can be extended by linearity to a map
from normal states (density matrices) on B(H1 H2 )
to normal states on B(H1 ) and gives rise to a
positivity-preserving and trace-preserving map.
One can in fact prove (Takesaki 1971) that any
conditioning for normal states of a von Neumann
algebra M is completely positive in the sense that it
remains positive after tensorization of M with B(K),
where K is an arbitrary Hilbert space.
It can also be proved that a partial converse is
true, that is, that every completely positive tracepreserving map on normal states of a von
Neumann algebra A B(H) can be written, for a
suitable choice of a larger Hilbert space K and
partial P
isometries Vk , in the form (Kraus form)
(a) = k Vk aVk .
@
H
@t
N
X
H
mk ihrk Ak 2
k1
N
X
k1
Vk xk
N
X
Vi;k xi xk
9
i6k;1
If some particles have of spin 1/2, the corresponding kinetic energy term should read (i
h
r)2 ,
where
k , k = 1, 2, 3, are the Pauli matrices and one
must add a term W(x) which is a matrix field with
values in Ck Ck and takes into account the
coupling between the spin degrees of freedom.
Notice that the local operator i
r is a square
root of the Laplacian.
A relativistic extension of the Schrodinger equation for a free particle of mass m
0 in dimension
3 was obtained by Dirac in a space of spinorvalued functions k (x, t), k = 0, 1, 2, 3, which carries
an irreducible representation of the Lorentz group.
In analogy with the electromagnetic field, for which
a linear partial differential equation (PDE) can be
written using a four-dimensional representation of
the Lorentz group, the relativistic Dirac equation is
the linear PDE
i
3
X
k0
@
@xk
m ;
x0 ct
a representation
where the k generate the algebra ofP
of the Lorentz group. The operator
(@=@xk )k is a
local square root of the relativistically invariant
dAlembert operator @ 2 =@x20 m I.
When one tries to introduce (relativistically
invariant) local interactions, one faces the same
problem as in the classical mechanics, namely one
must introduce relativistically covariant fields (e.g.,
the electromagnetic field), that is, systems with an
infinite number of degrees of freedom. If this field is
considered as external, one faces technical problems,
which can be overcome in favorable cases. But if one
tries to obtain a fully quantized theory (by also
quantizing the field) the obstacles become unsurmountable, due also to the nonuniqueness of the
representation of the canonical commutation relations if these are taken as the basis of quantization,
as in the finite-dimensional case.
In a favorable case (e.g., the interaction of a
quantum particle with the quantized electromagnetic
field) one can set up a perturbation scheme in a
parameter (the physical value of in natural units
is roughly 1/137). We shall come back later to
perturbation schemes in the context of the Schrodinger operator; in the present case one has been
able to find procedures (renormalization) by which
the series in that describe relevant physical
quantities are well defined term by term. But even
in this favorable case, where the sum of the first few
terms of the series is in excellent agreement with the
experimental data, one has reasons to believe that
the series is not convergent, and one does not even
know whether the series is asymptotic.
117
dUt
HUt
dt
10
1
Vx;
2m
X @x2
k
k
@x2k
11
a (a) (a )a for all a 2 M. The positive dissipation is called completely positive if it remains
positive after tensorization with B(K) for any
12
119
Ua eia^q ;
and therefore, setting
eiab=2 V(b)U(a) one has
Vb eibp
z = a ib
and
W(z)
WzWz0 ei!z;z =2 Wz z0
z 2 C;
!z; z0 Imz; z0
13
N
X
1
Nk ;
Nk
@2
x2k 1
@xk2
ah ; ak 0
Nh ; ak ak h;k
15
cfnk g fnk g ;
jcfnk g j2 < 1
fnk g
fnk g
121
16
class an operator
(D, x) (the Weyl symbol of the
function
) defined by
Z
D; xf ; g
; xWf ; g
; x d
dx
Z
p
p
dp
Wf ; g
; x ei
;p f x ; x
2
2
1.
2.
3.
4.
Note that (1) and (4) imply (2) and (3) through a
limit procedure.
We described the Weyl quantization as a correspondence between functions in the Schwartz class S
and a class of bounded operators. Weyls quantization can be extended to a much wider class of
functions. Operators that can be so constructed are
called Fourier integral operators. One uses the
notation
(D, x).
We have the following useful theorems (Robert
1987):
Theorem 4 Let l1 , . . . , lK be linear functions on RN
such that {li lk } = 0. Let P be a polynomial and let
(CalderonVaillancourt). If
0
jD
jjjj2N1
Dx
j < 1 the norm of the operator
(D, x) is bounded by
0 .
n=2
Z
1=2
jAzj dz
123
N
@ X
ihrk Ak 2 V H
@t
k1
17
2 L2 R3N
where Ak are vector-valued functions (vector potentials) and V = Vk (xk ) Vi, k (xi xk ) are scalarvalued function (scalar potentials) on R3 .
Typical problems in Schrodingers quantum
mechanics are:
1. Self-adjointness of H, existence of bound states
(discrete spectrum of the operator), their number
and distribution, and, in general, the properties
of the spectrum.
2. Existence, completeness, and continuity properties of the wave operators
W
s lim eitH0 eitH
1
18
@x; t
hx; t h1 Vxx; t
@t
19
In terms of the Wigner function Wh, the Schrodinger equation [19] takes the form
@f h
rx f h Kh f h 0
@t
h t 0 0
h
20
where
Kh
i
2N
hy
hy
h V x
V x
2
2
i ; y 1
125
21
1
X
k Ek ;
1
X
k k
22
N
X
0
k Ek RN ;
lim jRN j 1
N!1
Interpretation Problems
In this section we describe some of the conceptual
problems that plague present-day QM and some of
the attempts that have been made to cure these
problems, either within its formalism or with an
altogether different approach.
Approaches within the QM Formalism
n;m
127
D Bohm (1952) and put recently on firm mathematical grounds by Duerr et al. (1999). The approach is
radically different from the traditional one and it is
not clear at present whether it can give a solution to
the measurement problem and a description of all
the phenomena which traditional QM accounts for.
But it is very interesting from the point of view of
the mathematics involved.
We have remarked that the formulation of QM
that is summarized in the three axioms given earlier
has many unsatisfactory aspects, mainly connected
with the superposition principle (described in its
extremal form by the Schrodingers cat paradox)
and with the problem of measurement which
reveals, for example, through the EinsteinRosen
Podolski paradox, an intrinsic nonlocality if one
maintains that their objective properties can be
attributed to systems which are far apart. From the
very beginning of QM, attempts have been made to
attribute these features to the presence of hidden
variables; the statistical nature of the predictions
of QM is, from this point of view, due to the
incompleteness of the parameters used to describe
the systems. The impossibility of matching the
statistical prediction of QM (confirmed by experimental findings) with a local theory based on hidden
variables and classical probability theory has been
known for sometime (Kochen and Specker 1967),
also through the use of Bell inequalities (Bell
1964) among correlations of outcomes of separate
measurements performed on entangled system
(mainly two photons or two spin-1/2 particles
created in a suitable entangled state).
A proof of the intrinsic nonlocality of QM (in the
above sense) was given by L Hardy (see Haroche
(2003)).
While experimental results prove that one
cannot substitute QM with a naive theory of
hidden variables, more refined attempts may have
success. We shall only discuss the approach of Bohm
(following a previous attempt by de Broglie) as
presented in Duerr et al. (1999). It is a dynamical
theory in which representative points follow classical
paths and their motion is governed by a timedependent vector velocity field (in this sense, it is
not Newtonian). In a sense, Bohmian mechanics is a
minimal completion of QM if one wants to keep the
position as primitive observable. To these primitive
objects, Bohms theory adds a complex-valued function (the guiding wave in Bohms terminology)
defined on the configuration space Q of the particles.
In the case of particles with spin, the function is
spinor-valued. Dynamics is given by two equations:
one for the coordinates of the particles and one for
the guiding wave. If x x1 , . . . , xN describes the
129
rk
x;
x_ k h=mk Im
xk 2 R3
Further Reading
Alicki R (2004) Pure decoherence in quantum systems. Open Syst.
Inf. Dyn. 11: 5361.
Araki H (1988) In: Jorgensen P and Muhly P (eds.) Operator
Algebras and Mathematical Physics, Contemporary Mathematics 62. Providence, RI: American Mathematical Society.
Araki H and Ezawa H (eds.) (2004) Topics in the theory of
Schroedinger Operators. River Edge, NJ: World Scientific.
Bach V, Froelich J, and Sigal IM (1998) Quantum electrodynamics of constrained non relativistic particles. Advanced
Mathematics 137: 299395.
Bargmann V (1967) On a Hilbert space of analytic functions and
an associated integral transform. Communications of Pure and
Applied Mathematics 20: 1101.
Bell J (1966) On the problem of hidden variables in quantum
mechanics. Reviews of Modem Physics 38: 42474280.
Blanchard P and DellAntonio GF (eds.) (2004) Multiscale
methods in quantum mechanics, theory and experiments.
Trends in Mathematics. Boston: Birkhauser.
Blanchard P and Olkiewiz R (2003) Decoherence in the
Heisenberg representation. International Journal of Physics B
18: 501507.
Bohm D (1952) A suggested interpretation of quantum theory in
terms of hidden variables I, II. Physical Reviews 85: 161179,
180193.
Bohr N (1913) On the constitution of atoms and molecules.
Philosophical Magazine 26: 125, 476502, 857875.
Bohr N (1918) On the quantum theory of line spectra. Kongelige
Danske Videnskabernes Selskabs Skrifter Series 8, IV, 1, 1118.
Born M (1924) Uber quantenmechanik. Zeitschrift fur Physik 32:
379395.
Born M and Jordan P (1925) Zur quantenmechanik. Zeitschrift
fur Physik 34: 858888.
Born M, Jordan P, and Heisenberg W (1926) Zur quantenmechanik II. Zeitschrift fur Physik 35: 587615.
Cycon HL, Frese RG, Kirsh W, and Simon B (1987) Schroedinger
operators with application to quantum mechanics and geometry. Texts and Monogrphs in Physics. Berlin: Springer Verlag.
de Broglie L (1923) Ondes et quanta. Comptes Rendue 177:
507510.
DellAntonio GF (2004) On decoherence. Journal of Mathematical Physics 44: 49394955.
Dirac PAM (1925) The fundamental equations of quantum
mechanics. Proceedings of the Royal Society of London A
109: 642653.
Dirac PAM (1926) The quantum algebra. Proceedings of the
Cambridge Philosophical Society 23: 412428.
Dirac PAM (1928) The quantum theory of the electron. Proc.
Royal Soc. London A 117: 610624, 118: 351361.
Duerr D, Golstein S, and Zangh` N (1996) Bohmian mechanics as
the foundation of quantum mechanics. Boston Studies Philosophical Society 184: 2144. Dordrecht: Kluwer Academic.
Einstein A (1905) The Collected Papers of Albert Einstein, vol. 2,
pp. 347377, 564585. Princeton, NJ: Princeton University
Press.
Einstein A (19241925) Quantentheorie des einatomigen idealen
gases. Berliner Berichte (1924) 261267, (1925) 314.
Gell-Mann M and Hartle JB (1997) Strong decoherence.
Quantum-Classical Correspondence, pp. 335. Cambridge,
MA: International Press.
Gross L (1972) Existence and uniqueness of physical ground state.
Journal of Functional Analysis 19: 52109.
Haroche S (2003) Quantum Information in cavity quantum
electrodynamics. Royal Society of London Philosophical
131
(i) ;, X 2 T .
(ii) Let I be an index set. then
A 2 T ; 2 I )
[
2I
Tn
A 2 T
Introduction
(iii) Ai 2 T , i = 1, . . . , n )
",1,5,1,0,0pc,0pc,0pc,0pc>Essential
Concepts
Definition 1 Let X be a set. A collection T of
subsets of X is called a topology if the following are
satisfied:
i=1
Ai 2 T .
A
x 62 A}.
boundary of A as the set An
Definition 10 Let A be a subset of a topological
space X. A point x 2 X is called a limit point of A if
every open set containing x contains some point of
A other than x.
Definition 11
= X.
X if A
133
(i) d(x, y) 0,
(ii) d(x, y) = 0 , x = y,
(iii) d(x, z) d(x, y) d(y, z) (triangle inequality).
Metric Spaces
A special class of topological spaces plays an
important role: metric spaces.
Remarks
(i) The function d is called the metric, or distance
function, between the two points.
(ii) This concept of metric is what is generally
known as Euclidean metric in mathematical
physics. The distinguishing feature is the positive definiteness (and the triangle inequality).
One can, and does, introduce indefinite metrics
(for example, the Minkowski metric) with
various signatures. But these metrics are not
usually used to induce topologies in the spaces
concerned.
Definition 26 Given a metric space X and a point
x 2 X, we define the open ball centred at x with
radius r (a positive real number) as
Br x fy 2 X : dx; y < rg
Given a metric space X, we can immediately
define a topology on it by taking all the open balls in
X as a basis. We say that this is the topology
induced by the given metric. Then we can recover
our usual definition of continuity.
Proposition 4 Let f : X ! Y be a map from the metric
space X to the metric space Y. Then f is continuous
(with respect to the corresponding induced topologies)
at x 2 X if and only if given any > 0, 9 > 0 such that
d(x, x0 ) < implies d(f (x, ), f (x0 )) < .
Note that we do not bother to give two different
symbols to the two metrics, as it is clear which
spaces are involved. The proof is easily seen by
taking the relevant balls as neighborhoods. Equally
easy is the following:
Proposition 5
Examples
Examples
k1
Remarks Completeness is not a topological invariant. For example, the open interval (1, 1) and the
whole real line are homeomorphic (with respect to
the usual topologies) but the former is not complete
while the latter is. The homeomorphism can
conveniently be given in terms of the trigonometric
function tangent.
dx; y kx yk
In many cases, for example, the metrics defined in
example (iii) above, one can define the norm of a
vector as just the distance of it from the origin. One
obvious exception is the discrete metric.
A slightly more general concept is found to be
useful for spaces of functions and operators: that of
seminorms. A seminorm is one which satisfies the
last two of the conditions, but not necessarily the
first, for a norm, as listed above.
Definition 29 Given a metric space X, a sequence
of points {x1 , x2 , . . . } is called a Cauchy sequence if,
given any > 0, there exists a positive integer N
such that for any k, > N we have d(xk , x ) < .
Definition 30 Given a sequence of points
{x1 , x2 , . . . } in a metric space X, a point x 2 X is
called a limit of the sequence if given any > 0,
there exists a positive integer N such that for any
n > N we have d(x, xn ) < . We say that the
sequence converges to x.
135
n!1
n!1
kTk
kTxk
x 2X; kxk 1
Homotopy Groups
The most elementary and obvious property of a
topological space X is the number of connected
components it has. The next such property, in a
certain sense, is the number of holes X has. There
Remarks
(i) When we consider the gauge group G in a Yang
Mills theory, its fundamental group classifies the
monopoles that can occur in the theory.
(ii) For n 1, n (X) is a group, the group action
coming from the joining of two loops together
to form a new loop. On the other hand, 0 (X)
in general is not a group. However, when X is a
Lie group, then 0 (X) inherits a group structure
from X, because it can be identified with the
quotient group of X by its identity-connected
component. For example, the two components
of O(3) can be identified with the two elements
of the group Z2 , the component where the
determinant equals 1 corresponding to 0 in Z2
and the component where the determinant
equals 1 corresponding to 1 in Z2 .
(iii) For n 2, the group n (X) is always abelian.
(iv) Examples of nonabelian 1 are the fundamental
groups of some Riemann surfaces.
(v) Since 1 is not necessarily abelian, much of the
direct-sum notation we use for the homotopy
if n 3
if n 2
6 S3 Z12
(iii) From the theory of sphere bundles, we can
deduce:
i S2 i1 S1 i S3 if i 2
i S4 i1 S3 i S7 if i 2
i S8 i1 S7 i S15 if i 2
and the first of these relations give the following more succinct result:
i S3 i S2
if i 3
Remarks
(i) By considering closed loops in X and their
coverings in B it is easily seen that the
fundamental group 1 (X) acts on the coverings
of X. If we further assume that the action is
n2
Examples
(i) The real line R is a covering of the group U(1).
(ii) The group SU(2) is a double cover of the group
SO(3).
(iii) The group SL(2, C) is a double cover of the
Lorentz group SO(1, 3).
(iv) The group SU(2, 2) is a 4-fold cover of the
conformal group in four dimensions. This local
isomorphism is of great importance in twistor
theory.
137
i SO6 i SU4
Just for interest, and to show the richness of the
subject, some isomorphisms for homotopy groups
are shown in Table 1 and some homotopy groups
for low SU(n) and SO(n) are listed in Table 2.
Table 1 Some isomorphisms for homotopy groups
Isomorphism
Range
i (SO(n)) i (SO(m))
i (SU(n)) i (SU(m))
i (Sp(n)) i (Sp(m))
i (G2 ) i (SO(7))
i (F4 ) i (SO(9))
i (SO(9)) i (SO(7))
n, m i 2
n, m 12 (i 1)
n, m 14 (i 1)
2i 5
2i 6
i 13
SU(2)
SU(3)
SU(4)
SU(5)
SU(6)
SO(5)
SO(6)
SO(7)
SO(8)
SO(9)
SO(10)
4
5
6
7
8
9
10
Z2
0
0
0
0
Z2
0
0
0
0
0
Z2
Z
Z
Z
Z
Z2
Z
0
0
0
0
Z12
Z6
0
0
0
0
0
0
0
0
0
Z2
0
Z
Z
Z
Z
Z
Z
ZZ
Z
Z
Z2
Z12
Z24
0
0
0
Z24
Z2 Z2
Z2 Z2 Z2
Z2 Z2
Z2
Z3
Z3
Z2
Z
Z
0
Z2
Z2 Z2
Z2 Z2 Z2
Z2 Z2
Z Z2
Z15
Z30
Z120 Z2
Z120
Z3
Z120
Z120 Z2
Z24
Z24 Z24
Z24
Z12
Appendix: A Mathematicians
Basic Toolkit
The following is a drastically condensed list, most
of which is what a mathematics undergraduate
learns in the first few weeks. The rest is included
for easy reference. These notations and concepts
are used universally in mathematical writing. We
have not endeavored to arrange the material in a
logical order. Furthermore, given structures such as
sets, groups, etc., one can usually define substructures such as subsets, subgroups, etc., in a
straightforward manner. We shall therefore not
spell this out.
Sets
A [ B fx : x 2 A or x 2 Bg
A \ B fx : x 2 A and x 2 Bg
union
intersection
Maps
139
Fields
Rings
1.
2.
3.
4.
(x y) = x y (linearity),
( )x = x x,
()x = (x), and
1x = x.
2 F; a; b 2 A
Modules
Further Reading
1.
2.
3.
4.
(x y)r = xr yr,
x(r s) = xr xs,
x(rs) = (xr)s, and
x1 = x
A
Abelian and Nonabelian Gauge Theories Using Differential Forms
A C Hirshfeld, Universitat Dortmund, Dortmund,
Germany
2006 Elsevier Ltd. All rights reserved.
Introduction
Quantum electrodynamics is the theory of the
electromagnetic interactions of photons and electrons. When attempting to generalize this theory to
other interactions it turns out to be necessary to
identify its essential components. The essential
properties of electrodynamics are contained in its
formulation as an abelian gauge theory. The
generalization to include other interactions is then
reduced to incorporating the structure of nonabelian
groups. This becomes particularly clear when we
formulate the theory in the language of differential
forms.
Here we first present the formulation of electrodynamics using differential forms. The electromagnetic fields are introduced via the Lorentz force
equation. They are recognized as the components of
a differential 2-form. This form fulfills two differential conditions, which are equivalent to Maxwells
equations. These are expressed with the help of a
differential operator and its Hermitian conjugate,
the codifferential operator. We consider the effects
of charge conservation and introduce electromagnetic potentials, which are defined up to gauge
transformations. We finally consider Weyls argument for the existence of the electromagnetic
interaction as a consequence of the local phase
invariance of the electron wave function.
We then go on to present the nonabelian generalization. The gauge bosons appear in a theory with
fermions by requiring invariance of the theory with
respect to local gauge transformations. When the
fermions group into symmetry multiplets this gives
rise to a gauge group SU(N) involving N 2 1 gauge
bosons mediating the interaction, where N is the
dimension of the Lie algebra. The interaction arises
through the necessity of replacing the usual derivatives by covariant derivatives, which transform in a
natural way in order to preserve the gauge
Relativistic Kinematics
The trajectory of a mass point is described as x (),
where is the invariant proper time interval:
d 2 dt2 dx dx dt2 1 v2
1
dx dx
1
d 2
1
0
0C
C
0A
1
2
3
dm0 v
dt
5
dp0
f v 0
dt
6
or
dp0
dx
f v f
dt
dt
7
x ! x;
y ! y;
E ! E;
0
dp f dx dW
8
10
with
@
@
@
@
vx
vy
vz
@t
@x
@y
@z
Ey v By dy Ez v Bz dz
12
14
15
16
22
23
21
F E ^ dt B
19
p
np
n p!p!
f qiu F
18
9
f qE v; E v B
B!B
f qiu F
u
17
z ! z
24
25
and
26
143
d 2 / dd / d2 0
28
35
@Kj
@xj
36
g11 gpp sign dxp1 ^ ^ dxn
29
or
d dL ^ dxL
1
pnp
1
Ind g
Id
30
jK
d ^ 1
1
Indg
dI I
np1p1
1
d ^ 1
^ d
31
d ^
32
with
d 1np1 1Ind g d
40
41
where
We then have
d; ; d
@K
@xj
^ I IKL K L
Indg
1
39
Here we use
^ d
np1
@L L
@xj K
or
d^ d ^ 1p1 ^ d
p1
38
IKL
8
>
1
>
>
>
>
>
>
<
1
>
>
>
>
>
>
>
: 0
if (KL) is an even
permutation of I
if (KL) is an odd
permutation of I
42
otherwise
33
1
@xj
Z
I
jK @
1
K I
@xj
43
dF 0;
45
@
@t
46
dB 0
j
0 u
0 ;
0 v
; J
47
48
49
50
55
56
div B 0
54
where
is the charge density and J the current
density. This corresponds to a 1-form
j
dt Jx dx Jy dy Jz dz
In vector notation
@B
curl E
;
@t
@E
J
@t
Current Conservation
curl B
@B
dF dE
^ dt dB 0
@t
53
We then get
57
58
that is
dj d3 j2 ^ dt d3 dj2 ^ dt
@
div J dt ^ dx ^ dy ^ dz
@t
@
div J 0
@t
59
dV
J n dS
60
dt
dt V
@V
52
61
B dA
63
70
D d !
B curl A
64
65
66
2
4A d dA d F j
67
Gauge Invariance
In quantum mechanics, the electron is described by a
wave function which is determined up to a free
phase. Indeed, at every point in space this phase can
be chosen arbitrarily:
0
x ! x expfixg x
x ! 0 x x expfixg
68
145
72
74
1
SA
A J vol
4
M
76
We write now
Z
M
A J vol4 A; j
77
and
1
4
L0 i @ m ! L00
L0 i g@ g1
78
@ ! D @ !
80
A gauge transformation is
x g1 x x;
89
88
79
87
0 x xgx
90
91
92
81
Ja Aa
93
Aa iq!a
94
Ja Ta
95
with
with
gx exp fxg
82
83
For su(2) the basis elements are traceless and antiHermitian (see below), they are conventionally
expressed in terms of the Pauli matrices,
Ta
a
2i
84
and
with
1
3
0
1
1
0
;
0
0
1
1
2
i
85
86
96
147
Lie-Algebra-Valued p-Forms
97
98
; Ta ; Tb
^ b Tb ; Ta
^ b
; 1pq1 ;
Ta Tb a ^
Ta a ^ Tb
^
a
a b c
F
@ Aa @ Aa iqfbc
A A
0 g1 g
1
pq
102
103
^ Ta a
104
112
odd
^! !^
105
114
or
111
110
1pq Tb
109
0 dg1 !g dg1 ^ dg
and
; Ta ; Tb a ^
101
108
a
F i q 12 F
dx ^ dx Ta
100
or
99
a
where fbc
are the structure constants. It follows from
this that
107
106
d ! ^ ^ ! 0
115
116
trX; Y hX; Yi
21
! ^ 1
^ ! !;
117
or for X, Y, Z 2 G
1
^ g g dg 1 g ^ dg
adXY X; Y
g1 Dg dg1 ^ g 1 g ^ dg
119
y
T
eX eX eX
128
129
127
p 1
1
126
130
adXy adX
131
or
132
120
or
121
122
123
symmetric:
hX; Yi trX; Y trY; X hY; Xi
124
hX; Xi X X X X
jX j2
125
135
149
!; ! f! ^ ! ! ^ !g@ ; @
! ! ! ! ! ! ! !
!; ! 2! ; !
138
D d !;
139
and
S! D !; !; D 0
140
141
tr ! ! ; vol4
149
M
!; fd ! ; g
D J 0
143
S! D !; !; D
d ! !; !;
150
142
147
h! ; ! ; ivol4
148
146
and hence
136
d
1
S!tjt0 ;
dt
2
1
; ;
2
1
; d! !; ! ;
2
1
1
d! !; ! !; !;
2
2
S!
d ! !; !;
151
152
153
and
@ F @ F @ F iqfA ; F
144
A ; F A ; F g 0
154
F E;
F12 B3 ;
i 1; 2; 3
F31 B2 ;
F12 B3
155
156
157
158
159
@E
iqA0 E EA0
@t
iqA B B A
160
@B
iqfA E E A A0 B BA0 g
@t
161
div B iqA B B A
162
and
Further Remarks
The foundations of the mathematics of differential
forms were laid down by Poincare (1953). They
were applied to the description of electrodynamics
Further Reading
Abraham A, Marsden J, and Ratiu T (1983) Manifolds, Tensor
Analysis, and Applications. MA: Addison-Wesley.
Cartan E (1923) On manifolds with an Affine Connection and the
Theory of General Relativity. English translation of the French
original 1923/1924 (Bibliopolis, Napoli 1986).
Frankel T (1987) The Geometry of Physics, An Introduction.
Cambridge University Press.
Gell-Mann M (1972) Quarks: developments in the quark theory
of hadrons. Acta Physica Austriaca Suppl. IV: 733.
Glashow SL (1980) Towards a unified theory: threads in a
tapestry. Reviews of Modern Physics 52: 539.
Hehl FW and Obukhov YN (2003) Foundations of Classical
Electrodynamics. Boston: Birkhauser.
Higgs PW (1964) Broken symmetries and the masses of gauge
bosons. Physical Review Letters 13: 508.
tHooft G and Veltman M (1972) Regularization and renormalization of gauge fields. Nuclear Physics B 44: 189.
Poincare H (1953) Oeuvre. Paris: Gauthier-Villars.
Salam A (1980) Gauge unification of fundamental forces. Reviews
of Modern Physics 52: 525.
Weinberg SM (1980) Conceptual foundations of the unified
theory of weak and electromagnetic interactions. Reviews of
Modern Physics 52: 515.
Weyl H (1929) Elektron und gravitation. Zeitschrift fuer Physik
56: 330.
Yang CN and Mills RL (1954) Construction of isotopic spin and
isotopic gauge invariance. Physical Review 96: 191.
151
Introduction
For the purpose of this article, vortices are topological
solitons arising in field theories in (2 1)-dimensional
spacetime when a complex-valued field is allowed to
acquire winding at infinity, meaning that the phase of
(t, x), as x traverses a large circle in the spatial plane,
changes by 2n, where n is a nonzero integer. Such
winding cannot be removed by any continuous
deformation of (hence topological) and traps a
considerable amount of energy which tends to coalesce
into smooth, stable lumps with highly particle-like
characteristics (hence solitons). Clearly, the universe
is (3 1) dimensional. Nonetheless, planar field
theories are of physical interest for two main reasons.
First, the theory may arise by dimensional reduction of
a (3 1)-dimensional model under the assumption of
translation invariance in one direction. Vortices are
then transverse slices through straight tube-like objects
variously interpreted as magnetic flux tubes in a
superconductor or cosmic strings. Second, a crucial
ingredient of the standard model of particle physics is
spontaneous breaking of gauge symmetry by a Higgs
field. As well as endowing the fundamental gauge
bosons and chiral fermions with mass, this mechanism
can potentially generate various types of topological
solitons (monopoles, strings, and domain walls) whose
structure and interactions one would like to understand. Vortices in (2 1) dimensions are interesting in
this regard because they arise in the simplest field
theory exhibiting the Higgs mechanism, the abelian
Higgs model (AHM). They are thus a useful theoretical laboratory in which to test ideas which may
ultimately find application in more realistic theories.
This article describes the properties of abelian Higgs
vortices and explains how, using a mixture of
numerical and analytical techniques, a good understanding of their dynamical interactions has been
obtained.
A 7! A e1 @
1
21
:R
! R being any smooth function, F and
jj remain invariant, while D 7! ei D . Only
gauge-invariant quantities are physically observable
(classically).
With these conventions, the AHM has Lagrangian
density
L
1
F F D D 2 jj2 2
4
2
8
2
3
4
5
R!1 SR
R!1 0
dr2 r dr r2
2
2
d a 1 da
n a
2 0
dr2 r dr
7
= 1, a = n, yielding
p
qn
r 1
K0 r
8
2
ar n
mn
rK1 r
2
9
0.5
0.8
0.4
0.6
0.3
B
, a
153
0.4
0.2
n=1
0.2
n=5
0.1
4
r
(a)
4
r
(b)
0.6
n=1
Energy density
0.5
0.4
0.3
n=5
0.2
0.1
0
4
r
(c)
Figure 1 Static, radially symmetric n-vortices: (a) the 1-vortex profile functions
(r ) (solid curve) and a(r ) (dashed curve) for = 2, 1,
and 1/2, left to right; (b) the magnetic field B; and (c) the energy density of n-vortices, n = 1 to 5, left to right, for = 1.
Interaction Energy
1.4
=2
1.3
En /n
1.2
1.1
=1
1
0.9
=1
2
0.8
0.7
n
Figure 2 The energy per unit winding En =n of radially
symmetric n-vortices for = 1=2, 1, and 2.
1.75
2.42
Eint
Eint
2.38
1.7
2.34
1.65
2.3
10
10
s
(a)
(b)
Figure 3 The 2-vortex interaction energy Eint (s) as a function of vortex separation (solid curve), in comparison with its asymptotic
1
form Eint
(s) (dashed curve) for (a) = 1=2 and (b) = 2.
A0 ; A
p
q
K0 r
2
m
0; k rK0 r
2
11
12
@ @ 1A j
13
j0 ; j m0; k rx
14
Eint s E1
int s 2E1
p
q2
K0 s
2
m2
K0 s
2
15
p
q2
m2
K0 r
K0 r
2
2
16
Vortex Scattering
The AHM being Lorentz invariant, one can obtain
time-dependent solutions wherein a single n-vortex
travels at constant velocity, with speed 0 < v < 1
and Etot = (1 v2 )1=2 En , by Lorentz boosting the
static solutions described above. Of more dynamical
interest are solutions in which two or more vortices
undergo relative motion. The simplest problem is
vortex scattering. Two vortices, initially well separated, are propelled towards one another. In the
center-of-mass (COM) frame they have, as t ! 1,
equal speed v, and approach one another along
parallel lines distance b (the impact parameter)
apart, see Figure 4. If b = 0, they approach headon. Assuming they do not capture one another, they
interact and, as t ! 1, recede along parallel straight
lines having been deflected through an angle (the
scattering angle). If scattering is elastic, the exit lines
also lie b apart and each vortex travels at speed v as
t ! 1. The dependence of on v, b, and has
been studied through lattice simulations by several
authors, perhaps most comprehensively by Myers,
Rebbi, and Strilka (1992). We shall now describe
their results.
Note first that vortex scattering is actually
inelastic: vortices recede with speed < v because
some of their initial kinetic energy is dispersed by
the collision as small-amplitude traveling waves
(radiation). This energy loss can be as high as
80% in very fast collisions at small b. At small v the
energy loss is tiny, but can still have important
consequences for type I vortices: if v is very small,
they start with only just enough energy to escape
their mutual attraction. In undergoing a small b
collision they can lose enough of this energy to
become trapped in an oscillating bound state. In this
case they do not truly scatter and is ill-defined.
Myers et al. find that v 0.2 suffices to avoid
17
155
18
180
100
160
140
120
50
100
80
60
40
20
50
3
b
b
(b)
(a)
90
80
70
60
50
40
30
20
10
0
3
b
(c)
Figure 5 The 2-vortex scattering angle as a function of impact parameter b for v = 0.1 (5), v = 0.2 (4),
v = 0.3 (}), v = 0.4 (&), v = 0.5 (), and v = 0.9 (), as computed by Myers et al. (1992): (a) = 1=2; (b) = 2; (c) = 1. The
dotted curves are merely guides to the eye.
pThe solid curves in (b), (c) were computed using the point vortex model. Note that Myers
et al. use different normalizations, so b = 2bMRS and = MRS =2.
Critical Coupling
The AHM with = 1 has many remarkable properties, at which we have so far only hinted. These all
stem from Bogomolnyis crucial observation
(Manton and Sutcliffe 2004, pp. 197202) that the
potential energy in this case can be rewritten as
157
0.5
crit
0.4
0.3
0.2
0.1
2
Z (
1
2
B 1 jj
2
Z
jD1 iD2 j2 B d2 x i
1
E
2
R2
dD
19
jj2 B
20
21
n1
22
1
_ _
2
rs aar as
23
24
Manton originally proposed this geodesic approximation for monopoles, but it is now standard for all
topological solitons of Bogomolnyi type (where one
has a moduli space of static multisolitons saturating
a topological lower bound on E). Note that
geodesics are independent of initial speed, which
agrees with Myers et al: Figure 5c shows that (v, b)
25
@
12 h2
@t
26
Other Developments
The (critically coupled) AHM on a compact physical
space is of considerable theoretical and physical
interest. Bradlow showed that Mn () is empty unless
V = Area() 4n, so there is a limit to how many
vortices a space of finite area can accommodate
(Manton and Sutcliffe 2004, pp. 227230). Manton
has analyzed the thermodynamics of a gas of
vortices by studying the statistical mechanics of
geodesic flow on Mn (). In this context, spatial
compactness is a technical device to allow nonzero
vortex density n=V for finite n, without confining
the fields to a finite box, which would destroy the
Bogomolnyi properties. In the limit of interest,
n, V ! 1 with n/V fixed, the thermodynamical
properties turn out to depend on only through
V, so = S2 and = T 2 give equivalent results, for
example. The equation of state of the gas is
(P = pressure, T = temperature)
nT
P
V 4n
28
159
reference frame (the rest frame of the superconductor) so it is unsurprising that the Lorentz-invariant
AHM is inappropriate. Insofar as vortices move at
all, they seem to obey a first-order (in time)
dynamical system, in contrast to the second-order
AHM. Manton has devised a first-order system
which may have relevance to superconductivity, by
replacing Ekin with a ChernSimonsSchrodinger functional (Manton and Sutcliffe 2004, pp. 193197).
Rather than attracting or repelling, vortices now
tend to orbit one another at constant separation.
There is again a moduli space approximation to
slow vortex dynamics for
1, but it has a
Hamiltonian-mechanical rather than Riemanniangeometric flavor.
Finally, an interesting simplification of the AHM,
which arises, for example, as a phenomenological
model of liquid helium-4, is obtained if we discard the
gauge field A , or equivalently set the electric charge of
to e = 0. There is now no type I/II classification, since
may be absorbed by rescaling. The resulting model,
which has only global U(1) phase symmetry, supports
n-vortices =
(r)ein for all n, but these are not
exponentially spatially localized,
r 1
n2 n2 8 n2
Or6
22 r4
r2
29
Abbreviations
A
b
D
E
Ekin
F
L
L
S
Further Reading
Atiyah M and Hitchin N (1988) The Geometry and Dynamics of
Magnetic Monopoles. Princeton: Princeton University Press.
Fossheim K and Sudbo A (2004) Superconductivity: Physics and
Applications. Hoboken NJ: Wiley.
Jaffe A and Taubes C (1980) Vortices and Monopoles: Structure
of Static Gauge Theories. Boston: Birkhauser.
Nakahara M (1990) Geometry, Topology and Physics. Bristol:
Adam-Hilger.
Manton NS and Speight JM (2003) Asymptotic interactions of
critically coupled vortices. Communications in Mathematical
Physics 236: 535555.
Adiabatic Piston
Ch Gruber, Ecole Polytechnique Federale de
Lausanne, Lausanne, Switzerland
A Lesne, Universite P.-M. Curie, Paris VI, Paris,
France
2006 Elsevier Ltd. All rights reserved.
Introduction
Macroscopic Problem
N
p
T
N+
p+
T+
Adiabatic Piston
equation of state p = nkB T is satisfied at equilibrium, where n is the density of particles in the gas
and kB the Boltzmann constant. The adiabatic piston
is taken as a heavy particle of mass M m without
any internal degree of freedom. Using this same
model Feynman (1965) gave a qualitative analysis in
Lectures in Physics. He argued intuitively but
correctly that the system should converge first
toward a state of mechanical equilibrium where
p = p and then very slowly toward thermal
equilibrium. This approach toward thermal equilibrium is associated with the wiggles of the piston
induced by the random collisions with the atoms of
the gas. Of course, this stochastic behavior is not
part of thermodynamics and the evolution beyond
the mechanical equilibrium cannot appear in the
macroscopical framework assuming that the piston
does not conduct heat.
From a microscopical point of view, one is
confronted with two different problems: the
approach toward mechanical equilibrium in the
absence of any a priori friction (where the entropy
of both gases should increase) and, on a different
timescale, the approach toward thermal equilibrium
(where the entropy of one gas should decrease but
the total entropy increase).
The conceptual difficulties of the problem beyond
mechanical equilibrium come from the following
intuitive reasoning. When the piston moves toward
the hotter gas, the atoms of the hotter gas gain
energy, whereas those of the cooler gas lose energy.
When the piston moves toward the cooler side, it is
the opposite. Since on an average the hotter side
should cool down and the cold side should warm
up, we are led to conclude that on an average the
piston should move toward the colder side. On the
other hand, from p = nkB T, the piston should move
toward the warmer side to maintain pressure
balance.
In 1996, Crosignani, Di Porto, and Segev introduced a kinetic model to obtain equations describing
the adiabatic approach toward mechanical equilibrium. Starting with the microscopical model
introduced by Lebowitz, Gruber, Piasecki, and
Frachebourg, later joined by Lesne and Pache,
initiated in 1998 a systematic investigation of the
adiabatic piston within the framework of statistical
mechanics, together with a large number of numerical simulations. This analysis was based on the fact
that m=M is a very small parameter to investigate
expansions in powers of m=M (see Gruber and
Piasecki (1999) and Gruber et al. (2003) and
reference therein). An approach using dynamical
system methods was then developed by Lebowitz
et al. (2000) and Chernov et al. (2002). An
161
2
p
p O
V p p
M
8m p T p T
for jp p j 1
4
where
Microscopical Model
The system consists of two fluids separated by an
adiabatic piston inside a cylinder with x-axis,
length L, and area A. The fluids are made of N
identical light particles of mass m. The piston is a
heavy flat disk, without any internal degree of
freedom, of mass M m, orthogonal to the
x-axis, and velocity parallel to this x-axis. If the
piston is fixed at some position X0 , and if the two
fluids are in thermal equilibrium characterized by
(p
0 , T0 , N ), then they will remain in equilibrium
forever even if T0 6 T0 : it is thus an adiabatic
piston in the sense of thermodynamics. At a certain
time t = 0, the piston is let free to move and the
problem is to study the time evolution. To define the
dynamics, we consider that the system is purely
Hamiltonian, that is, the particles and the piston
move without any friction according to the laws of
mechanics. In particular, the collisions between the
particles and the walls of the cylinder, or the piston,
are purely elastic and the total energy of the system
is conserved. In most studies, one considers that the
particles are point particles making purely elastic
collisions. Since the piston is bound to move only in
the x-direction, the velocity components of the
particles in the transverse directions play no role in
this problem. Moreover, since there is no coupling
between the components in the x- and transverse
directions, one can simplify the model further by
assuming that all probability distributions are
v ! v 0 2V v v V
V ! V 0 V v V
2m
Mm
5
6
7
2mA
A
A 2m
Mm
M
8
mN Mgas
M
M
9
Adiabatic Piston
dX; P X; v ; X; V; t
surf
x; v; t 0 n
0 0 v L x x X0
1
12
15
V; t 0
V
Using the fact that = 2m=(M m) 1, we can
rewrite eqn [13] as a formal series in powers of :
@t V; t
1
X
1k k1 @ k e
Fk1 V; t 16
@V
k!
k1
~k V; t
F
v Vk
surf v ; V; tdv
V
1
14
1
163
v Vk
surf v ; V; tdv
17
k1
k1
n!
k!n k!
~k1 V; t
dV V nk F
18
1
if v > V
surf v; V; t
if v < V
surf v; tV; t;
19
where for L = 1,
surf (v; t) = n0 0 (v) and thus the
conditions to obtain eqn [18] are satisfied.
If L is finite, one can show that the factorization
property (eqn [19]) is an exact relation in the
thermodynamic limit for the piston (A ! 1,
M=A = cte). For finite L and finite A, we introduce
r1;
F3
20
1
dvv Vk
surf v; t
Fk V; t Fk V; t
and from eqn [18]
M d
hVi MhF2 V; ti
A dt
21
22
M d 2
hV i MhVF2 V;ti hF3 V;ti 23
A dt
= hV i then from eqns [12] and [20],
Introducing V
it follows that the (kinetic) energies satisfy
h
d hE i
M hF2 V; ti V
dt
A
V; ti
hV VF
2
i
hF3 V; ti
24
2
which implies conservation of energy.
From the first law of thermodynamics,
i
d h E i
1 h P!
PW PPQ!
dt
A
A
25
27
= hVi ,
and the expansion around V
t
Fn V; t
1
X
1
r!
r0
r
Fnr; VV
V
r2;
F2
r;
2F0
29
d hE i
M F2 V; t V
dt
A
X1
t 1
2r 3
F3 V;
2
2 r 2 r!
r1;
F2
V; tr
30
31
if v < Vt
Xt ; v; t Xt ; v 0 ; t;
if v > Vt
32
26
r;
3F2
t
F2 V; t F2 V;
X 2
r;
F0 2r
r!
r 0
dvv Vk
surf v; t
V
28
p
To lowest order in = m=M, and assuming
j1 p =p j is of order , one obtains from eqn [16]
the usual FokkerPlanck equation whose solution
gives
!
t2
1
1
V V
exp
0 V; t p
33
22 t
2 t
Adiabatic Piston
with
r
1
kB p
p
p p 1 et
Vt
p p
8m T
T
s
A 8m p
p
34
p p
M kB T
T
p
p
kB p
p T p T
p
p 1 e2t
TT
2 t
M
p T p T
(V V(t),
where the velocity V(t)
of the piston
tends exponentially fast toward stationary value
Vstat = V(1)
with relaxation time = 1 .
0
Let us note that for p = p , we have V(t)
and the evolution [33] is identical to the
OrnsteinUhlenbeck process of thermalization of
the Brownian particle starting with zero velocity
and friction coefficient . The analysis of [16] to
first order in yields then
"
#
3
X
k
V; t 1
ak tV Vt
0 V; t 35
r
9
p >
kB p
=
T T >
8m
if p p
p
>
k
>
B
2
TT ;
hV 2 istat hVistat
M
m
hVistat
M
p T p T
p
p
p p
8 p T p T 2
36
and
hV it
hVi2t
@ ~
F2 V; t
@V
39
(V) yields
V;t
V Vt;
that is; hV n it hVint
40
where
d
Vt F2 Vt; t;
dt
Vt 0 0
41
Moreover,
~2 V; t F2 V; tV; t
F
42
and
; P X; v; X; V; t x; v; t
X Xt
V Vt
43
1 2tet e2t
m
1
p p T p T
M TT
p
p!
o
p T p T
p
p 1 et 2
p T p T
38
k0
p p 1 et
165
r
m
2
t 1
37
2 ta2 t
M
2
surf (v; t), the functions F2 (V; t) and F2 (V; t) are
monotonically decreasing, we can introduce the
decomposition
M
^
p
2mF
V;
t
p
V; tV 44
surf
2
A
where the static pressure at the surface is
^ (t) = p
p
surf (V = 0; t) and the friction coefficients
45
its final velocity Vstat and one can solve eqn [12] to
obtain the evolution of the fluids.
Finite Cylinder (L < 1, M = 1)
2hE it
kB N
50
51
Infinite Cylinder (L = 1, M = 1)
with
^ t
n
^ t
n
1
dv
surf v; t
53
dv
surf v; t
and
Vt 0 0
47
49
^ p
^
^ kB T
n
54
p
For a time interval 1 = L m=kB T which is the time
for the shock wave to bounce back, the piston will
evolve as already discussed. In particular, if R is
sufficiently large, then after a time 0 = O((R )1 ) the
given by F2 (V,
t) = 0
piston will reach the velocity V
(eqn [47]). For t > 1 , F2 (V; t) depends explicitly on
time. For R sufficiently large, we can expect that for
all t the velocity V(t) will be a functional of
surf (v; t)
given by F2 [V(t);
surf (. ; t)] = 0, and thus the problem
is to solve eqn [12] with the boundary condition (eqn
[32]). Since V(t) so defined is independent of M=A,
the evolution will be independent of M=A if R is
sufficiently large. This conclusion, which we cannot
prove rigorously, will be confirmed by numerical
simulations.
To give a qualitative discussion of the evolution
for arbitrary values of R , we shall use the following
assumption already introduced in the experimental
measurement of cp =cv .
thermal equilibrium. If p
0 6 p0 , that is, n0 kB T0 6
n0 kB T0 , the piston will evolve monotonically to a
stationary state with constant velocity Vstat solution
of F2 (Vstat ) = 0. From [34], it follows that Vstat is a
function of n
0 =n0 , T0 , T0 but does not depend on
the value M=A. Moreover, the approach to this
stationary state is exponentially fast with relaxation
time 0 = 1=(V = 0). For Maxwellian distributions
0 (v), Vstat is a solution of
r q
q
8kB m
T0
kB n0 T0 n0 T0 Vstat
n0 T0 n
0
3
2
48
m n
Vstat
0 n0 O Vstat 0
Moreover,
01
r q
q
8kB m
n0 T0 n0 T0
Adiabatic Piston
167
56
4
ML
Xf
L Xf
with period
^
n
N
;
AXt
^
n
N
AL Xt
^ T t
T
av
55
N
2E0
Xf
1
Tf
A
AkB
L
where
2E0
AkB
N
N
T0
T0
A
A
and
s
s
s
A
A
AL
3
C
X3f
L Xf
N
N
2E0 kB
60
61
2
1
p
!0 R Rmax
66
nn 1 n2
V t2 t
2
67
but not thermal equilibrium Tf 6 Tf . Moreover, this
equilibrium state does not depend on M. Having
obtained the equilibrium point, we can then investigate the evolution close to the equilibrium point.
Linearizing eqn [51] around (Xf , Tf ) yields
2
Tf Xf
d
N
V kB
dt
M
X3
Tf L Xf 2
N
V 0V
M
L X3
1d
Vt F2 2 F0
dt
1d
2 t 42 F1 F3
dt
68
1d
hE it MF2 2 F0 V
dt
M=242 F1 F3
:
and 2 kB TP =M defines the temperature of the
piston.
Infinite Cylinder: Heat Transfer
63
For the infinite cylinder, the factorization assumption is an exact relation and in this case the
functions Fk (V; t) are independent of t. The solution
69
70
P
T T
A Q
M m
p p
p
p
72
p T p T
In conclusion, to first order in m=M, there is a heat
flux from the warm side to the cold one proportional to (T T ), induced by the stochastic
motion of the piston.
Finite Cylinder (L < 1, M < 1)
A O1
73
74
F3 F1 F3 F1
3 E0
2m
75
F3 F1
O
4F1
e 2 F3 O
4F1
We then introduce a (dimensionless) rescaled position for the piston
1 X
1 1
2 ;
76
2 L
2 2
which satisfies
d
2A F1 F1
kB T T
d
3E0
F1
The second perturbation approach above is supplemented with a slaving principle, expressing that at
each time of the slow evolution, that is, at fixed ,
the still present fast dynamics has reached a local
asymptotic state, slaved to the values of the slow
77
Adiabatic Piston
r
kB T
2m
2k
T
B
F3 V
F1 V VF2 V
m
F1 V
78
169
Numerical Simulations
As we have seen, the results were established under
the condition that m/M is a small parameter. Moreover for finite systems (L < 1, M < 1), it was
assumed that before collisions and to first order in
m/M, the factorization and the average assumptions
are satisfied. The numerical simulations are thus
essential to check the validity of these assumptions, to
determine the range of acceptable values m/M for the
perturbation expansion, to investigate the thermodynamic limit, and to guide the intuition.
In all simulation, we have taken kB = 1, m = 1,
T = 1 and usually T = 10. For L finite, we have
taken L = 60, X0 = 10, A = 105 , and N = N = N=2,
that is, p = R(M=A)(1=10) and p = 2p . The
number of particles N was varied from a few hundreds
to one or several millions; the mass M of the piston
from 1 to 105 . We give below some of the results
which have been obtained for L = 1 (Figures 2 and 3)
M=5
450
M = 10
400
350
0.5
M = 15
250
0.4
200
M = 25
150
Vstat
X(t )
300
0.3
0.2
100
M = 50
50
M = 100
0.1
0
0
(a)
500
1000
1500
t
2000
2500
3000
20
40
60
80
100
M
(b)
Figure 2 Evolution of the piston for L = 1, and p = p = 1 as observed in simulations (stochastic line in (a), dots in (b)) compared
with prediction: (a) position X(t ) for T = 10; and (b) stationary velocity for T = 10 (continuous line) and T = 100 (dotted line), as a
function of M.
0
0
2000
X (t )
2000
X(t)
4000
6000
4000
6000
8000
8000
10000
12000
0.5
1.5
2.5
10000
3.5
104
0.5
1.5
t
2.5
3
104
(b)
(a)
Figure 3 Evolution of the piston for L = 1, M = 104 , and p 6 p as observed in simulations (continuous line) compared with
predictions (dotted line): (a) p = 1, p = p p, from top to bottom p=p = 0.05, 0.1, 0.2, 1, 2, 3; and (b) p =
, p = 2
,
p=p = 1; X 0 =
X , t 0 =
t,
= 103 , 102 , 101 , 1, 10, 102 , 103 , 104 .
0.3
10
0.2
9.5
0.1
9
Xad
8.5
0.1
0.2
8
7.5
0.3
0
50
100
150
200
250
300
0.4
350
10
20
30
40
50
0.3
9.5
0.2
9
0.1
0
Xad
8
V(t )
X(t )
8.5
V
0.1
0.2
7.5
0.3
50
100
150
200
250
300
0.4
350
10
0.2
0
10
30
40
50
0.1
9.5
0.05
8.5
Xad
0.05
7.5
0.1
0.15
6.5
20
0.15
10
50
100
150
200
250
300
350
(a)
30
20
40
50
(b)
sim
Figure 4 Deterministic evolution toward mechanical equilibrium for L < 1, M = 105 : (a) position X(t); one finds Xad
= 8.3 whereas
th
sim = 0.343 whereas V
th = 0.3433. From top to bottom: R = 12: strong damping,
Xad
= 8.42 and (b) velocity V(t); one finds V
3
independent
pof R and M for R > 4 and M > 10 . R = 2: critical damping. R = 0.1: weak damping; damping coefficient increases with R
and !0 R for R < 1 but is independent of M for M > 103 .
Adiabatic Piston
105
2.8
2.8
2.6
2.6
2.4
2.4
2.2
2.2
1.8
1.8
1.6
1.6
1.4
1.4
0
Tav+ (t )
10
50
50
100
150
200
250
100
150
200
250
1.2
9.5
9
40
60
80
20
40
60
80
20
40
60
80
10
9.5
9
2
(t)
Tsurf
Tav (t )
20
8.5
2
1.5
1
10.5
+ (t)
Tsurf
1.2
105
psurf
p av
171
50
100
150
200
250
t
(a)
1.5
t
(b)
Figure 5 Same conditions as Figure 4, R = 12: (a) average pressure and temperature in the fluid: pav
(t) = 2E n =N ,
Tav
= E =N kB and (b) pressure and temperature at the surface of the piston. Prediction: Tad
= 1.54, Tad
= 9.46, pad
= pad
= 2.2.
= 1.52, Tad
= 9.48, pad
= pad
= 2.2.
Simulations: Tad
0.3
0.2
0.2
0.1
0.1
0
5
0.3
0.2
0.2
0.1
0.1
()
0.3
0.3
0.2
0.2
0.1
0.1
0
5
0.25
0.2
0
0.15
0.1
0
5
0
5
0.3
0
5
0.3
0
5
0.35
()
0.3
0.05
0.3
0.3
0.2
0.2
0.1
0.1
0
0
5
0
5
(a)
2.5
2.5
(b)
Figure 6 Velocity distribution in the left compartment. Same conditions as Figure 4, R = 12. Dotted line corresponds to Maxwellian
with T = 1.52: (a) t = 12, 24, 36, 48, 60, 92, 144, 240 from top to bottom and (b) t = 276460.
35
10
9
30
8
7
25
20
6
5
15
4
3
10
2
5
50
100
150
= t
200
250
300
(a)
50
100
150
= t
200
250
300
(b)
Adiabatic Piston
0.4
173
0.2
0.35
0.15
0.25
()/n
()/n
0.3
0.2
0.15
0.1
0.1
0.05
0.05
0
10
0
15
10
(a)
10
10
15
(b)
Tad
Further Reading
Callen HB (1963) Thermodynamics. New York: Wiley.
(Appendix C. See also Callen HB (1985) Thermodynamics
and Thermostatics, 2nd edn., pp. 51 and 53. New York:
Wiley.)
Chernov N, Sinai YaG, and Lebowitz JL (2002) Scaling dynamic
of a massive piston in a cube filled with ideal gas: exact
results. Journal of Statistical Physics 109: 529548.
Feynman RP (1965) Lectures in Physics I. New York: AddisonWesley.
Gruber Ch (1999) Thermodynamics of systems with internal
adiabatic constraints: time evolution of the adiabatic piston.
European Journal of Physics 20: 259266.
Gruber Ch and Lesne A (2005) Hamiltonian model of heat
conductivity and Fourier law. Physica A 351: 358.
Gruber Ch, Pache S, and Lesne A (2003) Two-time-scale
relaxation towards thermal equilibrium of the enigmatic
piston. Journal of Statistical Physics 112: 11991228.
AdS/CFT Correspondence
C P Herzog, University of California at Santa Barbara,
Santa Barbara, CA, USA
I R Klebanov, Princeton University, Princeton, NJ,
USA
2006 Elsevier Ltd. All rights reserved.
Introduction
The anti-de Sitter/conformal field theory (AdS/CFT)
correspondence is a conjectured equivalence
between a quantum field theory in d spacetime
dimensions with conformal scaling symmetry and a
quantum theory of gravity in (d 1)-dimensional
anti-de Sitter space. The most promising
approaches to quantizing gravity involve superstring theories, which are most easily defined in
10 spacetime dimensions, or M-theory which is
defined in 11 spacetime dimensions. Hence, the
AdS/CFT correspondences based on superstrings
typically involve backgrounds of the form AdSd1
Y9d while those based on M-theory involve backgrounds of the form AdSd1
Y10d , where Y are
compact spaces.
The examples of the AdS/CFT correspondence
discussed in this article are dualities between
(super)conformal nonabelian gauge theories and
superstrings on AdS5
Y5 , where Y5 is a fivedimensional Einstein space (i.e., a space whose
Ricci tensor is proportional to the metric,
Rij = 4gij ). In particular, the most basic (and maximally supersymmetric) such duality relates
N = 4 SU(N) super YangMills (SYM) and type IIB
superstring in the curved background AdS5
S5 .
There exist special limits where this duality is
more tractable than in the general case. If we take
the large-N limit while keeping the t Hooft coupling
= g2YM N fixed (gYM is the YangMills coupling
strength), then each Feynman graph of the gauge
theory carries a topological factor N , where is
the Euler characteristic of the graph. The graphs of
spherical topology (often called planar), to be
identified with string tree diagrams, are weighted by
N 2 ; the graphs of toroidal topology, to be identified
1
i1
AdS/CFT Correspondence
d1
X
Xi 2 L2
2
i1
1
gac gbd gad gbc
L2
3
175
Gauge theories in 3 1 dimensions play an important role in physics, and as explained above, parallel
D3 branes realize a (3 1)-dimensional U(N) SYM
L4
;
r4
f r 1
r0 4
r4
8
N
2
5
3
where vol(S
)=
p
is the volume of a unit 5-sphere,
and = 8G is the ten-dimensional gravitational
constant. It follows that
L4 5=2 N g2YM N0 2
11
2
12
2Ah 2 2
N V3 T 3
2
2
13
22 2
N V3 T 3
3
14
AdS/CFT Correspondence
22 2
N f V3 T 3
3
15
!1
3
4
17
177
18
Z
4
W
0 x exp d x
0 O
19
Zstring is the string theory path integral calculated as
a functional of
0 , the boundary condition on the
field
related to O by the AdS/CFT duality. In the
20
22
23
AdS/CFT Correspondence
1
hOxi
2 d
24
25
2 d
1
d=2 jx x0 j2
d=2
2
4
26
27
28
179
30
X
Y
U
Y
Y
V
Y
Y
Y
37
U
V
4,3
AdS/CFT Correspondence
za 2 0
39
a1
41
40
181
ac
3 N 2
4 volY
42
43
NovikovShifmanVainshteinZakharov (NSVZ)
beta function of each gauge group vanishes and
the R-charge of each superpotential term is 2.
For the Y p, q spaces mentioned above, one finds
that
p
q2 2p 4p2 3q2
3
volY p;q
p
2
2
2
2
2
3p 3q 2p p 4p 3q
44
The gauge theory consists of p q fields Z, p q
fields Y, 2p fields U, and 2q fields V. These fields all
transform in the bifundamental representation of a
pair of SU(N) gauge groups (the quiver diagram for
Y 4, 3 is given in Figure 2). The NSVZ beta function
and superpotential constraints determine the
R-charges up to two free parameters x and y. Let x
be the R-charge of Z and y the R-charge of Y. Then
the U have R-charge 1 (1=2)(x y) and the V
have R-charge 1 (1=2)(x y).
The technique of a maximization leads to the result
p
1
x 2 4p2 2pq 3q2 2p q 4p2 3q2
3q
p
1
y 2 4p2 2pq 3q2 2p q 4p2 3q2
3q
Thus, as calculated by Benvenuti et al. (2004) and
Bertolini et al. (2004)
aY p;q
3 N 2
4 volY p;q
45
z2a 2
46
a1
and introduces a warp factor so that the full tendimensional geometry has the form
ds10 2 h1=2 dx0 2
dxi 2 h1=2 d~s6 2
47
Conclusions
This article tries to present a logical path from
studying gravitational properties of D-branes to the
formulation of an exact duality between conformal
field theories and string theory in anti-de Sitter
backgrounds, and also sketches some methods for
breaking the conformal symmetry. Due to space
limitations, many aspects and applications of the
AdS/CFT correspondence have been omitted. At
the moment, practical applications of this duality
are limited mainly to very strongly coupled, large-N
gauge theories, where the dual string description is
well approximated by classical supergravity. To
understand the implications of the duality for more
general parameters, it is necessary to find better
Acknowledgments
The authors are very grateful to all their collaborators on gauge/string duality for their valuable
input over many years. The research of I R Klebanov
is supported in part by the National Science
Foundation (NSF) grant no. PHY-0243680. The
research of C P Herzog is supported in part by the
NSF under grant no. PHY99-07949. Any opinions,
findings, and conclusions or recommendations
expressed in this material are those of the authors
and do not necessarily reflect the views of the NSF.
See also: Brane Construction of Gauge Theories; Branes
and Black Hole Statistical Mechanics; Einstein Equations:
Exact Solutions; Gauge Theories from Strings; Large-N
and Topological Strings; Large-N Dualities; Mirror
Symmetry: A Geometric Survey; Quantum
Chromodynamics; Quantum Field Theory in Curved
Spacetime; Superstring Theories.
Further Reading
Aharony O, Gubser SS, Maldacena JM, Ooguri H, and Oz Y
(2000) Large N field theories, string theory and gravity.
Physics Reports 323: 183 (arXiv:hep-th/9905111).
Benvenuti S, Franco S, Hanany A, Martelli D, and Sparks J (2005)
An infinite family of superconformal quiver gauge theories with
SasakiEinstein duals. JHEP 0506: 064 (arXiv:hep-th/0411264).
Affine quantum groups are certain pseudoquasitriangular Hopf algebras that arise in mathematical physics
in the context of integrable quantum field theory,
integrable quantum spin chains, and solvable lattice
models. They provide the algebraic framework behind
the spectral parameter dependent YangBaxter equation
R12 uR13 u vR23 v
R23 vR13 u vR12 u
1
^) is a quantization of
A quantum affine algebra Uq (g
^
the enveloping algebra U(g) of an affine Lie algebra
^. So we start by introducing
(KacMoody algebra) g
affine Lie algebras and their enveloping algebras
before proceeding to give their quantizations.
Let g be a semisimple finite-dimensional Lie algebra
over C of rank r with Cartan matrix (aij )i,j = 1,..., r ,
symmetrizable via positive integers di , so that di aij is
symmetric. In terms of the simple roots i , we have
aij 2
i j
ji j2
and
k0
E
i
The
are referred to as Chevalley generators and
the last set of relations are known as Serre relations.
The generator D is known as the canonical derivation. We will denote the algebra obtained by
^0 .
dropping the generator D by g
^
In applications to physics, the affine Lie algebra g
often occurs in an isomorphic form as the loop Lie
algebra g[z, z1 ] C c with Lie product (for
^)
untwisted g
l
Hi
h
i
i
qH
i qi
;
E
E
ij
i
j
qi q1
i
D; Hi 0;
D; E
i;0 E
i
i
ji j2
:
di
2
P
We can introduce an 0 = ri = 1 ni i in such a way
that the extended Cartan matrix (aij )i,j = 0,..., r is of
affine type that is, it is positive semidefinite of
rank r. The integers ni are referred to as Kac indices.
Choosing 0 to be the highest root of g leads to an
untwisted affine KacMoody algebra while choosing
0 to be the highest short root of g leads to a twisted
affine KacMoody algebra.
^ corresponding
One defines the affine Lie algebra g
to this affine Cartan matrix as the Lie algebra
(over C) with generators Hi , E
i for i = 0, 1, . . . , r
and D with relations
h
i
Hi ; E
Hi ; Hj 0
aij E
j
i ;
h
i
E
ij Hi
2
i ; Ej
i;0 E
D; Hi 0;
D; E
i
i
1a
Xij
k
1aij k
1 aij
1k
Ej E
0; i 6 j
E
i
i
k
kl
3
1a
Xij
1
1 aij
k
k0
k 1aij k
E
Ej Ei
0;
i
4
i 6 j
qi
m
n
mq !
nq !m nq !
6
7
^) is a Hopf
The quantum affine algebra Uh (g
algebra with coproduct
D D 1 1 D
Hi Hi 1 1 Hi
Hi =2
H =2
E
qi i E
E
i
i qi
i
8
antipode
SD D;
SHi Hi
1
S Ei qi Ei
and co-unit
D Hi E
i 0
9
10
^) by
obtained from the ones given above for Uh (g
setting
Hi =2
qi
Ki ;
i 0; . . . ; r
11
for all x 2 A
13
where
: x y 7! y x exchanges the two factors in
the coproduct. In a quasitriangular Hopf algebra,
this element R satisfies
idR R13 R23
id R R13 R12
14
15
16
(with summation over repeated indices). The Yangian Y(g ) is the algebra generated by these and a
second set of generators Ja satisfying
12
R12
R
12 = 1 2 x
R
12 =
2 1 x R
Ia ; Jb fabc Jc
Ja Ja 1 1 Ja 12 fabc Ic Ib
Ja ; Jb ; Ic Ia ; Jb ; Jc abcdeg fId ; Ie ; Ig g
and
17
Ja ; Jb ; Il ; Jm Jl ; Jm ; Ia ; Jb
abcdeg flmc lmcdeg fabc Id ; Ie ; Jg
^0
V 2 V 1
V 3 V 2
V 1
V 1 V 2
V 3
=
V 1
V 2 V 3
abcdeg
1
f f f f ;
24 adi bej cgk ijk
fx1 ; x2 ; x3 g
xi xj xk
i6j6k
n
X
tij u eij
i;j1
Yangians
As remarked by Drinfeld (1986), for untwisted ^g the
quantum affine algebra Uh (^g 0 ) degenerates as h ! 0
into another quasipseudotriangular Hopf algebra,
the Yangian Y(g ) (Drinfeld 1985). It is associated
with R-matrices which are rational functions of the
additive spectral parameter u. Its representation ring
coincides with that of Uh (^g 0 ).
Consider a general presentation of a Lie algebra g ,
with generators Ia and structure constants fabc ,
so that
Ia ; Ib fabc Ic ;
where
Ia Ia 1 1 Ia
Ru v 1
P
;
uv
where P
n
X
eij eji
i;j1
n
X
k1
tik u tkj u
m
@ @ i
aij j
0 :j j
ni e
j1
Lt x; t
r
r
X
mX
Hi @t i
e
=2aij j E
i Ei
2 i1
2 i;j1
r
m X
=2a0j j
1
E
e
E
0
2 j1
0
r
r
X
mX
Hi @x i
e
=2aij j E
i Ei
2 i1
2 i;j1
r
mX
1
=2a0j j
e
E0 E0
2 j1
19
where
L1n (1 ) = Ln (1 ) id, L2n (2 ) = id Ln (2 ).
That R solves the YangBaxter equation follows
from the equivalence of the two ways of intertwining
Ln (1 ) Ln (2 ) Ln (3 ) with Ln (3 ) Ln (2 )
Ln (1 ).
To compute Ln (), one uses the canonical, equaltime commutation relations for the i and _ i . In
terms of the lattice fields
pi;n
qi;n
Z
Z
n1=2
_ i x dx
n1=2
n1=2
n1=2
e =2aij j x dx
qi;n i E
0 E0
i
!
X
exp
Hj pj;n O2
4 j
the expression used by the St Petersburg school and
by Jimbo. We now make the replacement
Hi =4 Hi =4
E
Ei q
, where q = exp(ih
2 =2), and
i 7! q
compute the O() terms in [19], which reduce to
RzHi 1 1 Hi
Hi 1 1 Hi Rz
Hi =2
qHi =2 E
Rz E
i q
i
Hi =2
qHi =2 E
E
q
Rz
i
i
H0 =2
qH0 =2 E
Rz z1 E
0 q
0
1
H0 =2
qH0 =2 E
z
E
q
Rz
0
0
^)
where z = 1 =2 . We recognize in these the Uh (g
coproduct and thus the intertwining relations, in the
homogeneous gradation. These equations were
solved for R in defining representations of
nonexceptional g by Jimbo (1986).
c2 , it was Kulish and Reshetikhin (1983)
For ^g = sl
who first discovered that the requirement that the
coproduct must be an algebra homomorphism forces
the replacement of the commutation relations of
b 2 ) by those of Uh (sl
b 2 ); more generally it requires
U(sl
the replacement of U(^g ) by Uh (^g ).
i
b
20
i
a
S 1R R1
An S-matrix will have poles at certain imaginary
rapidities ab
corresponding to the formation of
c
virtual bound states. This is graphically represented
in Figure 1b. The location of the pole is determined
by the masses of the three particles involved,
m2c m2a m2b 2ma mb cosiab
c
At the bound state pole, the S-matrix will project
matrix has to have
onto the multiplet V c . Thus, the R
this projection property as well and indeed, this turns
out to be the case. The bootstrap principle, whereby
the S-matrix for a bound state is obtained from the
S-matrices of the constituent particles,
c
(a)
d
b
d a
21
a
c
a
b
cab
(b)
or, graphically,
W
W
V1
V2
...
Vn
V1
V2
...
Vn
V2
V3 . . . Vn 1 Vn
22
23
Qi qi
Hi
E
i Ei i qi 1;
i 0; . . . ; r
24
26
Further Reading
27
1
V 1/
V 1
V 2
2
V 1/
V 1
V 2
and
~Jp Jp 1 fpiq Ii Iq Iq Ii
4
AharonovBohm Effect
M Socolovsky, Universidad Nacional Autonoma de
Mexico, Mexico DF, Mexico
2006 Elsevier Ltd. All rights reserved.
Introduction
In classical electrodynamics, the interaction of charged
particles with the electromagnetic field is local,
through the pointlike coupling of the electric charge
of the particles with the electric and magnetic fields, E
and B, respectively. This is mathematically expressed
by the Lorentz-force law. The scalar and vector
potentials, and A, which are the time and space
components of the relativistic 4-potential A , are
considered auxiliary quantities in terms of which
the field strengths E and B, the observables, are
expressed in a gauge-invariant manner. The homogeneous or first pair of Maxwell equations are a direct
consequence of the definition of the field strengths in
terms of A_ The inhomogeneous or second pair of
Maxwell equations, which involve the charges and
currents present in the problem, are also usually
written in terms of E and B ; however when writing
them in terms of A , the number of degrees of freedom
of the electromagnetic field is explicitly reduced from
six to four; and finally, with two additional gauge
transformations, one ends with the two physical
degrees of freedom of the electromagnetic field.
In quantum mechanics, however, both the
Schrodinger equation and the path-integral approaches
for scalar and unpolarized charged particles in the
presence of electromagnetic fields, are written in
terms of the potential and not of the field strengths.
Even in the case of the SchrodingerPauli equation
for spin 1=2 electrons with magnetic moment m
interacting with a magnetic field B, one knows that
the coupling m B is the nonrelativistic limit of the
Dirac equation, which depends on A but not on E and
B_ Since gauge invariance also holds in the quantum
domain, it was thought that A and were mere
auxiliary quantities, like in the classical case.
Aharonov and Bohm, in 1959, predicted a quantum interference effect due to the motion of charged
particles in regions where B(E) vanishes, but not
A(), leading to a nonlocal gauge-invariant effect
depending on the flux of the magnetic field in the
inaccessible region, in the magnetic case, and on the
difference of the integrals over time of time-varying
potentials, in the electric case. (The magnetic effect
was already noticed 10 years before by Ehrenberg
and Siday in a paper on the refractive index of
electrons.)
d
v
p = q E B
dt
c
1
where
mv
p = p
1 v2 =c2
is the mechanical momentum of the particle with
electric charge q, mass m, and velocity v = x (c is
the velocity of light in vacuum, and for jvj c the
left-hand side (LHS) of [1] is approximately mv); the
right-hand side (RHS) is the Lorentz force, where E
and B are, respectively, the electric and magnetic
fields at the spacetime point (t, x) where the particle
is located. Equation [1] is easily derived from the
EulerLagrange equation
d @L
@L
=0
2
dt @v
@x
with the Lagrangian L given by the sum of the free
Lagrangian for the particle,
r
v2
2
L0 = mc 1 2
3
c
and the Lagrangian describing the particlefield
interaction,
q
4
Lint = A v q
c
In [4], A and are, respectively, the vector potential
and the scalar potential, which together form the
4-potential A = (A0 , A) = (, Ai ), i = 1, 2, 3,
in terms of which the electric and magnetic field
strengths are given by
1@
E=
A r
c @t
5a
B = rA
5b
dt L0
dt Lint S0 Sint
6
t1
B2
A ! A0 = A r
7a
dF = d2 A = 0
7b
9
rE
1@
B=0
c @t
10a
10b
r E = 4
rB
1@
4j
E=
c @t
c
12a
12b
1@
! 0 =
c @t
At the level of A,
t1
B1
13
Z t0
i
1
Dxexp
d mx 2
h t
2
xtx
Z t0
iq
0
exp
A dx dx
hc t
Z t0
Z xt0 x0
i
1
Dxexp
d mx 2
h t
2
xtx
Z t0
iq
exp
dx A
hc t
xt0 x0
16
R
where the integral Dx() . . . is over all continuous
spacetime paths (, x()) which join (t, x) with (t0 , x0 ).
If one knows the wave function at (t, x), then the
wave function at (t0 , x0 ) is given by
t0 ; x0 =
d3 x Kt0 ; x0 ; t; x t; x
17
where
P=
@
q
L = p A
@v
c
i
hr A q
@t
2m
c
2
h
q2
r2
A2
2m
2mc2
i
hq
i
hq
15
rA
A r q
2mc
mc
The gauge transformation [7a] and [7b] is a
symmetry of this equation, if simultaneously to the
change of the 4-potential, the wave function transforms as follows:
t; x !
0
t; x = eiq=hc t; x
7c
ih
(iq=hc)
So, A and
obey [15]. At each (t, x), e
belongs to U(1), the unit circle in the complex plane.
In the path-integral approach, the kernel
K(t0 , x0 ;t, x), which gives the probability amplitude
for the propagation of the particle from the spacetime
point (t, x) to the spacetime point (t0 , x0 ) (t < t0 ), is
given by
Kt0 ; x0 ; t; x
Z xt0 x0
i
Dxexp S0 Sint
h
xtx
Z t0
Z xt0 x0
i
1
mx 2
Dxexp
d
h
2
xtx
t
q
A v q
c
@
@t
=
2 2
h
r
2m
18
then
t; x = exp
iq
hc
Ax0 dx0
0 t; x
19
x0
AharonovBohm Proposal
In 1959, Aharonov and Bohm proposed an experiment to test, in quantum mechanics, the coupling of
electric charges to electromagnetic field strengths
through a local interaction with the electromagnetic
potential A , but not with the field strengths
themselves. However, as we saw before, no physical
effect exists, that is, A can be gauged away, unless
magnetic and/or electric fields exist somewhere,
although not necessarily overlapping the wave function of the particles.
Consider the usual two-slit experiment as depicted
in Figure 1, with the additional presence, behind the
slits, of a long and narrow solenoid enclosing a
nonvanishing magnetic flux due to a constant and
homogeneous magnetic field B normal to the plane
I
S
ijej=hc
e
P
z
R
II
2
i=hS0 0 ijej=hc
R
I
ijej=hc
ijej=hc
e
f 0 g
e
y
ei=hS0
R
II
0
!
S
0
P I
II[I
ei=hS0
f 0 g
R
R
I
fg
R
ijej=hc
ijej=hc
0
P I
0
P II
e2i=0
0
P II
23
and
with
0
P II
= 2
ei=hS0
f 0 g
2
Pj
=j
2
0
P Ij
2
0
P IIj
2i=0 0
0
2Re e
P I P II
25
X1
m = 1
eim
0
P m
26
0
27
k0 =
P ;
k2Z
28
29
30a
Since R2
is homeomorphic to an open disk minus a
point (D20 )
, then the total space of the bundle is
homeomorphic to an open solid 2-torus minus a
circle, since (T02 )
= (D20 )
S1 . Then the AB
bundle has the topological structure
AB : S1 ! T02
! D20
30b
31
A; f ! A f 1 df
32
through
C0 G ! C0 ;
C0
fgauge equivalence classes
G
of flat connections on AB g
33
34
where
A0 = i
x dy y dx
2 C0
x 2 y2
35
36a
36b
36c
Di = i
@
j
Ai
@x
1
Z
Z
1
X
m1
36d
1
1
t
0
m
dm !
m
0
@
= X ij Aij s0i e0j
@x
36e
gt !
t
t
=0
dt
37
38
1
d1 !
1
2
d2 !
2
m1
0
where
AU i = Ai dx =
!U i
d !
gtg01 T exp
39
In particular, if is a product bundle, then is the
identity, and choosing g(0) = 1 gives
!!
Z ct
"
AU
40
c t = ct; T exp
c0
41
42
For c 2 (R2
;(x0 , y0 )), which turns n times
around the solenoid at (0, 0), eqn [40] gives
H
H
n A
in a
c x ; y ; e
c
c" x0 ; y0 ; e
0 0
H
x0 ; y0 ; e
ijejn
Adx
x0 ; y0 ; e2in=0
Z; 62 Q
43
nonabelian. Examples with YangMills and gravitational fields are considered in the literature.
Acknowledgment
The author thanks the University of Valencia,
Spain, where part of this work was done.
See also: Deformation Quantization and Representation
Theory; Fractional Quantum Hall Effect; Geometric
Phases; Moduli Spaces: An Introduction; Quantum
Chromodynamics; Variational Techniques for
GinzburgLandau Energies.
44
Further Reading
Aguilar MA and Socolovsky M (2002) AharonovBohm effect,
flat connections, and Greens theorem. International Journal
of Theoretical Physics 41: 839860.
Aharonov Y and Bohm D (1959) Significance of electromagnetic potentials in the quantum theory. Physical Review
15: 485491.
Berry MV (1984) Quantal phase factors accompanying adiabatic
changes. Proceedings of the Royal Society of London A 392:
4557.
Chambers RG (1960) Shift of an electron interference pattern by
enclosed magnetic flux. Physical Review Letters 5: 35.
Corichi A and Pierri M (1995) Gravity and geometric phases.
Physical Review D 51: 58705875.
Dirac PMA (1931) Quantised singularities in the electromagnetic field. Proceedings of the Royal Society of London A
133: 6072.
Kobe DH (1979) AharonovBohm effect revisited. Annals of
Physics 123: 381410.
Peshkin M and Tonomura A (1989) The AharonovBohm Effect.
Berlin: Springer.
Schulman LS (1971) Approximate topologies. Journal of Mathematical Physics 12: 304308.
Steenrod N (1951) The Topology of Fibre Bundles. Princeton, NJ:
Princeton University Press.
Sundrum R and Tassie LJ (1986) Non-abelian AharonovBohm
effects, Feynman paths, and topology. Journal of Mathematical Physics 27: 15661570.
Wu TT and Yang CN (1975) Concept of nonintegrable phase
factors and global formulation of gauge fields. Physical
Review D 12: 38453857.
Introduction
Quantum field theory may be understood as the
incorporation of the principle of locality, which is at
the basis of classical field theory, into quantum
199
0
idM idaM
(M)
O 2 KM
201
Hence, pointlike localized observables are necessarily singular. Actually, the Wightman formulation
of quantum field theory is based on the use of
distributions on spacetime with values in the algebra
of observables (as a topological -algebra). In spite
of technical complications whose physical significance is unclear, this formalism is well suited for a
discussion of the connection with the Euclidean
theory, which allows, in fortunate cases, a treatment
by path integrals; it is more directly related to
models and admits, via the operator-product expansion, a study of the short-distance behavior. It is,
therefore, an important question how the algebraic
approach is related to the Wightman formalism. The
reader is referred to the literature for exploring the
results on this relation.
Whereas these results point to an essential equivalence of both formalisms, one needs in addition a
criterion for the existence of sufficiently many Wightman fields associated with a given local net. Such a
criterion can be given in terms of a compactness
condition to be discussed in the next subsection. As a
benefit, one derives an operator-product expansion
which has to be assumed in the Wightman approach.
In the purely algebraic approach, the ultraviolet
structure has been investigated by Buchholz and
Verch. Small-scale properties of theories are studied
with the help of the so-called scaling algebras whose
elements can be described as orbits of observables
under all possible renormalization group motions.
There results a classification of theories in the scaling
limit which can be grouped into three broad classes:
theories for which the scaling limit is purely classical
(commutative algebras), those for which the limit is
essentially unique (stable ultraviolet fixed point) and
not classical, and those for which this is not the case
(unstable ultraviolet fixed point). This classification
does not rely on perturbation expansions. It allows
an intrinsic definition of confinement in terms of the
so-called ultraparticles, that is, particles which are
visible only in the scaling limit.
Phase-Space Analysis
A 2 AO
AO
it
AO;
t2R
203
M1
! AM1
#
DM2
#
! AM2
M2
Further Reading
Anomalies
205
Anomalies
S L Adler, Institute for Advanced Study, Princeton, NJ,
USA
Synopsis
1a
1b
1c
e20
F xF x
162
2
3a
206 Anomalies
3b
with
! 0 at the end of the calculation, one
observes that the divergence of eqn [3b] contains
an extra term with a factor of
. On careful
evaluation, one finds that the coefficient of this
factor is an expression that behaves as
1 , which
gives the anomaly in the limit of vanishing
. (4)
Finally, if the field theory is defined by a functional
integral over the classical action, the standard
Noether analysis shows that the classical action is
invariant under the chiral transformation of eqn
[1b], apart from the contribution of the mass term,
which gives the naive axial-vector divergence. However, as pointed out by Fujikawa, the chiral
transformation must also be applied to the functional integration measure, and since the measure is
an infinite product, it must be regularized to be well
defined. Careful calculation shows that the regularized measure is not chiral invariant, but contributes
an extra term to the axial-vector Ward identity that
is precisely the chiral anomaly.
A key feature of the anomaly is that it is
irreducible: a local polynomial counter term cannot
be added to the AVV diagram that preserves
vector-current conservation and eliminates the
anomaly. More generally, one can show that there
is no way of modifying quantum electrodynamics
so as to eliminate the chiral anomaly, without
spoiling either vector-current conservation (i.e.,
electromagnetic gauge invariance), renormalizability, or unitarity. Thus, the chiral anomaly is a new
physical effect in renormalizable quantum field
theory, which is not present in the prequantization
classical theory.
The abelian chiral anomaly is the simplest case of
the anomaly phenomenon. It was extended to
nonabelian gauge theories by Bardeen using a
point-splitting method to compute the divergence,
followed by adding polynomial counter terms to
remove as many of the residual terms as possible.
The resulting irreducible divergence is the nonabelian chiral anomaly, which in terms of YangMills
field strengths for vector and axial-vector gauge
potentials V and A ,
FV x @ V x @ V x iV x; V x
iA x; A x
FA x @ A x @ A x iV x; A x
iA x; V x
is given by
4a
4b
4c
Anomalies
207
6a
with the pion mass, f 131 MeV the chargedpion decay constant, and S a constant determined
by the constituent fermion charges and axial-vector
couplings. Taking the matrix element of eqn [6a]
between the vacuum state and a two-photon state,
and using the fact that the left-hand side has a
kinematic zero (the SutherlandVeltman theorem),
one sees that the 0 ! amplitude F is completely determined by the anomaly term, giving the
formula
p
F =2S 2=f
6b
For a single set of fractionally charged quarks, the
amplitude F is a factor of three too small to agree
with experiment; for three fractionally charged
for all ; ;
7
208 Anomalies
giving rise to the -vacuum of quantum chromodynamics, and a host of issues associated with (the lack
of) strong CP violation, the PecceiQuinn mechanism, and axion physics. Also, the fact that the
integral of eqn [8b] is nonzero means that the U(1)
chiral symmetry of quantum chromodynamics is
broken by instantons, which as shown by t Hooft
resolves the longstanding U(1) problem of strong
interactions, that of explaining why the flavor
singlet pseudoscalar meson 0 is not light, unlike its
flavor octet partners.
Anomaly Matching Conditions
9
A
A x
10a
X @
A x
10b
Anomalies
covariant derivative. In terms of the internalsymmetry component fields Aa and Va of the
YangMills potentials of eqn [4a], one introduces
operators
Xa x @
fabc Vb c
Aa x
A x
fabc Ab
Y a x @
Vc x
fabc Vb c
a
V x
V x
fabc Ab
11a
Ac x
with fabc the antisymmetric nonabelian group structure constants. The operators Xa and Y a are easily
seen to obey the commutation relations
11b
Y a x; Y b y fabc x yYc x
Let [V, A] be the effective action as a functional of
the fields V , A , constructed so that the vector
currents are covariantly conserved, as expressed
formally by
Y a V; A 0
12a
12b
Trace Anomalies
Xa x; Xb y fabc x yYc x
Xa x; Y b y fabc x yXc x
209
12c
13c
210 Anomalies
The focus above has been on anomalies in fourdimensional spacetime, but anomalies of various
types occur both in lower-dimensional quantum
field theories (such as theories in two- and threedimensional spacetimes) and in quantum field theories in higher-dimensional spacetimes (such as N = 1
supergravity in ten-dimensional spacetime). Anomalies also play an important role in the formulation
and consistency of string theory. The bosonic string is
consistent only in 26-dimensional spacetime, and the
analogous supersymmetric string only in ten-dimensional spacetime, because in other dimensions both
these theories violate Lorentz invariance after quantization. In the Polyakov path-integral formulation of
these string theories, these special dimensions are
associated with the cancellation of the Weyl anomaly,
which is the relevant form of the trace anomaly
discussed above. YangMills, gravitational, and
mixed YangMills gravitational anomalies make an
appearance both in N = 1 ten-dimensional supergravity and in superstring theory, and again special
dimensions play a role. In these theories, only when
the associated internal symmetry groups are either
SO(32) or E8 E8 is elimination of all anomalies
possible, by cancellation of hexagon-diagram anomalies with anomalous tree diagrams involving
exchange of a massless antisymmetric two-form
field. This mechanism, due to Green and Schwarz,
requires the factorization of a sixth-order trace
invariant that appears in the hexagon anomaly in
terms of lower-order invariants, as well as two
numerical conditions on the adjoint representation
generator structure, restricting the allowed gauge
groups to the two noted above.
Covariant versus Consistent Anomalies;
Descent Equations
Anomalies
211
Retrospect
The wide range of implications of anomalies has
surprised even astonished the founders of the
subject. New anomaly applications have appeared
within the last few years, and very likely the future
will see continued growth of the area of quantum
field theory concerned with the physics and mathematics of anomalies.
Acknowledgment
This work is supported, in part, by the Department of Energy under grant #DE-FG02-90ER40542.
See also: Bosons and Fermions in External Fields;
BRST Quantization; Effective Field Theories; Gauge
Theories from Strings; Gerbes in Quantum Field Theory;
Index Theorems; Lagrangian Dispersion (Passive
Scalar); Lattice Gauge Theory; Nonperturbative and
Topological Aspects of Gauge Theory; Quantum
Electrodynamics and Its Precision Tests; Quillen
Determinant; Renormalization: General Theory;
SeibergWitten Theory.
Further Reading
Adler SL (1969) Axial-vector vertex in spinor electrodynamics.
Physical Review 177: 24262438.
Adler SL (1970) Perturbation theory anomalies. In: Deser S,
Grisaru M, and Pendleton H (eds.) Lectures on Elementary
Particles and Quantum Field Theory, vol. 1, pp. 3164.
Cambridge, MA: MIT Press.
Adler SL (2005) Anomalies to all orders. In: t Hooft G (ed.) Fifty Years
of YangMills Theory, pp. 187228. Singapore: World Scientific.
Adler SL and Bardeen WA (1969) Absence of higher order
corrections in the anomalous axial-vector divergence equation.
Physical Review 182: 15171536.
Bardeen W (1969) Anomalous ward identities in spinor field
theories. Physical Review 184: 18481859.
Bell JS and Jackiw R (1969) A PCAC puzzle: 0 ! in the
-model. Nuovo Cimento A 60: 4761.
Bertlmann RA (1996) Anomalies in Quantum Field Theory.
Oxford: Clarendon.
De Azcarraga JA and Izquierdo JM (1995) Lie Groups,
Lie Algebras, Cohomology and Some Applications in Physics,
ch. 10. Cambridge: Cambridge University Press.
Fujikawa K and Suzuki H (2004) Path Integrals and Quantum
Anomalies. Oxford: Oxford University Press.
Golterman M (2001) Lattice chiral gauge theories. Nuclear
Physics Proceeding Supplements 94: 189203.
Green MB, Schwarz JH, and Witten E (1987) Superstring Theory.
vol. 2, sects. 13.313.5. Cambridge: Cambridge University Press.
Hasenfratz P (2005) Chiral symmetry on the lattice. In: t Hooft G
(ed.) Fifty Years of YangMills Theory, pp. 377398.
Singapore: World Scientific.
#fj : j g
!1
3
Introduction
The central objective in the study of quantum chaos
is to characterize universal properties of quantum
systems that reflect the regular or chaotic features of
the underlying classical dynamics. Most developments of the past 25 years have been influenced by
the pioneering models on statistical properties of
eigenstates (Berry 1977) and energy levels (Berry
and Tabor 1977, Bohigas et al. 1984). Arithmetic
quantum chaos (AQC) refers to the investigation of
quantum systems with additional arithmetic structures that allow a significantly more extensive
analysis than is generally possible. On the other
hand, the special number-theoretic features also
render these systems nongeneric, and thus some of
the expected universal phenomena fail to emerge.
Important examples of such systems include the
modular surface and linear automorphisms of tori
(cat maps) which will be described below.
The geodesic motion of a point particle on a
compact Riemannian surface M of constant negative curvature is the prime example of an Anosov
flow, one of the strongest characterizations of
dynamical chaos. The corresponding quantum
eigenstates j and energy levels j are given by the
solution of the eigenvalue problem for the Laplace
Beltrami operator (or Laplacian for short)
0;
AreanH
;
4
kkL2 M 1
1
2
Xj
AreanH
j
4
4
Figure 1 Image of the absolute-value-squared of an eigenfunction j (z) for a nonarithmetic surface of genus 2. The surface is
obtained by identifying opposite sides of the fundamental region.
Reproduced from Aurich and Steiner (1993) Statistical properties of
highly excited quantum eigenstates of a strongly chaotic system.
Physica D 64(13): 185214, with permission from R Aurich.
jj j2 g dA !
g dA;
j!1
6
Hyperbolic Surfaces
Let us begin with some basic notions of hyperbolic
geometry. The hyperbolic plane H may be abstractly
defined as the simply connected two-dimensional
Riemannian manifold with Gaussian curvature 1.
A convenient parametrization of H is provided by
the complex upper-half plane, H = {x iy: x 2
R, y > 0}, with Riemannian line and volume
elements
ds2
dx2 dy2
;
y2
dA
dx dy
y2
7
8
c d
A fundamental domain for the action of the
modular group PSL(2, Z) on H is the set
F PSL2;Z z 2 H : jzj > 1; 12 < Re z < 12
9
(see Figure 2). The modular group is generated by
the translation
1 1
: z 7! z 1
0 1
and the inversion
1
: z 7! 1=z
0
0
1
4dx2 dy2
1
x2
y2 2
dA
4dx dy
1 x2 y2 2
10
: ; 2 C; jj2 jj2 1
11
acting on D as above via fractional linear transformations. The fundamental group of the regular
octagon surface is the subgroup of all elements in
PSU(1, 1) with coefficients of the form
p
k l 2;
p
m n 2
q
p
1 2
12
13
2X
17
as X, L ! 1, L X.
The main tool in the attempts to prove the above
conjectures has been the Selberg trace formula. It
relates sums over eigenvalues of the Laplacians to
sums over lengths of closed geodesics on the
hyperbolic surface. The trace formula is in its
simplest form in the case of compact hyperbolic
surfaces; we have
Z
1
X
AreaM 1
hj
h tanh d
4
1
j0
1
XX
2H
gn
2
sinhn
=2
n1
18
23
24
25
26
28
27
N 0 ce=2
32
33
r ! 1 r b
b
lim
34
The Hecke operators are normal, form a commuting family, and in addition they commute with
the Laplacian . In the following, we consider an
orthonormal basis of eigenfunctions j of that
are simultaneously eigenfunctions of all Hecke
operators. We will refer to such eigenfunctions as
Hecke eigenfunctions. The above assumption is
automatically satisfied, if the spectrum of is
simple (i.e., no eigenvalues coincide), a property
conjectured by Cartier and supported by numerical
computations. Lindenstrauss work is based on the
following two observations. Firstly, all quantum
limits of Hecke eigenfunctions are geodesic-flow
invariant measures of positive entropy, and secondly, the only such measure of positive entropy
that is recurrent under Hecke correspondences is
the Lebesgue measure.
The first property is proved by Bourgain and
Lindenstrauss (2003) and refines arguments of
Rudnick and Sarnak (1994) and Wolpert (2001) on
the distribution of Hecke points (see Sarnak (2003) for
38
39
that is, pointwise values of jj j cannot be substantially larger than its sum over Hecke points. This,
and the observation that Hecke points for a large set
of integers n are sufficiently uniformly distributed
on nH as n ! 1, yields the estimate of positive
entropy with a quantitative lower bound.
Lindenstrauss proof of the second property,
which shows that Lebesgue measure is the only
quantum limit of Hecke eigenfunctions, is a result of
a currently very active branch of ergodic theory:
measure rigidity. Invariance under the geodesic flow
alone is not sufficient to rule out other possible limit
measures. In fact, there are uncountably many
measures with this property. As limits of Hecke
eigenfunctions, all quantum limits possess an additional property, namely recurrence under Hecke
correspondences. Since the explanation of these is
rather involved, let us recall an analogous result in a
simpler setup. The map 2 : x 7! 2x mod 1 defines a
hyperbolic dynamical system on the unit circle with
a wealth of invariant measures, similar to the case of
the geodesic flow on a surface of negative curvature.
Furstenberg conjectured that, up to trivial invariant
measures that are localized on finitely many rational
points, Lebesgue measure is the only 2-invariant
measure that is also invariant under action of
3 : x 7! 3x mod 1. This fundamental problem is
still unsolved and one of the central conjectures in
measure rigidity. Rudolph, however, showed that
Furstenbergs conjecture is true if one restricts the
statement to 2-invariant measures of positive
entropy (cf. Lindenstrauss (to appear)). In Lindenstrauss work, 2 plays the role of the geodesic
flow, and 3 the role of the Hecke correspondences.
Although here it might also be interesting to ask
whether an analog of Furstenbergs conjecture
1
X
41
n1
1
X
aj nns
42
n1
43
where
s ij
s ij
s; j
Ls; j
2
2
s
44
45
1
1 aj pps p12s
46
Ls; j
p prime
4 12 ; j1 j2 j3
48
49
M
2
1 X
j a a
N
50
with N() = #{j: j }; cf. [3]. Following a conjecture by FeingoldPeres and Eckhardt et al. (see Sarnak
(2003) for references) for generic quantum chaotic
systems, one expects a central-limit theorem for the
statistical fluctuations of the j (a), where the normalized variance N()1=2 V (a) is asymptotic to the
classical autocorrelation function C(a), see eqn [54].
Conjecture 4
we have
!
1 X j a a
p
g
N
V a
j
Z 1
1
2
! p
gte1=2t dt
2 1
51
at gagdg dt
54
nPSL2;R
55
1;
2i
1 X
N Q mod N
1 Q 2 Q
56
n2Z2
where b
f (n) are the Fourier coefficients of f, and
TN (n) are translation operators
Ca :
Z Z
as ! 1.
!1
where
57
52
53
Bj 12 L 12; j Cj j
t1 Q Q 1
t2 Q e2iQ=N Q
58
59
1=2
exp
Q0 mod N
0
02
QQ Q
2i 2
Q
N
Q0
62
kkL2 Z=NZ 1
63
64
Graffi, and Isola (1995). That is, [65] holds for all
j = 1, . . . , N. Rudnick and Kurlberg, and more
recently Gurevich and Hadani, have established
results on the rate of convergence analogous to
[49]. These results are unconditional. Gurevich and
Hadani use methods from algebraic geometry based
on those developed by Deligne in his proof of the
Weil conjectures (an analog of the Riemann hypothesis for finite fields).
In the case of quantum-cat maps, there are values
of N for which the number of coinciding eigenvalues
can be large, a major difference to what is expected
for the modular surface. Linear combinations of
eigenstates with the same eigenvalue are as well
eigenstates, and may lead to different quantum
limits. Indeed, Faure, Nonnenmacher, and De Bie`vre
(see De Bie`vre (to appear)) have shown that there
are subsequences of values of N, so that, for all
f 2 C1 (T2 ),
Z
1
1
66
f d f 0
hOpf Nj ; Nj i !
2 T2
2
that is, half of the mass of the quantum limit
localizes on the hyperbolic fixed point of the map.
This is the first, and to date the only, rigorous result
concerning the existence of scarred eigenfunctions in
systems with chaotic classical limit.
Acknowledgment
The author is supported by an EPSRC Advanced
Research Fellowship.
See also: Quantum Ergodicity and Mixing of
Eigenfunctions; Random Matrix Theory in Physics.
Further Reading
Aurich R and Steiner F (1993) Statistical properties of highly
excited quantum eigenstates of a strongly chaotic system.
Physica D 64(13): 185214.
Berry MV and Keating JP (1999) The Riemann zeros and
eigenvalue asymptotics. SIAM Review 41(2): 236266.
Bogomolny EB (2006) Quantum and arithmetical chaos. In:
Cartier PE, Julia B, Moussa P, and Vanhove P (eds.) Frontiers
in Number Theory, Physics and Geometry on Random
Matrices, Zeta Functions, and Dynamical Systems, Springer
Lecture Notes. Les Houches.
Bogomolny EB, Georgeot B, Giannoni M-J, and Schmit C (1997)
Arithmetical chaos. Physics Reports 291(56): 219324.
De Bie`vre S Recent Results on Quantum Map Eigenstates,
Proceedings of QMATH9, Giens 2004 (to appear).
Hejhal DA and Rackner BN (1992) On the topography of Maass
waveforms for PSL(2, Z). Experiment. Math. 1(4): 275305.
Keating JP (1991) The cat maps: quantum mechanics and classical
motion. Nonlinearity 4(2): 309341.
221
Introduction
A major motivation for studying the asymptotic
structure of spacetimes has been the need for a
rigorous description of what should be understood by
an isolated system in Einsteins theory of gravity.
As an example, consider a gravitating system somewhere in our universe (e.g., a galaxy, a cluster of
galaxies, a binary system, or a star) evolving according to its own gravitational interaction, and possibly
reacting to gravitational radiation impinging on it
from the outside. Thereby it will emit gravitational
radiation. We are interested in describing these waves
because they provide us with important information
about the physics governing the system.
To adequately describe this situation, we need to
idealize the real situation in an appropriate way, since
it is hopeless to try to analyze the behavior of the
system in its interaction with the rest of the universe.
We are mainly interested in the behavior of the
system, and not so much in other processes taking
place at large distances from the system. Since we
would like to ignore those regions, we need a way to
isolate the system from their influence.
The notion of an isolated system allows us to
select individual subsystems of the universe and
describe their properties regardless of the rest of the
universe so that we can assign to each subsystem
such physical attributes as its energymomentum,
angular momentum, or its emitted radiation field.
Without this notion, we would always have to take
into account the interaction of the system with its
environment in full detail.
In general relativity (GR) it turns out to be a rather
difficult task to describe an isolated system and the
reason is as always in Einsteins theory the fact
that the metric acts both as the physical field and as
the background. In other theories, like electrodynamics, the physical field, such as the Maxwell field,
is very different from the background field, the flat
metric of Minkowski space. The fact that the metric
in GR plays a dual role makes it difficult to extract
physical meaning from the metric because there is no
nondynamical reference point.
Imagine a system alone in the universe. As we
recede from the system we would expect its influence
to decrease. So we expect that the spacetime which
models this situation mathematically will resemble
the flat Minkowski spacetime and it will approximate
it even better the farther away we go. This implies
that one needs to impose fall-off conditions for the
curvature and that the manifold will be asymptotically flat in an appropriate sense. However, there is
the problem that fall-off conditions necessarily imply
the use of coordinates and it is awkward to decide
which coordinates should be good ones. Thus, it is
not clear whether the notion of an asymptotically flat
spacetime is an invariant concept.
What is needed, therefore, is an invariant definition of asymptotically flat spacetimes. The key
observation in this context is that infinity is far
away with respect to the spacetime metric. This
means that geodesics heading away from the system
should be able to run forever, that is, be defined
for arbitrary values of their affine parameter s.
Infinity will be reached for s ! 1. However,
suppose we do not use the spacetime metric g but a
metric ^g which is scaled down with respect to g, that
is, in such a way that ^g = 2 g for some function .
Then it might be possible to arrange in such a way
that geodesics for the metric ^g cover the same events
(strictly speaking, this holds only for null geodesics,
but this is irrelevant for the present plausibility
argument) as those for the metric g yet that their
affine parameter ^s (which is also scaled down with
respect to s) approaches a finite value ^s0 for s ! 1.
Then we could attach a boundary to the spacetime
manifold consisting of all the limit points corresponding to the events with ^s = ^s0 on the ^g-geodesics.
3
RV U
1
4
i+
vtr
i0
g du dv 14 v u2 d2
The coordinates u and v both take arbitrary real values
but they are restricted by the relation v u = 2r 0.
In order to see what happens at infinity, we introduce
the coordinates U and V by the relations
I
u tan U;
v tan V
1
4 cos2 U cos2 V
2
223
Definition 1 A spacetime (M, gab ) is called asymptotically simple if there exists a manifold-withc with metric ^gab and scalar field on
boundary M
c and boundary I = @M such that the following
M
conditions hold:
1.
2.
3.
4.
5.
c M = int M;
c
M is the interior of M:
2
^gab = gab on M;
and ^gab are smooth on all of M;
> 0 on M; = 0, ra 6 0 on I ; and
each null geodesic acquires both future and past
endpoints on I .
Some Consequences
There are several consequences of the existence of
the conformal boundary I . They all can be traced
back to the fact that this boundary can be used to
separate the geometric fields into a universal background field and dynamical fields which propagate
on it. The background is given by the boundary
points attached to an asymptotically flat spacetime
which always form a three-dimensional null hypersurface I with two connected components (in the
sequel, we restrict our attention to I only; I is
treated similarly), each with the topology of a
cylinder. And in each case, I is shear-free.
The BMS Group
c Since it
Now consider the Weyl tensor Ca bcd on M.
vanishes on I where = 0 we may form the
quotient
Ka bcd 1 Ca bcd
5
225
Further Reading
Ashtekar A (1987) Asymptotic Quantization. Naples: Bibliopolis.
Bondi H, van der Burg MGJ, and Metzner AWK (1962)
Gravitational waves in general relativity VII. Waves from
axi-symmetric isolated systems. Proceedings of the Royal
Society of London, Series A 269: 2152.
Frauendiener J (2004) Conformal infinity. Living Reviews in
Relativity, vol. 3. http://relativity.livingreviews.org/Articles/
lrr-2004-1/index.html.
Friedrich H (1992) Asymptotic structure of space-time. In: Janis AI
and Porter JR (eds.) Recent Advances in General Relativity.
Boston: Birkhauser.
Friedrich H (1998a) Einsteins equation and conformal structure.
In: Huggett SA, Mason LJ, Tod KP, Tsou SS, and Woodhouse
NMJ (eds.) The Geometric Universe: Science, Geometry and
the Work of Roger Penrose. Oxford: Oxford University Press.
Friedrich H (1998b) Gravitational fields near space-like and null
infinity. Journal of Geometry and Physics 24: 83163.
Geroch R (1977) Asymptotic structure of space-time. In: Esposito
FP and Witten L (eds.) Asymptotic Structure of Space-Time.
New York: Plenum.
Hawking S and Ellis GFR (1973) The Large Scale Structure of
Space-Time. Cambridge: Cambridge University Press.
Penrose R (1965) Zero rest-mass fields including gravitation:
asymptotic behaviour. Proceedings of the Royal Society of
London, Series A 284: 159203.
Penrose R (1968) Structure of space-time. In: DeWitt CM
and Wheeler JA (eds.) Battelle Rencontres. New York:
W. A. Benjamin.
Penrose R and Rindler W (1984, 1986) Spinors and Space-Time,
Cambridge: Cambridge University Press.
Sachs RK (1962) Gravitational waves in general relativity VIII.
Waves in asymptotically flat space-time. Proceedings of the
Royal Society of London, Series A 270: 103127.
Averaging Methods
A I Neishtadt, Russian Academy of Sciences,
Moscow, Russia
2006 Elsevier Ltd. All rights reserved.
fast oscillations. The most common field of applications of averaging methods is the analysis of the
behavior of dynamical systems that differ from
integrable systems by small perturbations.
Introduction
Averaging Principle
Averaging Methods
I_ "gI; ; ";
_ !I "f I; ; "
I I1 ; . . . ; In 2 R
1
227
G J 2m
I
Tm
g J; ; 0 d
2
#i ki ; ;
j krj ;
F J; # 2mr
3
f J; ; 0 d
T mr
@H1
@x
@H0
@H1
_
"
@I
@I
y_ "
4
p_ 0;
5
Tm
"v1 J; "2 v2 J;
6
7
G0 J g J; ; 0
8
i1
9
where gk , k 2 Zm , are Fourier coefficients of function g at " = 0, and u01 is an arbitrary function of J. It
is assumed that the denominators in [9] do not
vanish, and that the series in [9] converges and
determines a smooth function. In the same way,
from the other equations in [8] one can sequentially
determine F0 , v1 , . . . , Gi , u i1 , Fi , v i1 , i 1.
On truncating the series in [6] and [7] at the terms
of order " l , we obtain a truncated system of the lth
approximation. The equation for J is decoupled
from the other equations and can be solved
separately. Then the behavior of
is determined
by means of quadrature. The behavior of original
variable I in this approximation is a slow drift
(described by the equation for J), on which small
oscillations (described by transformation of variables)
are superimposed. The behavior of can be represented as a rotation with slowly varying frequency,
on which oscillations are also superimposed. For l = 1,
the truncated system coincides with the averaged
system [2].
If the sublattice Zm specifying possible
resonant relations is given, then in an analogous
manner one can construct a formal transformation
of variables (I, ) 7! (J, ) such that, in the new
variables, the fast phase will appear on the righthand sides of the differential equations for the new
variables only in combinations (k, ), with k 2
(see, e.g., Arnold et al. (1988)). Again, on truncating the series on the right-hand sides of the
differential equations for the new variables at the
terms of order " l , we obtain a truncated system of
the lth approximation. At l = 1, this truncated
system coincides with the partially averaged system
[3] (for some special choice of arbitrary functions
that are contained in the formulas for transformation
of variables). If the original system is a Hamiltonian
system of the form [4], then the transformation of
variables eliminating the fast phases from the righthand sides of the differential equations can be
chosen to be symplectic. The corresponding
procedures are called Lindstedt method and
Newcomb method (nonresonant case for n = m),
Delaunay method (resonant case for n = m), and
von Zeipel method (resonant case for n m) (see
Poincare (1957) and Arnold et al. (1988)).
The calculation of high-order terms in the
procedures of elimination of fast variables is rather
cumbersome. There are versions of these procedures
which are convenient for symbolic processors
(especially for Hamiltonian systems, e.g., the
DepritHori method; Giacaglia 1972).
The averaging method consists in using the
averaged system for the description of motion in
the first approximation and the truncated systems
Averaging Methods
229
x 2 R p; 0 < " 1
10
X0 x lim
Xt; x; 0 dt
11
12
y_ "gx; y; "
13
Averaging Methods
invariant measure
y and is ergodic for almost all
values of y. Introduce the averaged system
Z
_Y " GY; GY 1
gx; Y; 0d
Y
Y M M
According to the averaging principle, one should use
the solution Y(t) of the averaged system with initial
condition Y(0) = y(0) for approximate description of
slow motion y(t) in the original system. This
averaging principle is justified by the following
Anosov theorem [1]: for any positive the measure
of the set E(, ") of initial data (from a compact in
the phase space) such that
max j yt Ytj >
0 t 1="
tends to 0 as " ! 0.
The particular case when the original system is
a Hamiltonian system depending on slowly varying parameter = "t, and for almost all values of
the motion of the system with = const is
ergodic on almost all energy levels, is considered
in Kasuga (1961).
For the case when the has strong mixing properties, see Bakhtin (2004) and Kifer (2004).
For slowfast systems, there is also a generalization of approach of the previous section that uses
time averaging and the condition of uniform average
(Volosov 1962).
231
Further Reading
Anosov DV (1960) Averaging in systems of ordinary differential
equations with rapidly oscillating solutions. Izvestiya Akademii Nauk SSSR, Ser. Mat. 24(5): 721742 (Russian).
Arnold VI (1983) Geometrical Methods in the Theory
of Ordinary Differential Equations. New YorkBerlin:
Springer.
Arnold VI, Kozlov VV, and Neishtadt AI (1988) Mathematical
Aspects of Classical and Celestial Mechanics, Encyclopaedia
of Mathematical Sciences, vol. 3. Berlin: Springer.
Bakhtin VI (2004) Cramer asymptotics in the averaging method
for systems with fast hyperbolic motions. Proceedings of the
Steklov Institute of Mathematics 244(1): 79.
Bogolyubov NN (1945) On some statistical methods in mathematical physics. Akad. Nauk USSR. Lvov (Russian).
Bogolyubov NN and Mitropolskii YuA (1961) Asymptotic
Methods in the Theory of Nonlinear Oscillations. New York:
Gordon and Breach.
Giacaglia GEO (1972) Perturbation Methods in Nonlinear
Systems, Applied Mathematical Science, vol. 8. Berlin: Springer.
Kasuga T (1961) On the adiabatic theorem for the
Hamiltonian system of differential equations in the classical
mechanics I, II, III. Proceedings of the Japan Academy 37(7):
366382.
Kevorkian J and Cole JD (1996) Multiple Scale and Singular
Perturbations Methods, Applied Mathematical Sciences,
vol. 114. New York: Springer.
Kifer Y (2004) Some recent advances in averaging. In: Modern
Dynamical Systems and Applications, 403. Cambridge:
Cambridge University Press.
Lochak P and Meunier P (1988) Multiphase Averaging for
Classical Systems, Applied Mathematical Sciences, vol. 72.
New York: Springer.
Mitropolskii YuA (1971) Averaging Method in Nonlinear
Mechanics. Kiev: Naukova Dumka (Russian).
Poincare H (1957) Les Methodes Nouvelles de la Mecanique
Celeste, vols. 13. New York: Dover.
Sanders JA and Verhulst F (1985) Averaging Methods in
Nonlinear Dynamical Systems, Applied Mathematical
Sciences, vol. 59. New York: Springer.
Volosov VM (1962) Averaging in systems of ordinary differential
equations. Russian Mathematical Surveys 17(6): 1126.
VX !
Introduction
M
Axioms of a TQFT
An (n 1)-dimensional TQFT (V, ) over a scalar
field k assigns to every closed oriented n-dimensional manifold X a finite-dimensional vector space
V(X) over k and assigns to every cobordism
(M, X, Y) a k-linear map
M M; X; Y : VX ! VY
Here a cobordism (M, X, Y) between X and Y is a
compact oriented (n 1)-dimensional manifold M
endowed with a diffeomorphism @M X q Y (the
overline indicates the orientation reversal). All
manifolds and cobordisms are supposed to be
smooth. A TQFT must satisfy the following axioms.
1. Naturality Any orientation-preserving diffeomorphism of closed oriented n-dimensional manifolds f : X ! X0 induces an isomorphism f] : V
(X)! V(X0 ). For a diffeomorphism g between the
cobordisms (M, X, Y) and (M0 , X0 , Y 0 ), the following diagram is commutative:
VX
M
gjX ]
!
VY
gjY ]
!
VX0
M0
VY 0
VY
f]
!
VZ
M0
VY 0
There
are
functorial
VX q Y VX VY
V; k
such that the following diagrams are commutative:
VX q Y q Z VX VY VZ
#
#
VX q Y q Z VX VY VZ
VX q ;
#
VX
VX k
#
VX
233
circles S1 q S1 and one circle S1 ) defines a commutative multiplication on the vector space A = V(S1 ).
The 2-disk, considered as a cobordism between S1
and ;, induces a nondegenerate trace on the algebra
A. This makes A into a commutative Frobenius
algebra (also called a symmetric algebra). This
algebra completely determines the TQFT (V, ).
Moreover, this construction defines a one-to-one
correspondence between equivalence classes of twodimensional TQFTs and isomorphism classes of
finite dimensional commutative Frobenius algebras
(Kock 2003).
The formalism of TQFTs was to a great extent
motivated by the three-dimensional case, specifically, Wittens ChernSimons TQFTs. A mathematical definition of these TQFTs was first given
by Reshetikhin and Turaev using the theory of
quantum groups. The WittenReshetikhinTuraev
three-dimensional TQFTs do not satisfy exactly the
definition above: the naturality and the functoriality
axioms only hold up to invertible scalar factors
called framing anomalies. Such TQFTs are said to
be projective. In order to get rid of the framing
anomalies, one has to add extra structures on the
three-dimensional cobordism category. Usually one
endows surfaces X with Lagrangians (maximal
isotropic subspaces in H1 (X; R)). For 3-cobordisms,
several competing but essentially equivalent
additional structures are considered in the literature:
2-framings (Atiyah 1989), p1 -structures (Blanchet
et al. 1995), numerical weights (K Walker, V Turaev).
Large families of three-dimensional TQFTs are
obtained from the so-called modular categories.
The latter are constructed from quantum groups at
roots of unity or from the skein theory of links.
See Quantum 3-Manifold Invariants.
Additional Structures
The axiomatic definition of a TQFT extends in
various directions. In dimension 2 it is interesting to
consider the so-called openclosed theories involving
1-manifolds formed by circles and intervals and
two-dimensional
cobordisms
with
boundary
(G Moore, G Segal). In dimension 3 one often
considers cobordisms including framed links and
graphs whose components (resp. edges) are labeled
with objects of a certain fixed category C. In such a
theory, surfaces are endowed with finite sets of
points labeled with objects of C and enriched with
tangent directions. In all dimensions one can study
manifolds and cobordisms endowed with homotopy
classes of mappings to a fixed space (homotopy
quantum field theory, in the sense of Turaev).
Additional structures on the tangent bundles spin
Further Reading
Atiyah M (1989) Topological Quantum Field Theories. Publications Mathematiques de lIhes 68: 175186.
Bakalov B and Kirillov A Jr. (2001) Lectures on Tensor
Categories and Modular Functors. University Lecture Series
vol. 21. Providence, RI: American Mathematical Society.
Introduction
The term axiomatic quantum field theory subsumes a collection of research branches of quantum
field theory analyzing the general principles of
relativistic quantum physics. The content of the
results typically is structural and retrospective rather
than quantitative and predictive.
The first axiomatic activities in quantum field theory
date back to the 1950s, when several groups started
investigating the notion of scattering and S-matrix in
detail (Lehmann, Symanzik, and Zimmermann 1955
(LSZ-approach), Bogoliubov and Parasiuk 1957, Hepp
and Zimmermann (BPHZ-approach), Haag 195759
and Ruelle 1962 (HaagRuelle theory) (see Scattering,
Asymptotic Completeness and Bound States and
Scattering in Relativistic Quantum Field Theory:
Fundamental Concepts and Tools).
Wightman (1956) analyzed the properties of the
vacuum expectation values used in these approaches
and formulated a system of axioms that the vacuum
expectation values ought to satisfy in general. Together
with Garding (1965), he later formulated a system of
axioms in order to characterize general quantum fields
in terms of operator-valued functionals, and the two
systems have been found to be equivalent.
A couple of spectacular theorems such as the PCT
theorem and the spinstatistics theorem have been
obtained in this setting, but no interacting quantum
fields satisfying the axioms have been found so far
(in 1 3 spacetime dimensions). So, the development of alternatives and modifications of the setting
got into the focus of the theory, and the axioms
themselves became the objects of research. Their
role as axioms understood in the common sense
turned into the role of mere properties of quantum
fields. Today, the term axiomatic quantum field
theory is widely avoided for this reason.
In a long list of publications spread over the
1960s, Araki, Borchers, Haag, Kastler, and others
worked out an algebraic approach to quantum field
theory in the spirit of Segals postulates for general
quantum Mechanics (1947) (see Algebraic Approach
to Quantum Field Theory).
The Wightman setting was the basis of a framework into which the causal construction of the
S-matrix developed by Stuckelberg (1951) and
Bogoliubov and Shirkov (1959) has been fitted by
Epstein and Glaser (1973). The causality principle
fixes the time-ordered products up to a finite
number of parameters at each order, which are to
be put in as the renormalization constants.
Already in 1949, Dyson had seen that problems in
the formulation of quantum electrodynamics (QED)
could be avoided by just multiplying the time
variable and, correspondingly, the energy variable by
the imaginary unit constant (Wick rotation). Schwinger then investigated time-ordered Green functions of
QED in this Euclidean setting. This approach was
formulated in terms of axioms by Osterwalder and
Schrader (1973, 1975) (see Euclidean Field
Theory).
Other extensions of the aforementioned settings
are objects of current research (see Indefinite Metric,
Quantum Fields
Garding and Wightman characterized operatorvalued quantum fields on the Minkowski spacetime
R13 by a couple of axioms. Given additional
assumptions concerning the high-energy behavior,
the GardingWightman fields are in oneone correspondence with algebraic field theories.
Without specifying or presupposing these additional assumptions, the axioms will now be formulated and discussed in detail and compared to the
corresponding conditions in the algebraic setting.
Adjoint operators are marked by an asterisk, and
Einsteins summation convention is used.
Operator-valued functionals The components of a
field F are an n-tuple F1 Fn of linear maps that
13
assign to each test function 2 C1
) linear
0 (R
operators F1 () Fn () in a Hilbert space H with
domains of definition D(F1 ()) D(Fn ()). There
exists a dense subspace D of H with
D D(F ()) \ D(F () ) and F ()D [ F () D D
for all indices . Consider m such fields F 1 Fm
with components Fa , 1 a m, 1 na . Assume
there to be an involution : (1 m) ! (1 m) such
that Fa () = Fa () , where (x) : (x).
Quantum fields cannot be operator-valued functions on R13 if one wants them to exhibit (part of)
the properties to follow. But point fields can be
quadratic forms; typically this is the case for fields in
a Fock space.
For each component Fa and each open region
O R13 , the field operators Fa () with supp O
generate a -algebra F a (O) of operators defined on
D. These operators typically are unbounded, which
is one of the differences with the traditional setting
of the algebraic approach. There a C -algebra A (O)
is assigned to each open region O in such a way
that O P implies A (O) A (P). Each C -algebra
is a -algebra, but in contrast to a C -algebra,
a -algebra does not need to be endowed with a
norm. The fundamental observables in quantum
theory are bounded positive operators (typically, but
not always, projections), and these generate a C algebra.
There is no fundamental physical motivation for
confining the setting to fields with a finite number of
components, except that it includes most of the
fields known from daily life.
235
Covariance There exist strongly continuous unitary representations U and T of SL(2, C) and
(R13 , ),
respectively,
and
representations
D1 Dm of SL(2, C) in Cn1 Cnm , respectively,
such that
UgFa Ug Da g1 Fa g1
and
TyFa Ty Fa y;
where Da (g1 ) are the elements of the matrix
Da (g1 ). Dropping coordinate indices, this reads
UgF a Ug Da g1 F a g1
These distributions are called the N-point functions of the fields F 1 F m and yield the vacuum
expectation values of the theory. It is straightforward to deduce the following properties from the
GardingWightman axioms.
Microcausality (BoseFermi alternative) If i and
i1 have spacelike separated supports, then
ai ai1 aN
wa11
i i1 N 1 i i1 N
ai1 ai aN
wa11
i1 i N 1 i1 i N :
and
TyFa Ty F a y:
The representations U and T generate a representation of the universal covering of the restricted
Poincare group.
As it stands, this assumption is a very strong one,
since it manifestly fixes the action of the representation on the field operators. In the algebraic
approach, the covariance assumption is more modestly formulated. Namely, it is assumed that
U(g)A (O)U(g) = A ((g)O) and T(y)A (O)T(y) =
A (O y), leaving open how the representation acts
on the single local observables.
Vacuum vector There exists a unique (up to a
multiple) vector 2 D that is invariant under the
representations U and T and cyclic with respect to
the algebra F (R13 ) generated by all field operators
Fa (), that is, F (R13 ) = H.
Spectrum condition The joint spectrum of the
components of the 4-momentum, i.e., of the generators of the spacetime translations, has support in
the closed forward light cone V , that is, the set
{k2 0, k0 0}.
The existence of an invariant ground state called
the vacuum is standard in algebraic quantum field
theory as well.
aN
wa11
N 1 N
N-Point Functions
Consider the above fields F 1 F m . For each N 2 N
and each N-tuple (a1 aN ) of natural numbers m
(labeling fields), define families (F a1 aN ) :
aN
a1 aN
aN
(Fa11
: (wa11
N )i nai and w
N )i nai of dis13 N
tributions on (R ) by
aN
a1
aN
Fa11
N 1 N : F1 1 FN N
where
x 1 N1 : x; x 1 ; x 1 2 ; . . . ; x 1
N1 :
aN
are called the Wightman
The functions Wa11
N
functions, and they have the following property
because of the spectrum condition of the field.
237
aN
lim wa11 ...
x
N
0<!1
aL
aL1 aN
wa11
L wL1 N :
Hermiticity
!( ) = !():
Positivity !( ) 0:
To see Hermiticity, compute
!
; h; F
X
i
ij
wi j
h; Fi j
ij
h; Fi
j ; i
ij
and
i
hF ; i ! ;
i
i
j
i
i
Fj j i
ij
; : ; :
and an involution by
a1 aN
aN a1
:
1 N
N 1
Define an antilinear
S((R13 )N ) by
hFi i ; Fj j i
ij
and
involution
; : ;:
on
S N :
x1 xN : xN x1
for each N 2 N. Put S 0 : C and z : z for all
z 2 C.
S
Define S I N : S N I N , and S I : N S I N . For
13 N
each 2 I N , the set S : S((R ) ) L
{} is a
linear space. On the direct sum BI : 2 I S
define an associative product by
; ; : ;
2
X
Fi i
0:
i
Results
The mathematical and structural analysis of quantum fields has improved the understanding of
scattering theory in the different approaches mentioned above; see Bogoliubov et al. (1975) and the
relevant articles in this encyclopedia. Apart from
this, the following results deserve to be mentioned.
Evidently, many others have to be omitted for
practical reasons.
PCT Symmetry
One can derive the spectrum condition, the BisognanoWichmann symmetries/the Unruh effect, and
239
Bibliographic Notes
Important monographs on axiomatic quantum field
theory are those by Streater and Wightman (1964),
Jost (1965), Bogoliubov et al. (1975), and Bogoliubov
et al. (1990). Note that the books of Bogoliubov et al.
differ in setup fundamentally and that neither replaces
the other. For a lecture notes volume, see also Volkel
(1977), and for a review article, see Streater (1975).
A valuable discussion of the Wightman axioms can
also be found in the second volume of the series by
Reed and Simon (1970).
The first monograph on the algebraic approach to
quantum field theory is due to Haag (1992), a more
recent one has been written by Araki (1999).
Concerning the sufficient conditions for switching
between the GardingWightman and the algebraic
approach, see Wollenberg (1988) and the Ph.D.
thesis of Bostelmann (2000) and references given
there. Dynamical and thermodynamical foundation
of standard axioms, the BisognanoWichmann
symmetries (Unruh effect), and the spinstatistics
theorem, have been investigated by Kuckert (2002,
2005), see also the references given there for related
work.
In different formulations and at differing degrees of
mathematical sophistication, the causal approach to
perturbation theory can be found in the monographs
by Bogoliubov and Shirkov (1959), Scharf (1989,
2001), and Steinmann (2000). Two modern review
articles have been written by Brunetti and Fredenhagen
(2000) and by Dutsch and Fredenhagen (2004).
The reference original articles on the Euclidean
axioms are those of Osterwalder and Schrader (1973,
1975). Note that the first one contains an error. (cf.
also Zinoviev (1995)). A monograph on Euclidean
field theory and its relations to the other axiomatic
settings of quantum field theory and to statistical
mechanics is that by Glimm and Jaffe (1987).
A recent review on quantum energy inequalities is
due to Fewster (2003).
Acknowledgments
The author is a fellow of the Emmy-Noether
Programme (DFG). Thanks for discussions are due
to Professor D Arlt.
See also: Algebraic Approach to Quantum Field Theory;
C*-Algebras and Their Classification; Constructive
Quantum Field Theory; Dispersion Relations; Euclidean
Field Theory; Indefinite Metric; Perturbative
Renormalization Theory and BRST; Quantum Field
Theory: A Brief Introduction; Quantum Field Theory in
Curved Spacetime; Scattering, Asymptotic Completeness
and Bound States; Scattering in Relativistic Quantum
Field Theory: Fundamental Concepts and Tools;
Scattering in Relativistic Quantum Field Theory: The
Analytic Program; Symmetries in Quantum Field Theory:
Algebraic Aspects; Symmetries in Quantum Field Theory
of Lower Spacetime Dimensions; Thermal Quantum Field
Theory; TomitaTakesaki Modular Theory;
Two-Dimensional Models.
Further Reading
Araki H (1999) Mathematical Theory of Quantum Fields.
Oxford: Oxford University Press.
Bogoliubov NN, Logunov AA, and Todorov IT (1975) Introduction to Axiomatic Quantum Field Theory, (Russian original
edition: Nauka (Moskow) 1969). New York: Benjamin.
Bogoliubov NN, Logunov AA, Oksak AI, and Todorov IT (1990)
General Principles of Quantum Field Theory, (Russian
original edition Nauka (Moskow) 1987). DordrechtBoston
London: Kluwer.
Bostelmann H (2000) Lokale Algebren und Operatorprodukte am
Punkt (in German). Ph.D. thesis, Gottingen.
Brunetti R and Fredenhagen K (2004) Microlocal Analysis and
Interacting Quantum Field Theories: Renormalization on
Physical Backgrounds. Communications in Mathematical
Physics 208: 623.
B
Backlund Transformations
showed that four such solutions can be related in an
algebraic way:
0
0
~
~ w
a1 a2
w w
w
tan
4
tan
4
4
a1 a2
Introduction
Backlund transformations appeared for the first time
in the work of the geometers of the end of the
nineteenth century, for instance, Bianchi, Lie,
Backlund, and Darboux, when studying surfaces
of constant curvature. If on a surface in threedimensional Euclidean space, the asymptotic directions are taken as coordinate directions, then the
surface metric may be written as
ds2 dx2 2 cosw dx dy dy2
2
3b
a2
a1
~
w
w
a2
1
where w(x, y) is a function of the surface coordinates x, y. A necessary and sufficient condition for
the surface to be of constant curvature is that w
satisfies the nonlinear partial differential equation
w;xy sinw
5
a1
~
w
~ = 4 arctan 1 e[axy/a]
w
1 + e[axy/a]
w=
6
7
by assuming that
~ w
~ ;x ; w
~ ;y
w;x f w; w;
~ w
~ ;x ; w
~ ;y
w;y gw; w;
8
9
~ ;x
w;y w
10
11
12
m1
~; u
~; u
~;...; u
~ 0
F2 x; u
1 2
m2
13b
s1
j 1; 2; . . . ; n
s2
14
Backlund Transformations
15
M ;
L @x2 ux; t
M 4@xxx 3u@x @x u
16a
16b
17
18
x!1
19
where pn are the bound state parameters corresponding to isolated singularities of the reflection
coefficients on the imaginary positive k-axis corresponding to a solution n (x, t; pn ) of the spectral
problem vanishing for x ! 1 and such that
lim epn x n x; t; pn 1
x!1
22a
;t Vux; t; k
22b
23
2u 4k2
20
243
21
25a
~ ~ k2 ~
L
25b
~D
26
implies that LD
= Dk2 , that is,
~ DL
LD
27
28b
i
;x x; t; k
k ip
Fx x; t
x; t; k
Fx; t
Fx;t;xx
1
2
2
k
k2 2
2Fx;t
Fx x;t
x;t; k
Fx;t
30a
;x x;t; k
30b
~x; t; k
29b
1
~;xx q;xx q
~ q3
q
8
~x qx 2gx 2q
~ q
q
1 ~
q x qx 2
~q
2 q
32
R1
where q = x u0 (y, t) dy with u0 (x, t) = u(x, t)
g(x), the asymptotically bounded part of u(x, t),
and R g(x)
its
asymptotic
behavior,
and
1
q = x u 0 (y, t) dy with u 0 (x, t) = u(x,
t) g(x).
Once the Lax operator L is given, we can obtain
in a constructive way the operators M which
give the admissible nonlinear partial differential
Backlund Transformations
33a
k;t kg4k2 ; t
dRk; t
2ikf 4k2 ; tRk; t
dt
33b
F~
u u G 1 0
33c
2
2
~ t F4k 2ikG4k Rk; t
Rk;
F4k2 2ikG4k2
34a
n0
1
X
n0
n Ln xu;x 2u
38
n0
1
X
2 n1
3
n t4k
R
39
n0
34b
37
33d
245
ux; t ux; t
f x ~
u;x x; t u;x x; tf x ~
Z 1
~
uy; t uy; tf y dy
34c
k;
1
X
n 4k2 n k
40
n0
~
1 2 v0 v
1 2 1/2v0 ~
v
36
41
By a proper natural choice of the constant parameters n and n , one can define two infinite series
of symmetries. The first one is obtained by choosing
n = 0 and n =
n, m with m = 1, 2, . . . , 1 and can
be denoted as the isospectral series as k, = 0. This is
formed by commuting symmetries. The second one
is given by n = 0 and n =
n, m with m = 1, 2, . . . , 1
and can be denoted as the nonisospectral series as
k, 6 0. The nonisospectral symmetries have a
nonzero commutation relation among themselves
and with the isospectral ones.
exp 2ik
1
X
)#
2 n
n 4k
42
n0
43
Defining R(k,
t) = R(k, t; ), eqn [42] is equal to
eqn [43] iff
n
2
;
2n1 2n 1
n 0; 1; . . . ; 1
44
45
wn 1; t wn; t
46
2
a discrete integrable differentialdifference approximation to the sG equation (Hirota 1977, Orfanidis 1978).
As the nonlinear superposition formulas are
purely algebraic relations involving potentials associated with integrable nonlinear partial differential
equations, one can interpret them as difference
difference equations. In the case of the sG equation
from eqn [7], we have
wn1;m1 wn;m
a1 a2
wn;m1 wn1;m
4 arctan1
tan
a1 a2
4
47
t) = wn1, m , w0 (x, t) =
where
w(x, t) = wn, m , w(x,
0
wn, m1 , and w
(x, t) = wn1, m1 . In a similar manner,
from [36], one gets
vn1;m1 vn;m
1 2 vn1;m vn;m1
48
1 2 12 vn1;m vn;m1
a2
4
gives back eqn [2] (Rogers and Schief 1997). It is
worth mentioning that one can also use known
nonlinear lattice equations to construct BT for
nonlinear partial differential equations (Levi 1981).
See also: Integrable Systems and Discrete Geometry;
Integrable Systems: Overview; Painleve Equations;
Solitons and KacMoody Lie Algebras; Toda Lattices.
Further Reading
Backlund AV (1874) Einiges uber Curven und Flachentransformationen. Lund Universitets Arsskrift 10: 112.
Bianchi L (1879) Ricerche sulle superficie a curvatura costante e sulle
elicoidi. Annali della R. Scuola normale superiore di Pisa 2: 285.
Bruschi M and Ragnisco O (1980a) Existence of a Lax pair for
any member of the class of nonlinear evolution equations
associated to the matrix Schrodinger spectral problem. Lettere
al Nuovo Cimento 29: 321326.
Bruschi M and Ragnisco O (1980b) Extension of the Lax method
to solve a class of nonlinear evolution equations with
x-dependent coeffcients associated to the matrix Schrodinger
spectral problem. Lettere al Nuovo Cimento 29: 327330.
Bruschi M and Ragnisco O (1980c) Backlund transformations
and Lax technique. Lettere al Nuovo Cimento 29: 331334.
Chiu S-C and Ladik JF (1977) Generating exactly soluble
nonlinear discrete evolution equations by a generalized
BatalinVilkovisky Quantization
Wronskian technique. Journal of Mathematical Physics
18: 690700.
Clarin J (1903) Sur quelques equations aux derivees partielles du
second ordre. Annales de la Facult des Sciences de Toulouse pour
les Sciences Mathmatiques et les Sciences Physiques. Serie 2
5: 437458.
Coley A, Levi D, Milson R, Rogers C, and Winternitz P (eds.) (2001)
Backlund and Darboux transformations. The Geometry of
solitons. Proceedings of the AARMS-CRM Workshop, Halifax,
NS, June 49, 1999. CRM Proceedings and Lecture Notes, vol.
29. Providence, RI: American Mathematical Society.
Faddeev LD and Takhtajan LA (1987) Hamiltonian Methods in
the Theory of Solitons. Berlin: Springer.
Fokas AS (1980) A symmetry approach to exactly solvable evolution
equations. Journal of Mathematical Physics 21: 13181325.
Gardner CS, Greene JM, Kruskal MD, and Miura RM (1967)
Method for solving the Kortewegde Vries equation. Physical
Review Letters 19: 10951097.
Hirota R (1978) Nonlinear partial difference equations. III.
Discrete sine-Gordon equation. Journal of the Physical Society
of Japan 43: 20792086.
Ibragimov NH and Shabat AB (1980) Infinite LieBcklund algebras
(in Russian). Funktsional. Anal. i Prilozhen 14: 7980.
Lamb GL (1976) Backlund transformations at the turn of
the century. In: Miura RM (ed.) Backlund Transformations,
pp. 6979. Berlin: Springer.
Lax PD (1968) Integrals of nonlinear equations of evolution and
solitary waves. Communications in Pure and Applied Mathematics 21: 647690.
Levi D (1981) Nonlinear differential difference equations as
Backlund transformations. Journal of Physics A: Mathematical
and General 14: 10831098.
Levi D (1988) On a new Darboux transformation for the
construction of exact solutions of the Schroedinger equation.
Inverse Problems 4: 165172.
247
BatalinVilkovisky Quantization
A C Hirshfeld, Universitat Dortmund,
Dortmund, Germany
2006 Elsevier Ltd. All rights reserved.
Introduction
The BatalinVilkovisky formalism for quantizing
gauge theories has a long history of development. It
begins with the FaddeevPopov procedure for
quantizing YangMills theory, involving the Faddeev
Popov ghost fields (Faddeev and Popov 1967). It
continued with the discovery of BRST symmetry by
Becchi et al. (1976). Then Zinn-Justin (1975)
introduced sources for these transformations, and
a symmetric structure in the space of fields and
sources in his study of renormalizability of these
theories. Finally, Batalin and Vilkovisky (1981)
systematized and generalized these developments.
A more detailed account of this history can be
found in Gomis et al. (1994), where many worked
1
1 or 2 or . . . m
2
3
4
5
T
mod 2
10
12
13
and
ij
E i j mod 2
14
16
cyclic123
where
3A Tk
Rk T
T
1
Tk
Rk T
T
1 Tk
Rk T
T
6
7
ij
ji
and
3Bji Ejik Rk Eji T
1i
j
kj
j i i
R;k Eki
R;k E
1
1 ! ! 1
! !
As in the familiar FaddeevPopov procedure, it is
useful to introduce ghost fields C with opposite
Grassmann parities to the gauge parameters " :
C 1 mod 2
17
BatalinVilkovisky Quantization
T1 2 3 4 ...
ji
B; B 2
@B @B
@A @A
29
30
X; X; X 0
31
19
and for any X
18
249
20
za fA ; A g;
ghi 0
ghC 1
gh A ghA 1
22
32
@r X ab @l Y
!
@za
@zb
33
with
!
a 1; . . . ; 2N
ab
0
BA
BA
0
34
@r X @l Y
@r X @l Y
@A @A @A @A
24
26
27
28
@S @S
0
@A @A
36
a
@S
z
a0
37
@z
Given a classical solution 0 of S0 one stationary
point is
i i0 ;
Ca 0;
A 0
38
39
with
@l @r S
@zc @zb
Rab !ac
40
Rac Rab
0
41
@l @r S
r rank a b
42
@z @z
We then have r N. The relevant solutions of the
classical master equation are those for which r = N.
In this case the number of independent gauge
invariances of the type in eqn [39] equals the number
of antifields. When at a later stage the gauge is fixed,
the nonphysical antifields are eliminated.
To ensure the correct classical limit, the proper
solution must contain the classical action S0 in the
sense that
S A ; A
0 S0 i
43
A
44
@
@A
45
46
47
ghC 1;
gh
0
48
49
50
51
54
BatalinVilkovisky Quantization
@ @
@A @A
55
i
hW
57
58
59
60
251
62
@
@xa
63
Considering the commutator of two gauge transformations leads to (see eqns [8][11])
Z
2Pmi ;j Pnj Pji Pmn ;j Cm Cn 0
ZM
70
2Pjk i Dlj Pmk ;ij Am Pjl
M
kl
j
kl
Dm
i P ;m D X P ;ji Cl Ck 0
The Jacobi identity is
Pij ;m Pmk Ci Cj Ck 0
1
j
Ai Di Cj Xi Pji XCj Ci Pjk ;i XCj Ck
2
1 i j
kl
A A P ;ij XCk Cl
73
4
The gauge-fixing conditions are taken to be of the
form i (A, X), so that the gauge fermion [50] becomes
i i (A, X). The antifields are then fixed to be
=C
@j A; X
@Ai
j @j A; X
Xi C
@Xi
Ci 0
i A; X
C
j
Ai C
65
66
and
71
74
ij
@ X P Aj D X 0
67
68
@k A; X j
k @k A; X Pij Cj
Di Cj C
@Ai
@Xi
1 m @m A; X n @n A; X
C
C
Pkl ;ij X
4
@Ai
@Aj
Ck Cl
i i A; X
75
k
C
76
Bethe Ansatz
1
i @ C
j Pkl ;ij X
@ C
4
Ck Cl i @ Ai
77
78
79
80
253
Further Reading
Alexandrov M, Kontsevich M, Schwarz A, and Zaboronsky O
(1997) Geometry of the Master Equation. International
Journal of Modern Physics A12: 14051430.
Axelrod S and Singer IM (1991) ChernSimons Perturbation
Theory, Proceedings of the XXth Conference on Differential
Geometric Methods in Physics, Baruch College/CUNY, NY.
(hep-th/9110056).
Batalin IA and Vilkovisky GA (1977) Gauge algebra and
quantization. Physics Letters 69B: 309312.
Becchi C, Rouet A, and Stora R (1976) Renormalization of gauge
theories. Annals of Physics (NY) 98: 287321.
Cattaneo AS and Felder G (2001) On the AKSZ formulation of
the PoissonSigma model. Letters of Mathematical Physics
56: 163179.
Faddeev LD and Popov VN (1967) Feynman diagrams for the
YangMills field. Physics Letters 25B: 2930.
Gomis J, Paris J, and Samuel S (1994) Antibracket Antifields and
gauge-theory quantization. Physics Reports 269: 1145.
Hirshfeld AC and Schwarzweller T (2000) Path integral quantization of the PoissonSigma model. Annals of Physics (Leipzig)
9: 83101.
Schaller P and Strobl T (1994) Poisson structure induced
(topological) field theories. Modern Physics Letters A9:
31293136.
Witten E (1988) Topological sigma models. Communications in
Mathematical Physics 118: 411449.
Zinn-Justin J (1975) Renormalization of gauge theories. In:
Rollnik H and Dietz K (eds.) Trends in Elementary
Particle Physics, Lecture Notes in Physics, vol. 37. Berlin:
Springer.
Bethe Ansatz
M T Batchelor, Australian National University,
Canberra, ACT, Australia
2006 Elsevier Ltd. All rights reserved.
Introduction
The Bethe ansatz is a particular form of wave function
introduced in the diagonalization of the Heisenberg
spin chain. It underpins the majority of exactly solved
models in statistical mechanics and quantum field
L1
X
i1
hi;i1 hL;1 ;
hij 12 s i s j 1
1
hji ji
2
hji ji;
hji ji
E L 2 2 cos k
6
8
Diagonalization of Sectors
ax 1; x 1 ax; x 2
Case 1: n = 0
L
X
axj xi
3
x1
11
j x 1i
ikx
gives
13
Eax L 2ax ax 1 ax 1
12
Hj xi L 2j xi j x 1i
10
9
5
A12
1 2 eik1 eik1 k2
A21
1 2 eik2 eik1 k2
14
A12
A21
and
eik2 L
A21
A12
15
Bethe Ansatz
255
Case 4: n = 3
24
25
sij 1 2zj zi zj
Eax; y; z L 6ax; y; z ax 1; y; z
ax; y 1; z ax; y; z 1
By a(x 1, y, z), we mean
a(x 1, y, z), etc.
2. y = x 1 and z > y 1, with
17
a(x 1, y,z)
Eax; y; y 1
Eax; x 1; x 2 L 2ax 1; x 1; x 2
20
29
ax; x 1; x 3
28
L 4ax; y; y 1 ax 1; y; y 1
ax; y 1; y 1 ax; y; y 2
27
Eax; x 1; z
L 4ax; x 1; z ax 1; x 1; z
ax; x 2; z ax; x 1; z 1
26
23
30
31
zL1
32
33
34
35
spj ;pi
36
1i<jn
n
Y
sp ;p
or zLj n1
s
2 p1 ;p
n
Y
s;j
1
6j
sj;
37
n
X
2 cos kj 2
38
j1
uj 1/2i
uj 1/2i
39
and
n
X
u2
j1 j
1
1/4
with uj satisfying
n
Y
uj 1/2i L
uj u i
uj 1/2i
u u i
1 j
40
43
N
X
X
@2
2c
xi xj
@x2i
i1
1i<jN
44
45
EL
N
Y
kj k ic
1
which gives
42
Ap1 ;...;pn
Ap1 ;...;pn P
meeting conditions can be handled in terms of twobody interactions. To see this more clearly, the six
permutation pair equations [29] can be written in
the general form Aabc = Yab Abac and Aabc = Ybc Aacb ,
where Yab = sba =sab . Now there are two possible
paths to get from Aabc to Acba , namely
kj k ic
j 1; . . . ; N
46
N
X
k2j
47
j1
41
for j = 1, . . . , n.
All eigenvalues of the Heisenberg spin chain may
be obtained in terms of the Bethe ansatz solution.
For example, the distribution of roots uj for the
ground state are real and symmetric about the
origin. Excitations may involve complex roots.
Although obtained exactly in terms of the Bethe
roots, the Bethe ansatz wave function is
cumbersome.
We have thus seen how the Bethe ansatz works
for the Heisenberg spin chain. The underlying
mechanism is the way in which the collision or
BF Theories 257
Further Reading
Baxter RJ (1983) Exactly Solved Models in Statistical Mechanics.
London: Academic Press.
Baxter RJ (2003) Completeness of the Bethe ansatz for the sixand eight-vertex models. Journal of Statistical Physics
108: 148.
Bethe HA (1931) Zur Theorie der Metalle I. Eigenwerte und
Eigenfunktionen der linearen Atomkette. Zeitschrift fur Physik
71: 205226.
Gaudin M (1967) Un Syste`me a` Une Dimension de Fermions en
Interaction. Physics Letters A 24: 5556.
Gaudin M (1983) la Fonction donde de Bethe. Paris: Masson.
Korepin VE, Izergin AG, and Bogoliubov NM (1993) Quantum
Inverse Scattering Method and Correlation Functions.
Cambridge: Cambridge University Press.
Lieb EH and Liniger W (1963) Exact analysis of an interacting
Bose gas I. The general solution and the ground state. Physical
Review 130: 16051616.
Mattis DC (1993) The Many-Body Problem: An Encyclopaedia of
Exactly Solved Models in One-Dimension. Singapore: World
Scientific.
McGuire JB (1964) Study of exactly soluble one-dimensional
N-body problems. Journal of Mathematical Physics
5: 622636.
Sutherland B (2004) Beautiful Models: 70 Years of Exactly Solved
Quantum ManyBody Problems. Singapore: World Scientific.
Takahashi M (1999) Thermodynamics of One-Dimensional
Solvable Models. Cambridge: Cambridge University Press.
Yang CN (1967) Some exact results for the many-body problem
in one-dimension with repulsive Delta-function interaction.
Physical Review Letters 19: 13121315.
BF Theories
M Blau, Universite de Neuchatel, Neuchatel,
Switzerland
2006 Elsevier Ltd. All rights reserved.
Introduction
BF theories are a class of gauge theories with a
nontrivial metric-independent classical action. As
such these theories are candidate topological field
theories akin to the ChernSimons theory in three
dimensions, but in contrast to the ChernSimons
theory these exist and are well defined in arbitrary
dimensions.
The name BF theories derives from the fact
that, roughly (see [1] below and the subsequent
discussion for a more precise description),
the action
R
of the BF theory takes the form B ^ FA with FA the
curvature of a connection A and B a Lagrange
multiplier. The classical equations of motion imply
258 BF Theories
The classical action and equations of motion TypiTypically, the classical action of the BF theory takes
the form
Z
SBF A; B
trG B ^ FA
1
M
dA B 0
2
where
dA B dB A; B
6
7
8
one has
3
4
Diffeomorphisms and local symmetries One manifestation of the general covariance of the BF action
[1] is the on-shell equivalence of (infinitesimal)
diffeomorphisms and (infinitesimal) local symmetries. Diffeomorphisms are generated by the Lie
derivative LX along a vector field X. The action of
LX on differential forms is given by the Cartan
formula LX = diX iX d, where i(.) is the operation
of contraction. The action of the Lie derivative on
A and B can be written in gauge covariant form as
LX A iX FA dA X;
LX B iX dA B B; X dA 0 X
9
5
A dA ;
B B;
BF Theories 259
10
11
12
13
14
2
15
260 BF Theories
dA 0np2
20
np1
Cn; p HA M; g HA
M; g
21
p0
TA M
n
p
Y
p 1 p=2
det A
25
p0
dn xn Dx J eiKx
Rn
1
1
eiK:D J
det D
27
22
det A : e
24
BF Theories 261
29
where denotes collectively all the fields. Concretely, when n = 2 and p = 0 (or, equivalently, p = 1),
the quantum action is
Z
0
q
SA 2; 0 B0 dA C1 dA ? C1 c ? A c 35
Likewise, for n = 3 and p = 1 (the only other case
when the gauge symmetry is indeed irreducible),
both B1 and C1 require separate gauge fixing, and
the quantum action is
Z
0
q
SA 3; 1 B1 dA C1 dA ? C1 c ? A c
0
0 dA ? B1 c0 ? A c0
30
32
F det A
33
det 0
det DA
37
38
39
The partition function Following the finite-dimensional model, both the -function implementing the
gauge-fixing condition and the FaddeevPopov
determinant can be lifted into the exponential, the
former by a Lagrange multiplier [26], a Grassmann
even 0-form, and the latter by a pair of Grassmann
odd 0-forms c and c [28], the ghost and antighost
fields, respectively. The sum of the classical action
and these gauge-fixing and ghost terms defines the
q
(BRST-invariant) quantum action SA (n, p), and the
partition function is
Z
q
ZA n; p deiSA n;p
34
36
where
DA
?dA
dA
dA ?
det 0 2
det DA
40
!
: 0 M; g 1 M; g
! 0 M; g 1 M; g
41
42
262 BF Theories
ZA n; p TA n1
43
Since the dimension of is equal to the codimension of S0 = @0 , and S0 will generically intersect
transversally at isolated points, and we define the
linking number of S and S0 to be the intersection
number of and S0 , expressed in terms of de Rham
currents as
Z
Z
LS; S0
S0
S0
46
47
hei WS B ei
WS0 C i ei
LS;S
48
As expected, correlation functions of these topological field theories encode topological information.
BF Theories 263
g2 !0
51
It is invariant under the infinitesimal gauge transformations C = dC , 2 0 (M, h), and the gauge-invariant
equation of motion is the flatness condition FC = 0.
Now let H = TG be the tangent bundle group
TG G s g. This is a semidirect product group
with G acting on g via the adjoint and g regarded
as an abelian Lie algebra of translations. Thus, in
terms of generators (Ja , Pa ), where the Ja are
generators of G, the commutation relations are
c
c
Jc , [Ja , Pb ] = fab
Pc and [Pa , Pb ] = 0, and
[Ja , Jb ] = fab
the curvature of the TG-connection C = Ja Aa Pa Ba is
FC Ja FAa Pa dA Ba
53
SBF A; B
54
55
B B; dA 0
57
264 BF Theories
Relation with Gravity
Wilson loops and string topology has been investigated in Cattaneo et al. (2003).
See also: BatalinVilkovisky Quantization; BRST
Quantization; ChernSimons Models: Rigorous Results;
Gauge Theories From Strings; Knot Invariants and
Quantum Gravity; Loop Quantum Gravity; Moduli
Spaces: An Introduction; Nonperturbative and
Topological Aspects of Gauge Theory; Schwarz-Type
Topological Quantum Field Theory; Spin Foams;
Topological Quantum Field Theory: Overview.
Further Reading
Baez J (2000) An introduction to spin foam models of
quantum gravity and BF theory. Lecture Notes in Physics
543: 2594.
Birmingham D, Blau M, Rakowski M, and Thompson G (1991)
Topological field theory. Physics Reports 209: 129340.
Blau M and Thompson G (1989) A New Class of Topological
Field Theories and the RaySinger Torsion. Physics Letters B
228: 6468.
Blau M and Thompson G (1991) Topological gauge theories
of antisymmetric tensor fields. Annals of Physics
205: 130172.
Blau M and Thompson G (1994) Lectures on 2d gauge theories:
topological aspects and path integral techniques. In: Gava E,
Masiero A, Narain KS, RandjbarDaemi S, and Shafi Q (eds.)
Proceedings of the 1993 Trieste Summer School on High
Energy Physics and Cosmology, pp. 175244. Singapore:
World Scientific.
Carlip S (1998) Quantum Gravity in 2 1 Diemensions. Cambridge:
Cambridge University Press.
Cattaneo A and Rossi C (2001) Higher-dimensional BF theories in
the BatalinVilkovisky formalism: the BV action and generalized Wilson loops. Communications in Mathematical Physics
221: 591657.
Cattaneo A, Cotta-Ramusino P, Fucito F, Martellini M, and
Rinaldi M, et al. (1998) Four-dimensional YangMills theory
as a deformation of topological BF theory. Communications in
Mathematical Physics 197: 571621.
Cattaneo A, Pedrini P, and Frohlich J (2003) Topological field
theory interpretation of string topology. Communications in
Mathematical Physics 240: 397421.
Freidel L, Krasnov K, and Puzio R (1999) BF description of
higher-dimensional gravity theories. Advances in Theoretical
and Mathematical Physics 3: 12891324.
Horowitz GT (1989) Exactly soluble diffeomorphism invariant
theories.
Communications
in
Mathematical
Physics
125: 417437.
Schwarz AS (1978) The partition function of a degenerate
quadratic functional and RaySinger Invariants. Letters in
Mathematical Physics 2: 247252.
Schwarz AS (1979) The partition function of a degenerate
functional. Communications in Mathematical Physics
67: 116.
Witten E (1989) Quantum field theory and the Jones
polynomial. Communications in Mathematical Physics
127: 351399.
Witten E (1991) On quantum gauge theories in two dimensions.
Communications
in
Mathematical
Physics
141: 153209.
265
Gravity
Position
Momentum
Curved
Noncommutative
x2 = 12
Introduction
One of the sources of quantum groups is a
bicrossproduct construction coming in the case of
Lie groups from considerations of Planck-scale
physics in the 1980s. This article describes these
objects and their currently known applications. See
also the overview of Hopf algebras which provides
the algebraic context (see Hopf Algebras and
q-Deformation Quantum Groups).
The construction of quantum groups here is
viewed as a microcosm of the problem of quantization in a manner compatible with geometry. Here
quantization enters in the noncommutativity of the
algebra of observables and curvature enters as a
quantum nonabelian group structure on phase
space. Among the main features of the resulting
bicrossproduct models (Majid 1988) are
1. Compatibility takes the form of nonlinear matched
pair equations generically leading to singular
accumulation regions (event horizons or a maximum value of momentum depending on context).
2. The equations are solved in an equal and
opposite form from local factorization of a
larger object.
3. Different classical limits are related by observer
observed symmetry and Hopf algebra duality.
4. Nonabelian Born reciprocity re-emerges and is
linked to T-duality.
It has also been argued that noncommutative
geometry should emerge as an effective theory of the
first corrections to geometry coming from any
unknown theory of quantum gravity. Concrete
models of noncommutative spacetime currently
provide the first framework for the experimental
verification of such effects. The most basic of these
possible effects is curvature in momentum space or
cogravity. We start with this.
Cogravity
We recall that curvature in space or spacetime
means by definition noncommutativity among the
covariant derivatives Di . Here the natural momenta
are pi = i
hDi and the situation is typified by the
top line in Figure 1. There are also mixed relations
between the Di and position functions as indicated
Cogravity
[pi , pj ] = ih
Curved
Noncommutative
[xi , xj ] = 2i
Quantum
mechanics
ijk xk
ijk pk
P2 = 12
[xi , pj ] = ihij
eipx ;
p 2 R3
f p
xi xi i
2
dxi
t; xi ixi ;
xi ; xj 0
d x @ i dxi @ 0
i
pi
tri i
sinjpj
2
jpj
p;p0
eipx eip t ;
p;p0
p0 ;p00
p; p0 1 ep p; p0
which is also the action of the antipode S on the
abstract p generators.
In particular, the right-invariant Haar measure on
B in these coordinates is the usual d4 p so the
dtx x dt idx
@ i :
@0
L : x; t i : ep :
1
0 2
In
P thisi 2case the natural 4D Laplacian is L ((@ )
i (@ ) ), which acts on plane waves as
2
0
coshp0 1 p2 ep
2
where
p2 =
3
X
pi2
i1
267
L
1044 s 100 MeV 1010 y 1 ms
c
x ; x
i
p
eixk
eix:k ;
p
k x
k
k x
k0
ei=2k
0
k
kk0
p ; p
0
269
"
M M G!G
of each group on the set of the other, such that
s3e s;
e"u u;
s"e e;
e3u e
s
u
Ss
s
t u
t u
s (uv )
s u (s u) v
s u
(s u ) v
uv
e u
e
s 1
u
Figure 3 Bicrossproduct Hopf algebra showing horizontal
product and vertical coproduct as an unproduct.
CG. Hence
this, so is a semidirect coalgebra C(M)
the two together are denoted C(M)
CG. The dual
needs G finite and has the same form but with
vertical and horizontal compositions interchanged,
that is, a bicrossproduct CM C(G). Both Hopf
algebras have the above labeled squares as basis.
It is possible to generalize both bicrossproducts
and double cross products associated to matched
pairs to general Hopf algebras H1 H2 and
H1 H2 , respectively, where H1 , H2 are Hopf
algebras (see Majid 1990) and to relate the two in
general by dualization of one factor. Another
general result (Majid 1995) is that H1 H2 acts
covariantly on the algebra H1 from the right, or
H1 H2 acts covariantly on H2 from the left. A
third general result is that bicrossproducts solve the
extension problem
H1 ! H ! H2
meaning that such a Hopf algebra H subject to some
technical requirements (such as an algebra splitting
map H2 ! H) is of the form H H1 H2 . The
theory was also extended to include cocycle bicrossproducts at the end of the 1980s (by the author).
The finite group case, however, was first found by
Kac and Paljutkin (1966) in the Russian literature
and later rediscovered independently in Takeuchi
(1981) and in the course of Majid (1988).
e u = e u e
u
= s,e
u
s (t u)
u 1
s (t u)
s (uv)
(st ) u
ab = s b
s u
(st ) u
s
t
1=
uv
u s
s
=s
v
s e
s
e
s e
= s s
e
@
;
@x
^
x
i
@
1 eh p
h
@p
271
x
Sx x;
x x 1 1 x
1 p;
Sp pe
x p 0
x
p
1
x
x_ 1 e v1 1
m
1 x
at the classical level. We see that the particle takes
an infinite time to reach the origin, which is an
accumulation point. This can be compared with the
formula in standard radial infalling coordinates
!
1
x_ v1 1
c2 x
1 2GM
for distance x from the event horizon of a black hole
of mass M (here G is Newtons constant and c the
speed of light). So c2=GM and for the sake of
further discussion we will use this value. With a
little more work, one can then see that
mM
m2P
Cx
!
Cp
!
mM m2P
where mP is the Planck mass of the order of 105 g
and X = R
R is a nonabelian group. In the first
limit, the particle motion is not detectably different
from usual flat space quantum mechanics outside
the Compton wavelength from the origin. In the
second limit, the estimate is such that noncommutativity would not show up for length scales much
larger than the background curvature scale.
This Hopf algebra is also the simplest way to
extend classical position C[x] and momentum C[p]
in the sense above. In other words, requiring to
maintain observerobserved symmetry or Born
reciprocity throws up both quantum mechanics (in
the form of h) and something with the flavor of
Bicrossproduct Poincare
Quantum Groups
Another example from the 1980s in the same family
as the Planck-scale Hopf algebra is G = SU2 and
M = B , a nonabelian version of R3 with Lie algebra
b of the form
x3 ; xi ixi ;
xi ; xj 0
Mi ; Mj iij k Mk
M3 ; pj i3j k pk ;
Mi ; p3 ii3 k pk
Mi ; Mj iij k Mk
Ni ; Nj iij k Mk ;
p0 ; Mi 0;
Mi ; Nj iij k Nk
pi ; Mj ii jk pk ;
p0 ; Ni ipi
pi pi 1 ep pi
and the usual additive coproducts on p0 , Mi . This
time the Lorentz group orbits in B are deformed
hyperboloids rather than deformed spheres, and the
Casimir that controls this has the same form as
above but with in the cosh term, that is, the
model is a Lorentzian one. We know from the
general theory of bicrossproducts that this Hopf
3
algebra acts on U(b ) = R 1,
the spacetime in the
section Cogravity, and the Casimir induces the
wave operator as we have seen there.
Let us look a bit more closely at the deformed
hyperboloids. Because neither group here is compact, one expects from the general theory of
bicrossproducts to have limiting accumulation
regions. This is visible in the contour plot of p0
against jpj in Figure 4, where the p0 > 0 mass shells
are now cups with almost vertical walls, compressed
into the vertical tube
jpj < 1
In other words, the 3-momentum is bounded above
by the Planck momentum scale (if is the Planck
time). Indeed, the light-cone equation (setting the
Casimir to zero) reads jpj = 1 ep3 so this is
2
2
PoissonLie T-Duality
We have explained in Section 3 that the matched
pair equations are equivalent to a local factorization
of Lie groups, with the action and back-reaction
created equally and oppositely from this. For the
two models in the last section, these are SL2 (C)
factorizing as SU2 and a 3D B , and SO2, 3 locally as
SO1, 3 and a 4D B . The first of these examples is in
fact one of a general family based on the Iwasawa
decomposition GC = G G where G is a compact
Lie group with complexification GC and G a
certain solvable group. From this, one may construct
a solution (G, G ) of the matched pair equations and
bicrossproduct quantum group
CG
Ug
273
Other Bicrossproducts
There are also infinite-dimensional factorizations
such as the RiemannHilbert problem (see
RiemannHilbert Problem) in the theory of
integrable systems and hence infinite-dimensional
matched pairs and bicrossproducts linked to
H1
n ; y nn ;
1 1 1 1 1
x x 1 1 x 1 y;
y y 1 1 y
which we see has a semidirect product form where
n 3x = n1 , n 3y = nn . The coalgebra is also a
semidirect coproduct by means of a back-reaction of
H1 in B (expressed as a coaction). From the
bicrossproduct theory, we also have a dual model
CB
Udiff 0
Further Reading
Amelino-Camelia G and Majid S (2000) Waves on noncommutative spacetime and gamma-ray bursts. International Journal
of Modern Physics A 15: 43014323.
Beggs E and Majid S (2001) PoissonLie T-duality for quasitriangular Lie bialgebras. Communications in Mathematical
Physics 220: 455488.
Connes A and Moscovici H (1998) Hopf algebras, cyclic
cohomology and the transverse index theory. Communications
in Mathematical Physics 198: 199246.
Kac GI and Paljutkin VG (1966) Finite ring groups. Transactions
of the American Mathematical Society 15: 251294.
Kempf A, Mangano G, and Mann RB (1995) Hilbert space
representation of the minimal length uncertainty relation.
Physical Review D 52: 11081118.
Klimcik C (1996) PoissonLie T-duality. Nuclear Physics B (Proc.
Suppl.) 46: 116121.
Lukierski J, Nowicki A, Ruegg H, and Tolstoy VN (1991)
q-Deformation of Poincare algebra. Physics Letters B
268: 331338.
Majid S (1988) Hopf algebras for physics at the Planck scale.
Journal of Classical and Quantum Gravity 5: 15871606.
Majid S (1990) Physics for algebraists: non-commutative and
non-cocommutative Hopf algebras by a bicrossproduct
construction. Journal of Algebra 130: 1764.
Majid S (1990) Matched pairs of Lie groups associated to
solutions of the YangBaxter equations. Pacific Journal of
Mathematics 141: 311332.
Majid S (1990) On q-regularization. International Journal of
Modern Physics A 5: 46894696.
Majid S (1995) Foundations of Quantum Group Theory.
Cambridge: Cambridge University Press.
Majid S (2000) Meaning of noncommutative geometry and the
Planck-scale quantum group. Springer Lecture Notes in
Physics 541: 227276.
Majid S and Ruegg H (1994) Bicrossproduct structure of the
-Poincare group and non-commutative geometry. Physics
Letters B 334: 348354.
Oeckl R (2000) Untwisting noncommutative Rd and the
equivalence of quantum field theories. Nuclear Physics B
581: 559574.
Seiberg N and Witten E (1999) String theory and noncommutative geometry. Journal of High Energy Physics 9909: 032.
Snyder HS (1947) Quantized space-time Physical Review D
67: 3841.
Takeuchi M (1981) Matched pairs of groups and bismash products
of Hopf algebras. Communications in Algebra 9: 841.
Bifurcation Theory
275
Bifurcation Theory
M Haragus, Universite de Franche-Comte, Besancon,
France
G Iooss, Institut Non Lineaire de Nice, Valbonne,
France
2006 Elsevier Ltd. All rights reserved.
Introduction
Consider the following equation:
FX; 0
1
2
b2 4c
For < 0, there is no real solution of [2], while
there are two solutions x in the region > 0,
which merge when the distance between the point
(b, c) and the parabola = 0 tends towards 0. It is
then clear that a singularity occurs in the structure
of the set of solutions of [2] at the crossing of the
parabola = 0 or, in other words, a bifurcation
occurs in the parameter space (b, c) on the parabola
= 0. A point (0 , x0 ) 2 R 3 is then called a
bifurcation point if 0 = (b, c) satisfies = 0, and
x0 = b=2.
In the theory of differential equations, F(X, )
often represents a vector field. This study is then
concerned with the existence of equilibrium solutions to the differential equation
dX
FX;
dt
3
4
5
For example, if
@ f x0 ; 0 6 0;
it is possible to solve with respect to and obtain a
regular solution (x) such that (x0 ) = 0 and
f (x, (x)) 0. In addition, if the second order
derivative
( 0, 0)
@x2 f x0 ; 0 6 0
the picture of the solution set in the plane (, x), also
called bifurcation diagram, shows a turning point
with a fold opened to the left or to the right
depending upon the sign of the product @ f (x0 , 0 )
@x2 f (x0 , 0 ); see Figure 1. Notice that here the
bifurcation point (0 , x0 ) 2 R2 corresponds to the
appearance of a pair of solutions of [4] from
nowhere. This is the simplest example of a onesided bifurcation in which the bifurcating solutions
exist for either > 0 or < 0 .
A particularly interesting situation arises when the
equation possesses a symmetry. For example, assume
that in [4] the function f is odd with respect to x. This
implies that we always have the solution x = 0, for any
value of the parameter . Assume now that f satisfies
@x f 0; 0 0
6
and that
2
f 0; 0 6 0;
@x
@x3 f 0; 0 6 0
7
8
( 0, x 0)
(a)
(b)
(c)
Bifurcation Theory
277
X0 PX; X1 id PX
Xc PX; Xh id PX
Bifurcation Theory
9
10
Dy Ny; L y L Ny;
for all y and . This characterization allows to determine
the classes of possible normal forms for a given matrix L,
and also provides an efficient way to compute the
normal form for a given vector field g. As for the
reduction methods, normal form transformations can be
made to preserve the additional structure of the
problem, such as Hamiltonian structure or symmetries.
As an example, consider a differential equation of
the form [9] with x 2 Rn and 2 R, which supports a
Hopf bifurcation so that L has simple eigenvalues
i!, ! > 0, and no other eigenvalues with zero real
279
provides a
linear part
which it is
in complex
2k2
2
dA
i!A AQ A ; o A
dt
for A(t) 2 C, where Q is a complex polynomial of
degree k in jAj2 with Q(0, 0) = 0, or, equivalently, in
polar coordinates A = rei ,
dr
rQr r 2 ; o r 2k2
dt
d
! Q r 2 ; o r 2k1
dt
Qr and Q being the real and imaginary part of Q,
respectively. The radial equation for r truncated at
order 2k 1 decouples and admits a pitchfork bifurcation. The bifurcating steady solutions of this equation
then lead first to periodic solutions for the truncated
system, which are then shown to persist for the full
equation by a standard perturbation analysis.
A situation that occurs in a large class of problems
is when the problem possesses a reversibility
symmetry, which often comes from some reflection
invariance in the physical space, that is, when the
vector field F( , ) anticommutes with a symmetry
operator S. One of the simplest examples is the case
of a differential equation [9] when the matrix L has
a double eigenvalue in 0, no other eigenvalues with
zero real part, and a one-dimensional kernel which
is invariant by S. In this case, the center manifold
reduction provides a two-dimensional reduced reversible system, which can be put in the normal form
da
b
dt
db
a2 ojaj jbj3
dt
which
anticommutes
with
the
symmetry
(a, b) 7! (a, b). The above system undergoes a
reversible TakensBogdanov bifurcation and has
for > 0 a phase portrait as in Figure 5. There are
two equilibria, one a saddle, the other a center, and
a family of periodic orbits with the zero-amplitude
limit at the center equilibrium, and the infiniteperiod limit a homoclinic orbit, originating at the
saddle point. In concrete problems the bounded
orbits of such a reduced system determine the shape
of physically interesting solutions of the full system
of equations, such as, for example, in water-wave
theory where to homoclinic and periodic orbits
correspond solitary and periodic waves, respectively.
Further Reading
Arnold VI (1988) Geometrical Methods in the Theory of
Ordinary Differential Equations. Grundlehren der Mathematischen Wissenschaften, vol. 250. New York: Springer.
Buffoni B and Toland J (2003) Analytic Theory of Global
Bifurcation. Princeton: Princeton University Press.
Chossat P and Lauterbach R (2000) Methods in Equivariant
Bifurcations and Dynamical Systems. Advanced Series in
Nonlinear Dynamics, vol. 15. River Edge, NJ: World
Scientific.
Chow S-N and Hale JK (1982) Methods of Bifurcation Theory.
Grundlehren der Mathematischen Wissenschaften, vol. 251.
New York: Springer.
Golubitsky M and Schaeffer DG (1985) Singularities and Groups
in Bifurcation Theory, Vol. I. Applied Mathematical Sciences,
vol. 51. New York: Springer.
Golubitsky M, Stewart I, and Schaeffer DG (1988) Singularities
and Groups in Bifurcation Theory, Vol. II. Applied Mathematical Sciences, vol. 69. New York: Springer.
Guckenheimer J and Holmes P (1990) Nonlinear Oscillations,
Dynamical Systems, and Bifurcations of Vector Fields. Applied
Mathematical Sciences, vol. 42. New York: Springer.
Iooss G and Adelmeyer M (1998) Topics in Bifurcation Theory
and Applications, Advances Series in Nonlinear Dynamics,
2nd edn., vol. 3, Singapore: World Scientific.
Iooss G, Helleman RHG, and Stora R (eds.) (1983) Chaotic
behavior of deterministic systems. Session XXXVI of the
Summer School in Theoretical Physics held at Les Houches
June 29July 31, 1981. Amsterdam: North-Holland.
281
Introduction
Almost all classical hydrodynamical stability problems
are experiments or gedankenexperiment which have
been designed to understand and to extract special
phenomena in more complicated situations. Examples
are the TaylorCouette problem, Benards problem,
Poiseuille flow, or Kolmogorov flow.
The TaylorCouette problem consists in finding the
flow of a viscous incompressible fluid contained in
between two coaxial co- or counterrotating cylinders,
cf. Figure 1. If the rotational velocity of the inner
cylinder is below a certain threshold, the trivial
solution, called the Couette flow, is asymptotically
stable. At the threshold, this spatially homogenous
solution becomes unstable and bifurcates via a pitchfork bifurcation or a Hopf bifurcation into different
spatially periodic patterns, that is, depending on the
rotational velocity of the outer cylinder the basic
patterns are stationary (called the Taylor vortices) or
@t U U NU
1
Rest of spectrum
283
Es
k
Mc
Ec
Rest of spectrum
Mc fu c1 1 hc1 j
hc1 2 spanf2 ; 3 ; . . .gg
the so-called center manifold which is tangential to Ec ,
that is, kh(c1 )k Ckc1 k2 (Figure 8). The dynamics on
Mc is no longer trivial due to the nonlinear terms.
Due to the fact that real problems are considered
Re1 (kc ) = 0 implies Re1 (kc ) = 0, that is, in case
of 2=kc -periodic boundary conditions always two
eigenvalues cross the imaginary axis simultaneously.
For Benardss problem in a strip or for the Taylor
Couette problem in case of a bifurcation of fixed
points, the reduced system on the center manifold is
derived with the ansatz
U "A"2 teikc x c:c: O"2
where 0 < " 1 is the small bifurcation parameter,
cf. Figure 5. Then due to eikc x eikc x eikc x = eikc x the
complex-valued amplitude A satisfies the so-called
Landau equation
@T A A AjAj2 O"2
where the Landau coefficient 2 R is obtained by
classical perturbation analysis (Figure 9). The
reduced system is symmetric under the S1 -symmetry
Im
Re
Rest of
spectrum
Rest of
spectrum
Modulation Equations
The analysis of the last section is of no use in case of
a sideband instability occurring at the wave number
kc = 0, as it happens in the inclined-plane problem
or in the Kolmogorov flow problem. Moreover, in
case of an instability at a wave vector kc 6 0, based
on the above analysis, front solutions cannot be
described. In such situations, the method of modulation equations generalizes the role of the finitedimensional amplitude equations from the last
section.
The complex cubic GinzburgLandau equation in
normal form is given by
@T A 1 i@X2 A A 1 iAjAj2
where the coefficients , 2 R are real, and we have
X 2 R, T 0, and A(X, T) 2 C. The Ginzburg
Landau equation is a universal amplitude equation
that describes slowly varying modulations, in space
and time, of the amplitude of bifurcating spatially
periodic solutions in pattern-forming systems close
to the threshold of the first instability. Whenever the
instability drawn in Figure 5 occurs, that is, for the
TaylorCouette problem and Benards problem in a
strip, that is, d = 1, it can be derived by a multiple
scaling ansatz
ux; t
"A"x cg t; "2 teikc x!0 t c:c:
For instance, in case of = = 0, the Ginzburg
Landau equation possesses front solutions connecting the stable fixed point A = 1 with the unstable
fixed point A = 0. Such solutions correspond in the
TaylorCouette problem to modulating fronts
Im
Continuous spectrum
Re
Discrete eigenvalues
Figure 11 Spectrum for the flow around an obstacle.
Discussion
There is no satisfactory bifurcation analysis for situations where boundary layers play a role. The most
simple problem is the flow around some obstacle. The
difficulties are according to the fact that due to the
unbounded flow region there is always continuous
spectrum up to the imaginary axis. From the localized
obstacle discrete eigenvalues are created, (cf. Figure 11).
In such a situation, so far there is no mathematical
bifurcation theory available.
See also: Bifurcation Theory; Dynamical Systems in
Mathematical Physics: An Illustration from Water Waves;
Further Reading
Chandrasekhar S (1961) Hydrodynamic and Hydromagnetic
Stability. Oxford: Clarendon.
Chang H-C and Demekhin EA (2002) Complex Wave Dynamics
on Thin Films, Studies in Interface Science, vol. 14. Amsterdam:
Elsevier.
Chossat P and Iooss G (1994) The TaylorCouette Problem,
Applied Mathematical Sciences, vol. 102. Springer.
Chow S-N and Hale J (1982) Methods of Bifurcation Theory,
Grundlehren der Mathematischen Wissenschaften, vol. 251.
Berlin: Springer.
Golubitsky M and Schaeffer DG (1985) Singularities and Groups
in Bifurcation Theory I, Applied Mathematical Sciences,
vol. 51. Berlin: Springer.
Golubitsky M, Stewart I, and Schaeffer DG (1988) Singularities
and Groups in Bifurcation Theory II, Applied Mathematical
Sciences, vol. 69. Berlin: Springer.
Haken H (1987) Advanced Synergetics. Berlin: Springer.
Henry D (1981) Geometric Theory of Semilinear Parabolic Equations, Lecture Notes in Mathematics, vol. 840. Berlin: Springer.
Mielke A (2002) The GinzburgLandau equation in its role as a
modulation equation. In: Fiedler B (ed.) Handbook of Dynamical Systems II, pp. 759834. Amsterdam: North-Holland.
Ruelle D and Takens F (1971) On the nature of turbulence.
Communications in Mathematical Physics 20: 167192.
Temam R (1988) Infinite-Dimensional Systems in Mechanics and
Physics. Berlin: Springer.
Introduction
Bifurcation theory of periodic orbits relates to
modeling of quite diverse subjects. It appeared
classically in the field of celestial mechanics with
the contributions of H Poincare. Van der Pol (1926,
1927, 1928, 1931) observed the frequency-locking
phenomenon in electrical circuits. More recently,
Malkins theory (Malkin 1952, 1956, Roseau 1966)
was used to justify synchronization of weakly
coupled oscillators modeling the electrical activity
of the cells of the sinusal node in the heart. This
article provides the essential mathematical background necessary for existence of frequency locking.
Applications can be found, for instance, in Weakly
Coupled Oscillators.
whose leaves are the points with a given asymptotic phase. The asymptotic phase can be considered
as a coordinate function defined on the
neighborhood S().
If we consider now the particular case of a plane
system, this function can be completed with the
square of the distance function to the orbit into a
coordinate system called the amplitudephase
system and denoted as (, ).
di
fi ; ;
dt
di
Fi ; ; ;
dt
3
i 1; . . . ; m
4
The term oscillator has two meanings. A conservative oscillator is a plane vector field which
displays an open set of periodic orbits. It is said to
be isochronous if all orbits have same period. A
dissipative oscillator is a planar vector field which
displays an attractive limit cycle (attractive periodic
orbit).
We consider N dissipative oscillators:
dxi
f xi ; yi
dt
dyi
gxi ; yi
dt
dyi
gxi ; yi Gi x; y;
dt
5
where i = 1, . . . , m.
The dynamical system obtained by considering the
space of all the variables (xi , yi ), i = 1, . . . , m, displays an invariant torus full of periodic orbits that
we denote by T m (0).
Assume now that the N oscillators are weakly
coupled:
dxi
f xi ; yi Fi x; y;
dt
di
Fi 0; ; 0
dt
2
6
7a
dx
P t x
dt
7b
8
V0 I
X0 I
(t) to
[7a]. To be more specific, one can choose x
be the unique solution of [6] such that
y(0)k = 0, k = m 1, . . . , n, and j (t) solutions of
[7a], such that y(0)k = jk . With these notations,
x (t) is such that
y0k k ;
then,
X1 t V t
The solution of [8] can be written as
Z t
yt Xt y0 Xt
X1 uru du
k 1; . . . ; m
k k0
9
j1
m1sn
11b
14
k
@j
the
15a
so that
j1
with
k 1; . . . ; m
Vjk T jk yj 0
jm1
n Z
X
systems
dx
Pt x qt f x; t;
dt
j1
n
X
nonlinear
13
j0 ;
k 1; . . . m; j 1; . . . ; m
15b
is invertible.
Proceed as in previous section with the coordinate
change x = U (0)1 y. Equation [14] gets transformed into
dy
Qty rt Fy; t;
dt
16
j1
k 1; . . . ; m
17a
X
Ck ; ;
7! x t
Vjk T I j
jm1;...;n
n Z
X
j1
Vjk srj s ds
n Z
X
j1
x x t
k m 1; . . . ; n
17b
m
X
0 18
j j 0 x
j1
; 7! C; ;
19
k = k0 ,
Dfx x t;t
gx t;t; 0 F
;t;
dt
k
j 0
@j
20
which is nonzero.
The implicit-function theorem shows that the
differential equation [14] (and thus [16] as well)
has, for small enough, a unique T-periodic solution
which tends to x0 when tends to 0.
24
Set, furthermore,
rt gx t; t; 0
for = ,
k = m 1, . . . , n , = 0. This is
equal to the product of and the determinant of
@
23
Pt Dfx x t; t;
01 ; . . . ; 0m
of the m equations:
k
n Z
X
j1
Ujk ugj x u; u; 0 du 0;
0
k 1; . . . ; m
25a
such that
21
22
@k
j 0;
@j
k 1; . . . m; j 1; . . . ; m
25b
p 1 ; . . . ; m
Dfx x0 t; t
gx0 t; t; 0
dt
display a family of T-periodic solutions which
depend on m parameters = (1 , . . . ,m ). From
(13), one can write
t 1 1 t m m t
t
j1
28
@ 2 fj @zk @zl
ds
@zk @zl @q @r
0
jkl
2
XZ T
@ zj
d
Ujp s
ds
ds @q @@r
0
j
XZ T
@fj @ 2 zk
Ujp s
ds
@zk @q @r
0
jk
Ujp
X @gj
k
@zk
Ujp
x0 s; s; 0
@gj
x 0 s; s; 0
@
@ 2 fj @zk @zl
ds
@zk @zl @q @r
0
jkl
2
X Z T dUjp
@fj
@ zk
Ujp s
ds 0
ds
@z
@
q @r
k
0
j
is invertible.
Fj
; s; 0
XZ
Ujk sFj
s; s; 0 ds 0;
such that
Lemma 2
@zk @zl
l
k
@q
@q
@k
j 0;
@j
q
"
1 @ 2 fj
Ujp
2 @zk @zl
2
@ zj
d
dt @q @@r
k 1; . . . ; m
@ 2 fj @zk @zl
ds
@zk @zl @q @r
to equations
k
Ujp
#
@gj @zk
ds
@zk @q
0
0
1 ; . . . ; m
n Z
X
qkl
26
1X
q r
2 qrkl
27
n Z
X
j1
Ujp ugj x u; u; 0 du
We can write
Z T
dp
@Ujp
@gj @zk
gj Ujp
ds
dq
@q
@zk @q
0
Note that
d
j X @fj
r gj zt; 0 ; 0
dt
@zr
r
and we obtain
dp
dq
d
j X @fj
r
ds
@zr
0
r
@gj @zk
Ujp
ds
@zk @q
T
@Ujp
@q
2
XZ T
dp
@ fj @gj @zk
Ujp
ds
dq 0
@zk @zl l @zk @q
0
kl
This achieves the proof of the theorem. In the special
case of Hamiltonian systems, in the case of the
peturbations of an isochronous system, the method
explained is equivalent to Mosers averaging theory.
The reader is referred to other articles in this
encyclopedia for a discussion of other aspects of
synchronization, frequency locking, and phase locking.
See also: Bifurcation Theory; Fractal Dimensions in
Dynamics; Integrable Systems: Overview; Isochronous
Systems; LeraySchauder Theory and Mapping Degree;
LjusternikSchnirelman Theory; Singularity and
Bifurcation Theory; Symmetry and Symmetry Breaking in
Dynamical Systems; Synchronization of Chaos; Weakly
Coupled Oscillators.
Further Reading
Integration by parts yields
!
Z T
X @fj
dp
d @Ujp
r
j
ds @q
dq 0
@zr
0
r
Z T
@gj @zk
Ujp
ds
@zk @q
0
From the equation
dUjp X @fk
U 0
dt
@zj kp
k
we deduce that
X @fk @Ujp X @ 2 fk
d @Ujp
@zr
Ukp
dt @q
@zj @q
@zj @zr
@q
k
k
and thus this shows that
ut 14uxxx 6uux
Introduction
At the end of the 1960s, the theory of integrable
systems received a great boost by the discovery
(made by Gardner, Green, Kruskal, and Miura) of
the inverse-scattering method (see Integrable
Systems: Overview). It allows one to reduce the
1
291
3
i 1; . . . ; 2n
4
ij
P
0
I
I
0
qit
h
;
pi
pit
h
;
qi
i 1; . . . ; N
5
utt 13 uxxxx 4u2x 4uuxx
u2 t 2u1x u2xx
6
t
2
u x
u1 x
with h = (u1 )2 (1=9)(u2 )3 u1 u2x (1=3)(u2x )2 , and
0 @x
7
@x 0
is easily seen to be a Poisson operator. Thus, the
Poisson manifold associated with the Boussinesq
equation is the space of periodic C1 functions with
values in R2 . More generally, one can consider the
space Mn of C1 functions from the unit circle S1 to
Rn . If Pij , for i, j = 1, . . . , n, are the entries of a
constant skew-symmetric matrix and ui, x assigns to
the generic function v 2 Mn the value of its ith
components at a fixed point x, then
fui; x ; uj; y g Pij x y
defines a Poisson bracket on Mn . One can also let
the Pij depend on the uk in such a way that they
form Rthe components of a Poisson tensor on Rn . If
H = h dx is a functional on Mn with density h, the
associated Hamiltonian vector field gives rise to the
following system of partial differential equations:
uit
n
X
j1
Pij
h
;
u j
In particular, if n = 2N and
i 1; . . . ; n
8
j; y g gij
f
ui; x ; u
u0 x y ijk
u
ukx x y 10
where gij (
u) are the components of the metric in the
ij
new coordinates and the k are the contravariant
Christoffel symbols related to the usual Christoffel
symbols by
ij
k gil lk
11
Bi-Hamiltonian Formulation
of the KdV Equation
The KdV equation [1] has a lot of remarkable
properties, such as the Lax representation and the
existence of a -function. In this section, we recall a
geometrical feature of KdV, namely, the fact that it
293
Such relations are often called LenardMagri relations. Then the functionals Hk are in involution with
respect to both Poisson brackets. Indeed, for k > j,
one has
fHj ; Hk g0 fHj ; Hk1 g1 fHj1 ; Hk1 g0
Xu P0 dH2
fHk ; Hj g0
X2 P1 dH2
12
13
h2 14ux ;
h3 18uxx u2
1
h4 16
uxxx 4uux
1
uxxxx 6uuxx 5u2x 2u3
h5 32
fF; Gg0 u; c
0
1
f
g
dx dx; 0 ;
2
u
u
0
Z 0 00
g
f
@
f
g
;
dx
u
u
@x
u
u
Z 0 0
1
f
g
g
f
dx
2
u
u
u
u
15
16
ut 3uux cuxxx 0;
ct 0
17
The
corresponding Hamiltonian is H2 = (1=2)
R
(u3 cu2x ) dx. From this (Lie algebraic) point of
view, the compatibility between the two Poisson
brackets follows from the fact that the pencil { , } =
{ , } { , }0 is obtained from the LiePoisson
bracket { , } by applying the translation
dx dx; c
u dx dx; c 7! u
2
Other Examples
In the previous section, we have presented the biHamiltonian structure of the KdV equation and
some of its properties. Now we give two more
examples of equations the Boussinesq equation
and the CamassaHolm equation admitting a
bi-Hamiltonian formulation. We have seen in an
earlier section that the system [6] associated with
the Boussinesq equation [5] is Hamiltonian with
respect to the Poisson structure [7] and the
Hamiltonian
Z
H1 u1 ; u2
u1 2 19 u2 3 u1 u2x 13 u2x 2 dx
ux
295
my sinhy x dy
0
2 sinh1=2
1
0
1
my cosh y x
dy
2
with
A 2@x5 4u2 @x3 6u2x @x2 2u2 2 6u1x 6u2xx @x
3u1xx 2u2xxx 2u2 u2x
and
B 3@x4 3u2 @x2 9u1 6u2x @x 6u1x 3u2xx
It can be obtained by means of the Drinfeld
Sokolov reduction (or also by means of a
bi-Hamiltonian reduction) from the LiePoisson
structure (modified with the cocycle @x ) on the
space of C1 maps from S1 to the Lie algebra of
3 3 traceless matrices. This is the reason why it is
a Poisson structure, compatible with [7]. The system
[6] can be written as
0 1
!
u1t
h2 /u1
@ A P
h2 /u2
u2
t
19
P1 @xxx @x ;
P2 2m@x mx
20
Further Reading
Arnold VI and Khesin BA (1998) Topological Methods in
Hydrodynamics. New York: Springer.
Baszak M (1998) Multi-Hamiltonian Theory of Dynamical
Systems. Berlin: Springer.
Dickey LA (2003) Soliton Equations and Hamiltonian Systems,
2nd edn. River Edge: World Scientific.
Dorfman I (1993) Dirac Structures and Integrability of Nonlinear
Evolution Equations. Chichester: Wiley.
Drinfeld VG and Sokolov VV (1985) Lie algebras and equations
of Kortewegde Vries type. Journal of Soviet Mathematics
30: 19752036.
Dubrovin BA (1996) Geometry of 2D topological field theories.
In: Donagi R et al. (ed.) Integrable Systems and Quantum
Groups (Montecatini Terme, 1993), Lecture Notes in Mathematics, vol. 1620, pp. 120348. Berlin: Springer.
Dubrovin BA, Krichever IM, and Novikov SP (2001) Integrable
systems. I. In: Arnold VI (ed.) Encyclopaedia of Mathematical
Sciences. Dynamical Systems IV, pp. 177332. Berlin: Springer.
Faddeev LD and Takhtajan LA (1987) Hamiltonian Methods in
the Theory of Solitons. Berlin: Springer.
Magri F, Falqui G, and Pedroni M (2003) The method of Poisson
pairs in the theory of nonlinear PDEs. In: Conte R et al. (ed.)
Direct and Inverse Methods in Nonlinear Evolution Equations,
Lecture Notes in Physics, vol. 632, pp. 85136. Berlin: Springer.
Marsden JE and Ratiu TS (1999) Introduction to Mechanics and
Symmetry, 2nd edn. New York: Springer.
Olver PJ (1993) Applications of Lie Groups to Differential
Equations, 2nd edn. New York: Springer.
297
where Sn1 and Bn1 are the unit sphere and the unit
disk in Euclidean spaces.
Integrable Billiards
Let a plane billiard table be an ellipse with foci F1
and F2 . It is known since antiquity that a billiard
ball shot from F1 reflects to F2 . A generalization of
this optical property of the ellipse is the following
theorem: a billiard trajectory inside an ellipse
forever remains tangent to a fixed confocal conic.
More precisely, if a segment of a billiard trajectory
does not intersect the segment F1 F2 , then all the
segments of this trajectory do not intersect F1 F2 and
are all tangent to the same ellipse with foci F1 and F2 ;
and if a segment of a trajectory intersects F1 F2 ,
then all the segments of this trajectory intersect F1 F2
and are all tangent to the same hyperbola with foci
F1 and F2 .
It follows that confocal ellipses are the caustics of
the billiard inside an ellipse. In particular, a
neighborhood of the boundary of such a billiard
table is foliated by caustics. A long-standing
conjecture, attributed to Birkhoff, asserts that if a
neighborhood of a strictly convex smooth boundary
of a billiard table is foliated by caustics, then this
table is an ellipse. This conjecture remains open. The
best result in this direction is a theorem of Bialy: if
almost every phase point of the billiard ball map in a
strictly convex billiard table belongs to an invariant
circle, then the billiard table is a disk.
The multidimensional analogs of the optical
properties of an ellipse are as follows. Consider an
ellipsoid M in Rn given by the equation
x21 x22
x2n
1
a2n
a21 a22
1
a21
X vi xj vj xi 2
j6i
a2j a2i
i 1; . . . ; n
Periodic Orbits
Periodic billiard trajectories inside a planar billiard
table correspond to inscribed polygons of extremal
perimeter length. When counting periodic trajectories, one does not distinguish between polygons
obtained from each other by cyclic permutation or
reversing the order of the vertices. In other words,
one counts the orbits of the dihedral group Dn
acting on n-periodic billiard polygons.
An additional topological characteristic of a
periodic billiard trajectory is the rotation number
defined as follows. Assume that the boundary of a
billiard table is parametrized by the unit circle and
consider a polygon (x1 , x2 , . . . , xn ) inscribed in .
For all i, one has xi 1 = xi ti with ti 2 (0, 1). Since
the polygon is closed, t1 tn 2 Z. This integer,
that takes values from 1 to n 1, is called the
rotation number of the polygon and denoted by .
Changing the orientation of a polygon replaces the
5
3
5
299
4
1
n!1
Length Spectrum
The set of lengths of the closed trajectories in a
convex billiard M is called the length spectrum of M.
There is a remarkable relation between the length
spectrum and the spectrum of the Laplace operator
in M with the Dirichlet boundary condition:
Ln l
1
X
ci
n2i
i1
Acknowledgments
This work was partially supported by NSF.
See also: Adiabatic Piston; Hamiltonian Systems:
Obstructions to Integrability; Hyperbolic Billiards;
Integrable Discrete Systems; Integrable Systems and
Algebraic Geometry; Optical Caustics; Integrable
Systems: Overview; Polygonal Billiards; Semiclassical
Spectra and Closed Orbits; Separatrix Splitting; Stability
Theory and KAM.
Further Reading
Chernov N and Markarian R, Theory of Chaotic Billiards
(to appear).
Farber M and Tabachnikov S (2002) Topology of cyclic
configuration spaces and periodic orbits of multi-dimensional
billiards. Topology 41: 553589.
Introduction
Over the last 30 years, black holes have been
shown to have a number of surprising properties.
These discoveries have revealed unforeseen relations
between the otherwise distinct areas of general
relativity, quantum physics, and statistical
mechanics. This interplay, in turn, led to a number
of deep puzzles at the very foundations of physics.
Some have been resolved while others continue to
baffle physicists. The starting point of these
fascinating developments was the discovery of
laws of black hole mechanics by Bardeen,
Bekenstein, Carter, and Hawking. They dictate the
behavior of black holes in equilibrium, under small
perturbations away from equilibrium, and in fully
dynamical situations. While they are consequences
of classical general relativity alone, they have a
close similarity with the laws of thermodynamics.
The origin of this seemingly strange coincidence lies
in quantum physics. For further discussion,
see Asymptotic Structure and Conformal Infinity;
Loop Quantum Gravity; Quantum Geometry and
Its Applications; Quantum Field Theory in Curved
Spacetime; Stationary Black Holes.
The focus of this article is just on black hole
mechanics. The discussion is divided into three parts.
In the first, we will introduce the notions of event
horizons and black hole regions and discuss properties
Global Equilibrium
To capture the intuitive notion that black hole is a
region from which signals cannot escape to the
asymptotic part of spacetime, one needs a precise
definition of future infinity. The standard strategy is to
use Penroses conformal boundary J . A black hole
region B of a spacetime (M, gab ) is defined as B = Mn
I (J ), where I denotes chronological past. The
boundary @B of the black hole region is called the
event horizon and denoted by E. Thus, E is the
boundary of the past of J . It therefore follows that E is
a null 3-surface, ruled by future inextendible null
geodesics without caustics. If the spacetime is globally
hyperbolic, an instant of time is represented by a
Cauchy surface M. The intersection of B with M may
have several disjoint components, each representing a
black hole at that instant of time. If M0 is a Cauchy
surface to the future of M, the number of disjoint
components of M0 [ B in the causal future of M [ B
must be less than or equal to those of M [ B
1
301
2
Shor W
2
4
physically. These considerations call for a replacement of E by a quasilocal horizon which leads to a
first law involving only horizon attributes, and
which can grow only in response to the influx of
energy. Such horizons are discussed in the next two
sections.
Local Equilibrium
The key idea here is drop the requirement that
spacetime should admit a stationary Killing field and
ask only that the intrinsic horizon geometry be time
independent. Consider a null 3-surface in a
spacetime (M, gab ) with a future-pointing normal
field a . The pullback qab := gab of the spacetime
metric to is the intrinsic, degenerate metric of
with signature 0, , . The first condition is that it
be time independent, that is, L qab = 0 on .
Then by restriction, the spacetime derivative operator r induces a natural derivative operator D on .
While D is compatible with qab , that is, Da qbc = 0, it
is not uniquely determined by this property because
qab is degenerate. Thus, D has extra information,
not contained in qab . The pair (qab , D) is said to
determine the intrinsic geometry of the null surface
. This notion leads to a natural definition of a
horizon in local equilibrium. Let be a null, threedimensional submanifold of (M, gab ) with topology
S R, where S is compact and without boundary.
Definition 1 is said to be isolated horizon if it
admits a null normal a such that:
(i) L qab = 0 and [L , D] = 0 on and
(ii) T a b b is a future pointing causal vector on .
On can show that, generically, this null normal field
a is unique up to rescalings by positive constants.
Both conditions are local to . In particular, (M, gab )
is not required to be asymptotically flat and there is no
longer any teleological feature. Since is null and
L qab = 0, the area of any of its cross sections is the
same, denoted by a . As one would expect, one can
show that there is no flux of gravitational radiation or
matter across . This captures the idea that the black
hole itself is in equilibrium. Condition (ii) is a rather
weak energy condition which is satisfied by all
matter fields normally considered in classical general
relativity. The nontrivial condition is (i). It extracts
from the notion of a Killing horizon just a tiny part
that refers only to the intrinsic geometry of . As a
result, every Killing horizon K is, in particular, an
isolated horizon. However, a spacetime with an
isolated horizon can admit gravitational radiation
and dynamical matter fields away from . In fact, as a
family of RobinsonTrautman spacetimes illustrates,
gravitational radiation could even be present arbitrarily close to . Because of these possibilities, there are
many nontrivial examples and the transition from
event horizons of stationary spacetimes to isolated
horizons represents a significant generalization of
black hole mechanics. (In fact, the derivation of the
zeroth and the first law requires slightly weaker
assumptions, encoded in the notion of a weakly
isolated horizon (Ashtekar et al. 2000, 2001).)
An immediate consequence of the requirement
L qab = 0 is that there exists a 1-form !a on such
that Da b = !a b . Following the definition of on a
Killing horizon, the surface gravity () of (, ) is
defined as () = !a a . Again, under a ! ca , we have
(c) = c . Together with Einsteins equations, the
two conditions of Definition 1 imply L !a = 0 and
a D[a !b] = 0. The Cartan identity relating the Lie
and exterior derivative now yields
Da !b b Da 0
6
303
H J J1
I
I
1
1
J
!a a
Aa a ? F
8G S
4 S
7
()
is the ADM angular momentum at
where J1
infinity, S is any cross section of , and the area
()
element thereon. The term J
is independent of the
choice of S made in its evaluation and interpreted as
the horizon angular momentum. It has numerous
properties that support this interpretation. In particular, it yields the standard angular momentum
expression in KerrNewman spacetimes.
To define horizon energy, one has to introduce a
time-translation vector field ta . At infinity, ta must
tend to a unit time translation. On , it must be a
symmetry of qab . Since a and a are both horizon
symmetries, ta = ca a on , for some constants
c and . However, unlike a , the restriction of ta to
cannot be fixed once and for all but must be
allowed to vary from one phase-space point to
another. In particular, on physical grounds, one
expects to be zero at a phase-space point
representing a nonrotating black hole but nonzero
at a point representing a rotating black hole. This
freedom in the boundary value of ta introduces a
qualitatively new element. The vector field X(t) on G
defined by the Lie derivatives of gravitational and
matter fields does not, in general, satisfy LX(t) W = 0;
it need not be an infinitesimal canonical transformation. The necessary and sufficient condition is that
((c) =8G)a J (c) Q be an exact variation. That is, X(t) generates a Hamiltonian flow if
and only if there exists a function E(t)
on G such that
t
E
c
a J c Q
8G
8
Ht Et
1 E
9
Dynamical Situations
A natural question now is whether there is an analog of
the second law of thermodynamics. Using event
horizons, Hawking showed that the answer is in the
affirmative (see Hawking and Ellis (1973)). Let (M, gab )
admit an event horizon E. Denote by a a geodesic null
normal to E. Its expansion is defined as () := qab ra b ,
where qab is any inverse of the degenerate intrinsic
metric qab on E, and determines the rate of change of the
area element of E along a . Assuming that the null energy
condition and Einsteins equations hold, the Raychaudhuri equation immediately implies that if () were to
become negative somewhere it would become infinite
within a finite affine parameter. Hawking showed that,
if there is a globally hyperbolic region containing
I (J ) [ E that is, if there are no naked singularities
this can not happen, whence () 0 on E. Hence, if a
cross section S2 of E is to the future of a cross section S1 ,
we must have aS2 aS1 . Thus, in any (i.e., not
necessarily infinitesimal) dynamical process, the change
a in the horizon area is always non-negative. This
result is known as the second law of black hole
mechanics. As in the first law, the analog of entropy is
the horizon area.
JS
Kab a^rb d2 V
j d2 V 11
8G S
8G S
where Kab is the extrinsic curvature of H in (M, gab ) and
j() is interpreted as the angular momentum density.
Now, in the Kerr family, the mass, surface gravity, and
the angular velocity can be unambiguously expressed as
J) of the
J), (a, J), and (a,
well-defined functions m(a,
horizon area a and angular momentum J. The idea is to
use these expressions to associate mass, surface gravity,
and angular velocity with each cut of H. Then, a
surprising result is that the difference between the
horizon masses associated with cuts S1 and S2 can be
expressed as the integral of a locally defined flux across
the portion H of H bounded by H1 and H2 :
I
Z
1
1
d2 V
1
2 m
da
m
j
8G H
8G
S2
I
Z 2
I
d2 V
d
j d2 V 12
j
S1
1
305
Further Reading
Ashtekar A, Beetle C, and Lewandowski J (2001) Mechanics
of rotating black holes. Physical Review 64: 044016 (gr-qc/
0103026).
Ashtekar A, Fairhurst S, and Krishnan B (2000) Isolated horizons:
Hamiltonian evolution and the first law. Physical Review D
62: 104025 (gr-qc/0005083).
Ashtekar A and Krishnan B (2003) Dynamical horizons and their
properties. Physical Review D 68: 104030 (gr-qc/0308033).
Ashtekar A and Krishnan B (2004) Isolated and dynamical
horizons and their applications. Living Reviews in Relativity
10: 178 (gr-qc/0407042).
Bardeen JW, Carter B, and Hawking SW (1973) The four laws of
black hole mechanics. Communications in Mathematical
Physics 31: 161.
DeWitt BS and DeWitt CM (eds.) (1972) Black Holes.
Amsterdam: North-Holland.
Frolov VP and Novikov ID (1998) Black Hole Physics.
Dordrecht: Kluwer.
Hawking SW and Ellis GFR (1973) Large Scale Structure of
Space-Time. Cambridge: Cambridge University Press.
Hayward S (1994) General laws of black hole dynamics. Physical
Review D 49: 64676474.
Iyer V and Wald RM (1994) Some properties of noether charge
and a proposal for dynamical black hole entropy. Physical
Review D 50: 846864.
Wald RM (1994) Quantum Field Theory in Curved Spacetime and
Black Hole Thermodynamics. Chicago: University of Chicago
Press.
Introduction
Ludwig Boltzmann (1872) established an evolution
equation to describe the behavior of a rarefied gas,
starting from the mathematical model of elastic balls
and using mechanical and statistical considerations.
The importance of this equation is twofold. First, it
provides a reduced description (as well as the
hydrodynamical equations) of the microscopic
world. Second, it is also an important tool for the
applications, especially for dilute fluids when the
hydrodynamical equations fail to hold.
The starting point of the Boltzmann analysis is to
abandon the study of the gas in terms of the detailed
motion of molecules which constitute it because of
their large number. Instead, it is better to investigate
a function f (x, v), which is the probability density of
a given particle, where x and v denote its position
and velocity. Actually, f (x, v)dx dv is often confused
with the fraction of molecules falling in the cell of
the phase space of size dx dv around x, v. The two
concepts are not exactly the same, but they are
asymptotically equivalent (when the number of
particles is diverging) if a law of large numbers holds.
The Boltzmann equation is the following:
@t v rx f Qf ; f
1
S2
f x; v f x; v01 f x; vf x; v1
2
and
v0 v nn v v1
v01 v1 nn v v1
3
7
8
307
12
13
S2
f x; v f x nr; v02
f x; vf x nr; v2
15
16
20
Consequences
The Boltzmann equation provoked a debate involving
Loschmidt, Zermelo, and Poincare, who outlined
inconsistencies between the irreversibility of the equation and the reversible character of the Hamiltonian
dynamics. Boltzmann argued the statistical nature of
his equation and his answer to the irreversibility
paradox was that most of the configurations behave
as expected by the thermodynamical laws. However,
he did not have the probabilistic tools for formulating
in a precise way the statements of which he had a
precise intuition.
Grad (1949) stated clearly the limit N ! 1,
r ! 0, Nr2 ! const:, where N is the number of
particles and r is the diameter of the molecules, in
which the Boltzmann equation is expected to hold.
This limit is usually called the BoltzmannGrad limit
(BG limit in the sequel).
The problem of a rigorous derivation of the
Boltzmann equation was an open and challenging
problem for a long time. Lanford (1975) showed that,
although for a very short time, the Boltzmann equation
can be derived starting from the mechanical model of the
hard-sphere system. The proof has a deep content but is
relatively simple from a technical viewpoint.
Existence
The mathematical study of the Boltzmann equation
starts with the problem of proving the existence of
the solutions. One would like to be able to show that,
for all (or at least for a physically significant family
of) initial distributions (which are positive and
summable functions) with finite momentum, energy,
and entropy, there exists a unique solution to eqn [1]
with the same mass, momentum, and energy as of the
initial distribution. Moreover, the entropy should
decrease and the solution should approach the right
Maxwellian as t ! 1. The problem, in such a
generality, is still unsolved, but several results in this
direction have been achieved since the pioneering
works due to Carleman (1933) for the homogeneous
equation. Actually, there are satisfactory results for
some special situations, such as the homogeneous
solutions (independent of x) close to the equilibrium,
to the vacuum, or to homogeneous data. The most
general result we have up to now is, unfortunately,
not constructive. This is due to Di Perna and Lions
(1989), who showed the existence of suitable weak
solutions to eqn [1]. However, we still do not know
whether such solutions, which preserve mass and
momentum, and satisfy the H-theorem, are unique
and also preserve the energy.
Hydrodynamics
The derivation of hydrodynamical equations from
the Boltzmann equation is a problem as old as the
equation itself and, in fact, it goes back to Maxwell
and Hilbert. Preliminary to the discussion of the
hydrodynamic limit, we establish a few properties of
the collision kernel.
It is a well-known fact that the only solution to
the equation
Qf ; f 0
21
ejvuxj =2Tx
3=2
2Tx
vM u
3
1
v2 M dv T u2
2
2
25
26
27
30
1
@t u u ru rp 0
31
@t T u rT 23Tru 0
32
22
1
2
309
24
33
where f0 is the local Maxwellian, with the parameters , u, T satisfying the Euler equations. All the
other coefficients fj of the developments can be
determined by recurrence, inverting suitable operators. However, the series is not expected to be
convergent, so that the way to show the validity of
the hydrodynamical limit rigorously is to truncate
the expansion and to control the remainder. The
first result in this direction was obtained by Caflisch
(1980). However, this approach is based on the
regularity of the solutions to the Euler equations,
which is known to hold only for short times since
shocks can be formed. How to approximate the
shocks in terms of a kinetic description is still a
difficult and open problem.
Note that the hydrodynamical picture of the
Boltzmann equation just means that we are looking
at the solutions of this equation at a suitable
macroscopic scale. The rarefaction hypothesis
underlying the Boltzmann description is reflected in
the law of perfect gas, which states that the
particles, in the local thermal equilibrium, are free.
Stationary Problems
36
35
0
^
Vp
4
dx eipx
38
1 f x; v log1 f x; v
39
z
2
e=2jvuj
z
40
1
e=2v2
1
v
311
t ! "t
44
42
Further Reading
Balesku R (1978) Equilibrium and Nonequilibrium Statistical
Mechanics. Moscow: Mir (distributed by Imported Publications, Chicago, Ill).
Caflisch RE (1980) The fluid dynamical limit of the nonlinear
Boltzmann equation. Communications of Pure and Applied
Mathematics 33: 651666.
Caflisch RE and Nicolaenko B (1983) Shock waves and the
Boltzmann equation. Nonlinear partial differential equations.
Contemporary Mathematics 17: 3544.
Carleman T (1933) Sur la theorie de lequation integro-differentielle
de Boltzmann. Acta Mathematica 60: 91146.
Cercignani C (1998) Ludwig Boltzmann. The Man Who Trusted
Atoms. Oxford: Oxford University Press.
Cercignani C, Illner R, and Pulvirenti M (1994) The Mathematical Theory of Dilute Gases. Springer Series in Applied
Mathematics, vol. 106. New York: Springer.
BoseEinstein Condensates
F Dalfovo, L P Pitaevskii, and S Stringari,
Universita` di Trento, Povo, Italy
2006 Elsevier Ltd. All rights reserved.
Introduction
In 1924 the Indian physicist S N Bose introduced a new
statistical method to derive the blackbody radiation law
in terms of a gas of light quanta (photons). His work,
together with the contemporary de Broglies idea of
matterwave duality, led A Einstein to apply the same
statistical approach to a gas of N indistinguishable
particles of mass m. An amazing result of his theory was
the prediction that below some critical temperature a
finite fraction of all the particles condense into the
lowest-energy single-particle state. This phenomenon,
named BoseEinstein condensation (BEC), is a consequence of purely statistical effects. For several years,
such a prediction received little attention, until 1938,
when F London argued that BEC could be at the basis of
the superfluid properties observed in liquid 4 He below
2.17 K. A strong boost to the investigation of Bose
Einstein condensates was given in 1995 by the observation of BEC in dilute gases confined in magnetic traps
and cooled down to temperatures of the order of a few
nK. Differently from superfluid helium, these gases
allow one to tune the relevant parameters (confining
potential, particle density, interactions, etc.), so to make
them an ideal test-ground for concepts and theories on
BEC.
What Is BEC?
In nature, particles have either integer or halfinteger spin. Those having half-integer spin, like
electrons, are called fermions and obey the Fermi
Dirac statistics; those having integer spin are
called bosons and obey the BoseEinstein statistics. Let us consider a system of N bosons. In
order to introduce the concept of BEC on a
^ rr
^ 0
1
n1 r; r 0
^ y (r) and (r)
^
^ 0 0
r;
r
2
BoseEinstein Condensates
np eips=h
9
n1 s
V p60
V
In the s ! 1 limit, the sum on the right vanishes due
to destructive interference between different plane
waves, but the first term survives. One thus finds that,
in the presence of BEC, the one-body density matrix
tends to a constant finite value at large distances. This
behavior is named off-diagonal long-range order,
since it involves the off-diagonal components of the
density matrix. Its counterpart in momentum space is
the appearance of a singular term at p = 0:
X
np N0 p
np0 p p0
10
p0 60
313
11
12
15
and
N0 T N1 T=Tc 3
16
1
2
^ y r
^ y r 0 Vr r 0 r
^ 0 r
^
dr 0 dr
17
BoseEinstein Condensates
18
19
!
2 r2
h
2
Vext r gjr; tj r; t
2m
20
315
24
25
h
=Sr; t
m
29
BoseEinstein Condensates
30
and
m
@
mv2
v = Vext n
0
@t
2
31
317
Final Remarks
BoseEinstein condensates of ultracold atoms are
easily manipulated by changing and tuning the
external potentials. This means, for instance, that one
can prepare condensates in different geometries,
including very elongated (quasi-1D) or disk-shaped
(quasi-2D) condensates. This is conceptually important, since BEC in lower dimensions is not as simple as
in three dimensions: thermal and quantum fluctuations play a crucial role, superfluidity must be properly
re-defined, and very interesting limiting cases can be
explored (TonksGirardeau regime, Luttinger liquid,
etc.). Another possibility is to use laser beams to
produce standing waves acting as an external periodic
potential (optical lattice). Condensates in optical
lattices behave as a sort of perfect crystal, whose
properties are the analog of the dynamic and transport
properties in solid-state physics, but with controllable
spacing between sites, no defects and tunable lattice
geometry. One can investigate the role of phase
coherence in the lattice, looking, for instance, at
Josephson effects as in a chain of junctions. By tuning
the lattice depth one can explore the transition from a
superfluid phase and a Mott-insulator phase, which is
a nice example of quantum phase transition. Controlling cold atoms in optical lattice can be a good starting
point for application in quantum engineering, interferometry, and quantum information.
Another interesting aspect of BECs is that the key
equation for their description in mean-field theory,
namely the GP equation [20], is a nonlinear Schrodinger equation very similar to the ones commonly
used, for instance, in nonlinear quantum optics. This
opens interesting perspectives in exploiting the analogies between the two fields, such as the occurrence of
dynamical and parametric instabilities, the possibility
to create different types of solitons, the occurrence of
nonlinear processes like, for example, higher harmonic
generation and mode mixing.
A relevant part of the current research also involves
systems made of mixtures of different gases, BoseBose
or FermiBose, and many activities with ultracold
atoms now involve fermionic gases, where BEC can
Further Reading
Cornell EA and Wieman CE (2002) Nobel lecture: BoseEinstein
condensation in a dilute gas, the first 70 years and some recent
experiments. Reviews of Modern Physics 74: 875.
Introduction
In this article we discuss quantum theories which
describe systems of nondistinguishable particles
interacting with external fields. Such models are
of interest also in the nonrelativistic case (in
quantum statistical mechanics, nuclear physics,
etc.), but the relativistic case has additional,
interesting complications: relativistic models are
genuine quantum field theories, that is, quantum
theories with an infinite number of degrees of
freedom, with nontrivial features like divergences
and anomalies. Since interparticle interactions are
ignored, such models can be regarded as a first
approximation to more complicated theories, and
they can be studied by mathematically precise
methods.
Models of relativistic particles in external electromagnetic fields have received considerable attention
in the physics literature, and interesting phenomena
like the Klein paradox or particleantiparticle pair
creation in overcritical fields have been studied; see
Rafelski et al. (1978) for an extensive review. We
will not discuss these physics questions but only
Examples
Consider the Schrodinger equation describing a
nonrelativistic particle of mass m and charge e
1
ir eA2 e
2m
1
(we set
h = c = 1, @t = @=@t, and ,, and A can
depend on the space and time variables x 2 R 3 and
t 2 R). This is a standard quantum-mechanical
model, with
the one-particle wave function
allowing for the usual probabilistic interpretation.
One interesting generalization to the relativistic
regime is the KleinGordon equation
h
i
i@t e2 ir eA2 m2 0
2
with a C-valued function . There is another
important relativistic generalization, the Dirac
equation
i@t e ir eA a m 0
with a = (1 , 2 , 3 ) and
matrices satisfying the relations
i j j i ij ;
3
Hermitian 4 4
i i ;
2 1
4
319
Second Quantization:
Nonrelativistic Case
Consider a quantum system of nondistinguishable
particles where the quantum-mechanical description of one such particle is known. In general, this
one-particle description is given by a Hilbert space
h and one-particle observables and transformations which are self-adjoint and unitary operators
on h, respectively. The most important observable
is the Hamiltonian H. We will describe a general
construction of the corresponding many-body
system.
Example As a motivating example we take the
Hilbert space h = L2 (R3 ) of square-integrable functions f (x), x 2 R3 , and the Hamiltonian H in [1]. A
specific example for a unitary operator on h is the
gauge transformation (Uf )(x) = exp(i(x))f (x) with
a smooth, real-valued function on R 3 .
In this example, the corresponding wave functions
for N identical such particles are the L2 -functions
fN (x1 , . . . , xN ), xj 2 R3 . It is obvious how to extend
where
f ; g
d3 xf xgx
8f 2 h
11
(y)
f1 ^ f2 ^ ^ fN
f1
f2
fN
12
1jPj
N
Y
fj ; gPj
13
j1
P2SN
General Construction
y 3 x y;
10
f ; g 0
6
g f ; g
8f ; g 2 h
fN x1 ; . . . ; xj ; . . . ; xk ; . . . ; xN
x;
f ;
5
fN x1 ; . . . ; xk ; . . . ; xj ; . . . ; xN
qH;
f
Hf
14
N
X
f1 ^ f2 ^ ^ Hfj ^ ^ fN
15
j1
QU
Uf QU
16
17
j1
With that, all formulas after [9] hold true as they stand.
Given any one-particle Hilbert space h with inner
product ( , ), observable H, and transformation U, the
formulas above define the corresponding Fock spaces
F (h) and many-body observable q(H) and transformation Q(U). It is also interesting to note that this
construction has various beautiful general (functorial)
properties: the set of one-particle observables has a
natural Lie algebra structure with the Lie bracket given
by the commutator (strictly speaking: i times the
commutator, but we drop the common factor i for
simplicity). The definitions above imply that
qA; qB qA; B
19
for one-particle observables A, B, that is, the abovementioned Lie algebra structure is preserved under
this map q. In a similar manner, the set of oneparticle transformations has a natural group structure preserved by the map Q,
QUQV QUV; QU1 QU1
20
21
where
(y)
n
=
m;
(y)
(fn ) obey
m;n ;
y
n
m;
y
n
23
321
24
25
1
M
hN
26
N0
27
29
31
Pj ; Pk Pj ; Qk 0
Qyj
Pj ;
Qj
32
33
f f
are fulfilled for all f , g 2 h, with y the star
operation in A (h). The particular representation
of this algebra discussed in the last section will be
denoted by 0 , 0 ((y) (f )) = (y) (f ). Other representations P can be constructed from any projection
operators P on h, that is, any operator P on h
satisfying P
= P2 = P . Writing (y) (f ) short for
P ((y) (f )), this so-called quasifree representation
is defined by
^y f
P f P f
^f P f
P f
34
F P P
with
for bosons
35
D ir eA a m e 36
323
n<0
8n 0
and
^n
y
n
8n < 0
39
where
that is,
^B q
^A; B SA; B
^
qA; q
43
P
P
S(A, B) = m<0 n0 (Amn Bnm Bmn Anm ),
45
46
U
P1
f U P2
47
i@t K
;
y
48
C e
49
C
iB
i
2
with
B2 ir eA2 m2 ;
0 i
K JKJ; J
50
i 0
with
the Hilbert space adjoint. It is important to
note that J is a grading operator. Thus, we can
define a sesquilinear form
f ; gJ f ; Jg
8f ; g 2 h
51
QU
QJU
J
52
1 B1=2 iB1=2
1 0
T p
; F
53
1=2
0 1
iB1=2
2 B
(for simplicity, we restrict ourselves to the case C = 0
and B > 0; we use the calculus of self-adjoint operators
here) with the following remarkable properties:
T 1 JT
F
B 0
^
K
TKT 1
0 B
54
Tf ;
^f T 1 f
55
325
a P AP
56
Further Reading
Carey AL and Ruijsenaars SNM (1987) On fermion gauge
groups, current algebras and KacMoody algebras. Acta
Applicandae Mathematicae 10: 186.
DeWitt B (2003) The Global Approach to Quantum Field
Theory, International Series of Monographs on Physics, vols.
1 and 2, p. 114. New York: Oxford University Press.
Gracia-Bonda JM, Varilly JC, and Figueroa H (2001) Elements
of Noncommutative Geometry, Birkhauser Advanced Texts:
Basel Textbooks. Boston: Birkhauser.
Grosse H and Langmann E (1992) A superversion of quasifree second
quantization. Journal of Mathematical Physics 33: 10321046.
Kac VG and Raina AK (1987) Bombay Lectures on Highest
Weight Representations of Infinite-Dimensional Lie Algebras,
Introduction
There is a common practice in mathematics of placing a
boundary on an object which may not appear to come
naturally equipped with one; this is often thought of as
adding ideal points to the object. Perhaps the most
famous example is the addition of a single point at
infinity to the complex plane, resulting in the Riemann
sphere: this is a boundary point in the sense of providing
an ideal endpoint for lines and other endless curves in
the plane. Often, there is more than one reasonable way
to construct a boundary for a given object, depending
on the intent; for instance, the plane is sometimes
equipped, not with a single point at infinity, but with a
circle at infinity, resulting in a space homeomorphic to a
closed disk. Both these boundaries on the plane have
useful but different things to tell us about the nature of
the plane; the common feature is that, by bringing the
infinite reach of the plane within the confines of a more
finite object, we are better able to grasp the behavior of
the original object.
The general usefulness of the construction of
boundaries for an object is to allow behavior of
structures in the completed object to aid in
visualization of behavior in the original object,
such as by providing a degree of measurement or
other classification of processes at infinity. This
utility has not been overlooked for spacetimes. A
variety of purposes may be served by various
boundary construction methods: providing a locale
for singularities (as the spacetime itself is modeled
by a smooth manifold with a smooth metric, free of
singular points); providing a platform from which to
measure global properties such as total energy or
angular momentum; displaying in finite form the
causal structure at infinity; or providing a compact
(or quasicompact) topological envelope for the
spacetime while preserving the causal structure.
The simplest and most common method of constructing a boundary for a spacetime M is to find a suitable
manifold N (of the same dimension) and an appropriate map : M ! N which is a topological embedding, that is, a homeomorphism onto its image (M).
, the closure of (M) in N, as the
We can consider M
(M) as the
-completion of M, and @ (M) = M
-boundary. Typically, this embedding is chosen in
such a way that curves of interest in M such as
timelike or null geodesics or causal curves of bounded
acceleration which have no endpoints in M, do have
endpoints in @ (M); in other words, if c : [0, 1) ! M is
such a curve of interest, then limt!1 (c(t)) exists in N.
The common practice, initiated by Penrose in
1967, is to choose N to be another spacetime
often called the unphysical spacetime, while M is
considered the spacetime of physical interest and to
require the embedding to be a conformal mapping,
that is, carries the spacetime metric in M to a scalar
multiple of the spacetime metric in N. As conformal
maps preserve the local causal structure, leaving
unchanged the notions of timelike curve or null
inherits from N a causal
curve, this means that M
structure which, locally, is an extension of that of M.
This allows us to speak of causal relationships within
, closely related to those in M.
M
Minkowski Space
E2
i+
=0
=
+
i0
Image of L2
i0
=0
i
=
i
327
Unrolled E2
+
Image of L2
i
Figure 3 L
R 1 L1 = L2 :
Universal Constructions
B-Boundary
329
Figure 4 PIP P = I (x ).
c
Figure 5 TIP P = I c.
consists of i , I , TIFs Ft and Ft emanating from
(0, t) for 0 t < 1, and the TIF F1 emanating from
(0, 1). We probably want to make at least the
following identifications for each t with 0 < t < 1,
P
and P
t Ft
t Ft ; P1 F1 P1 ; and F0
P0 F0 . This results in a two-sided replacement
for the deleted segment; for some purposes, it might
be deemed desirable to identify the two sides as one,
but a universal boundary is probably a good idea,
leaving further identifications as optional quotients
of the universal object.
How best to define the appropriate identifications
in general is a matter of some controversy. GKP
defined a somewhat complicated topology on
^
= @(M)
M
[ @(M)
[ M, then used an identification
intended to result in a Hausdorff space. There are
significant problems with this approach in some
outre spacetimes, as pointed out by Budic and Sachs
(1974) and Szabados (1989), both of whom recommended a different set of identifications. But what is
Xn
P
y
I (x)
331
z
I
(x)
^
is the topological
in M, the point I [c] in @(M)
^
endpoint of c in M.
cn , then X is homeomorphic to the conformal
5. If X = L
image of Ln in En together with I and i ; in
^ n ) has the topology of a cone.
particular, @(L
Examples The future causal boundary with the
future-chronological topology can be calculated
with a fair degree of success. For instance, if M
is conformal to a simple product spacetime Q L1
^
(Q a Riemannian manifold), then @(M)
is much
n
^
like @(L ) in that it consists of null or timelike
lines factored over a particular boundary construction @(Q) on Q, coming together at a single point i
(the IP which is all of M); if Q is complete, then
these are all null lines, and together they may be
called I .
The elements of @(Q) are defined in terms of the
Lipschitz-1 functions on Q known as Busemann
functions: if c : [, !) ! Q is any endless unit-speed
curve (typically, ! = 1), then the Busemann function
bc : Q ! R is defined by bc (q) = lims!! (s d(c(s), q)),
where d is the distance function in Q; this function
is either finite for all q or infinite for all q. The set
B(Q) of finite Busemann functions has an R-action
defined by a bc = bac , where (a c)(s) = c(s a).
^
Then @(Q) = B(Q)=R. For any P 2 @(M),
the
1
boundary of P, as a subset of Q L Q R, is
the graph of a Busemann function (the function is
bc for P generated by a null curve projecting to c);
and a point x = (q, t) in M can be represented by
@(I (x)), which is the graph of the function
t d(, q). Thus, one could use the function^ in that
space topology on B(Q) to topologize M;
^
function-space topology @(M) is a cone on @(Q),
^ apart from i , is the topological product of
and M,
R with Q [ @(Q). The future-chronological topology is sometimes different from the function-space
topology, allowing more convergent sequences
than the function-space topology does. When this
happens, the result is non-Hausdorff, revealing
^
pairs of points in @(M)
which are more closely
related to one another than the function-space
^
topology reveals; but it is still the case that @(M),
apart from i , is fibered by R over @(Q).
If Q is a warped product Q = (a, b) K for a
compact manifold K with metric dr2 e(r) h with h
a metric on K, then one can calculate more precisely:
if, for instance, has a minimum in the interior of
(a, b) and has suitable growth on either end, then
@(Q) represents two copies of K (one for each end of
(a, b) K), the future-chronological topology is the
^ (apart
same as the function-space topology, and M
from i ) is a simple product of R with Q [ @(Q):
^
@(M)
is precisely a null cone over two copies of K.
This applies, for instance, to exterior Schwarzschild,
where K = S2 ; the boundary at one end of exterior
Schwarzschild is the usual I , and the boundary at
the other end is the null cone {r = 2m}, where
exterior attaches to interior Schwarzschild.
Calculations for the future-chronological topology
^
become much easier when @(M)
is purely spacelike,
^
that is, no P 2 @(M)
is contained in the past of any
^ For instance, if M is conformal
other element of M.
to a multiwarped product, Q1 Qm (a, b)
with metric f1 (t)2 h1 fm (t)2 hm dt2 , where hi
^
is a Riemannian metric on Qi , then @(M)
will be
purely spacelike if all theR Riemannian factors are
b
complete and for each i, b 1=fi (t) dt < 1; in that
^
case, @(M)
Q, where Q = Q1 Qm and
^ Q (a, b). This applies, for instance, to interM
ior Schwarzschild, where Q1 = R 1 and Q2 = S2 ,
yielding the topology of R 1 S2 for the Schwarzschild singularity.
There is a categorical universality for spacelike
boundaries and the future-chronological topology.
This means that any other reasonable way of
future-completing interior Schwarzschild must yield
R1 S2 or a topological quotient of that for the
singularity; and if the result is to be past-distinguishing,
R1 S2 is the only possibility.
Of course, all this can be done in the time-dual
fashion, using the past-chronological topology on
It would be desirable to combine the future and
M.
past causal boundaries with a suitable topology as
well as appropriate identifications. There has been
some work in that direction.
Causal Boundary: Revisited
Further Reading
Budic R and Sachs RK (1974) Causal boundaries for general relativistic
space-times. Journal of Mathematical Physics 15: 13021309.
Garca-Parrado A and Senovilla JMM (2003) Causal relationship:
a new tool for the causal characterization of Lorentzian
manifolds. Classical and Quantum Gravity 20: 625664.
Geroch RP, Kronheimer EH, and Penrose R (1972) Ideal points
in space-time. Proceedings of the Royal Society of London,
Series A 327: 545567.
Harris SG (1998) Universality of the future chronological
boundary. Journal of Mathematical Physics 39: 54275445.
Harris SG (2000) Topology of the future chronological boundary:
universality for spacelike boundaries. Classical and Quantum
Gravity 17: 551603.
Harris SG (2001) Causal boundary for standard static spacetimes.
Nonlinear Analysis 47: 29712981 (Special Edition: Proceedings of the Third World Congress in Nonlinear Analysis).
Harris SG (2004a) Boundaries on spacetimes: an outline. Classical
and Quantum Gravity 359: 6585.
Harris SG (2004b) Discrete group actions on spacetimes: causality
conditions and the causal boundary. Classical and Quantum
Gravity 21: 12091236.
Harris SG and Dray T (1990) The causal boundary of the trousers
space. Classical and Quantum Gravity 7: 149161.
Hawking SW and Ellis GFR (1973) The Large Scale Structure of
Space-Time. Cambridge: Cambridge University Press.
Marolf D and Ross SF (2003) A new recipe for causal
completions. Classical and Quantum Gravity 20: 40854118.
Schmidt BG (1972) Local completeness of the b-boundary.
Communications in Mathematical Physics 29: 4954.
Scott SM and Szekeres P (1994) The abstract boundary a new
approach to singularities of manifolds. Journal of Geometry
and Physics 13: 223253.
LB (1988) Causal boundary for strongly causal spaceClassical and Quantum Gravity 5: 121134.
LB (1989) Causal boundary for strongly causal spaceII. Classical and Quantum Gravity 6: 7791.
333
Review of CFT
Stress Tensor and Ward Identities
Two-dimensional CFTs are massless, local, relativistic renormalized quantum field theories.
Usually they are considered in imaginary time,
that is, on two-dimensional manifolds with
Euclidean signature. In this article, the metric is
also taken to be Euclidean, although the formulation of CFTs on general Riemann surfaces is also
of great interest, especially for string theory. For
the time being, the domain is the entire complex
plane.
Heuristically, the correlation functions of such a
field theory may be thought of as being given by
the Euclidean path integral, that is, as expectation
values of products of local densities with respect
to a Gibbs measure Z1 eSE ({ }) [d ], where the
{ (x)} are some set of fundamental local fields, SE
is the Euclidean action, and the normalization
factor Z is the partition function. Of course, such
an object is not in general well defined, and this
picture should be seen only as a guide to
formulating the basic principles of CFT which
can then be developed into a mathematically
consistent theory.
k
= 0. It should be
noted that the vanishing of the trace of the stress
tensor for a scale invariant classical field theory does
5
E
X
D Y
hj 0 zj zj @=@zj
j zj ; zj
6
hj
zj ; zj
z zj 2
1
@z zj ; zj O1 7
z zj j
c=2
z z1
2
z z1 2
Tz1 8
Tz ! f 0 z2 Tz0
c 0
fz ; zg
12
9
2i C
2i C
^
^0 L
10
L
0
where, because of analyticity, C can be any contour
encircling the origin.
This suggests that one define other operators
Z
1
^
^
Ln
11
zn1 Tzdz
2 C
^ . From the OPE [8] then follows
and similarly the L
n
the Virasoro algebra V:
^ n; L
^ m n mL
^ nm c nn2 1nm;0
L
12
12
^
n.
generated by the L
with an isomorphic algebra V
In radial quantization, there is a vacuum state j0i.
Acting on this with the operator corresponding to a
scaling field gives a state jj i ^j (0, 0)j0i which is
in fact,
an eigenstate of D:
^ 0 jj i hj jj i;
L
^ j i h
j jj i
L
0 j
13
rm 1 sm2 1
4mm 1
14
V r1 ;s1 V r2 ;s2
0
0
r1 r
2 1 s1 s
2 1
X
X
335
Z
;
tr eIm
HiIm
P
^
L0 c=24
tr qL0 c=24 q
19
and P and
using the above expressions for H
2i
introducing q e .
Through the decomposition [15] of H, the trace
sum can be written as
X
Z
;
nh;h h q h q
20
h;h
where
^
dh Nqhc=24N
21
V r;s
17
Modular Invariance
1/
Boundary CFT
In any field theory in a domain with a boundary,
one needs to consider how to impose a set of
consistent boundary conditions. Since CFT is formulated independently of a particular set of fundamental fields and a Lagrangian, this must be done in
a more general manner. A natural requirement is
that the off-diagonal component Tk? of the stress
tensor parallel/perpendicular to the boundary should
vanish. This is called the conformal boundary
condition. If the boundary is parallel to the time
axis, it implies that there is no momentum flow
across the boundary. Moreover, it can be argued
that, under the RG, any uniform boundary condition will flow into a conformally invariant one. For
a given bulk CFT, however, there may be many
possible distinct such boundary conditions, and it is
one task of BCFT to classify these.
To begin with, take the domain to be the upperhalf plane, so that the boundary is the real axis. The
conformal boundary condition then implies that
z) when z is on the real axis. This has the
T(z) = T(
are
immediate consequence that correlators of T
those of T, analytically continued into the lower-
hj
1
@z
z
zj j
z zj
j
!
E
DY
j
h
1
@zj
j zj ; zj
2
z zj
z zj
j
2
24
1
Figure 2 The annulus, with boundary conditions a and b on
either boundary.
27
^ jBi
^ n jBi L
L
n
31
337
^ ab
H
tr q
^ 0 c=24
L
28
0 N 1 dh N
1 dhX
X
X X
hh0 ; N 0 ; j0 j
N 0 0 j0 1 N0 j1
^
30
33
jh; N; ji jh; N; ji
but note that now the expression is linear. The nonnegative integers nhab give the operator content with
the boundary conditions (ab): the lowest value of h
with nhab > 0 gives the conformal weight of the bcc
operator, and the others give conformal weights of
the other allowed primary fields which may also sit
at this point.
On the other hand, the annulus partition function
may be viewed, up to an overall rescaling, as the
path integral for a CFT on a circle of unit
circumference, being propagated for (imaginary)
time 1 . From this point of view, the partition
function is no longer a trace, but rather the matrix
^
h0 h
1 dX
h N
X
e4=hNc=24
34
N0 j1
h0 h h e4=
35
Shh nhab
38
These are called the Cardy conditions. The requirements that the right-hand side of [37] should give a
non-negative integer, and that the right-hand side of
[38] should factorize in a and b, give highly
nontrivial constraints on the allowed boundary
states and their operator content.
For the diagonal CFTs considered here (and for
the nondiagonal minimal models) a complete solution is possible. It can be shown that the elements Sh0
of S are all non-negative, so one may choose
~ = (Sh )1=2 . This defines a boundary state
hhhj0i
0
X 1=2
~
Sh0
jhii
39
j0i
j0ii
1=4
2
16
2
2
2 2
+
1
1
~
j0ii
16
2
44
45
46
V 1 V1 V 1
47
V1 V1 V0
48
16
16
n01~
nhh0 0
h0 h
nhh0 h00
Shh0 Shh00
Sh0
~
1
16 16
n21~
~
1
16 16
n21~
~
1
16 16
n116
~ 1
~ 1
49
2 16
Other Topics
Boundary Entropy
50
51
where the first term is the usual extensive contribution. The other two pieces sa ln (haj0i) and sb
ln (hbj0i) may be identified as the boundary entropy
associated with the corresponding boundary states.
A similar definition may be made in massive QFTs.
It is an unproven but well-verified conjecture that
the boundary entropy is a nonincreasing function
along boundary RG flows, and is stationary only for
conformal boundary states.
BulkBoundary OPE
The boundary Ward identity [24] has the implication that, from the point of view of the dependence
of its correlators on zj and zj , a primary field
j (zj , zj ) may be thought of as the product of two
local fields which are holomorphic functions of zj
and zj , respectively. These will satisfy OPEs as jzj
zj j ! 0, with the appearance of primary fields on the
right-hand side being governed by the fusion rules.
These fields are localized on the real axis: they are
the boundary operators. There is therefore a kind of
bulkboundary OPE:
X
j zj ; zj
djk Im zj hj hj hk bk Re zj
52
339
Acknowledgments
Further Reading
Affleck I (1997) Boundary condition changing operators in
conformal field theory and condensed matter physics. Nuclear
Physics B Proceedings Supplement 58: 35.
Cardy J (1984) Conformal invariance and surface critical
behavior. Nuclear Physics B 240: 514532.
Cardy J (1989) Boundary conditions, fusion rules and the
Verlinde formula. Nuclear Physics B 324: 581.
di Francesco P, Mathieu P, and Senechal D (1999) Conformal
Field Theory. New York: Springer.
Kager W and Nienhuis B (2004) A guide to stochastic Loewner evolution
and its applications. Journal of Statistical Physics 115: 1149.
Lawler G (2005) Conformally Invariant Processes in the Plane.
American Mathematical Society.
Lewellen DC (1992) Sewing constraints for conformal field theories
on surfaces with boundaries. Nuclear Physics B 372: 654.
Petkova V and Zuber JB Conformal Boundary Conditions and What
They Teach Us, Lectures given at the Summer School and
Conference on Nonperturbative Quantum Field Theoretic
Werner W Random Planar Curves and SchrammLoewner Evolutions, Springer Lecture Notes (to appear), math.PR/0303354.
Introduction
Inverse problems are generally positioned as the
problems of determination of a system (its structure,
parameters, etc.) from its input ! output
correspondence.
The boundary-value inverse problems deal with
systems which describe processes (wave, heat, electromagnetic ones, etc.) occurring in media occupying a
spatial domain. The process is initiated by a boundary
source (input) and is described by a solution of a certain
partial differential equation in the domain. Certain
additional information about the solution, which can be
extracted from measurements on the boundary, plays
the role of the output. The objective is to determine the
parameters of the medium in particular, the coefficients in the equation from this information.
The boundary control (BC) method (Belishev
1986) is an approach to the boundary-value inverse
problems based on their links with the control
theory and system theory. The present article is a
version of the BC method which solves the problem
of reconstruction of a Riemannian manifold from its
boundary spectral or dynamical data.
Forward Problems
Manifold
in 0; T
1
u jt0 ut jt0 0
in
2
uf
on 0; T
3
GT : L2 0; T; H
r0
= T*
()
*
=T
()
*
x
T
=0
F G ! H;
f ;h
W ff ; hg : u ; T
341
T
U T! : Wvol
GT!
T
Wvol
: GT ! H
HA : fy 2 H j supp y Ag
T
T
Wbd
f : uf ;0 ; T; Wvol
h : u0;h ; T
on 0; T
1
5
Controllability
t0
6
cl U T! Hh!iT
7
The Dirichlet homogeneous boundary-value problem is to find nontrivial solutions of the system
g
0
in
on
10
11
1
X
f ; sTk F T k
12
k1
where
1=2
sTk ; t : k
h
i
1=2
sin T tk @ k
Inverse problems
General Setup
~ T : f; sT T g1
W
k1
bd
k F
13
r0
14
T
W bd
~T
W bd
F
~
H
T
T 1=2
Wbd
Wbd
T 1=2
~ T : CT 1=2
W
bd
15
~ T f ; W
~ T ftt gj f 2 C1 0; T 16
: cl fW
bd
bd
0
T
is also determined
(see [8]) and, therefore, L
by R2T . In the case T > T , the operator
~ ! H~ dual to W r , is represented
r : L2 ([0, r]; H)
W
vol
vol
in the form
Z r
h
i
r
~ T 1=2 sin r tL
~ T 1=2 t dt;
~
L
Wvol
0
r0
x (, )
x (, )
343
17
In a general sense, a coordinatization is a correspondence between points x of the studied set A and
elements x of another set A~ such that: (i) the
elements of A~ are accessible and distinguishable; (ii)
the map x 7! x is a bijection; and (iii) relations
between elements of A~ determine those between
points of A which are studied (H Weyl). Coordinatization enables one to study A via operations with
~
coordinates x 2 A.
The external observer investigating the manifold probes with waves initiated by sources on
. The relevant coordinatization of described
below uses such waves and is implemented in
three steps.
Step 1 (subdomains) Let x(, ) be the end point of
the geodesic of the length > 0 emanating from 2
in the direction (), and let " be a small
neighborhood shrinking to as " ! 0. If (),
then the family of subdomains
18
;;
>
Step 2 (wave subspaces) Pass from the subdomains
to the corresponding subspaces Hhi , Hh" i ,
Hh!" (, )ir , and represent them via reachable sets
by [7]:
Hhi cl Wbd
F;
Hh" i cl Wbd
F "
r
Hh!" ; ir cl Wvol
L2 0; r; H!" ;
r
cl Wvol
L2 0; r; Hhi
Hhi " \ Hh" i
r
cl Wvol
L2 0; r; cl Wbd
F
" "
cl Wbd
F \ cl Wbd
F "
Define
r
W r; : lim cl Wvol
L2 0; r; cl Wbd
F
"!0
" "
cl Wbd
F \ cl Wbd
F "
19
Amplitude Formula
21
22
x 0 , x 00 ), so
In view of [21], one has d(x0 , x00 ) = d(
d) is an isometric copy
that the metric space (,
of (, d) by construction. Thus, the correspondence
x 7! x (point 7! family) is an isometry and
satisfies the general principles (i)(iii) of
coordinatization.
is the end product of the
d)
The manifold (,
wave coordinatization. It represents the original
manifold as a collection of infinitesimal sources
interacting with each other via the waves which they
produce.
Solving Inverse Problems
d)
knowledge of or R2T (T > T ) can recover (,
up to isometry by the following procedure:
1. Construct the model corresponding to the given
bd ,
inverse data and determine the operators W
0 T by [13], [15]; then determine
L T , and W
r by [14] or [16], [17].
L,
vol
2. Replace on the right-hand side of [19] all
operators W without tildes by the ones with
r
tildes, and get the subspaces W~ (,) = UW r(,) ,
2 , 0, r 0.
r
3. Gather all nonzero families {W~ (,) jr 0} = : x in the
= {x}
set
and redenote the subspaces as
r
r
~
W x := W~ (,) 2 x;
endow the set with the metric
x 0 , x 00 ):= 2 inf{r > 0 j W~ r 0 \ W~ r 00 6 {0}} (see [22]),
d(
x
x
of the wave copy (,
d)
d).
and get a sample (,
T
T
lim Wbd
I P Wbd
f ; t
t!T 0
0< <T
where I is the identity operator and P is the
projection in H onto clWbd
F . The formula is
derived by the ray method going back to
J Hadamard, the derivation uses the controllability
[7].
Any model determines the right-hand side of the
T
last relation by the isometry: (Wbd
) (I P )
T
T
T
T
T
the identity operator, and P = UP U is the projec bd F . This leads to the
tion in H~ onto cl W
representation
uf ;0g
; T;
0< <T
~ T ~I P
~ W
~ T f ; t
lim W
bd
bd
t!T 0
23
345
Further Reading
Belishev MI (1988) On an approach to multidimensional inverse
problems for the wave equation. Soviet Mathematics. Doklady
36(3): 481484.
Belishev MI (1996) Canonical model of a dynamical system with
boundary control in the inverse problem of heat conductivity.
St. Petersburg Mathematical Journal 7(6): 869890.
Belishev MI (1997) Boundary control in reconstruction of
manifolds and metrics. Inverse Problems 13(5): R1R45.
Belishev MI (2001) Dynamical systems with boundary control:
models and characterization of inverse data. Inverse Problems
17: 659682.
Belishev MI (2002) How to see waves under the Earth surface
(the BC-method for geophysicists). In: Kabanikhin SI and
Romanov VG (eds.) Ill-Posed and Inverse Problems, pp. 6784.
Utrecht/Boston: VSP.
Belishev MI (2003) The Calderon problem for two-dimensional
manifolds by the BC-method. SIAM Journal of Mathematical
Analysis 35(1): 172182.
Belishev MI (2004) Boundary spectral inverse problem on a class
of graphs (trees) by the BC-method. Inverse Problems
20(3): 647672.
Belishev MI and Glasman AK (2001) Dynamical inverse problem
for the Maxwell system: recovering the velocity in the regular
zone (the BC-method). St. Petersburg Mathematical Journal
12(2): 279319.
Belishev MI and Gotlib VYu (1999) Dynamical variant of the
BC-method: theory and numerical testing. Journal of Inverse
and Ill-Posed Problems 7(3): 221240.
Belishev MI, Isakov VM, Pestov LN, and Sharafutdinov VA
(2000) On reconstruction of metrics from external electromagnetic measurements. Russian Academy of Sciences.
Doklady. Mathematics 61(3): 353356.
Belishev MI and Ivanov SA (2002) Characterization of data of
dynamical inverse problem for two-velocity system. Journal of
Mathematical Sciences 109(5): 18141834.
Belishev MI and Lasiecka I (2002) The dynamical Lame system:
regularity of solutions, boundary controllability and boundary
data continuation. ESAIM COCV 8: 143167.
Katchalov A, Kurylev Y, and Lassas M (2001) Inverse Boundary
Spectral Problems. Chapman and Hall/CRC Monographs and
Surveys in Pure and Applied Mathematics, vol. 123. Boca
Raton, FL: Chapman and Hall/CRC.
Introduction
Integrable equations are a special class of nonlinear
equations arising in the modeling of a wide variety
of physical phenomena. It has been argued that
integrable PDEs are in a certain, specific sense
universal models for physical phenomena involving weak nonlinearity. Indeed, integrable equations
are obtained by a procedure involving rescaling and
an asymptotic expansion from very large classes of
nonlinear evolution equations, which preserves
integrability while retaining in the limit weakly
nonlinear effects. For this reason, integrable equations are a very important class of PDEs. Important
examples are the nonlinear Schrodinger (NLS)
equation
iqt qxx 2jqj2 q 0;
1
1
2
1
3
or
4
347
5
x ik3 Q;
2
q
6
2
7
bk 1 0; k
ct; k
ak
bk
Rk
; t 2 0; T; k 2 D
ak
8
349
At; k e2ik t 2 t; k;
2ik2 t
1 t; k
Bt; k e
9
k2C
0
0; k
1
~
Q0;
t; k
10
jg0 tj2
2kg0 t ig1 t
2kg0 t ig1 t
jg0 tj2
g0 tg1 t g0 tg1 t 0
Further Reading
Ablowitz MJ and Segur HJ (1974) The inverse scattering
transform: semi-infinite interval. Journal of Mathematical
Physics 16: 1054.
Adler VE, Gurel B, Gurses M, and Habibullin IT (1997) Journal
of Physics A 30: 3505.
Bona J, Sun S, and Zhang BY (2001) A non-homogeneous boundary
value problem for the KortewegdeVries equation. Transactions of
the American Mathematical Society 354: 427490.
Boutet de Monvel A, Fokas AS, and Shepelsky D (2004) The
modified KdV equation on the half-line. Journal of the
Institute of Mathematics of Jussieu 3: 139164.
Boutet de Monvel A and Kotlyarov VP (2003) Generation of
asymptotic solitons of the nonlinear Schrodinger equation by
boundary data. Journal of Mathematical Physics 44: 31853215.
Colin T and Ghidaglia J-M (2001) An initial-boundary value problem
for the KortewegdeVries equation posed on a finite interval.
Advanced Differential Equations 6(12): 14631492.
351
Introduction
Tensor or monoidal categories are encountered in
various branches of modern mathematical physics.
First examples came without mentioning the name of a
monoidal category as categories of modules over a
group or a Lie algebra. The operation of a monoidal
product in this case is the usual tensor product X C Y
of modules (representations) X and Y. These categories
are symmetric: the modules X Y and Y X are
1 cc 11 c : X Y Z ! Z Y X
X Y Z W ! X Y Z W ! X Y Z W
X a#
"a W
X Y Z W
X Y Z W
1
FX FY FZ ! FX FY Z ! FX Y Z
a#
#Fa
FX FY FZ ! FX Y Z ! FX Y Z
F1 FX ! F1 X
f 1
"
1f
FX
"
#Fr
FX 1
1 FX
FX F1 ! FX 1
# F l;
l
FX
FX FY ! FX Y
#
GX GY ! GX Y
f
g 1 ! F1 ! G1
The f datum of a monoidal functor (F, , f ) is
uniquely determined by the (F, ) data, so we can
denote a monoidal functor as (F, ) or even F.
353
1coev
11
coev0 1
a1
ev1
_ coev1
1
_ a
_ 1ev
r 1 _
1coev0 _
X! X1 !
ev0 1
A morphism
f: X
Y
The braiding
cX,Y : X
c 1 : X
The evaluation
evX : X
The coevaluation
coevX : 1
by
by
by
X Y
X X
X X
Y
X
ft
f
Y
Y
X
X Y Y _ X_ ! X Y Y _
ev
X_ XevX_ X 1 X_ rX_ X X_ ! 1
_
_ ev
X Y Y X ! X Y X Y ! 1
The equation
Y X
X Y
by
Y X
1 _
by
f
Y
coevXY 1 ! Y Y Y 1 Y
Y _ X_ X Y
j 1
! X Y_ X Y
1coevX 1
__ ; j2 : C; ; 1 ! C; ; 1; X 7! X__ ; f 7! f tt
1
j1t
j
X ! _ X_
X Y Z 1c
1 X Z Y ! X Z Y
a #
# c
1 1
c
1
X Y Z ! Z X Y ! Z X Y
(one for c and one for c1 ).
c cX;Y : X Y ! Y X
c
In a rigid braided category, we can define
functorial isomorphisms using again the conventions
from Figure 1:
,
X
#u21
j2
X__ Y __ ! X Y__
The square of the monoidal functor (__ , j2 ) is
____ ; j4 : C; ; 1 ! C; ; 1;
X 7! X____ ;
f 7! f tttt
where
jtt
____
____
____ j2
__
__ __ 2
j4X;Y X
Y
! X Y ! X Y
2
1
__
1 u2
u2
X
0 u1 v
1 v : X !
X
X
,
XY
2 =
u1
X
=
u2
1
u21
Ribbon Categories
u12 =
u21
c2
Braided Categories
=
u2
1
Xf t
X X ! 1
X__ X_
1 X_
X X_
f :X!Y2C
XcoevX_
HHHH
HH
iX iX
m =
X X_ Y Y _
and
X Y X Y_
FF
9 m
iXY
Xc
iX iY
#iY
iX
XX
HcH
mm
#
_
Y Y_
f Y _
X Y_
__ ev
coev
g :1
coev
Modular Categories
e :X X
ev
X_
355
: 1 1 1_ ! F
The diagram corresponding
F : F ! F is given by
to
the
antipode
F
F =
F
The structure of the coend F as a Hopf algebra can
also be found directly from its universal property, as
in Majid (1993).
There is a pairing of Hopf algebras ! : F F ! 1 in C:
X
u20
X
coev
1u20
: 1 ! X _ X! X X_ ! F
iX
F
The morphisms X form a basis of the commutative algebra Inv F = HomC (1, F). The Grothendieck ring of the category C determines the
multiplication law in Inv F via the algebra
isomorphism C Z K0 (C) ! Inv F, [X] 7! X .
Any morphism F ! 1 can be represented as a
linear combination of the morphisms
prX
evX
: F ! X X_ ! 1
357
sX1 sX;YZ
X
Y
u02
The number
X
u20
Z
1u20
coev
ev
u2
0
u02
u02
Z
u02
=
X
sXY sXZ
2
u02
u02
Proof
u02
XY
u02
u20
dimq X
u02
End X C
&
6
0.
q
q
Y2S XY
The easy implication (ii) ) (iii) can be deduced
from the Verlinde formula. If the dimension
dimq (X) = sX1 of a simple object X vanishes, then
s2XY = 0 for all Y 2 Ob C. This contradicts to the
assumption of nondegeneracy of (sXY ).
Let us determine the coefficients Y of the integral
element
X
Y Y : 1 ! F
Y2S
Y Y
= XY
3
1 1
X Y
Y
u20
21
dimq Y2
Y2S
Hence,
2
1
X
2
dimq Y
!1
8
Y2S
Y
u02
= XY
Y2S
Y Y
X Y
1
4
Conjugation Properties
For X = 1, we get
X : Inv F ! C
5
6
y . ~n
= y
: Inv F ! Inv F;
X X_
u02 Y
X jY XY ;
7
=
Y
X; Y 2 S
dim HomX; Y _ Z X jY Z
= dimqY
Y
359
Further Reading
Bakalov B and Kirillov A Jr. (2001) Lectures on Tensor
Categories and Modular Functors, University Lecture Series,
vol. 21. Providence, RI: American Mathematical Society.
Bespalov Y, Kerler T, Lyubashenko VV, and Turaev VG (2000)
Integrals for braided Hopf algebras. Journal of Pure and
Applied Algebra 148(2): 113164 (arXiv:math.QA/9709020).
Drinfeld VG (1987) Quantum groups. In: Gleason A (ed.)
Proceedings of the International Congress of Mathematicians
(Berkeley, 1986), vol. 1, pp. 798820. Providence, RI:
American Mathematical Society.
Drinfeld VG (1989a) Quasi-Hopf algebras. Algebra i Analiz
1(6): 114148.
Drinfeld VG (1989b) Quasi-Hopf algebras and Knizhnik
Zamolodchikov equations. In: Problems of Modern Quantum
Field Theory, pp. 113. BerlinNew York: Springer.
Finkelberg M (1996) An equivalence of fusion categories.
Geometric and Functional Analysis 6(2): 249267.
Huang Y-Z and Lepowsky J (1999) Intertwining operator
algebras and vertex tensor categories for affine Lie algebras.
Duke Mathematical Journal 99(1): 113134 (arXiv:q-alg/
9706028) (arXiv:q-alg/9706028).
Joyal A and Street RH (1991) Tortile YangBaxter operators in
tensor categories. Journal of Pure and Applied Algebra
71: 4351.
Kerler T and Lyubashenko VV (2001) Non-Semisimple Topological Quantum Field Theories for 3-Manifolds with Corners,
Lecture Notes in Mathematics, vol. 1765, vi + 379 pp.
Heidelberg: Springer.
Mac Lane S (1971) Categories for the Working Mathematician,
GTM, vol. 5. New York: Springer.
Majid S (1993) Braided groups. Journal of Pure and Applied
Algebra 86(2): 187221.
Majid S (1995) Foundations of Quantum Group Theory.
Cambridge: Cambridge University Press.
Moore G and Seiberg N (1989) Classical and quantum conformal
field theory. Communications in Mathematical Physics
123: 177254.
Reshetikhin NY and Turaev VG (1991) Invariants of 3-manifolds
via link polynomials and quantum groups. Inventiones
Mathematicae 103(3): 547597.
Turaev VG (1994) Quantum Invariants of Knots and 3-Manifolds,
de Gruyter Stud. Math, vol. 18. BerlinNew York: Walter de
Gruyter.
Introduction
Branes appear in string theories and M-theory as
extended objects which contain some nonperturbative information about the theory, and, apart from
gravity, they can couple with gauge fields.
At low energies, M-theory can be approximated
with an 11-dimensional N = 1 supergravity, which in
fact is unique and contains a graviton field (the metric
g ), a spin 3/2 field (the gravitino) and a gauge field
consisting of a 3-form potential field c. The gauge
field, whose field strength is a 4-form G = dc, can then
couple electrically with two-dimensional extended
objects, called M2 membranes. Moving in spacetime,
an M2 membrane describes a three-dimensional world
volume W3 so that its coupling to the gauge field is
Z
S2 k
c
1
W3
Closed string
Open string
@ Xa 0;
d
g
X
SX
h
h
4 0
@ @
Z
1
@X @X
B
X
d ^ d
2
4 0
@ @
3
@ Xi 0;
a 0; . . . ; p
6
i p 1; . . . ; 9
7
9
Using the chosen boundary conditions, the variation of the full action contains the boundary terms
Z 0
J
I
@ i d
Sbound Ai Ai
1
Z
1
i @ Xi ; 0d
2 0 0
i
Xi ; 0 Xi 0; 0
2 0
J
I
0
2 Ai Ai
10
@ X ; 0d
12
2 0 0
Xa ; XaL XaL
2 0 pa Xa0
13
Xi ; XiL XiL
20 AJi AIi Xi0
14
9
X
A Ji AIi
2
15
ip1
Aa
Aa
Ai
Aa
Massless
NS5
v
x6
x
D4
Massive
16
R x v R
17
18
19
2
d4 xtrF
F
20
gD 4 R 4
where ,
= 0, 1, 2, 3. Thus, the gauge coupling
in
p
four dimensions appears to be g4 = (gD4 )= L. In our
case, where L goes to infinity, the gauge coupling
vanishes and the gauge degrees of freedom are
frozen. Moreover, an argument similar to the one
made for the stretched strings shows that the energy
of the D4 brane is very high and makes the
mechanism of gauge group breaking difficult. The
same is true for the NS5 brane, which also turns out
to be extremely massive and does not participate in
the dynamics. But this is what we want.
To solve the problem and restore gauge dynamics
in four dimensions, one must consider a stack of
4-branes of finite length in the x6 direction. This can
be achieved placing in x6 = L a second NS5 brane
parallel to the first one and in the same point in y
(Figure 5). In this way, the D4 branes can stretch
between the NS5 branes. If L is little enough, the
gauge dynamics is restored also requiring a small
value for gD4 , to ensure the gravitational coupling
(and the couplings with the KaluzaKlein and NS5
modes) to be negligible. However, L must be bigger
then the X6 fluctuations in order to avoid quantum
corrections.
NS5
NS5
v
x
D4
x6
L
Figure 5 N = 2 four-dimensional super YangMills theory, with
U(n) gauge group.
What we just obtained is an N = 2 supersymmetric classical U(n) gauge theory in four dimensions, without matter, and in the Coulomb branch.
Before considering quantization, let us briefly
discuss some possible generalizations. For example,
matter can be realized attaching to the left-hand side
NS5 brane, new D4 branes parallel to the previous
ones, but extended in the x6 direction from 1 to 0
(Figure 6). Considering strings stretched between
long and short branes, we obtain states whose halfgauge action, associated with the end connected to
the long brane, is frozen. The corresponding states
thus appear in the fundamental representation and
can be interpreted as matter states.
To consider the Higgs branch, one should be able
to break supersymmetry giving an expectation value
to y. As mentioned above, in the actual configuration this cannot happen because y is set to 0 by
Dirichlet conditions. Fortunately, as we said, one
can add 6-branes in the (x, y) directions. If we insert
such branes to stop the long D4 branes in a large but
finite value of x6 , say x6 = M with M L, then
long branes have Neumann conditions in the y
directions. Thus, fluctuations of the long branes can
give an expectation value to y, breaking supersymmetry and subsequently the Higgs branch can be
tuned, shifting 4-branes stretched between 6-branes
(Figure 7).
NS5
Matter
NS5
D4
x6
v
L
Figure 6 Adding matter.
x
y
NS5
Matter
D6
NS5
x6
v
Higgs
branch
L
Figure 7 Permitting Higgs phases.
nL
X
log jv vLi j k
i1
nR
X
log jv vRi j
21
i1
NS5
(n1, n2)
NS5
D4
n1
x6
n2
D4
Figure 8 N = 2 four-dimensional super YangMills theory with U(n1 ) U(n2 ) gauge group and matter. Strings crossing the central
NS5 brane give matter in the (n1 , n2 ) representation.
NS5
Matter
D6
x6
s s2
n
Y
v vi 0
NS5
(v, y)
(x, x 6)
Fv; t A2 vt2 A1 vt A0 v
x 10
Figure 10 In M-theory one can think as if at any ten-dimensional
spacetime point, there is attached an S 1 circle of ray R10 .
(x, x 6)
D4 brane
M5 membrane
24
i1
D4
v
s s1
(v, y)
x 10
Figure 11 D4 branes become M5 membranes in M-theory.
25
nL
X
log v vLi
i1
R10
nR
X
log v vRi
22
i1
Brane Worlds
Further Reading
Giveon A and Kutasov D (1999) Brane dynamics and gauge
theory. Reviews of Modern Physics 71: 983.
Johnson CV (2003) D-Branes.Cambridge: Cambridge University
Press.
Lerche W (1997) Introduction to SeibergWitten Theory and Its
Stringy Origin. Nucl. Phys. Proc. Suppl. B 55: 83.
367
Brane Worlds
R Maartens, Portsmouth University, Portsmouth, UK
2006 Elsevier Ltd. All rights reserved.
Introduction
At high enough energies, Einsteins classical theory
of general relativity breaks down, and will be
superseded by a quantum gravity theory. The
singularities predicted by general relativity in gravitational collapse and in the hot big bang origin of
the universe are thought to be artifacts of the
classical nature of Einsteins theory, which will be
removed by a quantum theory of gravity. Developing a quantum theory of gravity and a unified theory
of all the forces and particles of nature are the two
main goals of current work in fundamental physics.
The problem is that general relativity and quantum
field theory cannot simply be molded together.
There is as yet no generally accepted (pre-)quantum
gravity theory.
The quest for a quantum gravity theory has a long
and thus far not very successful history. Many
different lines of attack have been developed, each
having a different way of dealing with the classical
singularities that arise from point particles and
smooth spacetime geometry. String theory does
away with zero-dimensional point particles, and
particles are modeled as different states of new
fundamental objects, the one-dimensional strings. It
turns out, however, that there is a price to pay the
number of spacetime dimensions must be greater
than four for a consistent theory. When fermions are
included, which leads to superstring theory, the
required number of dimensions is ten one time and
nine space dimensions.
There are in fact five distinct 19-dimensional
superstring theories. In the mid-1990s, duality
transformations were discovered that relate these
superstring theories to each other and to the 110dimensional supergravity theory. This led to the
conjecture that all of these theories arise as different
limits of a single theory, which has come to be
known as M theory. It was also discovered that
Higher-Dimensional Gravity
Brane worlds can be seen as reviving the original
higher-dimensional ideas of Kaluza and Klein in the
1920s, but in a new context of quantum gravity. An
important consequence of extra dimensions is that
the four-dimensional Planck scale Mp M(4) =
1.2 1019 GeV is no longer the fundamental energy
scale of gravity. The fundamental scale is instead
M(4d) . This can be seen from the modification of
the gravitational potential. For an EinsteinHilbert
gravitational action,
Z
q
1
d4 x dd y 4d g
Sgravity 2
24d
h
i
4d R 24d
1
we have the higher-dimensional Einstein field
equations,
4d
8
M2d
4d
3
24d
r1d
4
5
6
7
Brane Worlds
KaluzaKlein Modes
The dilution of gravity via extra dimensions not
only weakens gravity, it also broadens the range of
graviton modes felt on the brane. The graviton is
more than just the four-dimensional massless mode
of four-dimensional gravity other modes, with an
effective mass on the brane, arise from the fact
that the graviton is a (4d)-dimensional massless
particle. These extra modes on the brane are
known as KaluzaKlein (KK) modes of the
graviton.
For simplicity, consider a flat brane with one flat
extra dimension, compactified through the identification y $ y 2nL, where n = 0, 1, 2, . . . . The
perturbative five-dimensional graviton is defined
via
5
8
h 0 ) &h @
2
yh
10
It follows that the KK modes satisfy a fourdimensional KleinGordon equation with an effective four-dimensional mass, mn :
&hn m2n hn ;
mn
n
L
11
369
The massless mode, h0 , is the usual fourdimensional graviton mode. But there is a tower
of
massive
modes,
L1 , 2L1 , . . . ,
which
imprint the effect of the five-dimensional gravitational field on the four-dimensional brane. Compactness of the extra dimension leads to
discreteness of the spectrum. For an infinite
extra dimension, L ! 1, the separation between
the modes disappears and the tower forms a
continuous spectrum.
6
62
2
12
13
14
yL$Ly
15
2
3 Mp
4 2
16
M2p
18
19
h
hm y e2y= Bm J2 mley=
i
Cm Y2 mley=
27
Cm
J1 m
Bm
Y1 m
28
30
GM
1
Vr
31
2
r
r
r
which simply reflects the fact that the potential
becomes truly five dimensional on small scales. For
r ,
GM
22
1 2
Vr
32
r
3r
which gives the small correction to four-dimensional
gravity at low energies from extra-dimensional effects.
21
h00m
4 0
h e2y= hm 0
m
22
23
24
h0 y B0 C0 e4y=
25
Brane Worlds
h
i
2
brane
1 brane
K
T
T
g
K
5
3
33
34
where
brane
T g
T
35
K
36
2
S E
z
38
where
2 24 1645
39
4 12 5 2
40
1
S 12
TT 14T T
1
24
g 3T
T
T 2
41
E 5 CACBD nC nD g A g B
42
and
43
371
Cosmological Dynamics
A (14)-dimensional spacetime with spatial
4-isotropy (four-dimensional spherical/ plane/
hyperbolic symmetry) has a natural splitting into
hypersurfaces of symmetry, which are (13)dimensional surfaces with 3-isotropy and
3-homogeneity, that is, FriedmannRobertson
Walker (FRW) surfaces. In particular, the AdS5
bulk of the RS2 brane world, which admits a
foliation into Minkowski surfaces, also admits an
FRW foliation since it is 4-isotropic. The generalization of AdS5 that preserves 4-isotropy and
solves the five-dimensional Einstein equation is
Schwarzschild AdS5 , and this bulk therefore
admits an FRW foliation. It follows that an
FRW cosmological brane world can be embedded
in Schwarzschild AdS5 spacetime.
The black hole in the bulk is felt on the brane
via the E term. The bulk black hole gives rise to
dark radiation on the brane via its Coulomb
effect. The FRW brane can be thought of as
moving radially along the fifth dimension, with the
junction conditions determining the velocity via
the Friedmann equation. Thus, one can interpret
the expansion of the universe as motion of the
brane through the static bulk. In the special case
of no black hole and no brane motion, the brane is
empty and has Minkowski geometry, that is, the
original RS2 brane world is recovered, in different
coordinates.
An intriguing aspect of the cosmological metric is
that five-dimensional gravitational wave signals can
take shortcuts through the bulk in traveling
44
where H = a=a
is the Hubble expansion rate, a(t) is
the scale factor, K is the curvature index, and m is
the mass of the bulk black hole.
The 2 = term is the high-energy term. When
, in the early universe, then H 2 / 2 . This means
that a given energy density produces a greater rate of
expansion that it would in standard four-dimensional gravity. As a consequence, inflation in the
early universe is modified in interesting ways, some
of which may leave a signature in cosmological
observations.
The m=a4 term in eqn [44] is the dark
radiation, so called because it redshifts with
expansion like ordinary radiation. But, unlike
ordinary radiation, it is not a form of detectable
matter, but the imprint on the brane of the
gravitational field in the bulk (the Coulomb effect
of the bulk black hole). This additional effective
relativistic degree of freedom is constrained by
nucleosynthesis in the early universe. Any extra
radiative energy not thermally coupled to radiation
affects the rate of production of light elements, and
observed abundances place tight constraints on
such extra energy. The dark radiation can be no
more than 3% of the radiation energy density at
nucleosynthesis:
3m
. 0:03
2 nuc
45
46
47
The high-energy regime during radiation domination is short-lived. Since 2 = decays as a8 during the
radiation era, it will rapidly drop below one, and the
universe will enter the low-energy four-dimensional
regime. However, traces of the high-energy era may be
left in the perturbation spectra that leave an imprint in
the cosmic microwave background radiation.
In conclusion, simple brane-world models of RS2
type provide a rich phenomenology for exploring
some of the ideas that are emerging from M theory.
The higher-dimensional degrees of freedom for the
gravitational field, and the confinement of standard
model fields to the visible brane, lead to a complex
but fascinating interplay between gravity, particle
physics, and geometry, which enlarges and enriches
general relativity in the direction of a quantum
gravity theory. High-precision astronomical data
mean that cosmology is a potential laboratory for
testing and constraining these brane worlds. The
models predict extra-dimensional signatures in the
cosmic microwave background and other observations, and these predictions can in principle be tested
against data.
See also: String Theory: Phenomenology; Supergravity;
Superstring Theories.
Further Reading
Brax P and van de Bruck C (2003) Cosmology and brane worlds:
a review. Classical and Quantum Gravity 20: R201 (arXiv:
hep-th/0303095) (arXiv: hep-th/0303095).
Cavaglia M (2003) Black hole and brane production in TeV
gravity: a review. International Journal of Modern Physics
A18: 1843 (arXiv:hep-ph/0210296).
Langlois D (2003) Cosmology in a brane-universe. Astrophysics
and Space Science 283: 469 (arXiv:astro-ph/0301022).
Maartens R (2004) Brane-world gravity. Living Reviews in
Relativity 7: 7 (arXiv:gr-qc/0312059).
Quevedo F (2002) Lectures on string/brane cosmology. Classical
and Quantum Gravity 19: 5721 (arXiv:hep-th/0210292).
Rubakov V (2001) Large and infinite extra dimensions. PhysicsUspekhi 44: 871 (arXiv:hep-ph/0104152).
Wands D (2002) String-inspired cosmology. Classical and
Quantum Gravity 19: 3403 (arXiv:hep-th/0203107).
373
Introduction
In classical general relativity, a black hole is a
solution of Einsteins equations with a region of
spacetime which is causally disconnected from the
asymptotic region at infinity. The boundary of such
a region is called the event horizon. The spacetime
around the simplest black hole in three space
dimensions is described by the Schwarzschild metric
2GM
ds2 1
dt2
rc2
2GM 1 2
1
dr r 2 d2
1
rc2
where G is Newtons gravitational constant, c is the
velocity of light, and we have used spherical
coordinates with d the line element on an S2 . A
nonrotating, uncharged star which is too massive to
form a neutron star will eventually collapse, and at
late times the metric will be given by [1]. The
horizon is a null surface S2 t and the radius of the
S2 is rhorizon = 2GM=c2 . The Schwarzschild solution
has generalizations to black holes with charge and
angular momentum and no-hair theorems guarantee
that a black hole has no other characteristic property.
All these solutions can be generalized to other
theories like supergravity in various dimensions.
In 1974, Hawking showed that due to pair
production of particles near the horizon, black
holes radiate thermally. Hawkings calculation is
valid for black holes whose masses are much larger
than the Planck mass: for such black holes, the
curvature at the horizon is weak and normal
semiclassical quantization is valid. Remarkably, the
properties of Hawking radiation are quite universal.
A black hole can be characterized by an entropy
called the BekensteinHawking entropy. The leading
result for the entropy SBH for all black holes in any
theory with the standard EinsteinHilbert action is
given by
SBH
AH
4G
2
2
3
k
i k dd k
e!=TH 1 2d
4
5
6
7
ds f r
f r
1=3
r02
1 2 dt2
r
dr 2
r 2 d23
1 r02 =r2
8
where
r 2 sinh2 1
f r 1 0 2
r
r 2 sinh2
1 0 2
r
r 2 sinh2 5
1 0 2
r
Q5
9
10
VR2
N
r 2 sinh 2
324 ls8 g2s 0
where V is the volume of the T 4 and R is the radius
of the circle S1 .
The ADM mass of the black hole is
MADM
SBH
RVr03
cosh 1 cosh 5 cosh
83 ls8 g2s
12
1
2r0 cosh 1 cosh 5 cosh
13
1 ; 5 ; ! 1
Q1 ; Q5 ; N fixed
14
375
RVr02
324 g2s ls8
cosh 21 cosh 25 cosh 2
gauge, the off-diagonal gauge fields and their supersymmetric partners (which include scalar fields in
the adjoint representation) are the low-energy
degrees of freedom of open strings which connect
different branes.
The mass density or tension Tp of a single Dp
brane is given by
Tp
p1
gs 2p ls
16
17
and the YangMills coupling constant for the collective theory on the brane world volume is given by
D-Branes
2
gYMDp
2p2 gs lsp3
18
377
theory
the real
It is well known that in ap
U(Q
p ) gauge
p
coupling constant is gYM QP gs Qp . This means
that the semiclassical limit corresponds to a strongly
coupled string-field theory which reduces to strongly
coupled gauge theory in the low-energy limit and the
picture of D-branes as a collection of open strings is
not very useful. In fact, known calculational methods
in gauge theory or open-string theory are not valid in
this regime.
Microscopic Entropy for Two-Charge Systems
1
;
ee=Ti 1
i L; R
19
20
T 2 TL TR
The extensive quantities, such as the energy E,
momentum P, and entropy S, then become the sum
of left- and right-moving pieces:
E EL ER ;
P P L PR ;
S SL SR
21
; i L; R
22
Lf fL
Since the total momentum P = PR PL = ER EL is
nonzero, the lowest-energy state is clearly the one in
which all the particles move in the same direction,
for example, right moving. This is a BPS state and
corresponds to the extremal solution in supergravity.
Then E = ER = P = PR . This approach to the black
hole entropy was initiated by Das and Mathur
(1996) and Callan and Maldacena (1996).
For our two-charge system, f = 8, P = 2Q1 =L,
and L = 2RQ1 Q5 . Using [22] we get
2-charge-II
Smicro
2
p
2Q1 Q5
23
24
26
379
2G10 Q1 Q5
e!=T 1
! !=2T
R 1e!=2TL 1
V
e
29
V
2RV
4G5 Sextremal AH
30
where G5 is the five-dimensional Newtons gravitational constant. We have used the relation [22] with
L = 2RQ1 Q5 and f = 4. The fact that in the nearextremal limit SR is simply the extremal entropy and
the fact that the extremal entropy reproduces the
BekensteinHawking formula has been used as well.
Thus, the microscopic cross section exactly reproduces
the semiclassical result at low energies. Even more
remarkably, the full cross section [29] agrees with the
semiclassical answer for the gray-body factor for
parameters which correspond to the dilute-gas regime,
as shown by Maldacena and Strominger.
It is rather surprising that the results for microscopic absorption cross section calculated at weak
coupling agree with the semiclassical answers, since
the relevant process involves states which are not
31
Geometry of Microstates
A satisfactory solution of the information-loss
paradox requires a much more detailed understanding of black holes in string theory. The discussion
above shows that black holes have microstates
which may be described well in the weak-coupling
regime. It is interesting to ask whether there is a
description of these microstates in the strongcoupling regime in terms of the effective geometry
perceived by suitable probes. This question has been
answered for the two-charge system in great detail
(see Mathur (2004)). It turns out that the D1D5
microstates can be described by perfectly smooth
metrics with no horizons, and they asymptote to
the standard two-charge metric discussed above.
The location of the erstwhile stretched horizon
marks the point where the different microstates
start differing from each other significantly. Since
each such geometry does not have a horizon, neither
does it have any entropy this is consistent with
their identification with nondegenerate microstates.
Indeed, the number of such microstates correctly
accounts for the microscopic entropy. Whether a
similar picture holds for the three-charge system
remains to be seen in detail, although there are some
indications that this may be true. In this approach, it
is not yet fully understood how a horizon emerges
and why the entropy scales as the horizon area.
Outlook
One key feature of the understanding of black hole
statistical mechanics from the dynamics of branes is
the fact that a problem in gravity is mapped to a
problem in a theory without gravity, for example,
open-string field theory. In fact, the closed strings in
the bulk are already contained in the spectrum of the
381
Glossary
ADM (ArnowittDeserMisner) mass Mass of a gravitational background which is asymptotically flat.
AdSn (anti-de Sitter space) A space (or spacetime) with
constant negative curvature in n dimensions.
BPS state (BogomolnyPrasadSommerfeld state) In a
theory of extended supersymmetry, a state that is
invariant under a nontrivial subalgebra of the full
supersymmetry algebra. These states always carry
conserved charges, and supersymmetry determines the
mass exactly in terms of the charges.
CalabiYau space Complex Kahler manifold with
vanishing first Chern class.
Compactify (n. compactification) To consider a field or
string theory in a spacetime some of whose spatial
dimensions are compact.
Dirichlet boundary condition The boundary condition
which fixes the value of a field on the boundary.
Threshold bound state A bound state which is marginally bound, that is, the binding energy is zero.
Tree level In a Feynman diagram expansion of a field
theory, terms which contribute to lowest order of the
Planck constant
h.
U(N) The group of N N unitary matrices. If the
determinant is unity, the subgroup is called SU(N).
Further Reading
Callan CG and Maldacena M (1996) D-brane approach to black
hole quantum mechanics. Nuclear Physics B 472: 591
(arXiv:hep-th/9602043).
Dabholkar A (2004) Exact counting of black hole microstates,
arXiv:hep-th/0409148.
Das SR and Mathur SD (1996) Comparing decay rates for black
holes and D-branes. Nuclear Physics B 478: 561 (arXiv:hepth/9606185).
Das SR and Mathur SD (2001) The quantum physics of black
holes: results from string theory. Annual Review of Nuclear
and Particle Science 50: 153 (arXiv:gr-qc/0105063).
David JR, Mandal G, and Wadia SR (2002) Microscopic
formulation of black holes in string theory. Physics Reports
369: 549 (arXiv:hep-th/0203048).
Dijkgraaf R, Moore GW, and Verlinde E (1996) Elliptic genera of
symmetric products and second quantized strings. Communications in Mathematical Physics 185: 197 (arXiv:hep-th/
9608096).
t Hooft G (1990) The black hole interpretation of string theory.
Nuclear Physics B 335: 138.
Johnson C (2003) D-Branes. Cambridge: Cambridge University
Press.
Maldacena JM (1996) Black holes in string theory, arXiv:hep-th/
9607235.
Maldacena J, Strominger A, and Witten E (1997) Black hole
entropy in M-theory. Journal of High Energy Physics
9712: 002 (arXiv:hep-th/9711053).
Mathur SD (2004) Where are the states of a black hole?,
arXiv:hep-th/0401115.
Mohaupt T (2000) Black hole entropy, special geometry
and strings. Fortschritte der Physik 49: 3 (arXiv:hep-th/
0007195).
Sen A (1995) Extremal black holes and elementary string states.
Modern Physics Letters A 10: 2081.
Strominger A and Vafa C (1996) Microscopic origin of the
BekensteinHawking entropy. Physics Letters B 379: 99
(arXiv:hep-th/9601029).
Susskind L (1993) Some speculations about black hole entropy in
string theory, arXiv:hep-th/9309145.
Wald R (1994) Quantum Field Theory In Curved Space-Time and
Black Hole Thermodynamics. Chicago, IL: University of
Chicago Press.
383
features of wave breaking (blow-up rate and blowup set for certain types of breaking waves). We
conclude the presentation with a discussion of the
way in which solutions to the CamassaHolm
equation can be continued after wave breaking.
Introduction
Watching the sea or a lake it is often possible to
trace a wave as it propagates on the waters surface.
One can roughly distinguish two types of breaking
waves. All waves break while reaching the shore but
certain waves break far from the shore. In the first
case, the change in water depth or the presence of an
obstacle (e.g., a rock) seems to cause wave breaking,
while for certain waves within the second category,
these factors appear not to be essential. It is a matter
of observation that for many waves that break in the
open water a drastic increase in their slope near
breaking is noticeable. This leads us to the following
mathematical definition: the wave profile gradually
steepens as it propagates until it develops a point
where the slope is vertical and the wave is said to
have broken (Whitham 1980). Throughout this
article, we are concerned with wave breaking that
is not caused by a drastic change of the topography
of the bottom; for a discussion of wave breaking at
the beach we refer to Johnson (1997). The governing
equations for water waves (see the next section) are
too difficult to be dealt with in their full generality.
Therefore, to gain some insight, one has to find
simpler models that are more tractable mathematically. Investigating the properties of the model,
certain predictions can be made. The conclusions
reached will reflect reality only to some limited
extent. The value of a model depends on the number
and the degree of accuracy of physically useful
deductions that can be made from it the truth of
the model is meaningless as all experiments contain
inaccuracies and effects other than those accounted
for (while deriving the model) cannot be totally
excluded. We intend to discuss the way in which a
recent model due to Camassa and Holm (1993) can
lead to a better understanding of breaking water
waves. Firstly we survey a few classical nonlinear
partial differential equations that model the propagation of water waves over a flat bed (within the
confines of the linear theory one cannot cope with
the wave breaking phenomenon) and discuss their
relevance to the study of breaking waves. We then
analyze the breaking of waves within the context of
the CamassaHolm equation: existence of breaking
waves, criteria that guarantee that a certain initial
shape develops into a breaking wave, specific
1
2
on y t; x
3
on y t; x
4
on y d
5
6
y v
t > 0; x 2 R
7
x2R
whereas
ft; xg < 1
t;x20;TR
inf fx t; xg ! 1
x2R
as
t"T
10
coshx
jsinhxj
x 6 0
jx t; xj K K
t 2 0; T;
sup
385
for
f 2 L2 R
x2R
By introducing a new set of independent and dependent variables it is possible to resolve all singularities
due to wave breaking in the sense that [11] is
transformed into a semilinear system, the unique
solution of which can be obtained as a fixed point of
a contractive operator (Bressan and Constantin 2005).
In terms of [11], a semigroup of global conservative
solutions (in the sense that the total energy
Z
1
2 x2 dx
2 R
Further Reading
Acheson DJ (1990) Elementary Fluid Dynamics. New York:
Oxford University Press.
Benjamin TB (1992) The stability of solitary waves. Proceedings
of the Royal Society of London Series A 328: 153183.
Bressan A and Constantin A (2005) Global conservative
solutions of the CamassaHolm equation, Preprints on
Conservation Laws 2005-016 (www.math.ntnu.no/conservation/2005/016) .
Camassa R and Holm DD (1993) A new integrable shallow water
equation with peaked solitons. Physical Review Letters
71: 16611664.
Constantin A (2000) Existence of permanent and breaking waves
for a shallow water equation: a geometric approach. Annales
de lInstitut Fourier (Grenoble) 50: 321362.
Constantin A (2001) On the scattering problem for the Camassa
Holm equation. Proceedings of the Royal Society of London
Series A 457: 953970.
Constantin A and Escher J (1998) Wave breaking for nonlinear
nonlocal shallow water equations. Acta Mathematica
181: 229243.
BRST Quantization
M Henneaux, Universite Libre de Bruxelles, Bruxelles,
Belgium
2006 Elsevier Ltd. All rights reserved.
Introduction
The BRST symmetry was originally introduced in the
seminal papers by Becchi et al. (1976) and Tyutin (1975)
for YangMills gauge theories as a tool for controlling
the renormalization of the models in a consistent (gaugeindependent) way. This symmetry was discovered as a
residual symmetry of the gauge-fixed action. It was
realized later that, in fact, the BRST construction is quite
general, in the sense that it covers arbitrary gauge
theories and not just YangMills gauge models.
Furthermore, it is intrinsic, in that no gauge choice is
actually necessary to define it.
The purpose of this review is to explain the general,
intrinsic features of the BRST formalism applicable to
any gauge theory. The proper setting for discussing
these issues is that of homological algebra (Stasheff
(1998), and references therein). This article first explains
the necessary algebraic material underlying the construction and then illustrates it in the cases of the
Hamiltonian BRST formalism and the Lagrangian
BRST formalism.
gh 1
ghs0 1
1
and
2 0;
s0 s0 0;
s20 ; s1
2
BRST Quantization
k>0
3
r 6 0 ) ;
4
5
6
ghsk 1
7
8
s2 0
9
Hk s; C H k s0 ; H0 ; C
10
387
11
k<0
12
A Geometric Application
Geometric Setting
13
14
15
16
17
18
In general, H k (D, L ) 6 0.
KoszulTate Differential
19
8f 2 C1 M;
ta fa
20
21
rtA 2
22
23
tA .
24
k>0
25
Combining with D
26
27
BRST Quantization
Hk ; C 0
k > 0
28
29
30
s D s1 ;
s2 0
31
such that
H 0 s; C C1 =O
32
389
33
34
35
36
C ; C
37
F2C
38
39
Spacetime Locality
In the context of local field theory, one is often
interested in a particular class of functions of the
field histories, namely the so-called space of local
functionals. A local functional is, by definition, the
integral of a local n-form (where n is the spacetime
dimension). A local n-form reads, in local
coordinates,
! f x dn x
40
BRST Quantization
The quantization of a dynamical system can proceed
along different lines. For gauge models, the pathintegral approach is most efficiently pursued in the
context of the antifield formalism. We shall briefly
outline here the general principles underlying the
BRST Quantization
2 0
41
42
or
H H 6 0
43
Some Applications
The number of applications of the BRST formalism
is so large that it would be out of place to try being
391
Modern string theory would be inconceivable without the BRST formalism. This started with the
pioneering paper by Kato and Ogawa (1983), where
the critical dimension of the bosonic string was
derived from the condition that 2 should vanish
(quantum mechanically), and where it was shown
that the string physical states could be identified
with the state BRST cohomology. The reader is
referred to excellent monographs on modern string
theory (see Further reading).
Deformations of Gauge Models
Further Reading
Barnich G, Brandt F, and Henneaux M (2000) Local BRST
cohomology in gauge theories. Physics Reports 338: 439.
Barnich G and Henneaux M (1996) Isomorphisms between the
BatalinVilkovisky antibracket and the Poisson bracket.
Journal of Mathematical Physics 37: 5273.
C
C-Algebras and their Classification
G A Elliott, University of Toronto, Toronto, Canada
2006 Elsevier Ltd. All rights reserved.
1
2
One reason that the GelfandNaimark axiomatization of C -algebras is important is that it underlines how natural it is to consider a C -algebra
abstractly, i.e., independently of any particular
representation. Indeed, while one of the fundamental phenomena of von Neumann algebra theory
(discovered by Murray and von Neumann) is that,
essentially in rather a strong sense there is only
one way to represent a given von Neumann algebra
on a Hilbert space (and there is even a canonical
way, called the standard representation!), it is an
equally fundamental phenomenon of C -algebra
theory that, except in extremely special cases, this
is no longer true.
For instance, although the C -algebra of compact
operators on a given Hilbert space has, up to unitary
equivalence, only a single irreducible representation
this is what underlies the fact, proved by von
Neumann, referred to as the uniqueness of the
Heisenberg commutation relations for a quantummechanical system with finitely many degrees of
freedom as soon as one considers a physical system
with infinitely many degrees of freedom, one finds that
the naturally associated C -algebra has infinitely
many indeed, uncountably many unitary equivalence classes of irreducible representations, and it is
impossible to parametrize these in any reasonable way.
This striking dichotomy presents itself also in
other contexts, more elementary perhaps than the
physics of infinitely many degrees of freedom.
Consider the dynamical system consisting of a circle
and a fixed rotation acting on it. If the rotation is of
finite order i.e., if the angle is a rational multiple
of 2 then the naturally associated C -algebra is
relatively easy to study. In the case of angle zero, it
is the unital commutative C -algebra with Gelfand
Naimark spectrum the torus. In the general case of a
rational angle, the space of unitary equivalence
classes of irreducible representations is still naturally
parametrized by the torus. (And this is the same as
the space of primitive ideals the kernels of the
irreducible representations with the Jacobson
topology.)
In the irrational case the case of a rotation by an
irrational multiple of 2 (still elementary from a
geometrical point of view; note that the calendar is
based on such a system!) the irreducible representations are no longer parametrized up to unitary
equivalence by the torus and the space of primitive
ideals consists of a single point the C -algebra is
simple. (But it is decidedly not simple to study!)
This fundamental dichotomy in the classification
of C -algebras conjectured by Gaarding and
Wightman in the quantum-mechanical setting and
by Mackey in the geometrical one was established
by Glimm. Glimm proved (in the setting of separability; most of his results were generalized later
to the nonseparable case) that a large number of
a priori different ways that a C -algebra could
behave well were in fact one and the same behavior:
either all present for a given C -algebra, or all
catastrophically absent!
Some of the properties considered by Glimm, and
shown to be equivalent (for a separable C -algebra)
were as follows. First of all, every representation of
the C -algebra on a Hilbert space should be of type
I, i.e., should generate a von Neumann algebra of
type I. (A von Neumann algebra was said by Murray
and von Neumann to be of type I if it contained a
minimal projection of central support one, i.e., a
projection not contained in a proper direct summand and minimal with this property.) Second, in
every irreducible representation (not necessarily
injective) on a Hilbert space, the image of the
395
3
4
K1 J ! K1 A ! K1 A=J
5
and
(K0 and K1 are functors!). There exist natural connecting maps K1 (A=J) ! K0 (J) and K0 (A=J) ! K1 (J) the
397
Further Reading
Davidson KR (1996) C -Algebras by Example. Fields Institute
Monographs, 6. Providence, RI: American Mathematical
Society.
Dixmier J (1969) Les C -Alge`bres et leurs Representations,
2nd edn. Paris: GauthierVillars.
Elliott GA (1995) The classification problem for amenable
C -algebras. In: Chatterji SD (ed.) Proceedings of the International Congress of Mathematicians, vols. 1, 2, pp. 922932.
(Zurich, 1994). Basel: Birkhauser.
Calibrated Geometry
Calibrated geometry, introduced by Harvey and
Lawson (1982), is the study of special classes of
minimal submanifolds N of a Riemannian manifold (M, g), defined using a closed form on M
called a calibration. For example, if (M, J, g) is a
Kahler manifold with Kahler form !, then complex
k-submanifolds of M are calibrated with respect to
= !k =k!. Another important class of calibrated
submanifolds are special Lagrangian submanifolds
in CalabiYau manifolds, which is the focus of the
section Special Lagrangian geometry.
x2N
x2N0
x2N0
399
1
2
The deformation theory of compact special Lagrangian manifolds was studied by McLean (1998), who
proved the following result:
Theorem 8 Let (M, J, g, ) be a CalabiYau
m-fold, and N a compact special Lagrangian
m-fold in M. Then the moduli space MN of special
Lagrangian deformations of N is a smooth manifold
of dimension b1 (N), the first Betti number of N.
Sketch proof. Suppose for simplicity that N is an
embedded submanifold. There is a natural orthogonal decomposition TMjN = TN , where ! N is
the normal bundle of N in M. As N is Lagrangian,
the complex structure J : TM ! TM gives an isomorphism J : ! TN. But the metric g gives an
isomorphism TN T N. Composing these two
gives an isomorphism T N.
Let T be a small tubular neighborhood of N in M.
Then we can identify T with a neighborhood of the
zero section in . Using the isomorphism T N, we
have an identification between T and a neighborhood of
the zero section in T N. This can be chosen to identify
the Kahler form ! on T with the natural symplectic
401
are mirror
The SYZ conjecture Suppose M and M
CalabiYau 3-folds. Then (under some additional
conditions), there should exist a compact topological 3-manifold B and surjective, continuous maps
! B, such that
f : M ! B and f : M
(i) There exists a dense open set B0 B, such that
for each b 2 B0 , the fibers f 1 (b) and f 1 (b) are
nonsingular special Lagrangian 3-tori T 3 in M
Furthermore, f 1 (b) and f 1 (b) are in
and M.
some sense dual to one another.
(ii) For each b 2 = BnB0 , the fibers f 1 (b) and
f 1 (b) are expected to be singular special
and f , f,
argue that for generic mirrors M, M
cannot be homeomorphic
the discriminants ,
and so do not coincide. This contradicts part (ii)
above.
A better way to formulate the SYZ conjecture
may be in terms of families of mirror CalabiYau
t and fibrations ft : Mt ! B, ft : M
t!
3-folds Mt , M
B for t 2 (0, ) which approach the large complex
structure limit as t ! 0. Then we could require the
t of ft , f to converge to some
discriminants t ,
t
common, codimension 2 limit 0 as t ! 0.
It is an important, and difficult, open problem to
construct examples of special Lagrangian fibrations
of compact, holonomy SU(3) CalabiYau 3-folds.
None are currently known.
See also: Minimal submanifolds; Mirror Symmetry:
A Geometric Survey; Moduli Spaces: An Introduction;
Riemannian Holonomy Groups and Exceptional Holonomy.
Further Reading
Gross M, Huybrechts D, and Joyce D (2003) CalabiYau
Manifolds and Related Geometries, Universitext Series, Berlin:
Springer.
Harvey R and Lawson HB (1982) Calibrated geometries. Acta
Mathematica 148: 47157.
Joyce DD (2000) Compact Manifolds with Special Holonomy.
Oxford: Oxford University Press.
Joyce DD (2003a) Special Lagrangian submanifolds with isolated
conical singularities. V. Survey and applications. Journal of
Differential Geometry 63: 279347, math.DG/0303272.
Joyce DD (2003b) Singularities of special Lagrangian fibrations
and the SYZ conjecture. Communications in Analysis and
Geometry 11: 859907, math.DG/0011179.
Joyce DD (2003c) U(1)-invariant special Lagrangian 3-folds in C3
and special Lagrangian fibrations. Turkish Mathematical
Journal 27: 99114, math.DG/0206016.
McLean RC (1998) Deformations of calibrated submanifolds.
Communications in Analysis and Geometry 6: 705747.
Strominger A, Yau S-T, and Zaslow E (1996) Mirror symmetry
is T-duality. Nuclear Physics B 479: 243259, hep-th/
9606040.
403
Introduction
Systems of CalogeroMoserSutherland (CMS) type
form a class of finite-dimensional dynamical systems
that are integrable both at the classical and at the
quantum level. The CMS systems describe N point
particles moving on a line or on a ring, interacting
via pair potentials that are specific functions of four
types, namely rational (I), hyperbolic (II), trigonometric (III), and elliptic (IV). They occur not only in
a nonrelativistic (Galilei-invariant), but also in a
relativistic (Poincare-invariant) setting. Thus, one
can distinguish a hierarchy of 16 physically distinct
versions (classical/quantum, nonrelativistic/relativistic, type IIV), the most general one being the
quantum relativistic type IV system.
The nonrelativistic systems date back to pioneering work by Calogero, Sutherland, and Moser in the
early 1970s. The pair potential structure of the
interaction can be encoded in the root system AN1 ,
and there also exist integrable versions for all of the
remaining root systems. The classical systems are
given by N Poisson commuting Hamiltonians with a
polynomial dependence on the particle momenta
p1 , . . . , pN . Accordingly, the quantum versions are
described by N commuting Hamiltonians that are
partial differential operators.
The relativistic systems were introduced in the
mid-1980s, at the classical level by Ruijsenaars and
Schneider, and at the quantum level by Ruijsenaars.
They converge to the nonrelativistic systems in the
limit c ! 1, where c is the speed of light. Again, the
systems can be related to the root system AN1 , and
they admit integrable versions for other root
systems. All of the commuting classical Hamiltonians depend exponentially on generalized momenta
p1 , . . . , pN . Hence, the associated commuting quantum Hamiltonians are analytic difference operators.
The above integrable systems can be further
generalized by allowing supersymmetry or internal
degrees of freedom (spins), coupled in quite
special ways to retain integrability. In this article,
however, the focus is on the 16 versions of the
AN1 -symmetric CMS systems without internal
degrees of freedom. The primary aim is to acquaint
the reader with their definition and integrability,
and with their most prominent features and interrelationships. Second, we intend to give a rough
sketch of the state of the art concerning explicit
solutions for the various versions. This involves a
concretization of the action-angle maps and eigenfunction transforms that simultaneously diagonalize
the commuting dynamics, paying special attention to
their remarkable duality properties.
It is beyond the scope of this article to review the
hundreds of papers specifically dealing with CMS
type systems, let alone the much larger literature
where they play some role. Indeed, the systems have
been encountered in a great many different contexts
and they are related to a host of other integrable
systems in various ways. Accordingly, they can be
studied from the perspective of various subfields of
mathematics and theoretical physics. First some of
these perspectives and relations to seemingly quite
different topics will be mentioned before embarking
on the far more focused survey.
Staying first within the confines of the CMS type
systems, some nonobvious limits yielding other
familiar finite-dimensional integrable systems will
be mentioned. To begin with, all of the AN1 type
systems give rise to systems with a Toda type
(exponential nearest neighbor) interaction via a
suitable limiting transition (basically a strongcoupling limit). This leads to integrable N-particle
systems with a classical/quantum, nonrelativistic/
relativistic, nonperiodic/periodic version; starting
from the quantum relativistic periodic Toda system,
the remaining seven versions can be obtained by
suitable limits.
Next, we recall that the quantum system of N
nonrelativistic bosons on the line or ring interacting
via a pair potential of -function type is soluble via a
Bethe ansatz, with the line version exhibiting
quantum soliton behavior (factorized scattering). It
has been shown that there exist scaling limits of
eigenfunctions for suitable CMS systems that give
rise to the latter Bethe type eigenfunctions for N = 2,
while convergence for N > 2 is plausible, but has
not been demonstrated thus far.
Via suitable analytic continuations preserving
reality/formal self-adjointness, one can arrive at
CMS systems with more than one species of particle
(particles and antiparticles). Likewise, analytic
continuations and appropriate limits of CMS systems associated with root sytems other than AN1
lead to a further proliferation of N-dimensional
integrable systems. Typically, such limits refer either
m>0
1
g>0
2
3
4
is a scattering state.
The next level is given by the hyperbolic choice
Vx g2 2 =m sinh2 x;
>0
II
5
1
!; i!0 > 0
IV
6
III
7
405
8
IV GIV R N ;
GIV fxN < < x1 ; x1 xN < 2!g
9
N
X
pj
10
j1
obviously Poisson commutes with the above defining Hamiltonians of the systems. For N = 2, therefore, integrability is plain. It is possible to write
down explicitly the higher commuting Hamiltonians
for N > 2 as well but, in the nonrelativistic setting,
it is more illuminating to characterize them as the
power traces or (equivalently) the symmetric functions of a so-called Lax matrix.
The Lax matrix is an N N matrix-valued
function on the phase space of the system. It plays
a pivotal role not only for understanding integrability, but also for setting up an action-angle
transformation. The latter issue is discussed again
later. Here the more conspicuous features of the Lax
matrix will be explained, focusing on the type II
system for expository ease. Then one can choose
Ljj pj ; Ljk ig=sinh xj xk ;
j; k 1; . . . ; N; j 6 k
11
tr L2 2mH
12
k 1; . . . ; N
13
ig 2
^1 ; . . . ; p
^N L1 ;
Lt diagp
t!1
19
Since the time evolution is a canonical transformation and the Poisson brackets {Hk , Hl } are time
independent (by the Jacobi identity), it now readily
follows from [19] that they vanish. (Indeed, Hk and
Hl reduce to power traces of L1 , and the asymptotic
momenta p 1 , . . . , p N Poisson commute.)
j 1; . . . ; N
20
l6j
m sinh xj xl
2
Mjk
14
m sinh2 xj xk
H
j 6 k
When the positions and momenta in L and M evolve
according to the H-flow, one has
L_ t Mt ; Lt
15
N
X
k Sk
16
k0
S1 P;
17
x
j
^N < < p
^1 ;
p
^j =m;
tp
j 1; . . . ; N; t ! 1
18
N
X
2 X
h
@j2
Vxj xk
2m j1
1j<kN
21
for any classical Hamiltonian [1]. Thus, the defining Hamiltonians of the above systems give rise to
well-defined partial differential operators (PDOs),
which act on suitable dense subspaces of the
Hilbert space L2 (G , dx), = I, . . . , IV, with GI and
GII given by G in [4], and GIII , GIV by [8] and [9],
respectively.
We recall that there is no general result ensuring that
a classically integrable system admits an integrable
quantum version. More precisely, when one substitutes [20] in N Poisson commuting Hamiltonians, it
need not be true that they commute as quantum
operators, even when no ordering ambiguities are
present. For the power trace Hamiltonians such
ambiguities do occur. (For example, [11] gives rise
to a term in H3 proportional to p1 =sinh2 (x1 x2 ).)
On the other hand, no noncommuting factors occur
in the quantization of S1 , . . . , SN . To verify this, one
need only note that Sk equals the sum of all k k
principal minors of L, cf. [16]; choosing a diagonal
element pj in a summand, one therefore has no
dependence on xj in the remaining factors, hence no
ordering ambiguity.
As a result, the prescription [20] yields N
unambiguous operators Sk (x, ihr), which are
moreover formally self-adjoint on L2 (G , dx) for
each of the four cases = I, . . . , IV. Although by no
means obvious, it is true that these operators do
commute. Thus, integrability is preserved under
quantization of the above systems. Now the power
traces of a matrix can be expressed as polynomials
in the symmetric functions (via the Newton
N
X
xj
22
j1
fH; Bg P;
fP; Bg Nm
23
N
X
cosh
j1
P mc
N
X
j1
sinh
p
j
mc
p
24
mc
407
N
X
cosh
j1
P mc
N
X
sinh
j1
Vj x
p
j
Vj x
mc
p
j
Vj x
mc
25
f xj xk
k6j
fH; Bg P;
26
27
c!1
lim P Pnr
c!1
29
N
X
pj ;
j1
1=mc
30
l 1; . . . ; N 1
32
and
H S1 S1 =2m2 ;
P S1 S1 =2
33
j; k 1; . . . ; N
34
ej expxj pj =2
f xj xl 1=2
35
l6j
Cjk expxj xk
sinhig
sinh xj xk ig
Y wj wk zj zk
1
w zj 1j<kN wj zk zj wk
j1 j
N
Y
36
37
exp
ih
X Y
@j
f
xj xk
j2I
where
Y
!0
38
40
j2I
k62I
do commute. In the elliptic case [27], this factorization involves the Weierstrass -function, and commutativity can be encoded in a sequence of
functional equations satisfied by the -function.
For the type IIII systems the pertinent factorization
of [28] is given by
8
1=2
>
I
< 1
ig=x
f
x sinh x
ig=sinh x1=2 II
41
>
:
III
sin x
ig=sin x1=2
(Here one has g > 0, and the choice of square root is
such that f
(x) ! 1 for g # 0.)
The nonrelativistic limit c ! 1 of the quantum
Hamiltonians [33] can be determined by expanding
S1 and S1 in a power series in = 1=mc. In this
way, one obtains once more [29], except for a small,
but crucial change in Hnr : instead of the coupling
constant dependence g2 in the potential energy, one
gets g(g h). The extra term arises from the action
of the term linear in in the expansion of the
exponential on the term linear in in the expansion
of the functions f
(x).
From the perspective of the nonrelativistic quantum CMS systems, the change g2 ! g(g h) appears
ad hoc. As it transpires, however, the different
42
k 1; . . . ; N
43
To be specific,
depend only on the action vector p.
they arise from Sk (x, p) by taking g = 0 (no interac
tion, hence no x dependence) and substituting p ! p.
Indeed, the actions p k are the t ! 1 limits of the
momenta pk (t), where the t dependence refers to the
defining Hamiltonian of the system.
As it happens, the Lax matrix L is of decisive
importance to concretize the action-angle map ,
409
44
k (x) of A(x)
Obviously, the symmetric functions D
yield an integrable system on , so the Hamiltonians
k
1 ^
^ D
^;
x; p
x; p
Dk ^
k 1; . . . ; N
45
^Irel IInr
^ rel IIrel
II
46
47
j 1; . . . ; N
48
49
50
51
2 SN
n 2 NN1
54
Here,
Wx1=2 Pn x;
52
Wx
wxj xk
55
1j<kN
56
(Note that q converges to 1 both in the nonrelativistic limit c ! 1 and in the classical limit h ! 0.)
For the IInr case, the joint eigenfunctions were
found and studied a couple of decades ago by
Heckman and Opdam, yielding a multivariable
hypergeometric transform. Indeed, for N = 2, the
eigenfunctions can be expressed in terms of the
hypergeometric function 2 F1 , as has been known
since the early days of quantum mechanics. Likewise, the arbitrary-N Inr joint eigenfunction transform (studied in detail by de Jeu) can be viewed as a
multivariable Hankel transform, the N = 2 kernel
being essentially a Hankel function.
Much less is known concerning IVnr eigenfunctions, and a fortiori for the associated transform
IV . For N = 2 the time-independent Schrodinger
equation amounts to the Lame equation. Hence,
solutions are Lame functions that can be studied in
particular via Fuchs theory (regular singularities). A
far more explicit form of the eigenfunctions dates
back to work by Hermite in the nineteenth century.
More precisely, provided the g dependence of the
57
411
To conclude, we mention that the soliton scattering behavior at the classical level is preserved under
quantization in all cases where this can be checked.
That is, no new momenta are created in the
scattering process and the S-matrix is factorized as
a product of pair S-matrices. Moreover, for the type
I cases, the S-matrix is a momentum-independent
(but g-dependent) phase, as a quantum analog of the
classical billiard ball scattering.
See also: Bethe Ansatz; Classical r-Matrices, Lie
Bialgebras, and Poisson Lie Groups; Functional
Equations and Integrable Systems; Integrable Discrete
Systems; Integrable Systems and Algebraic Geometry;
Integrable Systems in Random Matrix Theory; Integrable
Systems: Overview; Isochronous Systems; Ordinary
Special Functions; q-Special Functions; Quantum
CalogeroMoser Systems; SeibergWitten Theory;
Separation of Variables for Differential Equations;
Sine-Gordon Equation; Toda Lattices.
Further Reading
Babelon O, Bernard D, and Talon M (2003) Introduction to
Classical Integrable Systems. Cambridge: Cambridge University Press.
Calogero F (1971) Solution of the one-dimensional N-body
problem with quadratic and/or inversely quadratic pair
potentials. Journal of Mathematical Physics 12: 419436.
Calogero F (2001) Classical Many-Body Problems Amenable to
Exact Treatments. Berlin: Springer.
van Diejen JF and Vinet L (eds.) (2000) CalogeroMoser
Sutherland Models. Berlin: Springer.
Fock V, Gorsky A, Nekrasov N, and Rubtsov V (2000) Duality in
integrable systems and gauge theories. Journal of High Energy
Physics 7(28): 139.
Marshakov A (1999) SeibergWitten Theory and Integrable
Systems. Singapore: World Scientific.
Moser J (1975) Three integrable Hamiltonian systems connected
with isospectral deformations. Advances in Mathematics
16: 197220.
Olshanetsky MA and Perelomov AM (1981) Classical integrable
finite-dimensional systems related to Lie algebras. Physics
Reports 71: 313400.
Olshanetsky MA and Perelomov AM (1983) Quantum integrable
systems related to Lie algebras. Physics Reports 94: 313404.
Ruijsenaars SNM (1987) Complete integrability of relativistic
CalogeroMoser systems and elliptic function identities.
Communications in Mathematical Physics 110: 191213.
Ruijsenaars SNM (1999) Systems of CalogeroMoser type. In:
Semenoff G and Vinet L (eds.) Proceedings of the 1994 Banff
Summer School Particles and Fields, pp. 251352. Berlin:
Springer.
Ruijsenaars SNM and Schneider H (1986) A new class of
integrable systems and its relation to solitons. Annals of
Physics (NY) 170: 370405.
Sutherland B (1972) Exact results for a quantum many-body
problem in one dimension II. Physical Review A
5: 13721376.
Introduction
Lagrangian formulations of general relativity (GR)
were found by Hilbert and by Einstein himself,
almost immediately after the discovery of the theory.
The construction of Hamiltonian formulations of
GR, on the other hand, has taken much longer, and
has required decades of theoretical research.
The first such formulations were developed by
Dirac and by Bergmann and his collaborators, in the
1950s. Their cumbersome formalism was simplified
by the introduction of new variables: first by
Arnowit, Deser, and Misner in the 1960s and then
by Ashtekar in the 1980s. A large number of
variants and improvements of these formalisms
have been developed by many other authors. Most
likely the process is not over, and there is still much
to learn about the canonical formulation of GR.
A number of reasons motivate the study of
canonical GR. In general, the canonical formalism
can be an important step towards quantum theory;
it allows the identification of the physical degrees of
freedom, and the gauge-invariant states and observables of theory; and it is an important tool for
analyzing formal aspects of the theory such as its
Cauchy problem. All these issues are highly nontrivial, and present open problems, in GR.
In turn, the structural peculiarity and the conceptual novelty of GR have motivated re-analyses
and extensions of the canonical formalism itself.
The following sections discuss the source of the
peculiar difficulty of canonical GR, and summarize
the formulations of the theory that are most
commonly used.
413
1
0
i
Orbits
ph
2
3
4
which generate fixed-time spatial coordinate transformations and the Hamiltonian constraint
Cx 0
5
ADM Formalism
In Einsteins formulation, the Lagrangian variable of
GR is the metric field g (x, t) (here we use the
signature [ , , , ]). Arnowit, Deser, and
Misner have introduced the following change of
variables:
p
qab gab ; N 1= g00 ; N a qab ga0
6
where qab is the inverse of the three-dimensional
metric qab , used henceforth to raise and lower space
indices a, b = 1, 2, 3. This is equivalent to writing the
invariant interval in the form
ds2 N 2 dt2 qab dxa N a dtdxb N b dt
These variables have an interesting geometric interpretation. Consider a family of spacelike (ADM)
surfaces t defined by t = constant. qab is the 3-metric
induced on the surface. N is called the lapse function
and N a is called the shift function. Their geometrical
interpretation is illustrated in Figure 2.
When written in terms of these variables, the
action of GR takes the form
Z
p
Sqab ; N; N a d4 x qNR kab kab k2
where q = det qab and R are the determinant and the
Ricci scalar of the metric qab ;
kab
1
@t qab Da Nb Db Na
2N
t + dt
Tetrad Formalism
N a dt
(x, t + dt)
(x, t)
Figure 2 The geometrical interpretation of the lapse N(x , t)
and shift N a (x , t) fields. Two ADM surfaces, defined by the
values t and t dt, are displayed. N(x , t)dt is the proper length
of the vector joining the two surfaces, normal to the first surface
at (x , t). This is the proper time lapsed between the two surfaces
for an observer at rest on the first surface at (x , t). The quantity
dx a = N a (x , t)dt is the shift (the displacement) between the
endpoint of this vector and the point (x , t dt) having the same
spacial coordinates as (x , t).
8
9
ab
@p x; t
fpab x; t; Htg
@t
10
where
Ht
12
13
415
ab
d x Nx; tCqab x; t; p x; t
2 i 1 i
p e p
det e a 2 a
14
15
Ashtekar Formalism
The Ashtekar formalism simplifies the form of the
constraints and casts GR in a form having the same
kinematics as YangMills theory. With its variants, it
is widely used in nonperturbative quantum gravity, in
particular in the loop formulation (see Loop Quantum Gravity). It can be obtained from the tetrad
canonical formalism by the canonical transformation
Aia 12 ijk !ajk ikia
16
17
!ija
dei !k ^ ek 0
The electric field E is real, while the SenAshtekar
connection Ai = Aia dxa is complex and satisfies the
reality condition
Ai Ai 2i e
18
The connection Ai has a simple geometrical interpretation. It is the pullback Aai = !()
a0i on the t = 0
ADM surface of the self-dual part
1
i KL
19
i
Ca Fab
Eai 0
20
i
C ijk Fab
Eja Ekb 0
21
1
2
22
1 2
det ekab kab k2 0
4
Evolution
In the gauge-invariant canonical structure of GR, there
is no explicit time flow generated by a Hamiltonian. If
the formalism is utilized just in order to express the
Einstein equation in first-order canonical form, this is
not a difficulty, because evolution in the coordinate
time is generated by the constraints. On the other
hand, if we are interested in understanding the
structure of states, observables, and evolution of GR,
the situation appears to be puzzling. An additional
complication arises from the fact that virtually no
gauge-invariant observable Aph is known explicitly as
a function on the phase space. These issues become
especially relevant when the canonical formalism is
taken as a starting point for quantization. How is
physical evolution represented in canonical GR?
The first relevant observation is that the gaugeinvariant phase space ph is better understood as a
phase space in the sense of Lagrange: namely as the
space of the solutions of the equations of motion
modulo gauges, rather than a space of instantaneous
states. Recall that in GR the notion of instantaneous state is rather unnatural.
In the ADM formulation, for instance, an orbit on
the constraint surface of GR can be understood as
the ensemble of all possible values that the variables
417
Further Reading
Arnowitt R, Deser S, and Misner CW (1962) The dynamics of
general relativity. In: Witten L (ed.) Gravitation: An Introduction to Current Research, p. 227. New York: Wiley.
Ashtekar A (1991) Non-Perturbative Canonical Gravity. Singapore:
World Scientific.
Bergmann P (1989) The canonical formulation of general
relativistic theories: the early years, 19301959. In: Howard D
and Stachel J (eds.) Einstein and the History of General
Relativity. Boston: Birkhauser.
Dirac PAM (1950) Generalized Hamiltonian dynamics. Canadian
Journal of Mathematical Physics 2: 129148.
Dirac PAM (1958) The theory of gravitation in Hamiltonian form.
Proceedings of the Royal Society of London, Series A 246: 333.
Dirac PAM (1964) Lectures on Quantum Mechanics. New York:
Belfer Graduate School of Science, Yeshiva University.
Gotay MJ, Isenberg J, Marsden JE, and Montgomery R (1998)
Momentum maps and classical relativistic fields. Part 1:
Covariant field theory. Archives: physics/9801019.
Hanson A, Regge T, and Teitelboim C (1976) Constrained
Hamiltonian Systems. Rome: Academia Nazionale dei Lincei.
Henneaux M and Teitelboim C (1972) Quantization of Gauge
Systems. Princeton: Princeton University Press.
Isham CJ (1993) Canonical quantum gravity and the problem of
time. In: Ibort LA and Rodriguez MA (eds.) Recent Problems in
Mathematical Physics, Salamanca, Dordrecht: Kluwer Academic.
Lagrange JL (1808) Memories de la premie`re classe des sciences
mathematiques et physiques. Paris: Institute de France.
Rovelli C (2004) Quantum Gravity. Cambridge: Cambridge
University Press.
Souriau JM (1969) Structure des Systemes Dynamics. Paris:
Dunod.
Introduction
Shared entanglement between a sender and receiver
can significantly improve the usefulness of a
quantum channel for the communication of either
classical or quantum data. Superdense coding and
teleportation provide the most well-known examples
of this improvement; free entanglement doubles the
classical capacity of a noiseless quantum channel
and makes it possible for a noiseless classical channel
to send quantum data. In fact, the entanglementassisted classical and quantum capacities of a
quantum channel are in many senses simpler and
better behaved than their unassisted counterparts
(Holevo 1998, Schumacher and Westmoreland
1997, Devetak 2005). Most importantly, these
capacities can be calculated using simple formulas
and finite optimization procedures (Bennett et al.
1999, 2002). No such finite procedure is known for
either of the unassisted capacities. Moreover, the
entanglement-assisted classical and quantum capacities are related by a simple factor of 2. The
unassisted capacities, in contrast, have completely
different formulas. In fact, the simple factor of 2
generalizes to a statement known as the quantum
reverse Shannon theorem, which governs the rate at
which one quantum channel can simulate another
(Bennett et al. 2005). The answer is given by the
ratio of the entanglement-assisted capacities.
Notation
1
2
3
Entanglement-Assisted Classical
and Quantum Capacities
The entanglement-assisted classical capacity of a
quantum channel N : A0 ! B is the optimal rate at
which classical information can be communicated
through the channel while in addition making use of
an unlimited number of maximally entangled states.
The formal definition proceeds as follows. Alice
and Bob are assumed to share nS ebits in the form of
~~
a maximally entangled state jiAB of Schmidt rank
2nS . Conditioned on her message m 2 {1, 2, . . . , 2nR },
~ ! A0n .
Alice will apply an encoding operation E m : A
2nR
Bobs decoding is given by a POVM {m }m = 1 on the
~ n . The procedure is said to have
composite system BB
maximum probability of error if
n
max tr m N E m 1
4
m
An
Bn
{m}mM = 1
B
Figure 1 Circuit representation of the elements of an
entanglement-assisted classical code for the channel N . Alice
encodes message m by applying the operation E m to her half
of the shared entanglement. Bob decodes by applying the
POVM fm 0g on the output of the channel and his half of the
shared entanglement.
5
0
CE Dp 2 log2 d
d2 1
hd 2 1 p
d2
10
7
6
419
An
Bn
B
8
Sketch of Proof
The proof of a capacity theorem can usually be
broken into two parts, achievability and optimality.
The achievability part demonstrates the existence of
a sequence of codes reaching the prescribed rate
while the optimality part shows that it is impossible
to do better.
The main idea in the achievability proof can be
understood by studying the special
case where
d2n
0
0
A = A . Let dA0 = dimA0 and {Uj }j =A0 1 be a set of
Weyl operators for A0n . The relevant property of
these operators is that averaging over them implements the constant map: for all density operators ,
d2n0
A
1 X
0n
Uj Ujy A
2n
dA0 j1
11
13
0n
A A A A
14
HN A HA
15
17
CE N lim
18
19
20
21
22
QE N
CE N
CE id2
~ ! Bn
decoding will likewise be a TPCP map D : Bm B
acting on m copies of the output of the channel, and his
~ This procedure is
half of the shared entanglement, B.
A0 n
said to -simulate N n
on
(
)
if
2
~ ~
0n
0n
n
m
F N 2 AA ; D N 1 E AB AA
1
24
25
where
F is the mixed state fidelity F(, ) =
p
(tr 1=2 1=2 )2 . The entire procedure, illustrated in
Figure 3, is said to be a (2nS , m, n, ) entanglementassisted simulation of N 2 by N 1 . A rate R, measured
in copies of N 2 per copy of N 1 , is said to be
0
achievable for A if there exists a choice of S 0 and
a sequence of (2nS , mn , n, n ) entanglement-assisted
simulations with n=mn ! R while n ! 0.
The quantum reverse Shannon theorem states
that the entanglement-assisted capacity completely
governs the achievable simulation rates.
Theorem 2 (Winter 2004, Bennett et al.). Given
two channels N 1 : A0 ! B and N 2 : A0 ! B, R is an
achievable simulation rate for N 2 by N 1 and all
0
input states A if and only if
R
CE N 1
CE N 2
26
23
421
27
An
Am
m
1
Bn
Bn
An
Bn
(a)
(b)
Moreover, since two channels that are asymptotically equivalent without free entanglement will
surely remain equivalent if free entanglement is
permitted, eqn [26] gives essentially the only
possible nontrivial, single-parameter asymptotic
characterization of quantum channels. This is the
sense in which the entanglement-assisted capacity
should be regarded as the most important capacity
of a quantum channel.
The proof of the quantum reverse Shannon
theorem is quite involved, but some of its features
can be understood without much work. First, note
that by the optimality statement of the entanglementassisted classical capacity, the desired simulation can
exist only if eqn [26] holds. Otherwise, composing
the simulation of N 2 by N 1 with a sequence of codes
achieving CE (N 2 ) would result in a sequence of codes
beating the capacity formula for N 1 .
Similarly, note that one method to simulate a
channel N 1 using N 2 is to first use N 2 to simulate
the noiseless channel and then use the simulated
noiseless channel to simulate N 1 . Since the achievable rates for the first step are characterized by the
entanglement-assisted capacity theorem, proving the
achievability part of Theorem 2 reduces to finding
protocols for simulating a general noisy quantum
channel N 2 by a noiseless one. That perhaps sounds
like a strange goal, but nonetheless is the difficult
part of the quantum reverse Shannon theorem.
It is likely that the quantum reverse Shannon
theorem can be extended to cover other types of
0
inputs than the known tensor power states (A )n .
The most desirable form of the theorem would be
one valid for all possible input density operators on
A0n , providing a single simulation procedure
dependent only on the channels and not the input
state. It is known that without modifying the form
of the free entanglement, this most ambitious form
of the theorem fails, but it is conjectured that the
full-strength theorem does hold provided very large
amounts of entanglement are supplied in the form of
the so-called embezzling states (van Dam and
Hayden 2003).
can be
written trE U BE for some isometry U BE .0 Let
AA0
ji
be a pure state and jiABE = U BE jiAA the
corresponding purified channel output state. Careful
analysis of the entanglement-assisted classical communication protocol achieving the rate I(A; B) leads to
an entanglement-assisted quantum communication
protocol consuming entanglement at the rate
(1=2)I(A; E) ebits per use of N and yielding communication at the rate of (1=2)I(A; B) qubits per use N .
The protocol achieving this goal is known as the
father (Devetak et al. 2004).
If the entanglement consumed in the father were
actually supplied by quantum communication from
Alice to Bob, then the net rate of quantum
communication produced by the resulting protocol
would be (1=2)I(A; B) (1=2)I(A; E) qubits from
Alice to Bob, that is, the total produced minus the
total consumed.
This quantity, how much more information B has
about A than E does, can be simplified using an
interesting identity. Since jiABE is pure,
IA; E HA HE HAE
28
29
12IA; E HAjB
Ic AiB
30
E H
j
j
pj H N
A0
j
31
423
P p
0
and input state jiAA = j pj jjiA jjiA . If = Ne(),
then I(A; B) is equal to (E). Thus, there are protocols
consuming entanglement that achieve the classical
communications rate (E) for the modified channel
Ne. Because the channel Ne includes an orthonormal
measurement which destroys all entanglement between
A and B, however, it can be argued that any
entanglement used in such a protocol could be replaced
by shared randomness, which could then in turn be
eliminated by a standard derandomization argument.
The net result is a procedure for choosing rate (E)
codes for the channel N consisting of states of the form
j1 jn , which is the essence of the achievability
proof for the unassisted classical capacity.
This may seem like an unnecessarily cumbersome
and even circular approach to the unassisted
classical capacity given that the proof sketched
above for the entanglement-assisted classical capacity itself invokes the unassisted result in the form of
the HSW theorem. The approach becomes more
satisfying when one learns that simple and direct
proofs of the father protocol exist that completely
bypass the HSW theorem (Abeyesinghe et al. 2005).
Thus, the entanglement-assisted communication
protocols can be easily transformed into their
unassisted analogs, confirming the central place of
entanglement-assisted communication in quantum
information theory.
0
Acknowledgmnts
The author is grateful to the inventors of the
quantum reverse Shannon theorem for letting him
Further Reading
Abeyesinghe A, Devetak I, Hayden P, and Winter A (2005) Fully
quantum SlepianWolf (in preparation).
Bennett CH, Devetak I, Harrow AW, Shor PW, and Winter A (2005)
The quantum Reverse Shannon Theorem (in preparation).
Bennett CH, Shor PW, Smolin JA, and Thapliyal AV (1999)
Entanglement-assisted classical capacity of noisy quantum
channels. Physical Review Letters 83: 3081 (arXiv.org:quantph/9904023).
Bennett CH, Shor PW, Smolin JA, and Thapliyal AV (2002)
Entanglement-assisted capacity of a quantum channel and
the reverse Shannon theorem. IEEE Transactions on Information Theory 48(10): 2637 (arXiv.org:quant-ph/0106052).
Cerf N and Adami C (1997) Von Neumann capacity of noisy
quantum channels. Physical Review A 56: 3470 (arXiv.org:
quant-ph/9609024).
Devetak I (2005) The private classical capacity and quantum
capacity of a quantum channel. IEEE Transactions on
Information Theory 51(1): 44 (arXiv.org/0304127).
Devetak I, Harrow AW, and Winter A (2004) A family of
quantum protocols. Physical Review Letters 93: 230504
(arXiv.org:quant-ph/0308044).
DiVincenzo DP, Smolin JA, and Shor PW (1998) Quantum
channel capacity of very noisy channels. Physical Review A
57: 830 (arXiv.org:quantph/9706061).
Holevo AS (1998) The capacity of the quantum channel with
general signal states. IEEE Transactions on Information
Theory 44: 269273.
Schumacher B and Westmoreland MD (1997) Sending classical
information via noisy quantum channels. Physical Review A
56: 131138.
van Dam W and Hayden P (2003) Universal entanglement
transformation without communication. Physical Review A
67: 060302 (arXiv.org:quant-ph/0201041).
Winter A (2004) Extrinsic and instrinsic data in quantum
measurements: asymptotic convex decomposition of
positive operator valued measures. Communications in
Mathematical Physics 244(1): 157 (arXiv.org:quantph/
0109050).
Introduction
Any processing of quantum information, be it
storage or transfer, can be represented as a quantum
channel: a completely positive and trace-preserving
map that transforms states (density matrices) on the
senders end of the channel into states on the
receivers end. Very often, the channel S that sender
and receiver (conventionally called Alice and Bob,
respectively) would like to implement is not readily
available, typically due to detrimental noise effects,
limited technology, or insufficient funding. They
may then try to simulate S with some other channel
T, which they happen to have at their disposal. The
quantum channel capacity Q(T, S) of T with respect
to S quantifies how well this simulation can be
performed, in the limit of long input strings, so that
Alice and Bob can take advantage of collective preand post-processing (cf. Figure 1). Higher capacities
may result if Alice and Bob are allowed to use
additional resources in the process, such as classical
side channels or a bunch of maximally entangled
pairs shared between them.
Quantum capacity thus gives the ultimate benchmarks for the simulation of one quantum channel by
another and for the optimal use of auxiliary
resources. Together with the compression rate of a
quantum source (see Source Coding in Quantum
Resources
Decoding
Encoding
denotes p
the
norm of linear operators, and
k%k1 := tr % % is the trace norm on the space of
trace-class operators B (H).
We use base two logarithms throughout, and we
write ld x := log2 x and exp2 x := 2x .
!1 D;E
1
2
425
3
Related Capacities
This article is chiefly concerned with the quantum
capacity of a quantum channel. A variety of other
Elementary Properties
The capacity of a composite channel T1 T2 cannot
be bigger than the capacity of the channel with the
smallest bandwidth. This in turn suggests that
simulating a concatenated channel is in general easier
than simulating any of the individual channels. These
relations are known as bottleneck inequalities:
QT1 T2 ; S minfQT1 ; S; QT2 ; Sg
4
QT; S1 S2
maxfQT; S1 ; QT; S2 g
5
6
QT1
QT1 ; T2
QT1 Qid; T2
QT2
7
8
ld n
ld m
9
ld m
QT; idm
ld n
10
11
12
StateChannel Duality
Quantum capacity is closely related to the distillable
entanglement, which is the optimal rate m/n at
427
13
Coding Theorems
Computing channel capacities straight from Definition 1 is a tricky business. It involves optimization in
systems of asymptotically many tensor factors, and
can only be performed in special cases, like the
noiseless channels in the section Elementary properties. Coding theorems aspire to reduce this
problem to an optimization over a low-dimensional
space. They usually come in two parts: the converse
provides an upper bound on the channel capacity
(typically in terms of some entropic expression),
while the direct part consists of a coding scheme
that attains this bound. By Shannons celebrated
coding theorem, the classical capacity of a classical
noisy channel can be obtained from a maximization
of the mutual information over all joint input
output distributions.
For the quantum channel capacity, the relevant
entropic quantity is the coherent information,
Ic T; % : H T% H T idj % ih % j
14
where H denotes the von Neumann entropy:
H(%) = tr% ld%, and % 2 HA HA0 is a purification of the density operator % 2 A. The coherent
information does not increase under quantum
operations, Ic (S T, %) Ic (T, %) for any quantum
channel S and state % 2 A. This is the data
processing inequality (Barnum et al. 1998), which
shows that the regularized coherent information
provides an upper bound on the quantum channel
capacity: if Alice and Bob have a coding scheme for
the channel T with capacity Q(T), n channel uses
allow them to share a maximally entangled state of
size
exp2 n Q(T). The coherent information of this
state equals
n Q(T), and was no larger prior to
Bobs decoding.
Recently, Devetak (2005) developed a coding
scheme to show that this bound is in fact attainable.
Different proofs were outlined by Lloyd and Shor.
Theorem 1
n!1 n
max Ic T n ; %
%
15
16
l1
The net transfer rate for private classical information is then R
(TB ) (TE ), which is just the total
transfer rate for the channel Alice ! Bob reduced by
the transfer rate Alice P
! Eve.
Remarkably, if % = i pi j i ih i j is a decomposition of % 2 A into pure states, the private transfer
rate exactly equals the coherent information,
Ic TB ; % H TB % H TE %
TB TE
The so-called entropy exchange
H TE % = H TB idj
% ih %
19
429
20
X p
Mkl jiB j kliB1 B2
kl
21
tr
tr
S
Time
tr
Time
Acknowledgmnts
The author thanks the members of the quantum
information group at TU Braunschweig for their
careful reading of the manuscript and many helpful
suggestions. He also gratefully acknowledges the
funding from Deutsche Forschungsgemeinschaft
(DFG).
See also: Capacities Enhanced by Entanglement;
Channels in Quantum Information Theory; Entanglement;
Positive Maps on C -Algebras; Quantum Channels:
Classical Capacity; Quantum Error Correction and Fault
Tolerance; Source Coding in Quantum Information Theory.
Further Reading
Barnum H, Nielsen MA, and Schumacher B (1998) Information
transmission through a noisy quantum channel. Physical
Review A 57: 4153 (quant-ph/9702049).
Bennett CH, Devetak I, Shor PW, and Smolin JA (2004)
Inequalities and separations among assisted capacities of
quantum channels, quant-ph/0406086.
Bennett CH, DiVincenzo DP, Smolin JA, and Wootters WK
(1996) Mixed-state entanglement and quantum error correction. Physical Review A 54: 3824 (quant-ph/9604024).
Bowen G and Mancini S (2004) Quantum channels with a finite
memory. Physical Review A 69: 012306 (quant-ph/0305010).
Devetak I (2005) The private classical information capacity and
quantum information capacity of a quantum channel. IEEE
Transactions on Information Theory 51: 44 (quant-ph/0304127).
Devetak I, Harrow AW, and Winter A (2004) A family of
quantum protocols. Physical Review Letters 93: 230504
(quant-ph/0308044).
Devetak I and Winter A (2004) Relating quantum privacy and
quantum coherence: an operational approach. Physical
Review Letters 93: 080501 (quant-ph/0307053).
DiVincenzo DP, Shor PW, and Smolin JA (1998) Quantum
channel capacities of very noisy channels. Physical Review A
57: 830 (quant-ph/9706061).
Eisert J and Wolf MM Gaussian quantum channels. In Cerf N,
Leuchs G, and Polzik E (eds.) Quantum Information with
Capillary Surfaces
R Finn, Stanford University, Stanford, CA, USA
g
u0
1
d sin #
1
2
dr
R
2
p 2
2H
3
2 R1 R2
where H is the mean curvature of S at the point.
Although Young provided no formal justification for
this step, we can establish it with the aid of a general
formula from differential geometry that was not
known in his lifetime:
I
Z
2HN dS
n ds
4
S
n
p1
p 2H N dS 0
5
1 2
0
6
Solid
Gas
2
0
Liquid
p2
Figure 2 Pressure change across fluid element, balanced by
surface tension.
2 cos
;
a
g
7
In 1806, Laplace published the first analytical expression for the mean curvature of a surface u(x, y), and
showed that the expression can be written as a
divergence. He obtained the equation
div Tu 2H;
ru
Tu q
1 jruj2
8
S
1
from which
Z
div Tu u dx
I
Tu ds 0
15
q
with Tu ru= 1 jruj2 , and with the unit
exterior normal on . Choosing first
to have
compact support in , the boundary term vanishes,
and the fundamental lemma of the calculus of
variations yields
g=
Tu cos
16
17
18
div Tu u ;
jj cos
jj
19
20
2
(a)
1
0
0
2
(b)
(a)
(b)
(c)
45
45
45
45
(d)
Figure 8 Support configurations: (a) capillary tube, general
section; (b) horizontal plate; (c) convex surface appearing during
deformation of horizontal plate to capillary tube; and (d)
Nonuniqueness of configuration appearing during convex deformation. Reproduced from Mathematics Intelligencer 24(3) 2002
2133 with permission from Springer-Verlag Heidelberg.
Spoon (left)
Rotationally
symmetric
Potato chip
2
Spoon (left)
Potato chip
D1
(No graph)
D2
(D)
(Continuous)
D2
D1+
(No graph)
(I )
2 . P
2
1
g= > 0
22
Discontinuous Dependence II
1+
0
1
(a)
(b)
1 sin !
1 sin
1 "
cos
cos
23
!0
"
24
"
Existence Questions I
2
26
(a)
(b)
1
regions of Figure 12. Lancaster and Siegel and
independently Chen, Finn, and Miersemann showed
that if 2 1 2 2, then every solution
is bounded at the vertex. This result holds also for
the zero gravity eqn [18].
In the case of [18], Concus and Finn showed that
in the D
2 are discontinuous at P. A
number of attempts to prove or to disprove this
conjecture have till now been unsuccessful.
An existence theorem for [16][17] alternative to
that of Emmer was obtained independently by
Uraltseva, using a very different approach. This
procedure yielded smoothness estimates up to the
boundary, but required a hypothesis of boundary
smoothness, so that the result does not mesh with the
discontinuous dependence behavior as does that of
Emmer. Later versions of the existence result, again
under boundary smoothness requirements, were given
by Gerhardt, Spruck, and Simon and Spruck. In the
procedure introduced by Emmer, the boundary trace is
shown to exist only in a very weak sense (which,
however, suffices for a uniqueness proof). The later
work can be adapted to show that the Emmer
solutions are smooth on the smooth parts of @.
None of the above procedures provides existence for
the zero gravity case [18]. As we shall see in the
following section, that is not an accident of the
methods, but reflects subtle properties of the equations.
Existence Questions II
28
where is defined by
; jj j j cos 2H j j
29
there
is no bounded solution
1
of [16][17] or [18][17] as a graph. For [16][17],
unbounded solutions do however exist for such data
(see the section Existence questions I).
A
A
A
2
B
2
A
B
B
B
(a)
(b)
(c)
Figure 18 (a) Fan domains APA0 and BPB 0 of constant limiting
values; 2a < p so that the fans overlap when data are in R. (b)
2a > p; case 1. Fans APA0 and BPB 0 of constant radial limits
appear. Limiting value changes strictly monotonically as
approach direction changes from A0 P to B 0 P. (c) 2a > p; case 2.
In addition to the two fans adjacent to the sides of the
wedge, a half plane of constant radial limits appears.
2( )
D2
D1
(D)
(ID), (D), (I )
2
R
2( )
Continuous, (I), (D)
+
D1
D2
(DI ), (D), (I )
(I )
0
0
Figure 19 p < 2a < 2p. Possible modes of behavior. Reproduced with permission from the Pacific Journal of Mathematics.
2 regions, the
drop detaches from one side of the wedge and
becomes a spherical cap resting on a single planar
surface, in accord with the above conjecture. As D
1
is approached, the liquid becomes a drop of very
large radius that fills out a long thin region in the
wedge, and disappears to infinity as the boundary of
R is crossed. However, as D
1 is entered, the
configuration transforms smoothly into a spherical
liquid bridge, connecting the two faces of the wedge
without contacting the wedge line.
Stability Questions
(a) In R, near D1
(a) In R, near D2
2
X
Wj cos j jSj 3jV j
R
1
30
(b) In R, near D1
(c) In D1
(A)
(c) In R, near D2
(B)
D
2 ); (b) g = 90 (center of R); (c) g = 35 (near D 2 ). As D 2 is
entered, original boundary conditions can no longer be satisfied
by spherical drop, but configuration changes smoothly into drop
on single plane, with prescribed data for that plane. Reproduced
with permission from Concus P, Finn R and McCuan J (2001)
Liquid bridges, edge blobs, and Scherk-type capilliary surfaces.
Indiana University Mathematics Journal 50: 411441.
3jV j
R
31
Compressibility
Until very recently, all literature on capillarity was
based on a hypothesis that the body of the fluid
is incompressible. Indeed, from the point of view
of macroscopic mechanical measurements, most
liquids are nearly incompressible. But all liquids are
also to some extent compressible, and this property
was even conceptually essential in our characterization in the section Gauss contribution: the energy
method of the surface energy, even for the nominally incompressible case. It is as yet unclear to what
extent the compressibility properties of the bulk
liquid will influence the physical predictions of the
theory. In this connection, see the remarks at the end
of the section Uniqueness and nonuniqueness.
The Equations I
33
34
throughout .
Note that in this result, no growth condition is
imposed at the corner points. It can happen that
both u and v are unbounded at a corner point;
nevertheless, [34] holds uniformly over .
The solutions of [32] emulate many of the
characteristics of solutions of [16]. Notably, there is
again a dichotomy of behavior, depending on opening angle 2 at a corner point, with all solutions
either bounded, or unbounded with growth like 1=r.
The Equations II
0
p0
gu
e 1
g1 cos !
35
Closing Remarks
This brief survey is intended only as a general
indication of the current state of the theory; much
material of interest could not be included. Nor have
we addressed hysteresis effects on contact angle.
Detailed references to the material discussed and also
to further information can be found in the articles
listed below. More recent publications can be located
by following links in MathSciNet or Zentralblatt.
Acknowledgmnt
I owe a special debt of thanks to my colleague
Paul Concus, who read the material in detail and
provided many effectual suggestions, leading to a
much-improved exposition.
See also: Compressible Flows: Mathematical Theory;
Interfaces and Multicomponent Fluids; Newtonian Fluids
and Thermohydraulics.
Further Reading
References for text material and for further reading are cited in
the expository articles:
Finn R (2002a) Milan Journal of Mathematics 70: 123.
Finn R (2002b) Mathematical Intelligencer 24: 2133.
dGn t
aGn1 t Gn t
dt
">0
1
">0
2
F = F(x, t), x 2 R, t 2 R .
Equation [1] for the case of linear f 7! (f )
was called as Burgers equation by Hopf (1950),
who justified this by the remark: equation was
first
@f
@f
@2f
f
" 2
@t
@x
@x
introduced by J. M. Burgers (1940) as a simplest
model to the differential equations of fluid flow. In
fact, eqn [1] for linear (f ) was introduced earlier in
1915 by Bateman. Equation [1] for general (f )
appeared later in very different models, for example,
in the model for displacement of oil by water, in a
model of road traffic, etc.
For (f ) = a b f , Hopf and Cole have studied
[1] basing on the substitution
1
@g
f
a"
g
b
@x
reducing [1] to the heat equation
@y
@2g
" 2
@t
@x
This transformation (often called as the Hopf
Cole transform) appeared for the first time in 1906
in the book of Forsyth Theory of differential
equations.
by the substitution
a Gn t Gn1 t
Fn; t
b
Gn t
Equation [2] for general (F) was introduced by
Henkin, Polterovich (1991) for the description of a
Schumpeterian evolution of industry. For any " > 0,
one can consider [2] as the family of difference
differential equations, depending on the parameter
= {x="} 2 [0, 1), where {x="} denotes the fractional part of x=". For physical applications of [1]
(see Gelfand (1959), Landan and Lifschitz (1968),
Lax (1973)), the inviscid case (" = 0) is the most
interesting. But, for some special physical models
and for some social and biological applications (see
Henkin, Polterovich (1991), Serre (1999)), the
interesting case concerns eqn [2] with " = 1 and
x 2 Z.
The results considered in this article concern
mainly the Cauchy problem for eqns [1] and [2]
with initial data f(x, 0), F(x, 0) satisfying the
conditions
f x; 0 ! ; x ! 1
Z 0
jf x; 0 jdx
1
Z 1
j f x; 0jdx < 1
3
and correspondingly
Fk" " ! ;
0
X
k ! 1
jFk" "; 0 j
k1
1
X
4
k0
0
@t
"
5
F
6
@t
@x 2 @x
@x
where
f x; 0 !
Z
0
dy
;
y
x ! 1
447
From references one can deduce the following general properties of Cauchy problems [1], [3] and [2], [4].
Theorem 0
f x; t dx
f x; t dx
dt 0
1
Z
ydy
Moreover, if the initial value f(x, 0) is nondecreasing as a function of x, then solution f(x, t)
is nondecreasing as a function of x for all t 0.
(ii) There exists a unique solution F(x, t) x 2 R, t 2
R of the problem [2], [4]; this solution is
smooth for t > 0; besides, it satisfies the following conservation laws for t > 0 and 2 [0, 1):
Fk" "; t ! ;
k ! 1
Fk" "; t ! ; k ! 1
"
#
Z Fk"";t
1 Z
0
X
d X
dy
dy
dt k1 Fk"";t y k1
y
Moreover, if for some 2 [0, 1) the F(k" ", 0) is
nondecreasing as a function of k 2 Z then solution
F(k" ", t) is also nondecreasing as a function of
k 2 Z for all t 0 and the same .
if x > x
Fx; 0 F0 x;
if x 2 x ; x
8
x=t; x 2 t; t
t ! 1 (see Iljin and Oleinik (1960) and Henkin
and Polterovich (1991)). More precise result in
this case about convergence to the so-called
~f x ct ;
~ x Ct ;
F
if x ct
if x Ct
11
and correspondingly
< C <
12
t ! 1 13
x2R
Fx; 0 dx
Fx; 0 dx 0
D0
Remarks
(i) The statements of Theorem 2 give a positive
answer to Gelfands question for the case of
initial problem [1], [3] and [2], [4], admitting
strict shock profiles.
(ii) For linear (f ) = a bf , a > 0, a b > 0,
F for [1], [3] and
b < 0, the traveling waves f,
[2], [4] can be found explicitly:
1 expfpx ctg
b
b
c a ; p
2
2"
~
F
1 expfPx Ctg
a b
a b
C b ln
; P ln
" a b
a b
~f
where
a b
b b 1
a b
14
t!1
The function D0 (), 2 [0, 1], is determined
uniquely from relation
1
X
d0
1
~ Ct D0 fx="gj ! 0;
sup jFx; t Fx
x2R
~ D0 g 0
fFn; 0 Fn
(iii) For initial problems [1], [7] and [2], [8], >
, the asymptotic convergence statements
[13][15] admit the precise asymptotic estimates (see Iljin and Oleinik (1960) for [1], [7]:
sup jf x; t ~f x ct d0 j Oet
k1
x2R
where
F
dy
;
y
~<A
F < A; F
~ Ct D0 fx="gj Oet
sup jFx; t Fx
x2R
17
for
x > ct d0
t t0 ; " 0
sup
j f x; tj ! 0;
t!1
xctd0
Fx; t
15
j Fx; tj
sup
j f x; tj
xctd0
16
sup
449
for
x > Ct D0
18
t t0 ; " 0
xCtD0
sup
xCtD0
j Fx; tj ! 0;
t!1
@
@
@
~f 1 f
" 2
@t
@x
@x
where (x, t) is some smooth function of (x, t) with
values in [0, 1].
Besides, by conservation law of Theorem 0(i), we
have (x, t) ! 0, x ! 1, 8t 0.
Estimates basing on maximum principle and
appropriate comparison statements give that
(x, t) ) 0, x 2 R, t ! 1. It implies that
f x; t ~f x ct d0 ) 0;
x 2 R; t ! 1
n
X
and, correspondingly,
1
X
~
fFk"
" D0 Fk" "; 0g 1
1
Then
Z
~ Ct D0 g
fFn; t Fn
1
dn; t
1 F
dt
~
1 Fn
1; t n; t
where (n, t) is some function with values in [0, 1].
Besides, by conservation law of Theorem 0(ii), we
have
n ! 1; 8 t 0
Estimates, basing on generalized maximum principle and comparison statements, give that
(n, t) ) 0, n 2 Z, t ! 1. It implies that
~ Ct D0 ) 0;
Fn; t Fn
n 2 Z; t ! 1
jf x; t ~f x ct d0 jdx ! 0
1
n; t ! 0;
~
jFn; 0 Fnj
<1
1
and, correspondingly,
1
X
~ Ct D0 j ! 0;
jFn; t Fn
t!1
1
(HenkinShananinTumanov).
(i) Let the initial problem [1], [3] admit the nonstrict
ct) be a
( , )-shock profile [9] and f(x
corresponding traveling-wave solution. Let
0 6 0;
if
c
0 6 0;
if
c
t!1
x2R
where
0
8
0
>
< 1= ;
0
1= ;
>
:
1=0 1=0 ;
if > c
if c >
if c
(ii) Let the initial problem [2], [4] with = 1 admit the
Ct)
nonstrict ( , )-shock profile [10] and F(n
be a corresponding traveling-wave solution. Let
0 6 0;
0 6 0;
Fn; 0 Fn; 0 Fn 1; 0 0
Then there exist constants 0 and D0 such that
~ Ct 0 ln t D0 j ! 0;
sup jFn; t Fn
n2Z
t!1
where
if > C
if C >
if C
if C
0
8
C=20 ;
>
>
>
< C=20 ;
>
C=21=0
>
>
:
1=0 ;
451
if C
Remarks
(i) One could think that nonstrict shock profiles
as in Theorem 4 can appear only in exceptional
cases. But Proposition 2 and Theorem 5 below
and, correspondingly,
p
CtA
X t
~ Ct DA tg
fFk; t Fk
p
kCtA t
p
p
p
Ct A t Ct A tFCt A t 1; t
p
~
FCt
A t 1 Ct DA t 0
20
x2ctA t;ctA t
dy ! 0;
t!1
n
X
sup
fFk; t
p
p
x2CtA t;CtA t kCtApt
~ Ct DA tg ! 0; t ! 1
Fk
Fx; t Fx; t Fx 1; t 0
jFx; tj p ; x; t 2 0 ; t t0
Ct
Then
Fx; t
B
;
Ct
x; t 2 0 ; t t0
where
1 0
1 ln1 a2
B B 0 a2
d
C
d min
x a1 ; a2 x
and B0 is an absolute constant.
It is interesting to compare a priori estimate of
Proposition 1 with some similar (but less precise)
estimates in the theory of classical quasilinear
parabolic equations (Ladyzhenskaya et al. 1968).
We will formulate now the general conjecture
concerning asymptotic behavior of solutions of
and
u
dy
y
x=t
;
if l t x
<
x
gl
l1 t
>
t
>
>
if x > l1 t;
: l1 ;
l 0; 1; . . . ; L
and, correspondingly,
8
b ;
>
> m1
x=t;
<
x >
Gl
>
t
>
>
: am ;
if x < bm t
if bm t x
am1 t
if x > am1 t;
m 0; 1; . . . ; M
(ii) For any interval (l , l ) s and, correspondingly, (am , bm ) S there exist traveling waves
fl (x cl t) for [1] with overfall (l , l ) and,
l 0; . . . ; L 1
cl l ;
l 1; . . . ; L
m 0; . . . ; M 1
Cm am ;
m 1; . . . ; M
m m
L1
L1
X
X
x
~f x c t " t
g
l
l
l
l
l
t
l0
l0
L
X
l ;
l1
M
X
~ m n" Cm t "m t
F
m0
M
1
X
M
1
X
m0
bm
m0
M
X
am ;
Gm
n"
t
M1
m1
l
"
l
8
l
1
>
>
;
>
0
>
>
>
< 1 l
1
0
;
0
>
l
l
>
>
>
1
>
>
;
: 0
l
if m 0 < M; a0 6 b0
if 0 < m < M
if m M > 0; aM 6 bM
n"=t;
if C0 t n"
Fn"; t7!
>
C1 t
>
:~
F1 n" C1 t "1 t; if n" C1 t
The proof of Theorem 5 is of the same nature as
the proof of Theorem 4.
t!1
l
Cm
bm am
8
1
>
>
;
> 0
>
b
>
m
>
>
< 1
1
;
>
0 am 0 bm
>
>
>
>
1
>
>
: 0
;
am
L1
~
Fn";
t; 0 ; . . . ; M
l 0; 1; . . . ; L
t!1
~f x; t; 0 ; . . . ; L
0
m m < 1;
n2Z
Cm bm ;
l0
m ln t O1 m t m ln t O1
~
t; 0 ; 1 ; . . . ; M j ! 0;
sup jFn"; t Fn";
such that
and, correspondingly,
Z bm
1
dy
C1
m
bm am am y
L
X
453
if l 0 < L; 0 6 0
if 0 < l < L
if l L > 0; L 6 L
Remarks
(i) The proof of stronger Conjectures (ii) and (iv)
for L = 1 or M = 1 are in preparation.
(ii) The numerical results, Rykova and Spivak (preprint, 2004), confirm conjecture (iii) for M = 2.
(iii) The results of Weinberger (1990) and Henkin
and Polterovich (1999) confirm convergence
statements of Conjectures (i), (iii) for all L and
M, but only on the intervals of rarefaction
profiles: x 2 [(l )t, (l 1 )t] or, correspondingly, x 2 [(bm )t, (am 1 )t], t > 0.
The problem of finding asymptotics (t ! 1) of
solutions of (viscous) conservation laws has been
posed originally not only for generalized Burgers
equations but also for systems of conservation laws in
one spatial variable (see Gelfand (1959)). In this
direction many important results on existence and
asymptotic stability of viscous shock profiles (continuous and discrete) have been obtained and applied
(see Benzoni-Gavage (2004), Lax (1973), Serre
(1999), Zumbrun and Howard (1998) and references
therein). The results of type of Theorems 4,5 have not
yet been obtained for systems of conservation laws.
It is also very interesting to study asymptotic
behavior of scalar (viscous) conservation laws in
several spatial variables (continuous or discrete),
basing on the asymptotic properties of Burgers type
equations. In this direction there have been several
important results and problems (see Bauman and
Phillips (1986), Henkin and Polterovich (1991),
Hoff and Zumbrun (2000), Serre (1999),
Weinberger (1990), and references therein).
Further Reading
Bauman P and Phillips D (1986b) Large-time behavior of
solutions to a scalar conservation law in several space
dimensions. Transactions of the American Mathematical
Society 298: 401419.
Belenky V (1990) Diagram of growth of a monotonic function and
a problem of their reconstruction by the diagram. Preprint,
CEMI Academy of Science, Moscow, 144 (in Russian).
Benzoni-Gavage S (2002a) Stability of semi-discrete shock profiles
by means of an Evans function in infinite dimension. J.Dyn.
Diff. Equations 14: 613674.
Burgers JM (1940) Application of a model system to illustrate
some points of the statistical theory of free turbulence. Proc.
Acad. Sci. Amsterdam 43: 212.
Dafermos CM (1977) Characteristics in hyperbolic conservation
laws. A study of structure and the asymptotic behavior of
solutions. In: Knops RJ (ed.) Nonlinear Analysis and
Mechanics: HeriotWatt Symposium, vol. 17, pp. 158.
Research Notes in Mathematics, London: Pitman.
Gelfand IM (1959) Some problems in the theory of quasilinear
equations. Usp. Mat. Nauk 14: 87158 (in Russian). ((1963)
American Mathematical Society Translations 33).
Harten A, Hyman JM, and Lax PD (1976) On finite-difference
approximations and entropy conditions for shocks. Communications in Pure and Applied Mathematics 29: 297322.
Henkin GM and Polterovich VM (1991) Schumpeterian dynamics as
a nonlinear wave theory. Journal of Mathematical Economics
20: 551590.
Cellular Automata
M Bruschi, Universita` di Roma La Sapienza, Rome,
Italy
F Musso, Universita` Roma Tre, Rome, Italy
2006 Elsevier Ltd. All rights reserved.
c (1)
c (2)
c (3)
c (4)
c (5)
c (6)
c (7)
t=0
t=1
t=2
t=3
t=4
t=5
t=6
t=7
Figure 1 A seven time-step evolution of CA1 starting from a
given ID (t = 0). Note that a stable configuration has been
reached at t = 6.
1
0
0
0
0
0
0
1
1
0
1
0
1
0
1
1
0
1
0
0
1
1
0
1
1
1
1
0
0
1
1
1
1
2
Sn; t 1 Sn 1; t Sn; t Sn 2; t
Sn; tSn 2; t
Sn 1; tSn; tSn 2; t
3
Sn; t 1 Sn 1; t Sn; t Sn 1; t
2Sn 1; tSn 1; t
4
5
N
X
Sn ki ; i 1; 2;... ;N; ki 2 Z
6
i1
Extensions
Sl n; t
l1
Vector CA
L1
X
L
Y
!
Mk
7
k>l
S1 n; t 1 1 S1 n; tS1 n 1; t
M1 1S2 n 1; tS2 n; t c1
8
S2 n; t 1 2 S1 n 1; tS2 n; t
S1 n; tS2 n 1; t c2
9
Figure 11 A class-4 CA. Note the interacting moving structures on the left and on the right; note also the apparently
2
chaotic behavior in the center; M = 2,V = 0,R = 2,EL: S(n,t 1)=
S(n,t) S(n 1,t) S(n 1,t)S(n 2,t).
10
MooreConway neighborhood
13
14
15
16
17
18
12
9
8
7
6
19
1
1
X
X
Sn1 k1 ; n2 k2 ; t
12
k1 1 k2 1
22
!
11
10
1
2
5
20
21
4
3
0
3
4
21
20
5
2
1
10
11
22
11
18
17
16
15
14
13
14
13
19
6
7
8
9
12
M = 3, V = 0, R = 1. The EL is:
Sn; t 1 Sn 1; t Sn; t 1 Sn 1; t
15
(a)
(b)
(c)
(d)
(e)
(f)
16
M = 2, V = 0, R = 2. The EL is:
Sn; t 1 Sn 1; t 1Sn 2; t
Sn; t Sn 1; tSn 2; t
17
Example 10 (CA9)
The EL is:
A 6 CA: M = 2, V = 0, R = 1.
EL,
according
20
to
[19],
reads
2 ~
~Sn; ~t 1
Sn; ~t 5 ~Sn; ~t 1 ~Sn 1; ~t 2
~Sn 1; ~t 3
~Sn; ~t 2~Sn 1; ~t 2
~Sn; ~t 4~Sn 1; ~t 4
21
(a)
18a
where
j
i 1; 2; . . . ; N j ; ki 2 Z
j 0; 1; 2; . . . ; K 2
18b
19
(b)
Figure 17 CA9, a 6 CA: (a) a 50 time-step evolution from a
peculiar ID; (b) a 50 time-step evolution of the inverse EL, starting
from the last six configurations of Figure 17a (taken in inverse
order); the ID of Figure 17a is recovered (in inverse order).
Example 11 (CA10)
The EL is:
2
Sn; t 1 Sn; t
~j ; t 1
FSn ki ; t; Sn k
M
~
Sn; ~t 1 ~
Sn; ~t M 1
~j ; ~t
F~
Sn ki ; ~t 1; ~
Sn k
Sn 2; t 1Sn 1; t 1
Sn 1; tSn 2; t
22
~j
where ki
and k
are positive integers
~ and F is an arbitrary
(i = 1, 2, . . . , N; j = 1, 2, . . . , N)
(polynomial) function, then it is invertible and
the inverse NEL reads
24
~Sn 2; ~t 1~Sn 1; ~t 1
~Sn 1; ~t~Sn 2; ~t
23
A 1.5 CA, M = 2, V = 0, R = 3.
25
(a)
(b)
(c)
(d)
Figure 18 CA10: (a) 230 time-step evolution, then the inverse EL is applied for 230 further time step in order to recover the initial
configuration. (b) Collisions between different kinds of particle-like coherent moving structures. The last collision (on the right) is
a solitonic one: the interaction produces just a phase shift, preserving number, shape, and velocities of the involved particles.
(c) Particles moving with different velocities and interacting in complex ways (solitonic collisions, particle creations and annihilations).
(d) A particle goes through a nonhomogeneous medium and undergoes refraction by the medium itself.
Applications of CAs
CAs as Universal Constructors and
Turing Machines
c2 ^
y;
c3 ^
x;
c4 ^
y
26
CAs in Physics
(a)
Collisions
Free flight
(b)
Figure 20 (a) An example of configuration for the HPP model.
(b) Head on collisions and three particle collisions in the HPP
model.
27
Other Applications
28a
!2 n2 1 n3 n4 1 n1
1 n2 n3 1 n4 n1
28b
!3 n3 1 n4 n1 1 n2
1 n3 n4 1 n1 n2
28c
!4 n4 1 n1 n2 1 n3
1 n4 n1 1 n2 n3
28d
Further Reading
Berlekamp ER, Conway JH, and Guy R (1982) Winning Ways for
Your Mathematical Plays. London: Academic Press.
Boghosian BM (1999) Lattice gases and cellular automata. Future
Generation Computer Systems 16: 171185.
Boon JP, Dab D, Kapral R, and Lawniczak A (1996) Lattice gas
automata for reactive systems. Physics Reports 273: 55147.
Burks AW (ed.) (1970) Essay on Cellular Automata. Urbana:
University of Illinois Press.
Chopard B and Droz M (1998) Cellular Automata Modeling of
Physical Systems. Cambridge: Cambridge University Press.
Doolen G (ed.) (1990) Lattice Gas Methods for Partial
Differential Equations. New York: Addison-Wesley.
Gardner M (1983) Wheels, Life, and Other Mathematical
Amusements. New York: W H Freeman.
467
von Neumann J (1966) In: Burks AW (ed.) Theory of SelfReproducing Automata. Urbana: University of Illinois Press.
Wolfram S (2002) A New Kind of Science. Champaign: Wolfram
Media.
Introduction
We consider differentiable dynamical systems generated by a diffeomorphism or a vector field on a
manifold. We restrict to the finite-dimensional case,
although some of the ideas can also be developed in
the general case (Vanderbauwhede and Iooss 1992).
We also restrict to the behavior near a stationary
point or a periodic orbit of a flow.
Let the origin 0 of Rn be a stationary point of a C1
vector field X, that is, X(0) = 0. We consider the
linear approximation A = dX(0) of X at 0 and its
spectrum (A), which we decompose as (A) = s [
c [ u , where s resp. c resp. u consists of those
eigenvalues with real part < 0 resp. = 0 resp. > 0. If
c = ; then there is no central manifold, and the
stationary point 0 is called hyperbolic. Let Es , Ec ,
and Eu be the linear A-invariant subspaces corresponding to s resp. c resp. u . Then Rn = Es
Ec
Eu . We look for corresponding X-invariant
manifolds in the neighborhood of 0, in the form of
graphs of maps. More precisely:
Theorem 1 Let the vector field X above be of class
Cr (1 r < 1). There exist map germs ss : (Es , 0) !
Ec
Eu , sc : (Es
Ec , 0) ! Eu , uu : (Eu , 0) ! Es
Ec ,
cu : (Ec
Eu , 0) ! Es , and c : (Ec , 0) ! Es
Eu of
class Cr such that the graphs of these maps are
invariant for the flow of X. Moreover, these maps
are of class Cr , and their linear approximation at 0
is zero, that is, their graphs are tangent to,
respectively, Es , Es
Ec , Eu , Ec
Eu , and Ec . If X is
of class C1 then ss and uu are also of class C1 . If
X is analytic then ss and uu are also analytic.
The graph of c is called the (local) central
(or, center) manifold of X at 0 and it is often
denoted by W c . Thus, it is an invariant manifold
of X tangent at the generalized eigenspace of
dX(0) corresponding to the eigenvalues having zero
real part.
@
@
@
y x2 z2 0
@x
@y
@z
m
X
i1
~ i z1 ; . . . ; zm @
X
@zi
mp
X
zi
im1
n
X
@
@
zi
@zi imp1 @zi
m
X
i1
~ i z1 ; . . . ; zm @
Y
@zi
mp
X
zi
im1
n
X
@
@
zi
@zi imp1 @zi
Pm
@
@Xh
@
x; 0 y
@x
@y
@y
Normal Forms
The general idea of a normal form is to put a
(complicated) system into a form as simple as
possible by means of a change of coordinates. This
idea was already developed to a great extent by
H Poincare. Simple examples are: (1) putting a square
matrix into Jordan form, (2) the flow box theorem
(Arrowsmith and Place 1990) near a nonsingular
point. Depending on the context and on the purpose
of the simplification, this concept may vary greatly. It
depends on the kind of changes of coordinates that are
tolerated (linear, polynomial, formal series, smooth,
analytic) and on the possible structures that must be
preserved (e.g., symplectic, volume-preserving, symmetric, reversible etc.). Let us restrict to local normal
forms, that is, in the vicinity of a stationary point of a
vector field or a diffeomorphism (the latter can be
Let X be a Cr1 vector field defined on a neighborhood of 0 2 Rn , and denote A = dX(0) (its linear
approximation at 0). The Taylor expansion of X at
0 takes the form
Xx A x
r
X
Xk x Ojxjr1
r
X
P
the Taylor series of
(X) is A y 1
k = 2 gk (y). For
practical computations, it is often appropriate to
first simplify the linear part A and to diagonalize it
whenever possible. Hence, it is convenient to use a
complexified setting and to use complex polynomials or power series. One can show that all
involved changes of variables preserve the property
of being a complex system coming from a real
system, that is, at the final stage we can return to a
real system (see, e.g., Arrowsmith and Place (1990)
for a more precise mathematical description).
Hence, we can assume that A is an upper
triangular matrix. Let the eigenvalues be 1 , . . . , n .
It can be calculated that the eigenvalues of LA , as an
operator H k ! P
Hk , are then the numbers h, i j
n
where 2 N , nj= 1 j = k and 1 j n. Hence, if
these would all be nonzero then Bk = H k , and then
we have an ideal simplification, that is, all gk equal
to zero. However, if such a number is zero, that is,
h; i j 0
k2
gk y Ojyjr1
k2
where gk 2 Gk , k = 2, . . . , r.
Sometimes this theorem is applied to the restriction of a vector field to its central manifold, for
reasons explained in the last section. This is the
reason why we did not assume X to be C1 ; in the
latter case one can let r ! 1 and obtain a normal
form on the level of formal Taylor series (also called
1-jets). Using a theorem of Borel, we infer the
existence of a C1 change of variables such that
469
where A
is the adjoint operator. But this choice [3] is
not unique and is, from the computational point of
view, not always optimal, especially if there are
nilpotent blocks. This fact has been exploited by
many authors. A typical example is the case where
A = y@=@x. On the other hand, if A is semisimple we
can choose the complementary space to be ker(LA ), so
LA gk = 0; we can assume it to be the (complex)
diagonal[1 , . . . , n ]. In that case we can be more
explicit as follows. Let ej = @=@xj denote the standard
basis on Cn . For a monomial one can calculate that
LA x ej h; i j x ej
C
jmj
471
Concluding Remarks
The concept of central manifold can be extended to
more general invariant sets (see Chow et al. (2000)
and references therein). It can also be extended to
the infinite-dimensional case and can be applied to
partial differential equations (Vanderbauwhede and
Iooss 1992).
Concerning the generic divergence of normalizing
transformations, the reader is referred to Broer and
Takens (1989), Bruno (1989), Ilyashenko (1981), and
Ilyashenko and Pyartli (1991). Although the power
series giving the normalizing transformation generally
diverges, the study of the dynamics is often performed
by truncating the normal form at a certain order.
Recently, Iooss and Lombardi (2005) considered the
question as to what an optimal truncation is. It is
shown, in case dX(0) is semisimple, that the order of
the normal form can be optimized so that the remainder
satisfies some estimate shrinking exponentially fast to
zero as a function of the radius of the domain.
Concerning normal forms preserving the
Hamiltonian structure, see Birkhoff (1966) and
Siegel and Moser (1995) for a starting point; this is
an extended subject on its own, sometimes called
Birkhoff normal form, and it would require another
review article.
Further simplifications of the normal form can
sometimes be obtained by taking into account
nonlinear terms (instead of just A) in order to obtain
reductions of higher-order terms (see Gaeta (2002)
and especially the references therein).
Applications of normal forms and central manifolds to bifurcation theory have been explained in
Dumortier (1991).
See also: Averaging Methods; Bifurcation Theory;
Dynamical Systems and Thermodynamics; Dynamical
Systems in Mathematical Physics: An Illustration from
Water Waves; Finite Group Symmetry Breaking;
Kortewegde Vries Equation and Other Modulation
Equations; Multiscale Approaches; Normal Forms and
Semiclassical Approximation; Symmetry and Symmetry
Breaking in Dynamical Systems.
Further Reading
Arrowsmith D and Place C (1990) Dynamical Systems. Cambridge:
Cambridge University Press.
Aulbach B (1992) One-dimensional center manifolds are C1 .
Results in Mathematics 21: 311.
Introduction
Consider a typical quantum system such as a string
of ions in a trap. To process the quantum
information the ions carry, we have to perform in
general many steps of a quite different nature.
Typical examples are: free time evolution (including
unwanted but unavoidable interactions with the
environment), controlled time evolution (e.g., the
application of a quantum gate in a quantum
computer), preparations and measurements. Each
processing step can be described by a channel which
transforms input systems into output system of a
possibly different type (e.g., a measurement transforms quantum systems into classical information).
Systems, States, and Algebras
C
-algebras (which are, in our case, always finite
dimensional): quantum systems can be represented
in terms of the algebra B(H) of (bounded) operators
on the Hilbert space H = Cd ; for classical information we have to choose the set C(X) of (continuous),
complex-valued functions on the finite alphabet X;
and the tensor product of both B(H) C(X)
describes hybrid systems which are half-classical
and half-quantum. Assume now that A is one of
these algebras. Effects (i.e., yes/no measurements on
the system in question) are then described by A 2 A
satisfying 0 A 1, states are positive, normalized
linear functionals ! : A ! C, and the probability to
get the result yes during an A measurement on a
system in the state ! is given by !(A). Since A is
assumed to be finite dimensional, each state ! on
B(H) is represented by a density operator , that is,
!(A) = tr(A).P
Likewise, a state ! on C(X) has the
form !(A) = x A(x)px , where (px )x2X denotes a
probability distribution on X, and a state ! on
B(H) C(X) is described by a sequence (x )x2X of
positive
(trace-class) operatorsPon B(H) with
P
tr(
)
=
1 such that !(A) = x tr(x Ax ). Here
x
x
473
Quantum Channels
In this section we will discuss some basic properties
of CP maps which transform quantum systems into
quantum systems, in particular the Stinespring
theorem, which constitutes the most important
structural result. For a more detailed presentation,
including generalizations to more general input/
1
N
X
Vj AVj
2
j1
with operators Vj : H2 ! H1 .
To get a third representation of channels, consider
the Stinespring form [1] of T and a vector 2 K
such that U( ) = V() can be extended to a
unitary map U : H K ! H K. It is then easy to
see that the dual T of T can be written as:
Corollary 4 (Ancilla form). Assume that T : B(H) !
B(H) is a channel. Then there is a Hilbert space K, a
pure state 0 , and a unitary map U : H K ! H K
such that
T trK U 0 U
holds.
3
4
5
6
7
8
with
he0 ; T e0 i d tr
je0 ihe0 j T
1 (U) =
2 (U) = U. All channels of this type are of
the form
TA 1 #A #d 1 trA1
9
where e01 , . . . , e0d0 2 H0 denote an (arbitrary) orthonormal basis of H0 and the transposition of is
defined with respect to the basis e , = 1, . . . , d used
to define in [5].
From the definition of RT in eqn [6], it is obvious
that RT is positive, if T is CP. To see that the
converse is also true is not as trivial (because a
transposition is involved), but it requires only a
short calculation, which is omitted here. Hence, we
get:
Corollary 7 The operator RT is positive, iff the
map T is CP.
Examples
Let us return now to the general case (i.e., arbitrary
input and output algebras) and discuss several
examples.
Channels Under Symmetry
475
10
and T(A)
=
V
A
1V.
P
(ii) If T = T is a decomposition of T in CP and
covariant
summands, there is a decomposition
P
1 = F of the identity operator on K into
positive operators F 2 B(K) with [F ,
(g)]
=0
with # 2 0; d2 =d2 1
11
#
trA1 AT
d1
1 #
trA1 AT ;
d 1
# 2 0; 1
12
y xy x2X
ex y xy
13
14
tr Tx tr Tx 1 trT1 ex
Observables
Instruments
CX 3 f 7! T1 f 2 BK
18
17
It describes the operation performed by the instrument T if x 2 X was measured. More precisely, if a
measurement on systems in the state gives the
result x 2 X, we get (up to normalization) the state
Tx () after the measurement, while
21
Further Reading
Arveson W (1969) Subalgebras of C -algebras. Acta Mathematica
123: 141224.
477
Introduction
Chaos is a type of behavior that can be exhibited by
a large class of physical systems and their mathematical models. These systems are deterministic.
They are modeled by sets of coupled nonlinear
ordinary differential equations (ODEs):
x_ i
dxi
fi x; c
dt
1
called dynamical systems. The coordinates x designate points in a state space or phase space.
Typically, x 2 Rn or some n-dimensional manifold
for some n 3, and c 2 Rk are called control
parameters. They describe parameters that can be
controlled in physical systems, such as pumping
rates in lasers or flow rates in chemical mixing
reactions. The most important mathematical property of dynamical systems is the uniqueness theorem,
which states that there is a unique trajectory through
every point at which f (x; c) is continuous and
Lipschitz and f (x; c) 6 0. In particular, two distinct
periodic orbits cannot have any points in common.
The properties of dynamical systems are governed, in lowest order, by the number, stability, and
distribution of their fixed points, defined by
x_ i = fi (x; c) = 0. It can happen that a dynamical
system has no stable fixed points and no stable
limit cycles (x(t) = x(t T), some T > 0, all t). In
such cases, if the solution is bounded and recurrent
but not periodic, it represents an unfamiliar type of
attractor. If the system exhibits sensitivity to initial
conditions
(jx(t) y(t)j
et jx(0) y(0)j
for
jx(0) y(0)j = and > 0 for most x(0)), the
solution set is called a chaotic attractor. If the
Overview
The program described below has two objectives:
1. classify the global topological structure of strange
attractors in R3 ; and
2. determine the perestroikas (changes) that such
attractors can undergo as experimental conditions or control parameters change.
Boundary
layer
(d)
(c)
Squeeze
Stre
tch
(b)
(a)
Figure 1 A common stretch-and-fold mechanism generates
many experimentally observed strange attractors. The Topology
of Chaos; R Gilmore and M Lefranc; Copyright 2002, Wiley.
This material is used by permission of John Wiley & Sons, Inc.
479
t!1
jxt ytj ! 0
2
Stretch
Shrink
Shrink Shrink
Flow
Flow
xxx
Boundary
layer
Squeeze
Stretch
Flow
Flow
Squeeze
Branch
line
Figure 2 Left: The stretch mechanism is modeled by a twodimensional surface with a splitting point singularity. Right: The
squeeze mechanism is modeled by a two-dimensional surface
with a branch line singularity. The Topology of Chaos; R Gilmore
and M Lefranc; Copyright 2002, Wiley. This material is used
by permission of John Wiley & Sons, Inc.
ab
ba
ab
ba
481
dx = y z
dt
dy
= x + ay
dt
dz = b + z(x c)
dt
(a)
z(t )
4
2
0
x(t )
2
4
(b)
(c)
(e)
(d)
Figure 4 The Rossler dynamical system. (a) Rossler equations. (b) Time series z(t) and x(t) generated by these equations, and
(c) projection of the strange attractor onto the xy plane. (d) Caricature of the flow and (e) knot holder derived directly from the
caricature. Control parameter values (a, b,c) = (2:0, 4:0, 0:398): The Topology of Chaos; R Gilmore and M Lefranc; Copyright 2002,
Wiley. This material is used by permission of John Wiley & Sons, Inc.
50
40
z(t )
30
dx = x + y
dt
dy
= Rx y xz
dt
dz = bz + xy
dt
(a)
20
10
0
x(t )
10
20
(b)
(e)
(c)
(d)
Figure 5 (a) Lorenz equations. (b) Time series x (t) and z(t) generated by these equations, and (c) projection of the strange attractor
onto the xy plane. (d) Caricature of the flow and (e) knot holder derived directly from the caricature by rotating the right-hand lobe by
radians. Control parameter values (R, , b) = (26:0, 10:0, 8=3): The Topology of Chaos; R Gilmore and M Lefranc; Copyright 2002,
Wiley. This material is used by permission of John Wiley & Sons, Inc.
1
0
0
(a)
(b)
b
c
a
a
a
b
(c)
(d)
Parameter
values
ODEs
Rossler
x_ = y z
y_ = x ay
z_ = b z(x c)
Duffing
x_ = y
y_ = y x 3 x
A sin(!t)
x_ = by (c dy 2 )x
y_ = x A sin(!t)
(b, c, d , A, !) =
(0:7, 1:0, 10:0, 0:25, =2)
x_ = x y
y_ = Rx y xz
z_ = bz xy
Lorenz
Figure 7b, serve to determine the topological organization of all the unstable periodic orbits in any
strange attractor with either branched manifold.
The periodic orbits are identified by a repeating
symbol sequence of least period p, which is unique
up to cyclic permutation. The symbol sequence
consists of a string of integers, sequentially identifying the branches through which the orbit passes. For
a branched manifold with two branches, there are
two symbols. The number of orbits of period
p, N(p), obeys the recursion relation
pNp 2p
kp=2
X
1kjp
kNk
3
0 2 1
(a)
(b)
Period
p
Two
Three
branches branches Period
N2 (p)
N3 (p)
p
Two
Three
branches branches
N2 (p)
N3 (p)
1
2
3
4
5
6
7
8
9
10
2
1
2
3
6
9
18
30
56
99
186
335
630
1 161
2 182
4 080
7 710
14 532
27 954
52 377
3
3
8
18
48
116
312
810
2184
5880
11
12
13
14
15
16
17
18
19
20
16 104
44 220
122 640
341 484
956 576
2 690 010
7 596 480
21 522 228
61 171 656
174 336 264
483
Table 3 Linking numbers of orbits to period 5 in the Smale horseshoe branched manifold with zero global torsion
21
31
31
41
42
42
51
51
52
52
53
53
0
1
01
011
001
0111
0011
0001
01111
01101
00111
00101
00011
00001
21
31
31
41
42
42
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
2
1
1
2
2
2
2
1
1
0
1
1
2
2
3
2
2
4
4
3
3
2
2
0
1
2
2
3
4
3
3
5
5
5
5
3
3
0
1
2
3
2
4
3
3
5
5
4
4
3
3
0
2
3
4
4
5
4
4
8
8
7
7
4
4
0
1
2
3
3
4
3
4
5
5
5
5
4
4
0
1
2
3
3
4
4
3
5
5
5
5
4
4
dr A t1 dr B t2
4
jr A t1 r B t2 j3
This table of experimental integers is compared with
the table of linking numbers for orbits with the same
symbolic name on a trial branched manifold. This
procedure serves to identify the branched manifold
and refine the symbolic identifications of the
experimental orbits, if necessary. The procedure is
vastly overdetermined. For example, the linking
numbers of only three low-period orbits serve to
identify the four pieces of information required to
specify a branched manifold with two branches.
Since six or more surrogate periodic orbits can
typically be
extracted from experimental data,
providing 62 = 15 or more linking numbers, this
topological analysis procedure has built-in selfconsistency checks, unlike analysis procedures
based on geometric and dynamical tools.
51
0
2
4
5
5
8
5
5
8
10
9
9
5
5
51
0
2
4
5
5
8
5
5
10
8
8
8
5
5
52
52
53
53
0
2
3
5
4
7
5
5
9
8
6
7
5
5
0
2
3
5
4
7
5
5
9
8
7
6
5
5
0
1
2
3
3
4
4
4
5
5
5
5
4
5
0
1
2
3
3
4
4
4
5
5
5
5
5
4
0.70
ln 2
815
78
0.65
Forcing of horseshoe
orbits to period 8
64
0.60
810F 811R
85F 86F 8
8
0.50
Entropy
810R 811F
52
0.55
813R 814R
85R 86R
0.45
813F 814F
73F 74F
72
73R 74R
76R 77R
0.40
76F 77F
84R 87R
0.35
84F 87F
Other
finite order
Period
doubled
Well
ordered
61
82
21 41 81
71 51
812
63
83 31
75
62
42
89
53
65 79 816
(a)
U-sequence order
Wo :
1/2
3/7 2/5
2/7 1/4
3/8 1/3
1/5
1/6
1/7 1/8
Braids
PD
QOD : f
21 41 81 61 82 71 51 72 83 31 62 84 73 85 52 86 74 87 63 88 75 42
2/5
1/3
1/4
1/5
1/6
(b)
Figure 8 (a) Forcing diagram for orbits up to period 8 in the Smale horseshoe branched manifold. (b) The sequence (universal
order) in which orbits are created in the highly dissipative limit, which is the logistic map. The Topology of Chaos; R Gilmore and
M Lefranc; Copyright 2002, Wiley. This material is used by permission of John Wiley & Sons, Inc.
logistic map in the transition from simple, nonchaotic behavior to fully chaotic (hyperbolic)
behavior.
The orbits in the forcing diagram are organized
according to their one-dimensional entropy
(horizontal axis, U-sequence order) and their twodimensional entropy (vertical axis). Nonchaotic
(laminar) behavior occurs at the lower left of
this figure, where both entropies are zero. Fully
chaotic behavior occurs at the upper right, where
both entropies are ln 2. As control parameters
change, a dynamical system that can exhibit chaos
generated by a stretch-and-fold mechanism follows a
path in the forcing diagram from the lower left to
the upper right. Each such path is a route to
chaos. The Smale horseshoe mechanism exhibits
many different routes to chaos: each follows a
different path in the forcing diagram.
The state of a strange attractor at any stage in its
route to chaos can be specified by a basis set of
orbits. This is a set of orbits whose presence forces
the existence of all other orbits that can concurrently be found in the attractor, up to any finite
485
Bounding Tori
As experimental conditions or control parameters
change, strange attractors can undergo grosser
perestroikas than those that can be described by a
change in the basis set of orbits. This occurs when new
orbits are created that cannot be contained on the initial
branched manifold for example, when orbits are
created that must be described by a new symbol. This is
seen experimentally in the transition from horseshoe
type dynamics to gateau roule type dynamics. This
involves the addition of a third branch to the branched
manifold with two branches, as shown in Figures 7a
and 7b. Strange attractors can undergo perestroikas
described by the addition of new branches to, or
deletion of old branches from, a branched manifold.
These perestroikas are in a very real sense grosser
than the perestroikas that can be described by changes
in the basis sets of orbits on a fixed branched manifold.
There is a structure that provides constraints on
the allowed bifurcations of branched manifolds
(creation/annihilation of branches), which is analogous to the constraints that a branched manifold
provides on the bifurcations and topological organization of the periodic orbits that can exist on it. This
structure is called a bounding torus.
Bounding tori are constructed as follows. The semiflow on a branched manifold is inflated or blown
up to a flow on a thin open set in R3 containing this
branched manifold. The boundary of this open set is a
two-dimensional surface. Such surfaces have been
classified. They are uniquely tori of genus g; g = 0
(sphere), g = 1 (tire tube), g = 2, 3, . . . . The torus of
genus g has Euler characteristic = 2 2g. The flow is
into this surface. The flow, restricted to the surface,
exhibits a singularity wherever it is normal to the
surface. At such singularities the stability is determined
by the local Lyapunov exponents: 1 > 0 and 3 < 0,
since the flow direction (2 = 0) is normal to the
7
A
E
a
1
2
B
b
a
4
b
4
E
7
ABCBDED
abbacca
(a)
6
B
c
E
ABCDCBE
abccbaa
(b)
ABCBDBE
abbccaa
(c)
Figure 9 Three inequivalent canonical forms of genus 8 are shown. Each is identified by a period-7 orbit and its dual. Reprinted
figure with permission from Physical Review E, 69, 056206, 2004. Copyright (2004) by the American Physical Society.
N(g)
N(g)
N(g)
3
4
5
6
7
8
1
1
2
2
5
6
9
10
11
12
13
14
15
28
67
145
368
870
15
16
17
18
19
20
2 211
5 549
14 290
36 824
96 347
252 927
g!1
lnNg
ln 3
g1
5
Table 5 All known strange attractors of dimension dL < 3 are bounded by one of the standard dressed tori. Dual labels for the
bounding tori depend on g 1 symbols describing holes with or without singularities
Strange attractor
A
A
A
A
AB
AB
ABCD
ABCB
ABCB
AEBECEDE
AB N
AB (2N)
(AZ )(BZ ) (NZ )
A(B M)N(B M)1
Genus
aa
a2
a4
abba
ab(ab)1
a2 b 2 c 2 d 2
an
a 2n
a2 b 2 n 2
(ab m)(ab m)1
1
1
1
1
3
3
5
5
5
9
n1
2n 1
2n 1
2m 1
487
(a)
Further Reading
(b)
(c)
Characteristic Classes
P B Gilkey, University of Oregon, Eugene, OR, USA
R Ivanova, University of Hawaii Hilo, Hilo, HI, USA
S Nikcevic, SANU, Belgrade, Serbia and Montenegro
2006 Elsevier Ltd. All rights reserved.
Vector Bundles
Let Vectk (M, F) be the set of isomorphism classes of
real (F = R) or complex (F = C) vector bundles of
rank k over a smooth connected m-dimensional
manifold M. Let
[
VectM; F
Vectk M; F
k
:P!M
with fiber H is said to be a principal bundle if there
is a right action of H on P which acts transitively on
the fibers, that is, if P=H = M. If H is a closed
subgroup of a Lie group G, then the natural
projection G ! G=H is a principal H bundle over
the homogeneous space G=H. Let O(k) and U(k)
denote the orthogonal and unitary groups, respectively. Let Sk denote the unit sphere in Rk1 . Then
we have natural principal bundles:
Ok Ok 1 ! Sk
Uk Uk 1 ! S2k1
Let RPk and CPk denote the real and complex
projective spaces of lines through the origin in Rk1
and Ck1 , respectively. Let
Z2 f
Idg Ok
S1 f Id : jj 1g Uk
One has Z2 and S1 principal bundles:
Z2 ! Sk1 ! RPk1
S1 ! S2k1 ! CPk1
Frames
Characteristic Classes
P B Gilkey, University of Oregon, Eugene, OR, USA
R Ivanova, University of Hawaii Hilo, Hilo, HI, USA
S Nikcevic, SANU, Belgrade, Serbia and Montenegro
2006 Elsevier Ltd. All rights reserved.
Vector Bundles
Let Vectk (M, F) be the set of isomorphism classes of
real (F = R) or complex (F = C) vector bundles of
rank k over a smooth connected m-dimensional
manifold M. Let
[
VectM; F
Vectk M; F
k
Frames
The maps
: O \ O ! GL(F, k) satisfy
Id on O
on O \ O \ O
Spin Structures
1
2 GL R; k
for P 2 O \ O
;
h1 ; h
The associated vector bundle is then given by
P Fk : P Fk=
Clearly, { } are the transition cocycles of the
vector bundle P Fk .
Homotopy
L L l
N; M Vectk M; F ! Vectk N; F
If M is contractible, then the identity map is
homotopic to the constant map c. Consequently,
V = Id V is isomorphic to c V = M Fk . Thus, any
vector bundle over a contractible manifold is trivial.
In particular, if {O } is a simple cover of M and if
V 2 Vect(M, F), then VjO is trivial for each . This
shows that a simple cover is a trivializing cover for
every V 2 Vect(M, F).
Stabilization
for k > m
Pullback Bundle
f
ff s gP P;
ff
f Ps f P
f s gP
2
if k > m
g
Vectk M; C KCM
if 2k > m
M; Grk F; 1 Vectk M; F
k
g
Z fL l g=2k ZfL l g
KRRP
Vk F; n : f
; x 2 HomFn Fn :
x xg
for > m
6
3
M; Grk R; k Vectk M; R
5
H Grk C; 1; Z Zc1 ; . . . ; ck
7
swTRPk 1 xk1
1 kx
k 1k 2
x
2
8
1.
2.
3.
4.
pTCPk 1 x2 k1
Chern Classes
Line Bundles
HdeR M :
kerd : C1 p M ! C1 p1 M
imd : C1 p1 M ! C1 p M
If s =
Connections
i d!i !ki ^ !k
js
i j
~ dgg1 g!g1
!
and
~ gg1
rX s hX; rsi
1.
2.
3.
4.
rfX s = f rX s.
rX (fs) = X(f )s f rX s.
rX1 X2 s = rX1 s rX2 s.
rX (s1 s2 ) = rX s1 rX s2 .
0
Thus, can be regarded as a 2-form-valued element
of the Lie algebra of the structure group, O(V) in the
real setting or U(V) in the complex setting.
Projections
r : C1 p M
V ! C1 p1 M
V
can be extended by defining
r!p
s d!p
s 1p !p ^ rs
In contrast to ordinary exterior differentiation, r2
need not vanish. We set
s : r2 s
V dV dV
fs ddf
s df ^ rs df ^ rs f r2 s
f s
f P iVP
1 c1 ck
Again ci () is independent of the local gauge and
dci () = 0. The de Rham cohomology class [ci ()]
represents ci (V) in H 2i (M; C).
The Chern Character
2 !
ch0 ch1
Let ch(V) = [ch()] denote the associated de Rham
cohomology class; it is independent of the particular
connection chosen. We then have the relations
chV W chV chW
chV
W chVchW
The Chern character extends to a ring isomorphism from KU(M)
Q to H e (M; Q), which is a
natural equivalence of functors; modulo torsion,
K theory and cohomology are the same functors.
f x :
k
X
ex
1 td1 x
If V is a real vector bundle, we can define
some
additional
characteristic classes similarly. Let
p
and
x
2 sinh1=2x
and
^ i V 2 H 4i M; C
A
Summary of Formulas
(V),
12
2
24
p1 7p21 4p2
(V),
6.
A(V) = 1
24
5760
p1 7p2 p21
7.
L(V) = 1
(V),
3
45
8.
9.
10.
X "I;J Ri
I; J
Rim1 im jm jm1
8n n!
1 i2 j2 j1
e2
2 P, let
td(V W) = td(V)td(W),
A(V
W) = A(V)
A(W),
L(V W) = L(V)L(W).
V
: ker : T
P ! T
M and
The Euler Form
pn (A) = x21
x2n
is a perfect
e2n A : x1 xn
is the square root of pn . If V is an oriented vector
bundle of dimension 2n, then
e2n V 2 H 2n M; C
is a well-defined characteristic class satisfying
e2n (V)2 = pn (V).
If V is the underlying real oriented vector bundle
of a complex vector bundle W,
e2n V cn W
H
: V ?
Acknowledgmnts
Research of P Gilkey was partially supported by
the MPI (Leipzig, Germany), that of R Ivanova by
the UHH Seed Money Grant, and of S Nikcevic by
MM 1646 (Serbia), DAAD (Germany), and Dierks
von Zweck Stiftung (Esen, Germany).
See also: Cohomology Theories; Gerbes in Quantum
Field theory; Instantons: Topological Aspects; K-Theory;
Mathai-Quillen Formalism; Riemann Surfaces.
Further Reading
Besse AL (1987) Einstein manifolds. Ergebnisse der Mathematik
und ihrer Grenzgebiete (3) [Results in Mathematics and
Related Areas (3)], p. 10. Berlin: Springer-Verlag.
Introduction
The relationship between topological invariants and
functional integrals from quantum ChernSimons
theory discovered by Witten (1989) raised several
challenges for mathematicians. Most of the tremendous amount of mathematical activity generated by
Wittens discovery has been concerned primarily with
issues that arise after one has accepted the functional
integral as a formal object. This has left, as an
important challenge, the task of giving rigorous
meaning to the functional integrals themselves and to
rigorously derive their relation to topological invariants. The present article will discuss efforts to put the
functional integral itself on a rigorous basis.
3
R3
8
9
where X = S(R )
LG is the space of rapidly
decreasing functions with values in the Lie algebra
LG. Let
!1
d2
x2
T1 2
dx
4
as a linear operator on L2 (R 3 ), T2 = T1
3
I the
induced operator on L2 (R 3 )
LG, and T = T2 T2 .
Then, as described in the preceding section, we have
the space E and its dual E 0 . There is then the
standard Gaussian measure on E 0 , and the space
[E]0 of distributions over E 0 .
The normalized ChernSimons integral may be
viewed as a linear functional
Z
1
CS : F 7!
eik=4SCS A FADA
10
N E
where N is a normalizing factor. Rigorous meaning can be given to this by first formally working out
what the S-transform of CS ought to be. Calculation shows that S is indeed a holomorphic function
on E c of quadratic growth. The PotthoffStreit
theorem then implies that CS does exist as a
distribution in the space [E]0 . Let us examine this
in some more detail.
As before, we take A to be of the form
A = A0 dx0 A1 dx1 , with the component A2 equal
to 0. Integration by parts shows that
Z
k
k
SCS A
trA0 @2 A1 dvol
11
4
2 R3
d l
P t BlA tPlA t 0
dt A
13
The equation of parallel transport can be reformulated as a differential equation for a matrix-valued
path t 7! PlA (t) satisfying
14
and the initial condition PlA (t) = I. With this smearing, one can consider functions of the form
W L; A
n
Y
trPli A
15
i1
Acknowledgments
This research is supported by US NSF grant DMS0201683.
See also: BF Theories; Feynman Path Integrals;
Fractional Quantum Hall Effect; Knot Theory and
Physics; Large-N and Topological Strings; Large-N
Dualities; Quantum 3-Manifold Invariants; Quantum Hall
Effect; Spin Foams; String Field Theory; Topological
Quantum Field Theory: Overview; Twistor Theory: Some
Applications.
Further Reading
Albeverio S, Hahn A, and Sengupta AN (2003) ChernSimons
theory, Hida distributions, and state models. Infinite Dimensional Analysis Quantum Probability and Related Topics
6: 6581.
Albeverio S and Schafer J (1994) Abelian ChernSimons
theory and linking numbers via oscillatory integrals.
Journal of Mathematical Physics (N.Y.) 36 (suppl.
5): 21352169.
Albeverio S and Sengupta A (1997) A mathematical construction
of the non-Abelian ChernSimons functional integral. Communications in Mathematical Physics 186: 563579.
Altschuler D and Freidel L (1997) Vassiliev Knot invariants and
ChernSimons perturbation theory to all orders. Communications in Mathematical Physics 187: 261287.
Atiyah M (1990) The Geometry and Physics of Knot Polynomials. Cambridge: Cambridge University Press.
Bar-Natan D (1995) Perturbative ChernSimons theory. Journal
of Knot Theory and its Ramifications 4: 503.
Chern S-S and Simons J (1974) Characteristic forms and
geometric invariants. Annals of Mathematics 99: 4869.
501
Flag Manifolds
These homogeneous spaces F = G=P with semisimple (in our case with classical) groups G have
parabolic subgroups P as the isotropy subgroups.
The group G = GL(n; C) transitively acts on the
flag manifolds F(n1 , . . . , nr ), 0 < n1 < < nr < n,
whose elements are (n1 , . . . , nr )-flags sequences of
embedded subspaces in Cn of the dimensions
(n1 , . . . , nr ). The isotropy subgroup P = P(n1 , . . . , nr )
is the subgroup of blocktriangle matrices with the
diagonal blocks of sizes k1 , . . . , kr1 , kj = (nj
nj1 ), n0 = 0, nr1 = n. The flag manifolds are compact complex manifolds. The matrices proportional
to the unit matrix En act trivially and we can
consider instead of the action of G = GL(n; C) the
transitive action of G = SL(n; C).
Let us pay particular attention to two extremal
cases. The first one is the case of the maximal
flag manifold when we have the sequence of
all integers (1, 2, 3, . . . , n 1) complete flags; the
subgroup P in this case is called Borelian. Another
case is minimal flag manifolds with r = 1 (for them
the unipotent radical of the parabolic subgroups is
commutative). Then in the case of SL(n; C) the
sequence has only one element n1 = k < n and we
have Grassmannian manifolds GrC (k; n) = F(k) of
k-dimensional subspaces in Cn . If k = 1 or k = n 1,
we obtain the dual realizations of the complex
projective space CPn1 . We can interpret points
of GrC (k; n) also as (k 1)-dimensional planes in
CPn1 .
We can define points of the projective space
CPn1 by homogeneous coordinates as the
equivalency classes (z cz, z 2 Cn n {0}, c 2 C n 0).
For the Grassmannians we can similarly use matrix
homogeneous coordinates (Stiefels coordinates):
classes of (k n)-matrices Z 2 Mat(k, n) of the
maximal rank k relative to the equivalency
Z uZ;
u 2 GLk; C
503
Compact Classical
Homogeneous Manifolds
Compact classical groups U(n), SU(n), O(n), SO(n),
Sp(l) are maximal compact subgroups in the corresponding classical complex groups GL(n; C), SL(n; C),
O(n; C), SO(n; C), Sp(l; C). This condition defines
them up to an isomorphism. They are fixed subgroups
of some antiholomorphic involutive automorphisms.
The unitary groups U(n) and SU(n) are the groups
of unitary matrices (g
g = E,) correspondingly, of
unitary matrices with determinant 1. As the compact
orthogonal group we can take the intersection U(n) \
O(n; C). For the standard form I, it will be the group of
real orthogonal matrices: g> g = E (so the involution in
O(n; C) is the conjugation g 7! g). Similarly, we can
take Sp(l) = SU(2l) \ Sp(l; C) (then the involution is
g 7! JgJ).
Compact classical groups act on compact homogeneous Riemann manifolds. There are two mechanisms
connecting
compact
and
complex
homogeneous manifolds. We observe the first
possibility in the case of flag manifolds which are
Noncompact Riemannian
Symmetric Manifolds
This class of symmetric manifolds has the strongest
connections with classical mathematics. Let us
consider noncompact real semisimple Lie groups
real forms of complex semisimple Lie groups. They
correspond to antiholomorphic involutions in complex groups.
Between real forms of SL(C, n) there are real and
quaternionic unimodular groups SL(R, n), SL(H, n)
and pseudounitary groups SU(p, q) of complex
matrices preserving a Hermitian form H of the
signature (p, q). The complex orthogonal group has
as real forms, in particular, pseudoorthogonal
groups SO(p, q) of real matrices preserving a
quadratic form of the signature (p, q).
Let G be a real simple Lie group and K be its
maximal compact subgroup. Then X = G=K is a
Riemann symmetric manifold of noncompact type;
K is defined by an involutive automorphism of G.
Therefore, in irreducible situation there is a correspondence between noncompact Riemann symmetric manifolds and real simple noncompact Lie
groups. K-orbits on X are parametrized by points of
the orbit on X of a maximal abelian subgroup A
the Cartan subgroup of the symmetric space X. Its
dimension l is the important invariant of X its
rank. The algebraic base for geometry of X is the
Iwasawa decomposition
G KAN
where N is a maximal unipotent subgroup (in a
natural sense compatible with A). Then the parabolic subgroup P = AN is transitive on X.
505
u 2 GLk; C; a 2 Hermk
H2 H1 ;
Ek
2 iEk
Then the transform Z 7! Z transforms X2 in X1 . In
inhomogeneous coordinates it is the fractional linear
matrix transform
z 7! iz iEk 1 z iEk
It is the matrix version of the classical Cayley transform.
Similarly, we can write down the inverse transform.
If q 6 k, then there is also an analog of the tube
realization. Let r = q k > 0 and
0
1
0
0
iEk
H2 @iEk 0
0 A
0
0 Er
1
z z 0;
2i
z z>
of complex symmetric matrices with positive imaginary parts; it is called Siegels half-plane. This is
the third class of Cartans domains. There are Siegel
domains of second kind connecting with the cones
of positive symmetric matrices; some of them are
homogeneous, but they are never symmetric.
There are two more series of classical minimal flag
manifolds: the isotropic Grassmannians and quadrics.
They both contain the dual bounded symmetric
domains (Cartans domains of second and fourth
classes correspondingly). Some of these domains in
the isotropic Grassmannians admit the realizations as
tubes with the cone of positive Hermitian quaternionic
matrices and others as Siegel domains of the second
kind corresponding to the same cones.
Symmetric domains in quadrics can be realized as
tube domains with the Lorentzian (light) cones.
507
g
Qg Q
ZQZ
> 0
w u iv
SU2; 2 SO2; 3
509
2 Cn1 ; & 1
Further Reading
Akhiezer D (1990) Homogeneous complex manifolds. In: Gindikin
S and Henkin G (eds.) Several Complex Variables IV, vol. 10,
Encyclopaedia of Mathematical Science. New York: Springer.
DAtri J and Gindikin S (1993) Siegel domain realization of
pseudo-Hermitian symmetric manifolds. Geometriae Dedicata
46: 91126.
Introduction
The notion of classical r-matrices has emerged as
a by-product of the quantum inverse scattering
method (which was developed mainly by L D
Faddeev and his team in their work at the Steklov
Mathematical Institute in Leningrad); it has given a
new insight into the study of Hamiltonian structures
associated with classical integrable systems solvable
by the classical inverse scattering method and its
generalizations. Important classification results for
classical r-matrices are due to Belavin and Drinfeld.
Based on the initial results of Sklyanin, Drinfeld
introduced the important concepts of Poisson Lie
groups and Lie bialgebras which arise as a
semiclassical approximation in the study of quantum groups.
A Poisson group is a Lie group G equipped
with a Poisson bracket such that the multiplication m : G G ! G is a Poisson mapping. A
Poisson bracket on G with this property is called
multiplicative. More explicitly, let x , x be the
left and right translation operators in C1 (G) by
an element x 2 G, x (y) = (xy), x (y) = (yx).
2 FG;
x; y 2 G
f; g f; g
X; Y X I I X; Y
Y I I Y; X
2
operator
f; gx hxrx; r i
:G !
3
4
Equation [4] implies that (e) = 0, that is, a multiplicative Poisson structure is identically zero at the
unit element. Its linearization at this point induces
the structure of a Lie algebra on the cotangent space
Te G g ; namely, for any , 0 2 g , choose , 0 2
F(G) in such a way that re = , re 0 = 0 , and set
; 0 re f; 0 g
5
for all X 2 g
6
7
8
c const 6 0
9
10
Theorem 2
(i) There exists a unique structure of the Lie algebra
on d such that: (a) g, g
d are Lie subalgebras.
(b) The inner product [10] is invariant.
(ii) Let Pg , Pg be the projection operators onto
g, g
d parallel to the complementary subalgebra. Set rd = Pg , rd = Pg ; then (d, rd ) is a
factorizable Lie bialgebra.
(iii) The inclusion map (g, g ) V (d, d ) is a homomorphism of Lie bialgebras and the dual inclusion
map (g , g) V (d, d ) is an antihomomorphism.
Conversely, let a be a Lie algebra equipped with a
nondegenerate invariant inner product, a
a its Lie
subalgebras such that (i) a are isotropic with respect
to inner product, (ii) a = a. a as a linear space.
The triple (a, a , a ) is called a Manin triple. Let
P be the projection operators onto a in this
decomposition. Set r = P . Then (a, r ) is a
factorizable Lie bialgebra; moreover, a and a are
set into duality by the inner product in a and inherit
the structure of a Lie bialgebra, and a is their double.
If (g, g ) is itself a factorizable Lie bialgebra, its
double admits a simple explicit description. Set
d = g g (direct sum of Lie algebras); let us equip
d with the inner product
hhX; X0 ; Y; Y 0 ii hX; Yi hY; Y 0 i
Let g
d be the diagonal subalgebra; we identify
g with the embedded subalgebra ir (g )
d.
Proposition 3
(i) (d, g , ir (g )) is a Manin triple.
(ii) As a Lie algebra, d = g g is isomorphic to the
double of g.
Key examples of factorizable Lie bialgebras are
associated with semisimple Lie algebras and their
loop algebras.
1. Let k be a compact semisimple Lie algebra: g = kC
its complexification regarded as a real Lie algebra,
2 Aut g the Cartan involution which fixes k, and
g = k p the associated Cartan decomposition.
Fix a real split Cartan subalgebra a
p and the
associated Iwasawa decomposition g = k. a. n;
put s = a. n. Let B be the complex Killing form
on g; let us equip g with the real inner product
(X, Y) = Im B(X, Y), then (g, k, s) is a Manin
e ^ e
2
1X
Hi Hi
2 i
11
Let b = h. n be the opposite Borel subalgebras; the inner product in g sets them into
duality, and (b , b ) is a Lie sub-bialgebra
in (g, g ). Let G be the connected, simply
connected Lie group associated with g, B =
HN its opposite Borel subgroups which correspond to b . Let p : B ! B =N H be the
canonical projection. The associated factorization problem in G, g = b b1
, (b , b ) 2 B
B , p(b ) = p(b )1 , is closely related to the
Bruhat decomposition; it is solvable for all g in
the open Bruhat cell B N
G.
3. Let Lg = g C((z)) be the loop algebra of a finite
dimensional semisimple Lie algebra g, as usual we
denote the ring of formal Laurent series by C((z)).
Put Lg = g C[[z]], Lg = g z1 C[z1 ]. Fix an
invariant inner product on g and equip Lg with
the inner product
hhX; Yii Resz0 hXz; Yzi dz
Then (Lg, Lg , Lg ) is a Manin triple. The associated classical r-matrix is called rational r-matrix; in
tensor notation, it is represented by a singular kernel
rz; z0
t
z z0
...
" e2 i=n
12
z 7! z
: h2 Ad I2
Let GE
Lg be the subspace of Laurent expansions
at zero of the global meromorphic sections of P
with a unique pole at 0 2 E. Then (Lg, Lg , GE ) is
again a Manin triple. The associated classical
r-matrix is the kernel of a singular integral operator
which associates a meromorphic section of P to its
principal part at 0. Explicitly, it is given by
n1
1X
z z0
0
rz z
a b
n a;b0
n
13
Ad Ia;b I t
where is the Weierstrass zeta function.
5. Let g be an arbitrary semisimple Lie algebra
again. Let us equip the loop algebra Lg with the
inner product
hhX; Yii0 Resz0 hXz; Yziz1 dz
_ g zC[[z]], N = n
_ g
Set N = n
_ h
_ N , where
z1 C[z1 ]. We have Lg = N
we identify h, n
g with the corresponding
subalgebras of constant loops in Lg. Let P , P0
be the projection operators onto N , h in this
decomposition and r = P (1=2)P0 . The
classical r-matrices r define on Lg the structure
of a factorizable Lie bialgebra. The associated
tensor kernels are called the trigonometric classical r-matrices.
Classical r-matrices described above are associated
with factorization problems in the infinite-dimensional
loop groups: matrix Riemann problems or matrix
Cousin problems (in the elliptic case). Belavin and
r2g^g
L1 L I; L2 I L 15
ir (G )
G G may be characterized in terms of the
matrix coefficients of (h , h ) = ir (h), or of their
quotient h = h h1
. Explicitly, we get
h1 ; h2
1 2
r; h1 h2 ;
h ; h r ; h1 h2 16
fh1 ; h2 g rh1 h2 h1 h2 r h2 r h1 h1 r h2 ;
r 12 r r
17
18
GnD G
D
p. &p0
G D=G
G nD G
linear operators). Equations [19] give the compatibility conditions for the auxiliary linear system
m1
Lm
m;
d m
Mm
dt
m2Z
19
m 2 Z 20
m;
exptrTL0
0 1
m ;
Further Reading
Babelon O, Bernard D, and Talon M (2003) Introduction to Classical
Integrable Systems. Cambridge: Cambridge University Press.
Belavin AA and Drinfeld VG (1984) Triangle equations and simple
Lie algebras. In: Mathematical physics reviews, vol. 4, Soviet
Scientific Reviews Section C Mathematical Physics Reviews,
pp. 93165. Chur: Harwood Academic Publishers, Reprinted in
Introduction
Introductory and Historical Remarks
2
xy
xy
z
!2
= x2 y2 z2
519
p
1. Multiplication on the right by the quaternion
unit
p
i is realized as the multiplication (on the left) by
1. If j and k = ij are the other two quaternion
s) and sk = sij.
units and s 2 S, then one puts sj = C(
A real vector space S can be complexified by
forming the tensor product C R S = S iS.
The realification of a complex vector space S is the
real vector space having S as its set of vectors so that
dimR S = 2 dimC S. The complexification of a realification of S is the double S S of the original space.
Inner-Product Spaces and Their Groups
Algebras
3
4
5
H R H = R4
An algebra over R can be complexified by complexifying its underlying vector space; it follows from [5]
that C(2) is the complex algebra obtained by
complexification of the real algebra H.
The center of an algebra A is the set
ZA = fa 2 A j ab = ba 8 b 2 Ag
The center is a commutative subalgebra containing
K. An algebra over K is said to be central if its center
coincides with K. The algebras R(N) and H(N) are
central over R. The algebra C(N) is central over C,
but not over R.
Simplicity and representations Let B1 and B2
be subsets of the algebra A. Define B1 B2 = fb1 b2 j
b1 2 B1 , b2 2 B2 g. A vector subspace B of A is said
to be a left (resp., right) ideal of A if AB B (resp.,
BA B). A two-sided ideal or simply an ideal is
a left and right ideal. An algebra A 6 f0g is said to
be simple if its only two-sided ideals are f0g and A.
For example, the algebras R(N) and H(N) are
simple over R; the algebra C(N) is simple when
considered as an algebra over both R and C; every
associative, finite-dimensional simple algebra over R
or C is isomorphic to one of them.
A representation of an algebra A over K in a vector
space S over K is a homomorphism of algebras : A !
End S. If is injective, then the representation is said to
be faithful. For example, the regular representation :
A ! End A of an algebra A, defined by (a)b = ab
for all a, b 2 A, is faithful. A vector subspace T of
the vector space S carrying a representation of A
is said to be invariant for if (a)T T for every
a 2 A; it is proper if distinct from both f0g and S.
For example, a left ideal of A is invariant for the
regular representation. Given an invariant subspace
T of one can reduce to T by forming the
representation T : A ! End T, where T (a)s = (a)s
for every a 2 A and s 2 T. A representation is
irreducible if it has no proper invariant subspaces.
A linear map F : S1 ! S2 is said to intertwine the
representations 1 : A ! End S1 and 2 : A ! End S2 if
F1 (a) = 2 (a)F holds for every a 2 A. If F is an
521
Graded Algebras
7
C ! H;
and
H ! C2
523
Cf
!
!
f
CW; h
"
W
i : CV; g ! ^V
9
CV; g = T V=J V; g
Clifford Algebras
8
va = v ^ a gvca
10
so that [v, a] = 2g(v)ca, where [ , ] is the supercommutator. It defines a super Lie algebra structure in the
vector space K V. The quadratic form defined by g
need not be nondegenerate; for example, if it is the
"
K2
"
C0 K2 ; h !
11
= e1 e2
em
If
0 is the volume element associated with another
orthonormal frame e0 in the same orthogonal space,
then either
0 =
(e and e0 are of the same
orientation) or
0 =
(e and e0 are of opposite
orientation). For K = C, one has
2 = 1; for K = R
and g of signature (k, l) one has
2 = 11=2klkl1
13
KK
12
14
15
R !
6"
C
5"
H
4
Proposition (K)
Ck, l :
RR
HH
R
#1
C
#2
H
17
525
From general theory (Chevalley 1954) or by inspection of [14], [15], and [17], one has
Proposition (L) Let m be the dimension of the
orthogonal space (V, g) over K.
(i) If m is even (resp., odd), then the algebra
C(V, g) (resp., C0 (V, g)) over K is central simple.
(ii) If K = C and m is odd (resp., even), then the
algebra C(V, g) (resp., C0 (V, g)) is the direct
sum of two isomorphic complex central simple
algebras.
(iii) If K = R and m is odd (resp., even), then the
algebra C(V, g) (resp., C0 (V, g)) when
2 = 1 is
the direct sum of two isomorphic central simple
algebras and when
2 = 1 is simple with a
center isomorphic to C.
Representations
The Pauli, Cartan, Dirac, and Weyl
Representations
3 =
18
n
z
z
!
;
4 =
20
21
is in the commutant of
; it
The automorphism CC
is, therefore, proportional to I and, by a change of
= I for k l 0 or
scale, one can achieve CC
6 mod 8 and CC I for k l 2 or 4 mod 8.
The spinor sc C1s 2 S is the charge conjugate of
s 2 S. If : V ! S is a solution of the Dirac equation
141 <
< p 2n
@ iqA = 0
19
B = 11=2nn1 B
= 2 CC1
Re S = fs 2 S j sc = sg
is a real vector space of dimension 2n , the space of
DiracMajorana spinors. The representation
is
real: restricted to Re S and expressed with respect to
a frame in this space, it is given by real 2n 2n
matrices. For k l 0 mod 8 the representations
and
are both real: in this case there are
WeylMajorana spinors.
Odd dimensions On computing (
) one finds that
the conjugate representation is equivalent to
527
(resp., ) if
2 = 1 (resp.,
2 = 1). There is an
isomorphism C : S !
S such that
m= 1 2 3 4 5 6 7 8
m = 1 2 2 3 3 3 3 4
22
= I (resp., CC
= I) for k l 1 or 7 mod 8
and CC
(resp., k l 3 or 5 mod 8). For k l 1 mod 8, the
restriction of the Pauli representation to C0k, l is real
and the Pauli matrices are pure imaginary; for k l
7 mod 8, the Pauli representations of Ck, l are both real
and so are the Pauli matrices. In both these cases there
are PauliMajorana spinors.
23
The isomorphism A0 = A intertwines the representations
y and
; it can also be made Hermitian
by rescaling. The Hermitian form A(s, s) is definite
for k = 0 and A0 (s, s) is definite for l = 0; otherwise,
these forms are neutral. For example, in the familiar
representation [20], one has A =
4 , a neutral form.
For p = 0, 1, . . . , m = 2n, two spinors s and t 2 S
define the p-vector with components
A1 ...p s; t = hs; A
1 . . .
p ti
24
Inductive Construction
of Representations
An
inductive
representations
construction
of
: Cn1;n ! R2n1 ;
the
Pauli
n = 1; 2; . . .
n = 1; 2; . . .
is as follows.
1. In dimension 1, put 1 = 1.
2. Given 2 R(2n1 ), = 1, . . . , 2n 1, define
!
0
=
for = 1; . . . ; 2n 1
0
and
2n =
I
z
z
!
;
40 =
I
2 z " x
3 z " z ;
4 x " I
5 x x ";
6 x z "
26
7 " I I
For m = 4, 5, 6, and 7 the matrices 1 , . . . , m generate the representations of Cm, 0 in R8 . The eight
matrices = x , = 1, . . . , 7, and 8 = " I
I I give the required representation of C8, 0 in
R16 . By dropping the first factor in 1 , 2 , 3 , one
obtains the matrices generating a representation of
C3, 0 in R 4 , etc. The symmetric matrix
= 1
8 = z I I I anticommutes with all
the s and 2 = I. If the matrices 2 R(2(m) )
correspond to a representation of Cm, 0 , then the
m 8 matrices 1 , . . . , m , 1 I, . . . , 8 I
generate the required representation of Cm8, 0 .
m
X
x yv ev
;v = 0
Spinor Groups
Let (V, g) be a quadratic space over K. If u 2 V is
not null, then it is invertible as an element of
C(V, g) and the map v 7! uvu1 is a reflection in
the hyperplane orthogonal to u. The orthogonal
group O(V, g) = O(V, g) = fR 2 GL(V) j R g
R = gg is generated by the set of all such reflections.
A spinor group G is a subset of C(V, g) that is a
group with respect to multiplication induced by the
product in the algebra, with a homomorphism
: G ! GL(V) whose image contains the connected
component SO0 (V, g) of the group of rotations of
(V, g). In the case of real quadratic spaces, one
considers also spinor groups that are subsets of C
C(V, g) with similar properties. By restriction, every
28
1 ! Z2 ! SpinV; g ! SOV; g ! 1
29
529
p
as;
at
= Av1 ...vp s; tRv11 a1 . . . Rvpp a1
Consider Spin0 (V, g) and assume that either V is
complex of dimension 52 or real with k or l 5 2.
Then there are two unit orthogonal vectors
e1 , e2 2 V such that (e1 , e2 )2 = 1. The vector
u(t) = e1 cos t e2 sin t is obtained from e1 by rotation
in the plane span fe1 , e2 g by the angle t 2 R. The
curve t 7! e1 u(t), 0 t , connects the elements
1 and 1 of Spin0 (V, g). Its image in SO0 (V, g), that
is, the curve t 7! Ad(e1 u(t)), 0 t , is closed:
Ad(1) = Ad(1). This fact is often expressed by
saying that a spinor undergoing a rotation by 2
changes sign. There is no homomorphism not
even a continuous map f : SO0 (V, g) ! Spin0 (V, g)
such that Ad f = id.
Spinc Groups
Spin0 Groups
The connected components of spin groups associated with orthogonal spaces of dimension 46 are
isomorphic to classical groups. They can be explicitly described starting from the following
observations.
Spin3 C = SL2 C
Spin01;2 = SL2 R
Spin01;3 = SL2 C
Spin02;2 = SL2 R SL2 R
Spin01;4 = Sp1;1 H
Spin02;3 = Sp4 R;
Spin02;4 = SU2;2
Spin03;3 = SL4 R
Spin01;5 = SL2 H
Further Reading
Adams JF (1981) Spin (8), triality, F4 and all that. In: Hawking
SW and Rocek M (eds.) Superspace and Supergravity.
Cambridge: Cambridge University Press.
Atiyah MF, Bott R, and Shapiro A (1964) Clifford modules.
Topology 3(suppl. 1): 338.
Baez JC (2002) The octonions. Bulletin of the American
Mathematical Society 39: 145205.
Brauer R and Weyl H (1935) Spinors in n dimensions. American
Journal of Mathematics 57: 425449.
Budinich P and Trautman A (1988) The Spinorial Chessboard,Trieste Notes in Physics. Berlin: Springer.
Cartan E (1938) Theorie des spineurs. Actualites Scientifiques et
Industrielles, No. 643 et 701. Paris: Hermann (English
transl.:The Theory of Spinors. Paris: Hermann, 1966).
Chevalley C (1954) The Algebraic Theory of Spinors. New York:
Columbia University Press.
Clifford WK (1878) Applications of Grassmanns extensive
algebra. American Journal of Mathematics 1: 350358.
Clifford WK (1882) On the classification of geometric algebras.
In: Tucker R (ed.) Mathematical Papers by William Kingdon
Clifford, pp. 397401. London: Macmillan.
Dirac PAM (1928) The quantum theory of the electron.
Proceedings of the Royal Society of London A 117: 610624.
Eckmann B (1942) Gruppentheoretische Beweis des Satzes von
HurwitzRadon uber die Komposition quadratischer Formen.
Commentarii Mathematici Helvetici 15: 358366.
Karoubi M (1968) Alge`bres de Clifford et K-theorie. Annales
cole Normale Superieure 4e`me ser 1: 161270.
Scientifiques de lE
Lipschitz RO (1886) Untersuchungen uber die Summen von
Quadraten. Berlin: Max Cohen und Sohn.
Lounesto P (2001) Clifford Algebras and Spinors, 2nd edn.
London Math. Soc. Lecture Note Series, vol. 286. Cambridge:
Cambridge University Press.
Pauli W (1927) Zur Quantenmechanik des magnetischen
Elektrons. Z. Physik 43: 601623.
Penrose R and MacCallum MAH (1973) Twistor theory: an
approach to the quantisation of fields and space-time. Physics
Report 6C(4): 241316.
Porteous IR (1995) Clifford Algebras and the Classical Groups,
Cambridge Studies in Advanced Mathematics, vol. 50. Cambridge: Cambridge University Press.
Postnikov MM (1986) Lie groups and Lie algebras. Mir: Moscow.
Sudbery A (1987) Division algebras (pseudo)orthogonal groups
and spinors. Journal of Physics A17: 939955.
Trautman A (1997) Clifford and the square root ideas.
Contemporary Mathematics 203: 324.
Trautman A and Trautman K (1994) Generalized pure spinors.
Journal of Geometry and Physics 15: 122.
Wall CTC (1963) Graded Brauer groups. Journal fur die Reine
und Angewandte Mathematik 213: 187199.
Cluster Expansion
and
1
@
log Z; ; V
V!1 jVj @
; lim
6
Introduction
p;
1
X
bn n
7
nbn n
8
and
Mayer Expansion
r i r j
2m i; j1
i1
1
n1
;
n1
f r er 1
r i r j BN
Z; ; V
Z
1
X
N
N0
N!
N
Y
V N i;j1
1
X
N X
N0
2
N g2GN
Y 3
1 f r i r j
d ri
wg
10
where
e
4
N! V N
N0
In the second expression we absorbed the factor
resulting from the integration over impulses into
(configurational) activity = (2m=h2 )3=2 z. In particular, the pressure p and the density are defined
by the thermodynamic limits (with V ! 1 in the
sense of Van Hove)
1
1
log Z; ; V
p; lim
V!1 jVj
9
i;j1
1
X
5
wg
f r i r j
d3 r i
11
V N fi;jg2g
k
Y
wg
12
we can rewrite
Z; ; V
1
X
N X Y
N0
N!
wg
13
fgl g g2G
1
X
n X
wg
n! g2Cn
n1
14
1
X
n n
16
n1
X2X W
>
:
1 otherwise v and v0
21
connected by an edge from G
P
Here, G(X) is the graph with jXj =
jX(v)j vertices
induced from G[supp X] by replacing each of its
vertices v by the complete graphQon jX(v)j vertices
and X! is the multifactorial X! = v2supp X X(v)!. The
sum is over all connected subgraphs H G(X)
spanned by the set of vertices of G(X) and jE(H)j
is the number of edges of the graph H.
23
(ii)
X2X : suppX3v
v;v0 2W 0
1 rv0
1 Uv; v0 rv0
v0 6v
26
p lim
27
X
X:X\c 6;
aXwX
jX \ j
jXj
28
jxyj1
p
30
31
Here, the sum runs over all subsets B of the set B() of
all bonds in (pairs of nearest-neighbor sites from )
such that each site is contained in an even number of
bonds from B. Using (B) to denote the set of sites
contained in bonds from B, we say that B1 , B2 B()
are disjoint if (B1 ) \ (B2 ) = ;. Splitting now B into a
collection B = {B1 , . . . , Bk } of its connected components
called (high-temperature) polymers and using B() to
denote the set of all polymers in , we are getting
X Y
Z 2jj cosh jBj
tanh jBj 32
BB B2B
33
jB j
1
X
aX X
w
jXj
X:X3x
37
38
A
with fA (
) = exp {UA (
)} 1. Expanding the
product we will get a polymer representation with
polymers A consisting of connected collections
A = (A1 , . . . , Ak ) with weights
Z Y
Y
wA
fA
d
x
39
A2A
x2[A2A A
n n 1
35
n1
log 2 d logcosh
36
hi
d
x
41
Z
x2
P
with H (
) = A UA (
) and a function
depending only on variables
x on sites x from a
finite set S Zd .
A convenient way of evaluating the expectation starts
with introduction of the modified partition function
Z; Z Z; Z 1 hi
42
Clearly,
d log Z;
hi
d
0
43
d
x
x2[A2A1 [ [ Ak A[S
44
for A = (A1 , . . . , Ak ) 2 AS (), we get
exactly in the form [18],
X Y
Z;
w A
Z, ()
45
jaXjjwjX K
XAjAj
jSjKdistS;
49
50
I W S A2I
Low-Temperature Expansions
As a result, we have
log Z ;
aXwX
46
X2XW S
aXwX
47
48
jEj
exp
j
j
51
Z e
2
aX X
w
jAXj
X:AX30
53
54
with
A
fx 2 Zd j such that distx;
1=2g
55
8
<
aXw
:
X:AX\c 6;
Y
1 fX
9
=
X jAX \ j
jAXj ;
X:AX\c 6;
Noticing that
jAX \ j
1
fX exp aXwX
jAXj
56
Bibliographical Notes
Cluster expansions originated from the works of Ursell,
Yvon, Mayer, and others and were first studied in terms
of formal power series. The combinatorial and enumeration problems considered in this framework were
summarized in Uhlenbeck and Ford (1962). For related
topics in modern language, see Bergeron et al. (1998).
The convergence results for Mayer and virial expansions
for dilute gas were first proved in the works of Penrose,
Lebowitz, Groenveld, and Ruelle (see Ruelle (1969) for
a detailed survey). General polymer models on lattice
were discussed by Gruber and Kunz (1971) (see also
Simon (1993) for discussion of high-temperature and
low-temperature cluster expansions of lattice models).
Abstract polymer models were introduced in Kotecky
and Preiss (1986). An elegant proof of a general claim
presented by Dobrushin (1996) was further extended
and summarized by Scott and Sokal (2005). We follow
their reformulation of the Dobrushin condition. Cluster
expansions with a view on applications in quantum field
theory are reviewed in Brydges (1986).
See also: Phase Transitions in Continuous Systems;
PirogovSinai Theory; Wulff Droplets.
Further Reading
Bergeron F, Labelle G, and Leroux P (1998) Combinatorial
Species and Tree-Like Structures, Coll. Encyclopaedia of
Mathematics and Its Applications, vol. 67. Cambridge, MA:
Cambridge University Press.
Brydges DC (1986) A short course on cluster expansions. In:
Osterwalder K and Stora R (eds.) Critical Phenomena, Random
Systems, Gauge Theories, pp. 129183. Les Houches, Session
XLIII, 1984. Amsterdam/New York: Elsevier.
Dobrushin RL (1996) Estimates of semi-invariants for the Ising
model at low temperatures. In: Dobrushin RL, Minlos RA,
Shukin MA, and Vershik AM (eds.) Topics in Statistical and
Theoretical Physics, pp. 5981. Providence, RI: American
Mathematical Society.
Gruber C and Kunz H (1971) General properties of polymer
systems. Communications Mathematical Physics 22: 133161.
Kotecky R and Preiss D (1986) Cluster expansion for abstract polymer
models. Communications in Mathematical Physics 103: 491498.
Coherent States
Ruelle D (1969) Statistical Mechanics: Rigorous Results, The
Mathematical Physics Monograph Series. Reading, MA:
Benjamin.
Scott AD and Sokal AD (2005) The repulsive lattice gas, the
independent-set polynomial, and the Lovasz local lemma.
Journal of Statistical Physics 118: 11511261.
537
Coherent States
S T Ali, Concordia University, Montreal, QC, Canada
2006 Elsevier Ltd. All rights reserved.
Introduction
Very generally, a family of coherent states is a set of
continuously labeled quantum states, with specific
mathematical and physical properties, in terms
of which arbitrary quantum states can be expressed
as linear superpositions. Since coherent states are
continuously labeled, they form overcomplete
sets of vectors in the Hilbert space of states.
Originally these states were introduced into physics
by Schrodinger (1926), as a family of quantum
states in terms of which the transition from quantum
to classical mechanics could be conveniently studied.
These states have the minimal uncertainty property,
in the sense that they saturate the Heisenberg
uncertainty relations. The name coherent state was
applied when these states were rediscovered in the
context of quantum optical radiation by Glauber,
Klauder, and Sudarshan. It was demonstrated that in
these states the correlation functions of the quantum
optical field factorize as they do in classical optics,
so that the optical field has a near-classical behavior,
with the optical beam being coherent. In this article,
we shall refer to these originally studied coherent
states as canonical coherent states (CCS).
The canonical coherent states, apart from their
use in quantum optics, have also been found to be
extremely useful in computations in atomic and
molecular physics, in quantum statistical mechanics,
and in certain areas of mathematics and mathematical physics, including harmonic analysis, symplectic geometry, and quantization theory. Their wide
applicability has prompted the search for other
families of states sharing similar mathematical and
physical properties. These other families of states are
usually called generalized coherent states, even when
there is no link to optical coherence in such studies.
jzi ejzj
=2
1
X
zn
p jni
n!
n0
1
2
4
and
a; ay aay ay a I
5
a ay
p
i 2
6
7
Fz hjzi e
f z
Z
jf zj2 dz; z
1=2
9
C
2
=2 zay
Uz e
e j0i Uzj0i
zay
za
10
General Definition
Let H be an abstract, separable Hilbert space over
the complexes, X a locally compact space and d a
measure on X. Let jx, ii be a family of vectors in H ,
defined for each x in X and i = 1, 2, 3, . . . , N, where
N is usually a finite integer, although it could also
be infinite. We assume that this set of vectors
possesses the following properties:
1. For each i, the mapping x 7! jx, ii is weakly
continuous, that is, for each vector ji in H , the
function i (x) = hx, iji is continuous (in the
topology of X).
2. For each x in X, the vectors jx, ii, i = 1, 2, . . . , N,
are linearly independent.
3. The resolution of the identity
N Z
X
jx; iihx; ijdx IH
12
i1
11
Coherent States
i; j 1; 2; . . . ; N
14
15
16
and
N Z
X
1
539
Some Examples
We present in this section some of the more
commonly used types of coherent states, as illustrations of the general structure given above.
A large class of generalizations of the canonical
coherent states [1] is obtained by a simple modification of their analytic structure. Let x1 x2
xn be an infinite sequence of positive numbers
(x1 6 0). Define xn ! = x1 x2 xn and by convention
set x0 ! = 1. In the same Fock space in which the CCS
were described, we now define the related deformed
or nonlinear coherent states via the analytic
expansion
jzi N jzj2 1=2
1
X
zn
p jni
xn !
n0
18
r 2n dr; n 0; 1; 2; . . .
20
2
0
This means that once the quantities xn ! are specified,
the measure d is to be determined by solving the
generalized coherent states arise from representations of the group SU(1, 1) belonging to the discrete
series, each irreducible representation being labeled
by a specific value of the index . The associated
Hilbert space of functions, analytic on the unit disk,
is a subspace of L2 (D, d ), with
Kz; z0 hzjz0 i
1
h
i1=2 X
zz0 n
21
N jzj2 N jz0 j2
xn !
n0
22
24
am
d z; z 2 1
1 r2 22
r dr d
z rei
K ; K 2K3
27
28
K3 jni njni
Thus, K j0i = 0 and
1
jni p Kn j0i
n!2n
The BarutGirardello coherent states jzi are now
defined as the formal eigenvectors of the ladder
operator K :
K jzi zjzi;
z2C
29
30
Coherent States
X
J e
p jni
jJ;
i N J1=2
n !
k0
32
33
34
541
36
j bkj2
dk < 1
jkj
R
jxi Ugxj i
40
jh jUb; a ij2
GAff
db da
<1
a2
b; a 2 GAff
41
jb; aihb; aj
GAff
db da
I
a2
42
on L2 (R, dx).
In the signal-analysis literature a vector satisfying
the admissibility condition [40] is called a mother
wavelet and the generalized coherent states [41] are
called wavelets. Signals are then identified with
vectors ji in L2 (R, dx) and the function
Fb; a hb; aji
43
44
45
Coherent States
543
x hxjT 1 i
in terms of the generalized coherent states.
The canonical coherent states satisfy the minimal
uncertainty relation [7]. It is possible to build
families of coherent states by generalizing from this
condition. To do this, one typically starts with two
self-adjoint generators in the Lie algebra of a
particular group representation and then looks for
appropriate eigenvectors of a complex combination
of these two generators. For two self-adjoint
operators B and C on a Hilbert space H , satisfying
the commutation relation [B, C] = iD and any
normalized vector in H , one can prove the
Heisenberg uncertainty relation
B2 C2
hDi2
4
47
z hBi ihCi
49
2
Y
dxkj ^ dykj
k;j1
2
bkI2 ; k; 0; 1; 2; . . . ; 1
51
I2 being the 2 2 identity matrix and
bk
k 3!
2k 1k 2
k 1; 2; 3; . . . ;
52
b0 1
53
2 X
1
X
Y ik Z 0y Y ik Z y y
i1 k0
2 X
1
X
Z 0yk Z k
i1 k0
54
bk
2 X
1
X
jy Z k i j
p jY k i
bk
j1 k0
0y
55
0y
N
AA qA A q
58
jxihxjd!x IH
f xjxihxjd!x
59
60
Further Reading
Ali ST, Antoine J-P, and Gazeau J-P (2000) Coherent States,
Wavelets and Their Generalizations. New York: Springer.
Ali ST and Englis M (2005) Quantization methods a guide for
physicists and analysts. Reviews in Mathematical Physics
17: 391490.
Brif C (1997) SU(2) and SU(1,1) algebra eigenstates: a unified
analytic approach to coherent and intelligent states. International Journal of Theoretical Physics 36: 16511682.
Klauder JR and Sudarshan ECG (1968) Fundamentals of
Quantum Optics. New York: Benjamin.
Klauder JR and Skagerstam BS (1985) Coherent States
Applications in Physics and Mathematical Physics. Singapore:
World Scientific.
Perelomov AM (1986) Generalized Coherent States and their
Applications. Berlin: Springer.
Schrodinger E (1926) Der stetige Ubergang von der Mikro- zur
Makromechanik. Naturwissenschaften 14: 664666.
Sivakumar S (2000) Studies on nonlinear coherent states. Journal
of Optics B: Quantum Semiclass. Opt. 2: R61R75.
Zhang W-M, Feng DH, and Gilmore RG (1990) Coherent states:
theory and some applications. Reviews of Modern Physics
62: 867927.
Cohomology Theories
U Tillmann, University of Oxford, Oxford, UK
2006 Elsevier Ltd. All rights reserved.
Introduction
The origins of cohomology theory are found in
topology and algebra at the beginning of the last
century but since then it has become a tool of nearly
every branch of mathematics. Its a way of life!
Naturally, this article can only give a glimpse at the
rich subject. We take here the point of view of
algebraic topology and discuss only the cohomology
of spaces.
Cohomology reflects the global properties of a
manifold, or more generally of a topological space.
It has two crucial properties: it only depends on the
homotopy type of the space and is determined by
local data. The latter property makes it in general
computable.
@i
@1
Ci1 ! Ci ! Ci1 ! C0
1
ker@i
im@i1
H0 Mg ; Z Z
H1 Mg ; Z Z2g
H2 Mg ; Z Z
Hi Mg ; Z 0 for i 3
4
Note that the Euler characteristic can be recovered as the alternating sum of the rank of the
homology groups:
M
dim
XM
1i rk Hi M; Z
5
i0
2
3
Singular Homology
n
X
1i jv0 ;...;^vi ;...;vn
6
i0
0
for i 2
For the one point space we have H0 (pt; R) = R. Define
~ (X; R) := ker(H (X; R) !
reduced homology by H
H (pt; R)).
~ i (pt; R) = 0 for all i.
2. Dimension axiom. H
More generally, it follows immediately from the
definition of simplicial homology that the homology
of any n-dimensional manifold is zero in dimensions
larger than n.
We mentioned in the introduction that homology
depends only on local data. This is made precise
by the
3. MayerVietoris theorem. Let X = A [ B be the
union of two open subspaces. Then the following
sequence is exact:
!Hn A \ B; R ! Hn A; R Hn B; R
@
! Hn X; R! Hn1 A \ B; R
! ! H0 X; R ! 0
On the level of chains, the first map is induced by the
diagonal inclusion, while the second map takes the
difference between the first and second summands.
Finally, @ takes a cycle c = a b in the chains of X
that can be expressed as the sum of a chain a in A
7
~ 0 X; Z Zk1
H
where k is the number of connected components in
X. Also, Sn Sn1 n S0 . Thus, by [7],
~ Sn ; Z 0 for 6 n
Hn Sn ; Z Z and H
8
! Hn1 Y; R ! ! H0 X; Y; R ! 0
This and the MayerVietoris sequence give two ways of
breaking up the problem of computing the homology of
a space into computing the homology of related spaces.
An iteration of this process leads to the powerful tool of
spectral sequences (see Spectral Sequences).
Relation to Homotopy Groups
9
Theorem If a map f : X ! Y between two simplicial complexes with trivial fundamental groups
induces an isomorphism on all homology groups,
then it is a homotopy equivalence.
Warning: This does not imply that two simply
connected spaces with isomorphic homology groups
are homotopic! The existence of the map f inducing
this isomorphism is crucial and counterexamples can
easily be constructed.
ker @ i1
im @ i
10
11
Singular Cohomology
the
formula
Steenrod Algebra
Htor
by
Sq Sq
i=2
X
jk1
k0
i 2k
!
Sqijk Sqk
for i 2j
The mod-2 Steenrod algebra A is then the free
Z2 -algebra generated by the Steenrod squares
Sqi , i 0, subject only to the Adem relations. With
the help of Adems relations, Serre and Cartan found
a Z2 -basis for A:
fSqI : Sqi1 Sqin jij 2ij1
for all jg
14
Hodge Decomposition
Rn : fC1 functions on Rn g R
L
The algebra (R n ) = nq = 0 q (Rn ) is naturally
graded by degree. There is a differential operator
d : q (Rn ) ! q1 (R n ) defined by
P
n
0
1. if f 2
P(R ), then df = P(@f =@xi )dxi
2. if ! =
fI dxI , then d! =
dfI dxI
I stands here for a multi-index. For example, in R3
the differential assigns to 0-forms ( = functions) the
gradient, to 1-forms the curl, and to 2-forms the
divergence. An easy exercise shows that d2 = 0 and
the qth de Rham cohomology of Rn is the vector space
q
Hde
Rn
R
ker d : q Rn ! q1 Rn
im d : q1 Rn ! q Rn
c R
e1 ^ ^ eq :
ep1 ^ ^ en
where one takes if the orientation defined by
{e1 , . . . , en } is the same as the given one, and
otherwise. Using local coordinate charts this definition can be extended to M. Clearly, depends on the
chosen metric and orientation of M. If M is
compact, we may define an inner product on the
q-forms by
Z
0
!; ! :
! ^ !0
M
15
if and only if
d! 0 and ! 0 16
Examples
We list the cohomology of some important
examples.
H CPn ; Z Zy=yn1
and more generally, infinite Grassmannian manifolds are classifying spaces for linear groups. When
G is a compact connected Lie group,
H HPn ; Z Zz=zn1
H BSO2k 1; Z1=2
Z1=2p1 ; p2 ; . . . ; pk
H BSO2k; Z1=2
Lie Groups
a3 ; a7 ; . . . ; a4k1
Z
H BUk; Z Zc1 ; c2 ; . . . ; ck
where the Pontryagin, Euler, and Chern classes have
degree jpi j = 4i, jek j = 2k, and jci j = 2i, respectively.
a1 ; a7 ; . . . ; a4k5 ; a2k1
Z
H Uk; Z
^
Hfree
SO2k; Z
^
Z1=2p1 ; p2 ; . . . ; pk1 ; ek
Moduli Spaces
a1 ; a3 ; . . . ; a2k1
Classifying Spaces
The geometric objects representing elements in complex K-theory K0 (X) are isomorphism classes of finite
dimensional complex vector bundles E over X. Vector
bundles E, E0 can be added to form a new bundle
E E0 over X, and K0 (X) is just the group completion
of the arising monoid. Thus, for example, for the point
space we have K0 (pt) = Z. Tensor product of vector
bundles E E0 induces a multiplication on K-theory
making K (X) into a graded commutative ring.
In many ways K-theory is easier than cohomology. In particular, the groups are 2-periodic: all even
degree groups are isomorphic to the reduced
~ 0 (X) := coker(K0 (pt) = Z ! K0 (X)),
K-theory group K
and all odd degree groups are isomorphic to
~ 0 (X).
K1 (X) := K
The theory of characteristic classes gives a close
relation between the two cohomology theories. The
Chern character map, a rational polynomial in the
Chern classes, defines
ch : K0 X Z Q ! H even X; Q
: H 2k X; Q
k0
Elliptic Cohomology
Combinatorics: Overview
553
Further Reading
Adams F (1978) Infinite Loop spaces. Annals of Mathematical
Studies 90: PUP.
Bott R and Tu L (1982) Differential Forms in Algebraic
Topology. Springer.
Galatius, Madsen, Tillmann, Weiss (2005).
Hatcher A (2002) Algebraic Topology, (http://www.math.cornell.
edu). Cambridge: Cambridge University Press.
Madsen, Weiss (2004).
Mosher R and Tangora M (1968) Cohomology Operations and
Applications in Homotopy Theory. Harper and Row.
Viehweg (1999) Aspects of Mathematics E33.
Witten (1998) Journal of Higher Energy Physics 12.
Witten (1998) Springer Lecture Notes in Mathematics, vol. 1326.
Combinatorics: Overview
C Krattenthaler, Universitat Wien, Vienna, Austria
2006 Elsevier Ltd. All rights reserved.
Introduction
Combinatorics is a vast field which enters particularly
in a crucial way in statistical physics. There, it is
particularly the enumerative problems that are of
importance. Therefore, in this article, we shall mainly
concentrate on the enumerative aspects of combinatorics. We first recall the basic terminology, in
particular the basic combinatorial objects and numbers, together with the simplest facts about them. We
then provide introductions into the most important
techniques of enumeration: the generating function
n
X
k0
sn; kxk
n;k0
yn
1 yx
n!
Bn
xn
x
ee 1
n!
Sn; kxk
n;k0
yn
y
exe 1
n!
or, explicitly, by
Sn; k
of n with
n is 2n1 . The number of compositions
n1
A partition of a
exactly k summands is k1.
positive integer n is a representation of n as a sum
n = 1 2 k of other positive integers i ,
where the order of the summands does matter. Thus,
we may assume that the summands are ordered,
1 2 k > 0. This is the motivation
to write partitions most often in the form of
tuples (1 , 2 , . . . , k ) the entries of which are
weakly decreasing. The summands of a partition
are called the parts of the partition. Let p(n) denote
the number of partitions of n. These numbers are
given by
1
X
n0
pn; kxn Qk
i
i1 1 x
Finally, if p(n, k, m) denotes the number of partitions of n into at most k parts, all of which are at
most m, then
X
pn; k; mxn
n0
k
1X
k n
1kj
j
k! j0
j
1
i
i1 1 x
pnxn Q1
Combinatorics: Overview
555
2
3
Fn xn
1
1 x x2
4
5
4
FA x
xjaj
a2A
3
1
(a)
(b)
Generating Functions
Generating functions are the very basic tools of
enumeration. For introductions to this technique,
from different points of view, the reader is referred
to Bergeron et al. (1998), Flajolet and Sedgewick
(chapter 1 in the reference listed in Further reading section), and Stanley (1998, chapter 1; 1999,
chapter 4).
Let A be a set of (unlabeled) objects. Each object
a in A has a certain size, jaj, which is a non-negative
integer. Let us also assume that there is only a finite
number of objects from A of a given size. Let an be
the number of objects from A of size n. The
1
X
an x n
n0
x
1x
1
1x
5
6
7
Combinatorics: Overview
8
X xjaj
a2A
jaj!
1
X
an
n0
xn
n!
1
1x
1
ECycles x log
1x
10
11
12
557
13
14
Since, in the labeled world, objects come automatically with atoms, the substitution of two sets A and
B of objects can now always be defined. The
substitution of B in A, denoted by A(B), is the set
of objects which arises by replacing the atoms of
objects from A by objects from B in all possible
ways, and labeling the substituted
objects in all
P
possible ways by 1, 2, . . . , b jbj (the sum being
over the objects from B which were put in the places
of the atoms) that are consistent with the original
labelings of the objects from B. The size of an object
from A(B) is the sum of the sizes of the objects from
B that it contains. In order that A(B) contains only a
finite number of objects of a given size, we must
assume that B contains no elements of size 0. Then
we have
EAB x EA EB x
15
An example of a composition is
Permutations SetsCycles
Thus, from [15] we have
EPermutations x ESets ECycles x
corresponding to the identity
1
explog 1=1 x
1x
Another manifestation of the composition rule is, for
example, the fact (which is sometimes called the
exponential principle) that, if one takes the log of
the partition function for some maps, the result is
the partition function for the connected maps among
them.
and
EA x
X
a2A
wa
xjaj
jaj!
19
20
21
n!
n0
2S
18
16
ZA[B x1 ; x2 ; . . . ZA x1 ; x2 ; . . .
ZB x1 ; x2 ; . . .
22
ZAB x1 ; x2 ; . . . ZA x1 ; x2 ; . . .
ZB x1 ; x2 ; . . .
23
ZAB x1 ; x2 ; . . . ZA ZB x1 ; x2 ; x3 ; . . .;
ZB x2 ; x4 ; x6 . . .;
ZB x3 ; x6 ; x9 ; . . .; . . . 24
Similar to the theory of generating functions
surveyed in the last section, one can also develop a
weighted version of the cycle index series. Given a set
of labeled objects A, where each object a is assigned a
weight w(a), one changes the definition [16] insofar as
fix
P (A) gets replaced by the weighted sum
(a) = a w(a), where (a) means the object arising
from a by permuting the labels according to . Then all
the above formulas remain true in this weighted setting.
Cycle index series are instrumental in the enumeration of colored objects. The basic situation is
that we have given a set A~ of unlabeled objects so
that every object of size n comes with n atoms
(nodes). For example, we may think of A~ as the set
of cycles. We are now going to color each atom by a
Combinatorics: Overview
c2C
25
27
or, alternatively, by
c2C
559
28
n
29
We want
the coefficients in the expansion
P to know
n
T(x) = 1
t
x
=n!.
Since, by [29], T(x) is the
n
n=0
compositional inverse of x exp (x), substitution of
x exp (x) instead of x gives
x
1
X
tn
n!
n0
x expxn
33
x
Fu; x 1 xuFu; x Fu; x F0; x 30
u
since a path can be empty (this explains the term 1),
it can end by a step (1,1) (this explains the term
xuF(u)), or it can end by a step (1,1). The latter
can only happen if the path before that last step did
not end at height 0. The generating function for
these paths is F(u, x) F(0, x), and this explains the
third term in the eqn [30]. In fact, we may replace
[30] by
x
Fu; x 1 xuFu; x Fu; x F1 x 31
u
Combinatorics: Overview
P
n
power series f (z) = 1
n = 0 fn z =n! instead, or one
must apply methods other than singularity analysis
or the saddle point method. In the latter case,
depending on the nature of the coefficients fn , this
may be the EulerMaclaurin or the Poisson summation formulas, the Mellin transform technique, or
other direct methods. The reader is referred to
Odlyzko (1995) and the Flajolet and Sedgewick
reference.) The idea is then to consider f (z) as a
complex function in z (and extend the range of f
beyond the disk of convergence about the origin),
and to study the singularities of f (z). (The point at
infinity can also be a singularity.) The upshot is that
the singularities of f (z) with smallest modulus
dictate the asymptotic behavior of the coefficients
fn . These singularities of smallest modulus are called
the dominating singularities.
If there is an infinite number of dominant
singularities, then one has to try the circle method.
We refer the reader to Andrews (1976) and Ayoub
(1963) for details of this method.
If there is a finite number of dominant singularities, then there can be again two different situations, depending on whether these are small or
large singularities. Roughly speaking, a singularity
is small if the function f (z) grows at most
polynomially when z approaches the singularity,
otherwise it is large. A typical example of a small
singularity is z = 1=4 in (1 4z)1=2 , whereas a
typical example of a large singularity is z = 1 in
exp (x) or z = 1 in exp (1=(1 z)).
The method to apply for small singularities is the
method of singularity analysis as developed by
Flajolet and Odlyzko. (Singularity analysis implies
Darbouxs method, which occurs frequently in the
literature, and, thus, supersedes it.) For the sake of
simplicity, we consider first only the case of a
unique dominant singularity. We shall address the
issue of several dominant singularities shortly.
Furthermore, we assume the singularity to be
z = 1, again for the sake of simplicity of presentation. The general result can then be obtained by
rescaling z.
The basic idea is the transfer principle:
If f z z Oz
z!1
fn n On
n!1
then
34
P
n
where (z) = 1
n = 0 n z is a linear combination of
standard functions of the form
P1 (1 z)n , or logarithmic variants, and (z) = n = 0 n z also lies in
the scale (see sections VI.3,4 of the Flajolet and
Sedgewick reference for the exact statement). The
expansion for f (z) in [34] is called the singular
561
n1
C1
log n 1
1! log n
!
C2 1
2! log n2
35
fn zn
1
p
1 z 1 4z
27
(We stopped the expansion after three terms. The
farther we go, the more terms can we compute
of the asymptotic expansion for fn .) Hence, we
obtain
1=2
1
1
n 4 n
1
fn 4
3 1=2
8n 128n2
4 n3=2
3
1
9 1=2
8n
5=2
4 n
7=2
O n
27 3=2
4n 4
1
11
1
O
n3
n 3 18n 288n2
If there are several small dominant singularities
(but only a finite number of them), then one simply
applies the above procedure for all of them and, to
obtain the desired asymptotic expansion, one adds
up the corresponding contributions.
as
n!1
37
B3
B2
B4
B5
B2
B6
B7
Combinatorics: Overview
d4
563
3
2
1
0
Figure 9 Bijection between Motzkin paths and heaps of
monomers and dimers.
the beginning to the end. Whenever we read a levelstep at height h, we make it into a monomer with
x-coordinate h, whenever we read a down-step from
height h to height h 1, we make it into a dimer
whose endpoints have x-coordinates h 1 and h.
Up-steps are ignored. Figure 9 shows an example. In
the figure, the heap is not in standard fashion, in
the sense that the x-axis is not shown as a horizontal
line but as a vertical line (cf. Figure 7). But it could
be easily transformed into standard fashion by a
simple reflection with respect to a line of slope 1.
Lattice animals on the triangular lattice and on the
quadratic lattice are also in bijection with heaps, this
time with heaps consisting entirely out of dimers.
Given an animal, one simply replaces each vertex of
the animal by a dimer, see Figures 10 and 11. While
in the case of animals on the triangular lattice this
gives a constraintless bijection (see Figure 10), in the
case of the quadratic lattice this sets up a bijection
with heaps of dimers in which two dimers of the
same type can never be placed directly one over the
other (see Figure 11). For example, two dimers d5 ,
one placed directly over the other (as they occur in
Figure 10), are forbidden under this rule.
Next we make heaps into a monoid by introducing a composition of heaps. (A monoid is a set with
a binary operation which is associative.) Intuitively,
given two heaps H1 and H2 , the composition of H1
and H2 , the heap H1 H2 , is the heap which results
d3
d2
d1
0 1 2 3 4 5 6 7 8
m0
m1
m2
m3
m4
m5
m6
m7
0 1 2 3 4 5 6 7
0 1 2 3 4 5 6 7
0 1 2 3 4 5 6 7 8
Figure 11 Bijection between animals and heaps of dimers.
0 1 2 3 4 5 6 7 8
Figure 12 The composition of the heaps in Figure 8.
38
where T (B, R) denotes the set of all trivial heaps
with pieces from B. In particular, the generating
function for all heaps is given by
X
1
wH P
39
jTj
H2HB;R
T2T B;R 1 wT
AGn xn
n0
u;v2V
n0
In AGx1
where In is the n n identity
In other words,
Pmatrix.
1
the generating functions
w
(u,
v)xn for the
n
n=0
walk numbers between u and v form the entries of a
matrix which is the inverse matrix of In A(G)x. By
elementary linear algebra,
1
X
wn u; v xn
n0
0 1 2 3 4 5 6 7
(a)
0 1 2 3 4 5 6
(b)
41
Combinatorics: Overview
Lattice Paths
Recall from the section on basic combinatorial
terminology that a lattice path P in Zd is a path in
the d-dimensional integer lattice Zd which uses only
points of the lattice, that is, it is a sequence
(P0 , P1 , . . . , Pl ), where Pi 2 Zd for all i. The vectors
!
P0 P1 , P1 P2 , . . . , Pl1 Pl are called the steps of P. The
number of steps, l, is called the length of P.
The enumeration of lattice paths has always
been an intensively studied topic in statistics,
because of their importance in the study of
random walks, of rank order statistics for nonparametric testing, and of queueing processes. The
reader is referred to Feller (1957) and particularly
Mohantys (1979) book, which is a rich source for
enumerative results on lattice paths, albeit in a
statistical language. We review the most important
results in this section. Most of these concern twodimensional lattice paths, that is, the case d = 2.
To begin with, we consider paths in the integer
plane Z2 consisting of horizontal and vertical unit
steps in the positive direction. Clearly, the number
of all (unrestricted) paths from
the origin to (n, m) is
the binomial coefficient nm
n . By the reflection
principle, which is commonly attributed to D Andre
(see, e.g., Comtet (1974) p. 22), it follows that the
number of paths from the origin to (n, m) which do
not pass above the line y = x t, where m n t, is
given by
n m n m
42
n
nt1
Roughly, the reflection principle sets up a bijection between the paths from the origin to (n, m)
which do pass above the line y = x t and all paths
565
Combinatorics: Overview
52
rs
rs
K1
53
567
cm
m
54
m
kr 1
ks 1
sin
K2
K2
55
kr 1
ks 1
sin
K2
K2
56
fi;jg2m
58
Combinatorics: Overview
!
adjacency matrix A(G ) has the block form of the
matrix on the left-hand side of [59] and, hence, the
Pfaffian reduces to a determinant. More precisely, let
G be a bipartite graph with vertex set V = U [ W,
U = {u1 , u2 , . . . , un } and W = {w1 , w2 , . . . , wn }, with
edges connecting some ui to some wj . Given a
!
Pfaffian orientation G , we build the signed bipartite
!
!
adjacency matrix B(G ) = (bi, j )1i, jn of G by setting
bi, j = w(ei, j ) if there is an edge from ui to wj oriented
that way, w(ei, j ) if there is an edge from uj to wi
oriented that way, and 0 if there is no edge between ui
and wj . Then we have
!
detBG Mw G
In particular, this holds for any bipartite planar
graph. See Robertson et al. (1999) for a structural
description about which (not necessarily planar)
bipartite graphs admit a Pfaffian orientation.
Kasteleyns construction in the planar case has
been generalized to graphs on surfaces of any genus
g in Dolbilin et al. (1996), Galluccio and Loebl
(1999), and Tesler (2000), independently. As predicted by Kasteleyn, the solution is in terms of a
linear combination of 4g Pfaffians.
With the help of his method, Kasteleyn computed
the number of dimer coverings of an m n
rectangle. (A dimer is a 2 1 rectangle. Thus, this
is equivalent to counting the number of perfect
matchings on the m n grid graph. The formula
was independently found by Temperley and Fisher.)
The result is
m Y
n
Y
p
i
j
2 1 cos
2 cos
m1
n1
i1 j1
For even m and n, the formula can be rewritten as
m=2
n=2
YY
i
j
4 cos2
4 cos2
m1
n1
i1 j1
There is a similar rewriting if one of m or n is odd.
(If both m and n are odd, there is no dimer
covering.)
For further reading and references see Dimer
Problems and Kuperberg (1998).
Nonintersecting Paths
Let G = (V, E) be a directed acyclic graph with
vertices V and directed edges E. Furthermore, we are
given a function w which assigns a weight w(x) to
every vertex or edge x. Let usQdefine the
Q weight w(P)
of a walk P in the graph by e w(e) v w(v), where
the first product is over all edges e of the walk P and
the second product is over all vertices v of P. We
569
sgn GFP u ! v; w
det
1i;jn
GFPuj ! vi ; w
60
1i;jn
GFPuj ! vi ; w 61
Pf
1i;j2n
X
GFPua ! vi ;wGFPub ! vj ;w
a<b
GFPub ! vi ; wGFPua ! vj ;w
62
6
E4
A4
E3
A3
3
4
A2
A1
(a)
E2
2
4
5
E1
2 3 4 6
4 4 6
5 6
(b)
Combinatorics: Overview
(a)
(b)
(b)
(a)
(a)
571
(b)
(b)
u2
(c)
Figure 18 (a) A rhombus tiling. (b) Bonds in rhombi.
(c) A perfect matching of a hexagonal graph.
ijk2
63
hu
64
aij1
ij1
1ijN
65
Combinatorics: Overview
3i 1!
n i!
i0
(a)
0 1 0
1 1 1
0
1
0
66
573
(b)
F
;
z
r s
b1 k bs k k!
k0
b1 ; . . . ; bs
where ()k = ( 1)( 2) ( k 1) for k >
0, and ()0 = 1. The symbol ()k is called the
Pochhammer symbol or shifted factorial. For indepth treatments of the subject, we refer the reader
2
6
6
6
6
7 F6 6
6
4
7
;1 7
7
7
7
7
5
a=2; 1 a b; 1 a c; 1 a d;
a b c d n; a 1 n
1 an 1 a b cn 1 a b dn 1 a c dn
1 a bn 1 a cn 1 a dn 1 a b c dn
a;b;c;n
6
6
3
7
;1 7
5
4 F3 4
e;f ;1 a b c e f n
e an f an
en f n
2
n;a;1 a c e f n;1 a b e f n
6
6
4 F3 4
3
7
;1 7
5
1 a b c e f n;1 a e n;1 a f n
68
where n is a non-negative integer, and
2
6
6
7 F6 4
a; 1 2a ; b; c; d; e; n
a
2 ; 1 a b; 1 a c; 1 a d; 1 a e; 1 a n
1 an 1 a d en
1 a dn 1 a en
2
1 a b c; d; e; n
6
4 F3 6
4
1 a b; 1 a c; a d e n
7
;17
5
3
7
;17
5
69
Combinatorics: Overview
70
575
71
Further Reading
http://algo.inria.fr This site includes, among its libraries, the
Maple program gdev.
Andrews GE (1976) The Theory of Partitions, Encyclopedia of
Mathematics and Its Applications, vol. 2. (reprinted by Cambridge
University Press, Cambridge, 1998). Reading: AddisonWesley.
Andrews GE, Askey RA, and Roy R (1999) In: Rota GC (ed.)
Special Functions, Encyclopedia of Mathematics and Its
Applications, vol. 71. Cambridge: Cambridge University Press.
Ayoub R (1963) An Introduction to the Analytic Theory of
Numbers. Mathematical Surveys, vol. 10, Providence, RI:
American Mathematical Society.
Bergeron F, Labelle G, and Leroux P (1998) Combinatorial Species
and Tree-Like Structures. Cambridge: Cambridge University Press.
Bousquet-Melou M and Jehanne A (2005), Polynomial equations
with one catalytic variable, algebraic series, and map
enumeration. Preprint, ariv:math.CO/0504018.
Bressoud DM (1999) Proofs and Confirmations The Story of
the Alternating Sign Matrix Conjecture. Cambridge: Cambridge University Press.
de Bruijn NG (1964) Polyas theory of counting. In: Beckenbach
EF (ed.) Applied Combinatorial Mathematics, New York:
Wiley, (reprinted by Krieger, Malabar, Florida, 1981).
Comtet L (1974) Advanced Combinatorics. Dordrecht: Reidel.
Dolbilin NP, Mishchenko AS, Shtanko MA, Shtogrin MI, and
Zinoviev YuM (1996) Homological properties of dimer
configurations for lattices on surfaces. Functional Analysis
and its Application 30: 163173.
Feller W (1957) An Introduction to Probability Theory and Its
Applications, vol. 1, 2nd edn. New York: Wiley.
Fisher ME (1984) Walks, walls, wetting and melting. Journal of
Statistical Physics 34: 667729.
Flajolet P and Sedgewick R, Analytic Combinatorics, book
project, available at http://algo.inria.fr.
Di Francesco P, Zinn-Justin P and Zuber J.-B. (2004), Determinant formulae for some tiling problems and application to
fully packed loops, Preprint, ariv:math-ph/0410002.
Di Francesco P and Zinn-Justin P (2005), Quantum Knizhnik
Zamolodchikov equation, generalized RazumovStroganov
sum rules and extended Joseph polynomials. Preprint,
ariv:math-ph/0508059.
finite groups,
quotient groups Tn = Rn =Zn , or more generally,
V=L, where V is a finite-dimensional real vector
space and L is a lattice in V, that is, a discrete
subgroup generated by some basis in V groups
of this type are called tori; it is known that
every commutative connected compact group is a
torus;
unitary groups U(n) and special unitary groups
SU(n), n 2;
orthogonal groups O(n) and SO(n), n 3; and
the groups U(n, H), n 1, of unitary quaternionic
transformations, which are isomorphic to Sp(n) :=
Sp(n, C) \ SU(2n).
The groups O(n) have two connected components,
one of which is SO(n). The groups SU(n) and Sp(n)
are connected and simply connected.
The groups SO(n) are connected but not simply
connected: for n 3, the fundamental group of
SO(n) is Z2 . The universal cover of SO(n) is a
simply connected compact Lie group denoted by
Spin(n). For small n, we have isomorphisms:
Spin(3) SU(2), Spin(4) SU(2) SU(2), Spin(5)
Sp(4), and Spin(6) SU(4).
577
presentation
is
Recall that for a real Lie algebra g, its complexification is gC = g C with obvious commutator. It
is also well known that gC is semisimple or
reductive iff g is semisimple or reductive, respectively. There is a subtlety in the case of simple
algebras: it is possible that a real Lie algebra is
simple, but its complexification gC is only semisimple. However, this problem never arises for Lie
algebras of compact groups: if g is a Lie algebra of a
real compact Lie group, then g is simple if and only if
gC is simple.
The notion of complexification for Lie groups is
more delicate.
Definition 1 Let G be a connected real Lie group
with Lie algebra g. A complexification of G is a
connected complex Lie group GC (i.e., a complex
manifold with a structure of a Lie group such that
group multiplication is given by a complex analytic
map GC GC ! GC ), which contains G as a closed
subgroup, and such that Lie(GC ) = gC . In this case,
we will also say that G is a real form of GC .
It is not obvious why such a complexification
exists at all; in fact, for arbitrary real group it may
not exist. However, for compact groups we do have
the following theorem.
Theorem 2 Let G be a connected compact Lie
group. Then it has a unique complexification GC G.
Moreover, the following properties hold:
(i) The inclusion G GC is a homotopy equivalence. In particular, 1 (G) = 1 (GC ) and the
quotient space GC =G is contractible.
(ii) Every complex finite-dimensional representation
of G can be uniquely extended to a complex
analytic representation of GC .
Bn , n 2
Cn , n 3
Dn , n 4
SU(n 1)
Spin(2n 1)
Sp(n)
Spin(2n)
Maximal Tori
Main Properties
2R
where
a fx 2 aj ad h:x h; hix 8h 2 hg
R f 2 h f0gja 6 0g h
The set R is called the root system of a with
respect to Cartan subalgebra h; elements 2 R are
called roots. We will also frequently use elements
_ 2 h defined by h_ , i = 2(, )=(, ) where ( , )
is a nondegenerate invariant bilinear form on a and
h , i is the pairing between a and a . It can be shown
that so defined _ does not depend on the choice of
the form ( , ).
Theorem 6 Let G be a connected compact Lie
group with Lie algebra g, and let T G be a
maximal torus in G, t = Lie(T) g. Let gC , GC be
the complexification of g, G as in Theorem 2.
Let h = tC gC . Then h is a Cartan subalgebra in
gC , and the corresponding root system R it .
Conversely, every Cartan subalgebra in gC can be
obtained as tC for some maximal torus T G.
Weights and Roots
579
P fjh_ ; i 2 Z 8 2 Rg it
P_ ftjht; i 2 Z 8 2 Rg it;
are
not
lattices:
Q Zn1 ,
!
" exp
"
g
AutV
EndV
581
for some k 2 Z
t 2 t; 2 XT
u2 v2
;
2
u2 v2
;
2i
z iuv
we have
2
0
1
!
Rea2 b2 2Imab Imb2 a2
a b B
C
@ 2Imab jaj2 jbj2 2Reab A
b a
Ima2 b2 2Reab Rea2 b2
i
i
!
2g
i i
sn=2
n2s
s0
Fourier Transform
n
Y
dtk
2it
k
k1
tk e2i
k
R
G, we get an intertwining operator hAi = G A(g)dg.
Comparing this fact with the Schur lemma, one
obtains the following fundamental results.
Let (, V) be any unirrep of a compact group G.
Choose any orthonormal basis {vk , 1 k dim V}
V
in V and denote by tkl
, or tkl
, the function on G
defined by
V
g gvl ; vk
tkl
V
The functions tkl
are called matrix elements of the
unirrep (, V).
b of
In particular, this theorem implies that the set G
equivalence classes of unirreps is countable. For an
f.d. representation (, V) we introduce the character
of as a function
g trg
dim
XV
tkk
g
k1
583
b as the space
We introduce the Hilbert space L2 (G)
b whose value at a point
of matrix-valued functions on G
b belongs to Matd() (C). The norm is defined as
2G
X
kFk2 2 b
d
trFF
L G
b
2G
For a function f on G define its Fourier transform e
f
b
as a matrix-valued function on G:
Z
e
f
f g1 gdg
G
d
tre
f e
f : ke
f k2 2 b
L G
b
2G
(iii) The following inversion formula expresses f in
terms of e
f:
X
f g
d
tre
f g
b
2G
(iv) The Fourier transform sends the convolution to
the matrix multiplication:
f1 f2 e
g
f1
e
f2
where the convolution product is defined by
Z
f1 f2 h
f1 hgf2 g1 dg
G
or
g
X
b
2G
d g
R
where (g) is Diracs delta-function:
G f (g)
(g) dg = f (e). Thus, we get a presentation of Diracs
delta-function as a linear combination of characters.
Classification of Finite-Dimensional
Representations
In this section, we give a classification of unirreps of
a connected compact Lie group G.
Weight Decomposition
V ;
6
2XT
mult;V mult;V w
for any w 2 W.
Classification of Unirreps
However, this representation can be infinite dimensional; moreover, it may not be possible to lift it to a
representation of G.
Definition 5 A weight 2 X(T) is called dominant if h, _i i 2 Z for any simple root i . The set
of all dominant weights is denoted by X (T).
Theorem 17
(i) All weights of L() are of the form = ni i ,
ni 2 Z .
(ii) Let 2 X . Then the irreducible highest-weight
representation L() is f.d. and lifts to a
representation of G.
(iii) Every irreducible f.d. representation of G is of
the form L() for some 2 X .
Thus, we have a bijection {unirreps of G} $ X .
Example 7 Let G = SU(2). There is a unique simple
root and the unique fundamental weight !, related
by = 2!. Therefore, X = Z
! and unirreps are
indexed by non-negative integers. The representation with highest weight k
! is precisely the
representation k constructed in the subsection
Examples of representations.
Example 8 Let G = U(n). Then X = Zn , and X =
{(1 , . . . , n ) 2 Zn j 1
n }. Such objects are
well known in combinatorics: if we additionally
assume that n 0, then such dominant weights are
in bijection with partitions with n parts. They can
also be described by Young diagrams with n rows
(see Fulton and Harris (1991)).
Explicit Construction of Representations
g 2 GC ; b 2 B
Then
585
mult e
10
2XT
xk1 xk1
x x1
k
x xk2
xk ;
k
x e!
Y h_ ; i
h_ ; i
2R
11
Further Reading
Bump D (2004) Lie Groups. New York: Springer.
Brocker T and tom Dieck T (1995) Representations of Compact
Lie Groups, Graduate Texts in Mathematics, vol. 98.
New York: Springer.
Introduction
Superstring theories and M-theory, at present the best
candidate quantum theories which unify gravity,
YangMills fields, and matter, are directly formulated in ten and eleven spacetime dimensions. To
obtain a candidate theory of our four-dimensional
universe, one must find a solution of one of
these theories whose low-energy physics is well
described by a four-dimensional effective field theory
(EFT), containing the well-established standard
model (SM) of particle physics coupled to Einsteins
general relativity (GR). The standard paradigm for
finding such solutions is compactification, along the
lines originally proposed by Kaluza and Klein in the
context of higher-dimensional general relativity. One
postulates that the underlying D-dimensional spacetime is a product of four-dimensional Minkowski
spacetime, with a (D 4)-dimensional compact and
small Riemannian manifold K. One then finds
that low-energy physics effectively averages over K,
leading to a four-dimensional EFT whose field
content and Lagrangian are determined in terms of
the topology and geometry of K.
Of the huge body of prior work on this subject, the
part most relevant for string/M-theory is supergravity
compactification, as in the limit of low energies, small
curvatures and weak coupling, the various string
theories and M-theory reduce to ten- and elevendimensional supergravity theories. Many of the qualitative features of string/M-theory compactification, and
a good deal of what is known quantitatively, can be
General Framework
Let us assume we are given a D- (=d k) dimensional field theory T . A compactification is then a
D-dimensional spacetime which is topologically
the product of a d-dimensional spacetime with an
k-dimensional manifold K, the compactification or
internal manifold, carrying a Riemannian metric
and with definite expectation values for all other
fields in T . These must solve the equations of motion,
and preserve d-dimensional Poincare invariance (or,
perhaps another d-dimensional symmetry group).
q
@Gij @Gkl
det GGik Gjl
@ @
@ @
1
587
for AdS compactifications). The remaining perturbations can be divided into massless fields, corresponding to zero modes of the linearized equations of
motion on K, and massive fields, the others. General
results on eigenvalues of Laplacians imply that the
masses of massive fields depend on the diameter of
K as m 1=diam(K), so at energies far smaller than
m, they cannot be excited (this is not universal;
given strong negative curvature on K, or a rapidly
varying warp factor, one can have perturbations of
small nonzero mass). Thus, the massive fields can be
integrated out, to leave an EFT with a finite
number of fields. In the classical approximation, this
simply means solving their equations of motion in
terms of the massless fields, and using these
solutions to eliminate them from the action. At
leading order in an expansion around a solution,
these fields are zero and this step is trivial; nevertheless, it is useful in making a systematic definition
of the interaction terms in the EFT.
As we saw in pure GR, the configuration space
parametrized by the massless fields in the EFT, is the
moduli space of compactifications obtained by
deforming the original solution. Thus, from a
mathematical point of view, low-energy EFT can
be thought of as a sort of enhancement of the
concept of moduli space, and a dictionary set up
between mathematical and physical languages. To
give its next entry, there is a natural physical metric
on moduli space, defined by restriction from the
metric on the configuration space of the theory T ;
this becomes the sigma-model metric for the scalars
in the EFT. Because the theories T arising from
string theory are geometrically natural, this metric is
also natural from a mathematical point of view, and
one often finds that much is already known about it.
For example, the somewhat fearsome two derivative
terms in eqn [1], are (perhaps) less so when one
realizes that this is an explicit expression for the
WeilPetersson metric on the moduli space of Ricciflat metrics. In any case, knowing this dictionary is
essential for taking advantage of the literature.
Another important entry in this dictionary is that
the automorphism group of a solution translates
into the gauge group in the EFT. This can be either
continuous, leading to the gauge symmetry of
Maxwell and YangMills theories, or discrete,
leading to discrete gauge symmetry. For example, if
the metric on K has continuous isometry group G,
the resulting EFT will have gauge symmetry G, as in
the original example of Kaluza and Klein with K S1
and G U(1). Mathematically, these phenomena
of enhanced symmetry are often treated using the
languages of equivariant theories (cohomology,
K-theory, etc.), stacks, and so on.
CalabiYau Compactification
of the Heterotic String
Contact with the SM requires finding compactifications
to d = 4 either without supersymmetry, or with at most
N = 1 supersymmetry, because the SM includes chiral
fermions, which are incompatible with N > 1. Let us
start with the E8 E8 heterotic string or HE theory.
This choice is made rather than HO because only in this
case can we find the SM fermion representations as
subrepresentations of the adjoint of the gauge group.
Besides the metric, the other bosonic fields of the HE
supergravity theory are a scalar called the dilaton,
YangMills gauge potentials for the group G E8
E8 , and a 2-form gauge potential B (often called the
NeveuSchwarz or NS 2-form) whose defining
characteristic is that it minimally couples to the
heterotic string world-sheet. We will need their gauge
field strengths below: for YangMills, this is a 2-form
a
FIJ
with a indexing the adjoint of Lie G, and for the NS
2-form this is a 3-form HIJK . Denoting the two
MajoranaWeyl spinor representations of SO(1, 9) as
S and C, then the fermions are the gravitino I 2
S
V, a spin 1/2 dilatino 2 C, and the adjoint
gauginos a 2 S. We use I to denote Dirac matrices
contracted with a zehnbein, satisfying {I , J } =
2GIJ , and IJ = (1=2)[I , J ], etc.
A local supersymmetry transformation with parameter
is then
DI
18HIJK JK
1
@I I
12
HIJK IJK
a IJ
a FIJ
2
3
4
1 X a
F ^ Fa
30 a
5
6
589
7
Holonomy
M, II
M
Torus
SU(2)
SU(3)
G2
Sp(4)
SU(4)
Spin(7)
SU(2)
SU(3)
G2
SU(2)
SU(3)
G2
Torus
SU(2)
SU(3)
G2
IIa
IIb
HE, HO, I
Ns
Any
7
5
4
3
3
3
6
4
3
6
4
3
Any
6
4
3
32
16
8
4
6
4
2
16
8
4
16
8
4
16
8
4
2
Max
1
1
1
3
2
1
(1, 1)
2
2
(0, 2)
2
2
Max/2
1
1
1
Ns = 32
Given the supersymmetry algebra, if such a supergravity exists, it is unique. Thus, toroidal compactifications of d = 11 supergravity, IIa and IIb
supergravity lead to the same series of maximally
supersymmetric theories. Their structure is governed by the exceptional Lie algebra E11d ; the
gauge charges transform in a fundamental representation of this algebra, while the scalar fields
parametrize a coset space G=H, where G is the
maximally split real form of the Lie group E11d ,
and H is a maximal compact subgroup of G.
Nonperturbative duality symmetries lead to a
further identification by a maximal discrete subgroup of G.
Ns = 16
8
591
9
First, extended supersymmetry can serve to eliminate many corrections. This is analogous to the
familiar result that the superpotential in d = 4, N = 1
supersymmetric field theory does not receive (or is
protected from) perturbative corrections, and in
many cases follows from similar formal arguments.
In particular, supersymmetry forbids corrections to
the potential and two derivative terms in the
Ns = 32 and Ns = 16 Lagrangians.
In Ns = 8, the superpotential is protected, but the
two derivative terms can receive corrections. However, there is a simple argument which precludes
many such corrections since vector multiplet and
hypermultiplet moduli spaces are decoupled, a
correction whose control parameter sits in (say) a
vector multiplet, cannot affect hypermultiplet moduli space. This fact allows for many exact computations in these theories.
As an example, in IIb on CY3 , the metric on
vector multiplet moduli space is precisely eqn [1] as
obtained from supergravity (in other words, the
WeilPetersson metric on complex structure moduli
space). First, while in principle it could have been
corrected by world-sheet instantons, since these
depend on Kahler moduli which sit in hypermultiplets, it is not. The only other instantons with the
requisite zero modes to modify this metric are
wrapped Dirichlet branes. Since in IIb theory these
wrap even-dimensional cycles, they also depend on
Kahler moduli and thus leave vector moduli space
unaffected.
As previously discussed, for K3-fibered CY3 , this
theory is dual to the heterotic string on K3 T 2 .
There, the vector multiplets arise from Wilson lines
on T 2 , and reduce to an adjoint multiplet of N = 2
supersymmetric YangMills theory. Of course, in
the quantum theory, the metric on this moduli space
receives instanton corrections. Thus, the duality
allows deriving the exact moduli space metric, and
many other results of the SeibergWitten theory of
N = 2 gauge theory, as aspects of the geometry of
CalabiYau moduli space.
In Ns = 4, only the superpotential is protected,
and that only in perturbation theory; it can receive
nonperturbative corrections. Indeed, it appears that
this is fairly generic, suggesting that the effective
potentials in these theories are often sufficiently
complicated to exhibit the structure required for
supersymmetry breaking and the other symmetry
breakings of the SM. Understanding this is an active
subject of research.
We now turn from corrections to novel physical
phenomena which arise in these regimes. While this
is too large a subject to survey here, one of the basic
principles which governs this subject is the idea that
593
3
@A
2A
^ tr A
3
Z
^ H 3 WNP
10
Further Reading
Aharony O (2000) A brief review of little string theories.
Classical and Quantum Gravity 17: 929938.
Aspinwall PS (1996) K3 surfaces and string duality, 1996
preprint, arXiv:hep-th/9611137.
Bachas C et al. (eds.) (2002) Les Houches 2001: Unity from
Duality: Gravity, Gauge Theory and Strings. Berlin:
Springer.
de Boer J et al. (2002) Triples, fluxes, and strings. Advances in
Theoretical and Mathematical Physics 4: 995.
595
e; p; e ; S; p ; S; ; S
6
2
Introduction
The Euler equations for compressible fluids consist of
the conservation laws of mass, momentum, and energy:
@t
rx m 0;
x 2 Rd
m
m
rx p 0
m
E p 0
@t E rx
@t m rx
1
2
3
x 2 Rd ; u 2 Rn
4
1 jmj2
e
2
5
p p
; S
eS=cv
p
e
1
p
R
7
8
@t e v =2 @x pv 0
9
10
>1
11
@t m @x m2 = p 0
@t E @x mE p= 0
12
13
@t v @x p 0
15
~(), = 1=.
respectively, where pressure p = p() = p
The solutions of [12] and [13], as well as [14] and
[15], are equivalent even for entropy solutions with
vacuum where = 0.
The potential flow is well known in transonic
aerodynamics, beyond the isentropic approximation [10] from [1] to [3]. Denote Dt = @t
P
d
k = 1 vk @xk the convective derivative along fluid
particle trajectories. From [1] to [3], we have
Dt S 0
16
@t v @x p 0
@t 12jrx j2 h K
18
597
@t D rx Drx 0
; v; Sjt0 0 ; v0 ; S0 x
19
20
0 x > 0; x 2 R3
21
22
Define
Z
Pt
pt; x; St; x1= p
; S1= dx
3
ZR
vt; x x dx
Ft
R3
with 16 =3
23
() = (
1 , . . . ,
d )() be a unit normal to S 0 . Define
the piecewise smooth initial values for respective
domains D
0 and D0 on either side of the hypersurface S 0 as
u0 x; x 2 D
0
u0 x
25
u
x 2 D
0
0 x;
It is assumed that the initial jump in [25] satisfies the
RankineHugoniot condition, that is, there is a
smooth scalar function () so that
u
0 u0
f u
0
26
0 f u0
s
(D
Sobolev space Hul
0 ), while (0 , v0 , E0 )(x) belongs
to the Sobolev space H s (D
),
for
some
fixed s 10.
0
Assume also that there is a function () 2 H s (S 0 )
so that [26] and [27] hold, and the compatibility
conditions up to order s 1 are satisfied on S 0 by
the initial data, together with the entropy condition:
q
v
p
0
0 ; S0 <
q
< v
p
29
0
0 ; S0
and that (
0 , v0 , E0 )(x) belongs to the uniform local
; mjt0 0 ; m0 x
31
jmt; xj Ct; x
32
33
599
rq; m r; mrf ; m
TVR ; v; St; C TVR 0 ; v0 ; S0
with
r2 ; m 0
and
jV 0
34
This result specially includes that for the barotropic case (Nishida 1968, NishidaSmoller 1973,
DiPerna 1973). Some efforts in the direction of
relaxing the requirement of small total variation
have been made. Some extensions to the initialboundary value problems have also been made. In
addition, an entropy solution in BV with periodic
data or compact support decays when t ! 0.
Furthermore, even for a general hyperbolic system
[4] for x 2 R, we have:
If the initial data functions u0 (x) and v0 (x) have
sufficiently small total variation and u0 v0 2 L1 (R),
then, for the corresponding exact Glimm, or wavefront tracking, or vanishing viscosity solutions u(t, x)
and v(t, x) of the Cauchy problem [4] and [8], there
exists a constant C > 0 such that
kut; vt; kL1 R Cku0 v0 kL1 R
for all t > 0
35
Now we present some multidimensional timedependent problems with a simplifying feature that
the data (domain and/or the initial data) coupled
with the structure of the underlying equations
obey certain geometric structure so that the multidimensional problems can be reduced to lowerdimensional problems with more complicated
couplings. Different types of geometric structure
call for different techniques.
The Euler equations for compressible fluids
with geometric structure describe many important
fluid flows, including spherically symmetric flows
and self-similar flows. Such geometric flows
are motivated by many physical problems such as
shock diffractions, supernovas formation in stellar
dynamics, inertial confinement fusion, and underwater explosions. For the initial data with large
amplitude having geometric structure, the required physical insight is: (1) whether the solution
has the same geometric structure globally and
(2) whether the solution blows up to infinity in a
finite time. These questions are not easily understood in physical experiments and numerical simulations, especially for the blow-up, because of the
limited capacity of available instruments and
computers.
601
36
37
rf k urf l u rf l urf k u
for all k; l 1; 2; . . . ; d
38
x 2 Rd
mm
rx p rx S
39
40
m
m
E p rx
S rx q 41
@t E rx
Here, S = S(rx v, , ) is the viscous stress tensor
which is symmetric from the conservation of angular
momentum and q is the heat flux. If the fluid is
isotropic and the viscous tensor S is a linear function
of rx v and invariant under a change of reference
frame (translation and rotation), then we deduce
from elementary algebraic manipulations that
necessarily
S ; rx v 2; D
42
2
43
0;
2=d 0
603
The inviscid limits from the NavierStokes equations to the Euler equations have been established as
long as the solutions of the Euler equations are
smooth, when the viscosity and heat conductivity
coefficients tend to zero (Klainerman-Majda 1982).
It is completely open for general entropy solutions,
even in the one-dimensional case.
See also: Breaking Water Waves; Capillary Surfaces;
Fluid Mechanics: Numerical Methods; Geophysical
Dynamics; Incompressible Euler Equations:
Mathematical Theory; Inviscid Flows;
Magnetohydrodynamics; Newtonian Fluids and
Thermohydraulics; Non-Newtonian Fluids; Partial
Differential Equations: Some Examples; Stability of
Flows; Viscous Incompressible Fluids: Mathematical
Theory.
Further Reading
Bressan A (2000) Hyperbolic Systems of Conservation Laws: The
One-Dimensional Cauchy Problem. Oxford: Oxford University Press.
Chen G-Q (2005) Euler equations and related hyperbolic
conservation laws. In: Dafermos CM and Feireisl E (eds.)
Handbook of Differential Equations II: Evolutionary Differential Equations, Chapter 1, pp. 1104. Amsterdam: Elsevier.
Chen G-Q and Wang D (2002) The Cauchy problem for the
Euler equations for compressible fluids. In: Friedlander S
and Serre D (eds.) Handbook of Mathematical Fluid
Dynamics, vol. 1, ch. 5, pp. 421543. Amsterdam: Elsevier
Science B.V.
Courant R and Friedrichs KO (1948) Supersonic Flow and Shock
Waves. New York: Springer.
Dafermos CM (2005) Hyperbolic Conservation Laws in Continuum Physics (2nd edn). Berlin: Springer.
Feireisl E (2004) Dynamics of Viscous Compressible Fluids.
Oxford: Oxford University Press.
Glimm J (1965) Solutions in the large for nonlinear hyperbolic
system of equations. Communications on Pure and Applied
Mathematics 18: 95105.
Glimm J and Majda A (1991) Multidimensional Hyperbolic
Problems and Computations. New York: Springer.
Lax PD (1973) Hyperbolic Systems of Conservation Laws and
the Mathematical Theory of Shock Waves. Philadelphia:
SIAM.
Lions PL (1996, 1998) Mathematical Topics in Fluid Mechanics,
vols. 12. New York: Oxford University Press.
Liu T-P (2000) Hyperbolic and Viscous Conservation Laws,
CBMS-NSF RCSAM, vol. 72. Philadelphia: SIAM.
Majda A (1984) Compressible Fluid Flow and Systems of
Conservation Laws in Several Space Variables. New York:
Springer.
Introduction
The numerical analysis of general relativity, or
numerical relativity, is concerned with the use of
computational methods to derive approximate solutions to the Einstein field equations
G 8T
1
2G
M 1027 kg m1
c2
2
Here, and only here, G and c Newtons gravitational constant and the speed of light, respectively
have been explicitly reintroduced. The fact that
RM =R is about 106 and 109 at the surfaces of the
sun and earth, respectively, is a reminder of just how
3
4
605
5
6
1=2
<1
jXj ij xi xj
7
8
9
10
11
13
respectively.
Given the Christoffel symbols, the components of
the spatial Riemmann tensor, denoted here Rijk l , are
computed using
Rijk l @j l ik @i l jk m ik l mj
m jk l mi
14
15
R ij Rij
16
dt
dx i
dx
(t )
12
and
Di Wj @i Wj k ij Wk
idt
(t + dt )
607
17
19
20
21
gij ij
22
completely covariant, but otherwise arbitrary, spacetime tensor, Q , constitute the components of a
completely covariant spatial tensor.
A straightforward calculation, which provides a
good exercise in the use of the 3 1 calculus,
yields the following equally useful identifications for
various pieces of the inverse spacetime metric: g
g00 2
23
24
gij ij 2 i j
25
n n T
ji n T
Sij Tij
26
27
28
29
31
1
@t ij Di j Dj i
2
32
R Kij K K 16
35
609
37
38
N
X
A
T
39
A1
36
p
their solutions as X ij xi xj ! 1 encodes the
conserved mass and linear momentum (four numbers)
that can be defined in asymptotically flat spacetimes.
In a general 3 1 coordinate system, and with an
appropriate choice of variables, the constraints can
be written as a set of quasilinear elliptic equations
for four of the {ij , Ki j } (or, more properly, for
certain algebraic combinations of the {ij , Ki j }).
Thus, especially for 2D and 3D calculations, the
setting of initial data for the Cauchy problem in
general relativity is itself a highly nontrivial mathematical and computational exercise. Readers
wishing more details on this subject are directed to
the comprehensive review by Cook (2000).
A
where T
is the stress tensor corresponding to the
matter field A . Choose a topology for (0) (e.g., R3
with asymptotically flat boundary conditions; T 3 ,
with no boundaries, etc.) This also fixes the
topology of M to be Rthe topology of (0).
Next, freely specify eight of the 12 {ij (0, xk ),
i
K j (0, xk )}, as well as initial values, A (0, xk ), for the
matter fields. Then determine the remaining four
dynamical gravitational fields from the constraints
[35] and [36]. This completes the initial data
specification.
One must now choose a prescription for the
kinematical (coordinate) functions, and i , so that
either explicitly or implicitly, they are completely fixed;
for the case of implicit specification, this may well
mean that the coordinate functions themselves will
satisfy PDEs, which, furthermore, can be of essentially
any type in practice (i.e., elliptic, hyperbolic, parabolic, . . .). Finally, with consistent initial data,
{ij (0, xk ), Ki j (0, xk ); A (0, xk )}, in hand, and with a
prescription for the coordinate functions, the evolution
45
and
1
lim K j O
X!1
X2
i
Further Reading
41
where X is defined by
X
q
ij xi xj
42
X!1
1
X
44
Conformal Geometry see Two-dimensional Conformal Field Theory and Vertex Operator Algebras
Constrained Systems
M Henneaux, Universite Libre de Bruxelles,
Brussels, Belgium
2006 Elsevier Ltd. All rights reserved.
Introduction
Consider a dynamical system with coordinates
qi (i = 1, . . . , n) and Lagrangian L(qi , q i ) (field theory
is formally covered by regarding the spatial coordinates as a continuous index). When going to the
Hamiltonian formulation, it is usually assumed that
the Legendre transformation between the velocities
q i and the momenta
pi
@L
@ q_ i
1
Gauge invariance A system possesses gauge symmetries if it is invariant under transformations that
involve arbitrary functions of time (gauge transformations). In that case, the solution of the
equations of motion with given initial data is not
unique, since it is always possible to perform a
gauge transformation in the course of the evolution
without changing the initial data. It is then clear
that the Legendre transformation cannot be invertible, for if it were, one could rewrite the equations
q2 ! q2 ";
! "_
3
p2 q_ 2 ;
0
4
1
p2 0
5
7
@H
@m
um
@pi
@pi
p_ i
@H
@m
um
@qi
@qi
m q; p 0
8
9
10
11
General Theory
Dirac Algorithm
Primary constraints When the Legendre transformation [1] cannot be inverted, the momenta pi s do
not span an n-dimensional space but are constrained
by relations
m q; p 0;
@m
@pi ; qi
m 1; . . . ; M
6
G; F
@G @F @G @F
@qi @pi @pi @qi
12
m ; H um m ; m0 0
13
j 1; . . . ; M K J
14
j 1; . . . ; J
15
16
17
18
we can now investigate more precisely the restrictions on the variables um . These read
j ; H um j ; m 0;
j 1; . . . ; J
F_ F; H 0 ua F; Vam m
19
H 0 H U m m
20
with
[Vam m ,
j ] 0.
j; j0 1; . . . ; J
21
a ; b 0;
a ; 0;
; C
22
23
It is the generator of
(with a summed from 1 to A).
the time evolution in which the complete gauge
symmetry is fully displayed.
Elimination of second-class constraints Dirac
brackets Second-class constraints do not generate
permissible canonical transformations, since they do
not map the constraint surface on itself. For this
reason, it is convenient to eliminate them. This can
consistently be done by using the Dirac brackets
instead of the Poisson brackets. By definition, the
Dirac bracket [F, G]D of two phase-space functions
F and G is given by
F; DD F; G F; C ; G
24
F arbitrary
25
26
Examples
First example (see eqn [2]). There is here one
primary constraint, namely = 0. The canonical
Hamiltonian is (1=2)((p1 )2 (p2 )2 ) (p1 p2 ).
The consistency algorithm yields the secondary
constraint p1 p2 = 0 and no condition on the us.
The constraints are first class. They generate the
gauge transformations q1 ! q1 ", q2 ! q2 ",
and !
, which coincide with the Lagrangian
gauge transformations if one identifies
with "_
(" and "_ are, of course, independent at any given
time). One can fix the gauge by means of the gauge
conditions = 0, q1 q2 = 0. The reduced phase
space is two-dimensional and the observables can
be identified with the functions of the gaugeinvariant variables (1=2)(q1 q2 ) and p1 p2 ,
which are conjugate. Any other gauge condition
leads to the same reduced phase space.
Quantization
The phase space of physical interest is the reduced
phase space and the physical algebra is the algebra
of the observables. The quantization of the theory
then amounts to quantizing the algebra of the
observables. This can be achieved along two
different lines:
1. Reduce then quantize: In this direct approach,
one represents as quantum operators only the
reduced phase-space functions. There is no
operator associated with non-gauge-invariant
functions.
2. Quantize then reduce: In this approach, one
represents as quantum operators the bigger algebra of functions of all the phase-space variables.
One must then take into account the constraints.
The second-class constraints are enforced as
operator equations, which is consistent with the
correspondence rule that the commutator in the
quantum theory is ih times the Dirac bracket,
AB BA ihA; BD
27
28
Geometric Description
We defined above first-class and second-class
constraints through algebraic means. It turns out
that these definitions also have a geometrical
interpretation, which sheds considerable insight
into their nature.
The phase-space symplectic 2-form induces, by
pullback, a 2-form on the constraint surface .
While is of maximal rank, this may not be the case
for the induced , which may be degenerate. In
fact, the rank of fails to be equal to the
maximum rank 2n J (where J is the total number
of firstof constraints) by precisely the number A
class constraints.
Indeed, the Hamiltonian vector fields Xa associated
with the first-class constraints are tangent to the
constraint surface and are null eigenvectors of ,
Xa ; Y 0
8Y tangent to
29
Further Reading
Anderson JL and Bergmann PG (1951) Constraints in covariant
field theories. Physical Review 83: 1018.
Barnich G, Henneaux M, and Schomblond C (1991) On the
covariant description of the canonical formalism. Physical
Review D 44: 939.
Dirac PAM (1950) Generalized Hamiltonian dynamics. Canadian
Journal of Mathematics 2: 129.
Dirac PAM (1967) Lectures on Quantum Mechanics. New York:
Academic Press.
Flato M, Lichnerowicz A, and Sternheimer D (1976) Deformations of Poisson brackets, Dirac brackets and applications.
Journal of Mathematical Physics 17: 1754.
Hanson A, Regge T, and Teitelboim C (1976) Constrained
Hamiltonian Systems. Rome: Accad. Naz. dei Lincei.
Henneaux M and Teitelboim C (1992) Quantization of Gauge
Systems. Princeton: Princeton University Press.
Henneaux M, Teitelboim C, and Zanelli J (1990) Gauge
invariance and degree of freedom count. Nuclear Physics B
332: 169.
Marsden JE and Weinstein A (1974) Reduction of symplectic
manifolds with symmetry. Reports on Mathematical Physics
5: 121.
617
x, h =
x
(N)
dPN , which in the physics literature is called a
h
propagator, given by
N
Cx;h
X Z eipxhnL
N jpjdd p
2 m2
d
p
2 n2Zd
1
2
m2 2N 1
p2 2N m2
3
ZN ; f
ZN ; 0
N!1
ZN ; f
ZN ; 0
m
d
n2Z
so that (N)
and its propagator C(N)
can be reprex
x, h
sented, see [2], [3], as
N
N
X
hd2=2 zh x
h1
N
Cx;h
N
X
7
hd2
h
Ch x;h h
h1
n2Z
619
Perturbation Theory
The naive approach to the problem is to fix N
> 0 and to develop ZN (, f ) or, more conveniently
and equivalently, (1=jj) log ZN (, f ) in powers of .
If one fixes a priori N , N independent of N,
however, even a formal power series is not possible:
this is trivially due to the divergence of the
coefficients of the power series, already to second
order, for generic f in the limit N ! 1. Nevertheless
it is possible to determine N (), N () as functions
of N and so that a formal power series exists (to
all orders in ): this is the key result of renormalization theory.
To find the perturbative expansion, the simplest is
to use a graphical representation of the coefficients of
the power expansion in , N , N , f and the Gaussian
integration rules which yield (after a classical
computation) that the coefficient of n pN fx1 . . . fxr is
obtained by considering the graph elements shown in
Figure 1, where the segments will be called half-lines
and the graph elements will be called, respectively,
coupling or 4 -vertex, mass vertex, vacuum
vertex, and external vertex.
The half-lines of the graph elements are considered distinct (i.e., imagine a label attached to
distinguish them). Then consider all possible connected graphs G obtained by first drawing, respectively, n, p, r graph elements in Figure 1, which are
not vacuum vertices, with their nodes marked by
points in named x1 , . . . , xn , xn1 , . . . , xnpr ; and
form all possible graphs obtained by attaching pairs
of half-lines emerging from the vertices of the graph
elements. These are the nontrivial graphs.
Furthermore, consider also the single trivial
graph formed just by the third graph element and
consisting of a single point. All graphs obtained in
this way are particular Feynman graphs.
Given a nontrivial graph G (there are many of
them) we define its value to be the product
WG x1 ; . . . ; xn ; xn1 ; . . . ; xnpr
Q
n pN fxnpj Y N
npr
1
Cx ;h
n!p!r!
11
a constant (N) .
1 X
jj G
WG x1 ; . . . ; xnpr
npr
Y
dxj
12
j1
or
N
N
13
which if d = 3 diverge as N ! 1 as or, respectively, as N; the second factor does not diverge in
dimension d = 2 while the first still diverges as N. The
divergences arise from the fact that as x h ! 0 the
propagator behaves as jx hjN if d = 3 or as
log jx hj if d = 2, all the way until saturation
occurs at distance jx hj m1 N : for this reason
the latter divergences are called ultraviolet
divergences.
However, if we set N 6 0, then for every graph
containing a subgraph like those in Figure 2 there
is another one identical except that the points
a, b are connected via a mass vertex, see Figure 1,
with the vertex in x, by a line ax and a line xb;
the new graph value receives a contribution from
the mass vertex inserted in x between a and b
simply given by a factor N . Therefore if we fix,
for d = 3,
42 3! 2
N
N 6Cxx
2
Z
def
N3
N
Cxh dh 6Cxx N
14
C(N)3
(C(N)
C(N)
) dh. If d = 2, we only
C(N)
hb
ax
xh
xb
need to define N as the first term on the right-hand
side (RHS) of [14] and we can leave the subgraphs like
the second in Figure 2 as they are (without any
renormalization).
Graphs without external lines are called vacuum
graphs and there are a few such graphs which are
divergent. Namely, if d = 3, they are the first three
drawn in Figure 3; furthermore, if N is set to the
above nonzero value a new vacuum graph, the
fourth in Figure 3, can be formed. Such graphs
contribute to the graph value, respectively, the terms
in the sum
Z
4!
23 3!3 3
N2
N4
3Cx ;x 2 Cx x dx2
1 1
1 2
2
3!
Z
N2 N2 N2
N
Cx x Cx x Cx x dx2 dx3 N Cx x
15
1 2
2 3
3 1
1 1
1
1
2
3
621
p
k=0
1
2
3
4
5
6
7
8
9
1
2
3
4
5
6
7
8
9
h
f
t
leads to
k=0
h
p
1
q
2
f
4
hv hv0 nv rv 1
X
hv kninner
v
v>root
16
v>root
v>root
with
5
v>root
def
e
4nv rv nev
ninner
v
hv hv0 e
ninner
v
17
v d 4 dnv rv
d2 d2 e
nv
2
2
623
18
Xx
N1
zN x d2=2 Xx
if d = 2 this becomes
N
Xx
1 N N 11=2 N1
zN
Xx
N 1=2 x
N 1=2
N 2
N d2N Xx
N fx d2=2N Xx
dx
jj
19
2QN
3
N fx 2N Xx
dx
jj
20
where
def
N 6cN 2 N N c0N ;
def
32
N
Xx
dx
jj
where
Vj; N1 is the Taylor expansion of
R
log eVN dP(z(N) ) in powers of (hence essentially
in the very small parameter (4d)N ) truncated at
order j, that is,
V1;N1 hVN i1
"
2QN
21
V2;N1 hVN i
where N = 6cN and N = 3c2N and cN , computable from [13], admits a limit as N ! 1.
The fields z(N) and X(N1) can be considered
constant over boxes 2 QN : z(N)
= s , X(N1)
= x
x
x
for x 2 and the s can be considered statistically
independent on the scale of the lattice QN .
Therefore, [20] and [21] show that integration over
z(N) in the integral defining ZN (, f ) is not too
different from the computation of a partition function of a lattice continuous spin model in which the
spins are s and, most important, interact extremely weakly if N is large. In fact, the coupling
constants are of order of a power of jX(N1) j times
O( N ) if d = 3 (O(N 2 2N ) if d = 2), or of order
O( N(d2)=2 max jfx j), no matter how large and f.
This says that the smallest scale fields are
extremely weakly coupled. The fields X(N1) can be
regarded as external fields of size that will be called
BN1 , of order 1 or even allowed to grow with a
power of N, see [6]. Their presence in VN does not
affect the size of the couplings, as far as the analysis
of the integral over z(N) is concerned, because the
couplings remain exponentially small in N, see [20]
and [21], being at worst multiplied by a power of
BN1 , i.e., changed by a factor which is a power of N.
The smallness of the coupling at small scale is a
property called asymptotic freedom. Once fields
and coordinates are correctly scaled, the real size
of the coupling becomes manifest, that is, it is
extremely small and the addends in VN proportional
to the counter-terms N , N , which looked
V3;N1 hVN i
def
"
#2
2
hVN
i hVN i2
2!
2
hVN
i hVN i2
2!
#
2
2
hVN hVN
i hVN i2 i hVN hVN
i hVN i2 3
3!
; ...
23
j
where []
denotes truncation to order j in ,
and R(j, N) is a remainder (depending on (N1)
)
x
which can be expected to be estimated, for d = 2, 3, by
jRj; Nj Rj; N
def
4j
Cj BN N 2 4dN j1 dN
24
625
Chh0 Chh0
2! 2!
2
h
h
h h0
dhdh0
which is conveniently expressed in terms of a
nonlocal field
h
h def
Yhh0
rel
as Vh
rel;1
Vh
h jh h0 j4
rel;2
Vh
with
X Z
rel;2 def
2 2h
Vh
h
h0
;0 2Q
0 h
ec
jhh j
h2
0
Yhh0 Ahh0
dhdh0
jj j0 j
25
where
h
0<a
0
Ahh0
h jh
h0 j31=2
!
< a0
N
h nk
Xx
q
Y
h n0
Yh 0 h0k0
k0 1
Wx1 ; . . . ; h0q
p
Y
k0
e
h 0
Remarks
(i) Checking scale
R independence for j = 1 is just
checking that P(dz(h) )V1; h = V1; h1 . Note that
Z
def
h4
h h2
h2
V1;h
x
6C00 x
3C00
dx
q
Y
dhk0 dh0k0
dxk
jk j k0 1 j1k0 j j2k0 j
k1
26
p
(1q
2q ) of boxes 2 Qh , and
d(x1 , . . . , h0q ) is the length of the shortest tree
graph that connects all the p 2q > 0 points, the
exponents n, t are 2, and t is 3 if q > 0;
the kernel W depends on all coordinates
x1 , . . . , h0q
Qq
0
and it is bounded
P above
Pby Cj k0 = 1 Ahk0 hk0 for some
Cj ; the sums
nk n0k0 cannot exceed 4j. The
test functions f do not appear in [26] because by
assumption they are bounded by 1: but W depends
on the f s as well.
The field-independent part is simply the value
of log ZN (, f ) computed by the perturbation
analysis in the section Perturbation theory up to
order j in but using as propagator (C(N) C(h) ):
thus, E(j, h) is a constant depending on N but
uniformly bounded as N ! 1 (because of the
renormalizability proved in the section Perturbation theory).
If d = 2, there is no need to introduce the nonlocal
fields Y (h) and in [26] one can simply take q = 0,
and the relevant part also can be expressed by
omitting the term Vh(rel, 2) in [25]: unlike the d = 3
case, the estimate on the kernels W by an
N-independent Cj holds uniformly in h without
having to introduce Y. For d = 2, it will therefore be
supposed that Vh(rel, 2) 0 in [25] and q = 0 in [26].
It is not necessary to have more information on
the structure of Vj; h even though one can find simple
graphical rules, closely related to the ones in the
section Perturbation theory, to construct the
coefficients W in full detail. The W depend, of
course, on h but the uniformity of the bound on W
is the only relevant property and in this sense the
effective potentials are said to be (almost) scale
independent.
The above bounds on the irrelevant part can
be checked by an elementary direct computation if
j 3: in spite of its elementary character, the
uniformity in h N is a result ultimately playing an
essential role in the theory together with the
dominance of the relevant part over the irrelevant
one which, once the fields are properly scaled, is
much smaller (by a factor of order h , see [26]),
at least if h is large.
Nonperturbative Renormalization:
Small Fields
Having introduced the notion of effective potential
Vj; h , of order j and scale h, satisfying the bounds
(described after [26]) on the kernels W representing
it, the problem is to estimate the remainder in [22]
and find its relation with the value [24] given by the
heuristic Taylor expansion. Assume < 1 to avoid
distinguishing this case from that with 1 which
would lead to very similar estimates but to different
-dependence on some constants.
Define B (z(h) ) = 1 if kz(h) k Bh2 for all 2 Qh ,
see [8], and 0 otherwise; then the following lemma
holds:
Lemma 1 Let kX(h) k be defined as [8] with z
replaced by X and suppose kX(h) k Bh4 for all
then, for all j 1, it is
Z
eVj;h1 B zh1 dPzh1
0
eVj;h R j;h1jj
27
Rj; h 1 c ec B
h12
627
Y
2 2
c ec BN ec
00
4dN N2 BN 4 4
#RN
29
2RN
constants:
this
is
00
1 c ec B
N4
ec
00
629
7.
8.
9.
10.
Further Reading
Aizenman M (1982) Geometric analysis of 4 -fields and Ising
models. Communications in Mathematical Physics 86: 148.
Balaban T (1983) (Higgs)3, 2 quantum fields in a finite volume. III.
Renormalization. Communications in Mathematical Physics
88: 411445.
Benfatto G, Cassandro M, Gallavotti G et al. (1978) Some
probabilistic techniques in field theory. Communications in
Mathematical Physics 59: 143166.
Benfatto G, Cassandro M, Gallavotti G et al. (1980) Ultraviolet
stability in Euclidean scalar field theories. Communications in
Mathematical Physics 71: 95130.
Benfatto G and Gallavotti G (1995) Renormalization Group,
pp. 1143. Princeton: Princeton University Press.
Benfatto G, Giuliani A, and Mastropietro V (2003) Low
temperature analysis of two dimensional Fermi systems with
symmetric Fermi surface. Annales Henry Poincare 4: 137193.
De Calan C and Rivasseau V (1981) Local existence of the Borel
transform in euclidean 44 . Communications in Mathematical
Physics 82: 69100.
Frohlich J (1982) On the triviality of 4d theories and the
approach to the critical point in d 4 dimensions. Nuclear
Physics B 200: 281296.
Contact Manifolds
Gallavotti G (1978) Some aspects of renormalization problems in
statistical mechanics. Memorie dell Accademia dei Lincei
15: 2359.
Gallavotti G (1981) Elliptic operators and Gaussian processes. In:
Aspects Statistiques et Aspects Physiques des Processus Gaussiens, pp. 349360. Colloques Internat. C.N.R.S, St. Flour.
Publications du CNRS, Paris.
Gallavotti G (1985) Renormalization theory and ultraviolet
stability via renormalization group methods. Reviews of
Modern Physics 57: 471569.
Gawedzky K and Kupiainen A (1983) Block spin renormalization
group for dipole gas and (@)4 . Annals of Physics 147: 198243.
Gawedzky K and Kupiainen A (1985a) GrossNeveu model
through convergent perturbation expansion. Communications
in Mathematical Physics 102: 130.
Gawedzky K and Kupiainen A (1985b) Massless lattice 44 theory:
rigorous control of a renormalizable asymptotically free model.
Communications in Mathematical Physics 99: 197252.
Giuliani A and Mastropietro V (2005) Anomalous universality in
the anisotropic AshkinTeller model. Communications in
Mathematical Physics 256: 681735.
Glimm J, Jaffe A, and Spencer T (1973) Velo G and Wightman A
(eds.) Constructive Field theory, Lecture Notes in Physics,
vol. 25, pp. 132242. New York: Springer.
631
Contact Manifolds
J B Etnyre, University of Pennsylvania,
Philadelphia, PA, USA
2006 Elsevier Ltd. All rights reserved.
(e.g., thermodynamics, fluid dynamics, holomorphic curves, and open book decompositions)
are provided in the Further reading section.
Introduction
A hyperplane field
on a manifold M is a codimension-1 sub-bundle of the tangent bundle TM. Locally,
a hyperplane field can always be described as the
kernel of a 1-form. In other words, for every point in
M there is a neighborhood U and a 1-form defined
on U such that the kernel of the linear map
x : Tx M ! R is
x for all x in U. The form is called
a local defining form for
. A contact structure on a
(2n 1)-dimensional manifold M is a maximally
nonintegrable hyperplane field
. The hyperplane
field
is maximally nonintegrable if for any (and hence
every) locally defining 1-form for
the following
equation holds:
^ dn 6 0
1
n
X
pi dqi
2
i1
3
Local Theory
The natural equivalence between contact structures
is contactomorphism. Two contact structures 0 and
Contact Manifolds
633
4
5
and restricting d to this subspace yields a symplectic form !, where is the Liouville form on T M
defined in Example 2. A choice of contact form
fixes an identification of X with the sub-bundle of
T M under which d(t) is taken to d.
The vector field v = @=@t on (X, !) is a symplectic
dilation that is transverse to M {1} X. Clearly,
v !jM{1} = . Thus, we see that any co-orientable
contact manifold can be realized as a hypersurface
of contact type in a symplectic manifold. In
summary, we have the following theorem.
Theorem 4 If (M, ) is a co-oriented contact
manifold, then there is a symplectic manifold
Symp(M, ) in which M sits as a hypersurface of
contact type. Moreover, any contact form for
gives an embedding of M into Symp(M, ) that
realizes M as a hypersurface of contact type.
We also note that all the hypersurfaces of contact
type in (X, !) look locally, in X, like a contact
manifold sitting inside its symplectification.
Theorem 5 Given a compact hypersurface M of
contact type in a symplectic manifold (X, !) with the
symplectic dilation given by v, there is a neighborhood of M in X symplectomorphic to a neighborhood of M {1} in Symp(M, ) where the
symplectization is identified with M (0, 1) using
the contact form = v !jM and = ker .
and
v d 0
6
Contact Manifolds
2n
X
v2j iav
7
j1
Hamiltonian Dynamics
Given a symplectic manifold (X, !), a function
H : X ! R will be called a Hamiltonian. (Only
autonomous Hamiltonians are discussed here.) The
unique vector field satisfying
vH ! dH
is called the Hamiltonian vector field associated to
H. Many problems in classical mechanics can be
formulated in terms of studying the flow of vH for
various H.
Example 8 If (X, !) = (R 2n , d), where is from
Example 2, then the flow of the Hamiltonian vector
field is given by
@H
;
@p
@H
@q
635
Geometric Optics
In this section, we study the propagation of light (or
various other disturbances) in a medium (for the
moment, we do not specify the properties of this
medium). The medium will be given by a threedimensional manifold M. Given a point p in M and
t > 0, let Ip (t) be the set of all points to which light
can travel in time t. The wave front of p at time t
is the boundary of this set and is denoted as
p (t) = @Ip (t).
Theorem 8 (Huygens principle). p (t t0 ) is the
envelope of the wave fronts q (t0 ) for all q 2 p (t).
This is best understood in terms of contact
geometry. Let : (T Mn{0}) ! P M be the natural
projection (see Example 3) and let S be any smooth
sub-bundle of T Mn{0} that is transverse to the radial
vector field in each fiber and for which jS : S ! P M
is a diffeomorphism. The restriction of the Liouville
form to S gives a contact form and a corresponding
Reeb vector field v. Given a subset F of M with a welldefined tangent space at every point set
LF fp 2 S : p 2 F and pw 0 for all
w 2 Tp Fg
8
q_
p_
Further Reading
Aebisher B, Borer M, Kalin M, Leuenberger Ch, and Reimann
HM (1994) Symplectic Geometry, Progress in Mathematics,
vol. 124. Basel: Birkhauser.
Arnold VI (1989) Mathematical Methods of Classical Mechanics,
Graduate Texts in Mathematics, vol. 60, xvi516, pp. 163179.
New York: Springer.
Arnold VI (1990) Contact Geometry: The Geometrical Method of
Gibbss Thermodynamics, Proceedings of the Gibbs Symposium.
(New Haven, CT, 1989), pp. 163179. Providence, RI: American
Mathematical Society.
Beals R and Greiner P (1988) Calculus on Heisenberg manifolds.
Annals of Mathematics Studies 119.
Eliashberg Y, Givental A, and Hofer H (2000) Introduction to
Symplectic Field Theory, GAFA 2000 (Tel Aviv, 1999), Geom.
Funct. Anal. 2000, Special Volume, Part II, pp. 560673.
Etnyre J. Legendrian and transversal knots. Handbook of Knot
Theory (in press).
Etnyre J (1998) Symplectic Convexity in Low-Dimensional
Topology, Symplectic, Contact and Low-Dimensional Topology (Athens, GA, 1996), Topology Appl., vol. 88, No. 12,
pp. 325.
Introduction
Control Theory is an interdisciplinary research area,
bridging mathematics and engineering, dealing with
physical systems which can be controlled, that is,
whose evolution can be influenced by some external
agent. A general model can be written as
yt At; y0; u
1
637
y2
y2
u=
1)
y1
u(y) = 1
y1
u(y) = +1
2
4
(u = +
1)
yt; 0 0
5
yt 0; y1
6
7
Control of ODEs
For most proofs we refer to Agrachev and Sachkov
(2004) and Sontag (1998).
Controllability
u 2 U; y0 y0
8
fi yui
y_ 1 u1 ;
y_ 2 u2 ;
y_ 3 u1 y2 u2 y1
y_ f0 y
9
i1
Alternatively,
Hamiltonian
one
can
define
the
639
maximized
Vt;
y inf Jy; u
Hy; p maxhp; f y; ui
Consider a general
y_ f y; u
10
u2U
yT; u;
y0
y
11
!2U
13
@Hy ; p; u
;
@p
p_
14
15
16
u2U
_
vt
Dy f y t; u t vt
@Hy ; p; u
@y
y_ f t; y; u;
ut 2 U
a.e.
17
FG
F+G
u = +1 u = 1
Control of PDEs
The theory for control of models governed by PDEs
is, as expected, much more ramified and much less
complete. An exhaustive resume of the available
results is not possible in short space, thus we focus
on Example 2 and few others to illustrate some
techniques to treat control problems and give
various references (see also Fursikov and Imanuvilov
(1996), Komornik (1994), and Lasiecka and Triggiani
(2000), and references therein).
Besides the variety of control problems illustrated
in the Introduction, for PDE models one can consider
different ways of applying the control, for example:
Boundary control One consider the system [3]
(with F independent of u) and impose the condition
y(t, x) = u(t, x) to hold for every time t and every x in
some region. Usually, we assume y(t) to be defined
bounded region and the control acts on some set
@. Obviously, also Neumann conditions are
natural as @ y = u where is the exterior normal to .
Internal control One consider the system [3]
with F depending on u. Thus, the control acts on the
equation directly.
Other controls There are various other control
problems one may consider as Galerkin-type
approximation and control of some finite family of
modes. An interesting example is given by Coron
(2002), where the position of a tank is controlled to
regulate the water level inside.
Control of a Vibrating String
z0; t z1; t 0
19
on
641
Further Reading
Agrachev A and Sachkov Y (2004) Control from a Geometric
Perspective. Springer.
Bardi M and Capuzzo Dolcetta I (1997) Optimal Control and
Viscosity Solutions of HamiltonJacobiBellman Equations.
Boston: Birkhauser.
Bloch AM (2003) Nonholonomic Mechanics and Control, with
the collaboration of J. Baillieul, P. Crouch and J. Marsden,
with scientific input from P. S. Krishnaprasad, R. M. Murray
and D. Zenkov. New York: Springer.
Boscain U and Piccoli B (2004) Optimal Synthesis for Control
Systems on 2-D Manifolds. Springer SMAI, vol. 43. Heidelberg:
Springer.
Introduction
Convexity is an important notion in nonlinear
optimization theory as well as in infinitedimensional functional analysis. As will be seen
below, very simple and powerful tools will be
derived from elementary duality arguments (which
are by-products of the MoreauFenchel transform
and HahnBanach theorem). We will emphasize on
applications to a large range of variational problems. Some arguments of measure theory will be
skipped.
Remark 3
(i) The following sequential notion can be also
used: f is -sequentially l.s.c. at x0 if
Let f : X ! R [ {1} be
(i) f is -l.s.c.,
(ii) 8r 2 R, f 1 ((1, r]) is -relatively compact.
2 X such that f (
Then there is x
x) = inf f and
argmin f := {x 2 Xjf (x) = inf f } is -compact.
In practice, the choice of the topology is ruled
by the condition (ii) above. For example, if X is a
reflexive infinite-dimensional Banach space and if f
is coercive (i.e., limkxk ! 1 f (x) = 1), we may take
for the weak topology (but never the normed
topology). This restriction implies in practice that
the first condition in Theorem 4 may fail. In this
case, it is often useful to substitute f with its lowersemicontinuous (l.s.c.) envelope.
Definition 5 Given a topology , the relaxed function
f (=f ) is defined as
f x supfgxjg : X ! R [ f1g;
g is -l:s:c:; g f g
It is easy to check that f is -l.s.c. at x0 if and only
if f (x0 ) = f (x0 ). Futhermore,
f x sup inf f ;
U2V u U
x; x 2 X X
Let f : X ! R, be defined by
f x
MoreauFenchel Conjugate
643
1
p
kxkX ;
p
1 < p < 1
then,
f x
1 p0
kx kX ;
p0
with
1 1
1
p p0
f = B , where
Example 2 Let A 2 R dsym be a symmetric positivedefinite matrix and let f (x) := (1=2)(Ax j x)(x 2 R d ).
Then, for all y 2 Rd , we have f (y) = (1=2)(A1 y j y).
Notice that if A has a negative eigenvalue, then
f 1.
Particular examples on Rd are also very popular.
For instance:
Minimal surfaces
q
f x 1 jxj2
( q
2
f y 1 jyj
1
if jyj 1
Duality Arguments
otherwise
f x C c suphxjx i
Entropy
x log x
f x
1
Example 3
Then,
if x 2 R
;
otherwise
f y expy 1
x2C
support function of C
Notice that if M is a subspace of X, then
(M ) = M? . We specify now a particular case of
interest.
Let be a bounded open subset of Rn . Take
R d ) to be the Banach space of continuX = C0 (;
with values in Rd .
ous functions on the compact )
As usual, we identify the dual X with the space
R d ) of R d -valued Borel measures on
with
Mb (;
finite total variation. Let K be a closed convex of
Rd such that 0 2 K. Then 0K () := sup {( j z): z 2 K}
is a non-negative convex l.s.c. and positively
1-homogeneous function on R d (e.g., K is the
Euclidean norm if K is the unit ball of Rd ). Let us
define C := { 2 X: (x) 2 K, 8x 2 }. Then, we
have
Z
C 0K
Z
d
dx
1
: 0K
d
where
is any non-negative Radon measure such
that
(the choice of
is indifferent). In the case
where K is the unit ball, we recover the total
variation of .
Then
f g x =
inf
x1 x2 = x
(i)
8f 2 X ;
Rkx kX M
645
{f x1 g x2 }
(the equality holds in R).
(ii) If both sides of the equality in (i) are finite, then
the infimum in the right-hand side is achieved.
Proof Without any loss of generality, we may
assume that x = 0 (we reduce to this case by
substituting g with g h , x i). We let
hp infff x p gxjx 2 Xg
Noticing that (p, x) 7! f (x p) g(x) is convex, we
infer that h(p) is convex as well. As h is majorized
by the function p 7! f (x0 p) g(x0 ), which by [2]
continuous at 0, we deduce from Theorems 1 and 11
that h(0) = h (0) and that h achieves its infimum.
Now h(0) = inf(f g) = (f g) (0) and
h p supfhp; p i hp: p 2 Xg
supfhp; p i f x p gx: x 2 X; p 2 Xg
g p f p
The assertions (i), (ii) follow since h (0) =
min h = min {g (p ) f (p )}.
&
Proposition 13 (Composition). Let X, Y be two
Banach spaces and A : X 7! Y a linear operator with
dense domain D(A). Let : Y ! R [ {1} be a
Ku; p p
&
u; 2 GA ()u 2 W1;2 ; 9
2 L2 ; Rn :
r u
;
x 2 T x?
u2X
2Y
Fu
f x; rudx
f x; ux; ruxdx
1
0
if ux > t
otherwise
Lemma 18
if AT y c 0; y 0
otherwise
Theorem 19
Ax0 b, AT y0 c 0.
such that
hp : inffcjx: x 0; Ax p bg
8y 2 Rm ;
h y bjy
1
647
is
Theorem 20 Assume that [5] holds. Then x
optimal for (P) if and only if there exist Lagrangian
multipliers 1 , 2 , . . . m in R such that
!
X
2 argmin f
x 0; 8j
j gj ; j gj
x
X
Notice that the existence of such a solution x
n
and
if,
for
some
is ensured if, for example,
X
=
R
P
k > 0, the function f k j gj is coercive.
jtrzj2 jzj2
2
After some computations, we may write the supremum appearing in Proposition 14 as our dual
problem
Z
2
P
sup j x; dx: 2 L2 ; Rnsym ;
div f on ; n g on 1
where j is the MoreauFenchel conjugate with
respect to the second argument and n(x) denotes
the exterior unit normal on . The matrix-valued
map is called the stress tensor and j the stress
potential. Note that the boundary conditions for n
have to be understood in the sense of traces.
Theorem 21 The problems (P) and (P ) have
solutions and we have the equality: inf(P) = sup (P ).
Futhermore, a pair (
u, ) is optimal if and only if it
satisfies the following system:
div f
on
x 2 @jx; e
u a:e: on
0
u
a:e: on 0
on 1
n g
equlibrium
constitutive law
XY
where
y : inffcx; y x: x 2 Xg
Proof The convexity property is obvious and the
properness follows from the fact that
Z
F
cx; y dxdy
XY
F
Let us compute now F (). We have
Z
cx; ydxdy
F inf
XY
Z
d: 2 PX; 2 ;
Z X
inf
cx; y xdxdy:
XY
2 ;
Z
c y dy
Tc ; F
Z
Z
d
c d: 2 C0 X
sup
Y
Z X
Z
sup
d
d: ; 2 F c
X
Tc ;
&
We observe that, for every 2 C0 (X), the function c introduced in Lemma 22 is continuous (use
the uniform continuity of c) and therefore the pair
(, c ) belong to the class
F c : ; 2 C0 X C0 Y:
x y cx; yg
Proof
XY
649
deduce that
Z
Z
c
n d
n d
sup7
lim
n
Y
Z X
Z
c
n cn d n cn d
lim
n
X
Y
Z
Z
c
d
d
Z
cx; y x
y dxdy
XY
&
Remark 24
(i) In their discrete version (i.e., , are atomic
measures), problems [6] and [7] can be seen as
particular linear programming problems (see the
section Finite-dimensional case).
(ii) The case X = Y Rn and c(x, y) = (1=2)jx yj2
is important. In this case, the notion of c-concavity
is linked to convexity and the Fenchel transform
since, for every 2 C0 (X), one has
!
2
j j2
j
j
c
2
2
c ) is a solution of [7], we find that
Then if (,
0 x :
jxj2
x
2
Tc ; max
X
As it appears, Tc (, ) depends only on the difference f = , which belongs to the space M0 (X) of
signed measure on X with zero average. Defining
N(f ) := Tc (f , f ) provides a seminorm (Kantorovich norm) on M0 (X) (it turns out that M0 (X) is
not complete and that in general its completion is a
strict subspace of the dual of Lip(X)).
We will now specialize to the case where X is a
compact manifold equipped with a geodesic distance. This will allow us to link the original problem
to another primaldual formulation closer to that
considered in the section Primaldual formulation
in mechanics and yielding to a connection with
partial differential equations. As a model example,
where is a bounded
let us assume that K = ,
connected open subset of Rn with a Lipschitz
be a compact subset (on
boundary. Let
which the transport will have zero cost) and define
cx;y: inf H1 S n :
S Lipschitz curve joining x to y; S
9
10
Z
R n ;
j j: 2 Mb ;
n
div f on
12
jruj 1 on u 0 on
To prove that = Tc (, ) = sup (11), we consider a
in [11] and prove that it can be
maximizer u
approximated uniformly by a sequence {un } of
which satisfy the same confunctions in C1 ()
straints. This technical part is done by truncation
and convolution arguments (we refer to Bouchitte
&
et al. (2003) for details).
Remark 26 By localizing the integral identity
associated with [12], it is possible to deduce
the optimality conditions which characterize optimal
for [11], [12] (without requiring any
pairs (
u, )
regularity). This is done by using a weak notion
of tangential gradient with respect to a measure
(see Bouchitte et al. (1997) and Bouchitte and
Fragala` (2002)). If = dx where 2 L1 (; Rn )
and if @, then we find that = ar
u, where the
pair (
u, a) solves the following system:
divar
u f
jr
uj 1
u0
@u
0
@n
on
diffusion equation
a.e. on fa > 0g eikonal equation
a.e. on
on
651
Sxy
xy
on n
on
the sequence {p } does converge weakly-star to ,
the unique minimizer of the problem
inffE : solution of 12
g
The general case, in particular when all optimal
measures are singular, is open.
Remark 29 Variational problems [11], [12] have
important counterparts in the theory of elasticity
and in optimal design problems (see Bouchitte and
Buttazo (2001)). They read, respectively, as
Z
max
u df: u 2 \p>1 W 1;p ; R n ;
rux 2 K a:e: on ; u 0 on
Z
R n2 ;
min
0K : 2 Mb ;
sym
div f on n
2
on
Further Reading
Alberti G, Bouchitte G, and Dal Maso G (2003) The calibration
method for the MumfordShah functional and free-discontinuity problems. Calculus of Variations and Partial Differential
Equations 16(3): 299333.
Ambrosio L (2003) Lecture notes on optimal transport problems.
In: Mathematical Aspects of Evolving Interfaces (Funchal
2000), Lecture Notes in Mathematics, vol. 1812, pp. 152.
Berlin: Springer.
Borwein M and Lewis SA (2000) Convex Analysis and Nonlinear
Optimization. Theory and Examples, CMS Series. Berlin:
Springer.
Bouchitte G and Buttazzo G (2001) Characterization of optimal
shapes and masses through MongeKantorovich equations.
Journal of the European Mathematical Society 3: 139168.
Bouchitte G, Buttazzo G, and De Pascale L (2003) A p-Laplacian
approximation for some mass optimization problems. Journal
of Optimization Theory and Applications 118: 125.
Bouchitte G, Buttazzo G, and Seppecher P (1997) Energies with
respect to a measure and applications to low dimensional
structures. Calculus of Variations and Partial Differential
Equations 5: 3754.
Bouchitte G and Dal Maso G (1993) Integral representation and
relaxation of convex local functionals on BV. Annali della
Scuola Superiore di Pisa 20(4): 483533.
Bouchitte G and Fragala` I (2002) Variational theory of weak
geometric structures: the measure method and its applications.
Variational Methods for Discontinuous Structures, Ser.
PNLDE, vol. 51, pp. 1940. Basel: Birkhauser.
Bouchitte G and Fragala` I (2003) Second order energies on thin
structures: variational theory and non-local effects. Journal of
Functional Analysis 204(1): 228267.
Bouchitte G and Valadier M (1988) Integral representation of
convex functionals on a space of measures. Journal of
Functional Analysis 80: 398420.
Ekeland I and Temam R (1976) Analyse convexe et proble`mes
variationnels. Paris: Dunod-Gauthier Villars.
Evans LC (1997) Partial differential equations and Monge
Kantorovich mass transfer. In: Bott R, Jaffe A, Jerison D,
Lutsztig G, Singer I, and Yau JT (eds.) Current Developments
in Mathematics, pp. 65126. Cambridge.
Michell AGM (1904) The limits of economy of material in frame
structures. Philosophical Magazine and Journal of Science
6: 589597.
Rockafellar RT (1970) Convex Analysis. Princeton: Princeton
University Press.
Villani C (2003) Topics in Optimal transportation, Graduate
studies in Mathematics, vol. 58. Providence, RI: AMS.
Introduction
Mathematical cosmology focuses on the geometrical
and mathematical aspects of the study of the
universe as a whole. Because the structure of
spacetime (with metric tensor gab (xj )) is governed
by gravity, with matter and energy causing spacetime curvature according to the nonlinear gravitational field equations of the theory of general
relativity, it has its roots in differential geometry. It
is to be distinguished from the three other major
aspects of modern cosmology, namely astrophysical
cosmology, high-energy physics cosmology, and
observational cosmology; see Peacock (1999) for
these aspects.
The Einstein field equations (EFEs) are
Rab 12 Rgab gab Tab
1
2
3
(note that realistic cosmological models are nonempty), whereas all ordinary matter has a positive
gravitational mass density:
3p > 0
4
Exact Properties
We can split the equations into spacelike and
timelike parts relative to the 4-velocity ua , obtaining the (1 3) covariant dynamical equations and
identities in terms of the fluid shear ab , vorticity
!ab , expansion = ua ;a , and acceleration ab =
ua;b ub (Ehlers 1961, Ellis 1971, Ellis and van Elst
1999). The energy density of a perfect fluid obeys
the conservation equation
_ 3 p
S_
S
5
S
3 2!2 2 ab;b 3p
S
2
6
= 3S=S.
This is the basis of the fundamental
singularity theorem: if in an expanding universe
! = 0 = ab and the combined matter present satisfies
[4], with 0, then there was a singularity where
0 being
S ! 0 a finite time t0 < 1=H0 ago, H0 = (S=S)
the present value of the Hubble constant. The energy
density will diverge there, so this is a spacetime
singularity: an origin of physics, matter, and spacetime itself. However, the deduction does not follow if
there is rotation or acceleration, which could
conceivably avoid the singularity, so this result is by
itself inconclusive for realistic cosmologies.
The vorticity obeys conservation laws analogous
to those in Newtonian theory (Ehlers 1961).
Vorticity-free solutions (! = 0) occur whenever the
fluid flow lines are hypersurface-orthogonal in
spacetime, that is, there exists a cosmic time
function for the comoving observers, which will
measure proper time along the flow lines if
additionally the fluid flow is geodesic. The rate of
change of shear is related to the conformal curvature
(Weyl) tensor, which represents the free gravitational field, and which splits into an electric part Eab
and a magnetic part Hab in close analogy with
electromagnetic theory. Shear-free solutions ( = 0)
are very special because they strongly constrain the
Weyl tensor; indeed if the flow is shear free and
geodesic, then it either does not expand ( = 0), or
does not rotate (! = 0) (Ellis 1967). The set of
cosmological observations associated with generic
cosmological models has been characterized in
power series form by Kristian and Sachs (1966),
and that result has been extended to general models
by Ellis et al. (1985).
The local regularity of the theory is expressed in
existence and uniqueness theorems for the EFEs,
provided the matter behavior is well defined through
prescription of suitable equations of state (Hawking
and Ellis 1973). However, in general the theory
breaks down in the large, and this feature is
specified by the HawkingPenrose singularity theorems, predicting the existence of a geodesic incompleteness of spacetime under conditions applicable
to realistic cosmological models satisfying the energy
conditions given by eqns [3] and [4] (Hawking and
Ellis 1973, Tipler et al. 1980). However, the
conclusion does not follow if the energy conditions
are not satisfied. Furthermore, the deduction follows
8
Approximate FL Solutions
The real universe is, of course, not exactly FL, and
studies of structure formation depend on studies of
solutions that are approximately FL models they
are realistic (lumpy) universe models. These
enable detailed studies of observable properties
such as CBR anisotropies and gravitational lensing
induced by matter inhomogeneities, and of the
development of those inhomogeneities from quantum fluctuations in the very early universe that then
get expanded to very large scales by inflation.
have locally rotationally symmetric (LRS) solutions, with precisely one preferred spacelike direction at a generic point (Ellis 1967). When q = 0, the
solutions are anisotropic in that there can be no
continuous group of rotations leaving the solution
invariant; however, there can be discrete isotropies
in some special cases.
When t = 4,we have spacetime homogeneous solutions, with all physical quantities constant; they cannot
expand (by [5] and [3]). Nevertheless, two cases are of
interest. For q = 1 (r = 5) we find the Godel universe,
rotating everywhere with constant vorticity, which
illustrates important causal anomalies (Godel 1949,
Hawking and Ellis 1973). For q = 3 (r = 6), we find
the Einstein static universe (Einstein 1917), the
unique nonexpanding FL model with k = 1 and > 0.
It is of interest because it could possibly represent the
asymptotic initial state of nonsingular inflationary
universe models (Ellis et al. 2003). The highersymmetry models (de Sitter and anti-de Sitter
universes with higher-dimensional isotropy groups)
are not included here because they do not obey the
energy condition [3] they are empty universes,
which can be interesting asymptotic states but are
not by themselves good cosmological models.
When t = 3, we have spatially homogeneous
evolving universe models. For q = 0 (r = 3), there
are a large family of Bianchi universes, spatially
homogeneous but anisotropic, characterized into
nine types according to the structure constants of
the Lie algebra of the three-dimensional symmetry
group G3 . These can be orthogonal: the fluid flow
is orthogonal to the surfaces of homogeneity, or
tilted; the latter case can have fluid rotation or
acceleration, but the former cannot. They exhibit a
large variety of behaviors, including power-law,
oscillatory, and nonscalar singularities (Tipler et al.
1980). A vexed question is whether truly chaotic
behavior occurs in Bianchi IX models. The behavior
of large families of these models has been characterized in dynamical systems terms (Wainwright and
Ellis 1996), showing the intriguing way that highersymmetry solutions provide a skeleton that guides
the behavior of lower-symmetry solutions in the
space of spacetimes. Many Bianchi models can be
shown to isotropize at late times, particularly if
viscosity is present; thus, they are asymptotic to the
FL universes in the far future. In some cases, Bianchi
models exhibit intermediate isotropization: they are
much like FL models for a large part of their life, but
are very different from it both at very early and very
late stages of their evolution. These could be good
models of the real universe. An important theorem
by Wald (1983) shows that a cosmological constant
will tend to isotropize Bianchi solutions at late
Further Reading
Bardeen JM (1980) Physical Review D 22: 1882.
Bondi H (1947) Monthly Notices of the Royal Astronomical
Society 107: 410.
Bonnor WB and Ellis GFR (1986) Monthly Notices of the Royal
Astronomical Society 218: 605.
Ehlers J (1961) Abh Mainz Akad Wiss u Lit (translated in Gen
Rel Grav 25: 1225, 1993).
Ehlers J, Geren P, and Sachs RK (1968) Journal of Mathematical
Physics 9: 1344.
Ehlers J and Rindler W (1989) Monthly Notices of the Royal
Astronomical Society 238: 503.
Einstein A (1917) Sitz Ber Preuss Akad Wiss (translated in The
Principle of Relativity, 1993). Dover.
Ellis GFR (1967) Journal of Mathematical Physics 8: 1171.
Introduction
The general symplectic reduction theory (see
Symmetry and Symplectic Reduction) becomes
much richer and has many applications if the
symplectic manifold is the cotangent bundle
(T Q, Q = dQ ) of a manifold Q. The canonical
1-form Q on T Q is given by Q (q )(Vq ) =
q (Tq Q (Vq )), for any q 2 Q, q 2 Tq Q, and
659
dv
ev B
dt
_ y,
_ z_ ) = q_ is the
where e is the charge and v = (x,
velocity of the particle. What is the Hamiltonian
description of these equations?
There are two possible answers to this question.
To formulate them, associate to the divergence free
vector field B the closed 2-form B = Bx dy ^ dz
By dx ^ dz Bz dx ^ dy. Also, write B = curl A for
some other vector field A = (Ax , Ay , Az ) on R 3 ,
called the magnetic potential.
Answer 1 Take on T R3 the symplectic form
B = dx ^ dpx dy ^ dpy dz ^ dpz eB,
where
(px , py , pz ) = p := mv is the momentum of the
particle, and h = mkvk2 =2 = m(x_ 2 y_ 2 z_ 2 )=2 is the
Hamiltonian, the kinetic energy of the particle. A
direct verification shows that dh = B (Xh , ), where
Xh x_
@
@
@
@
y_
z_ eBz y_ By z_
@x
@y
@z
@px
_
eBx z_ Bz x
@
@
_
eBy x_ Bz x
@py
@pz
661
1
1
kpk2 p2
2m
2
1
1
kpk2 2
2m
2
qh t
2
0 k
Q qh sk
and
_
qt
exp h;
i
Z
0
q_ h t
ds
k
Q qh sk2
h;
i
k
Q qh sk2
Q qh t
663
!
D
Z T
ds
3
2
0 k
Q qh sk
The first terms in both formulas are the so-called
geometric phases because they carry only geometric
information given by the connection, whereas the
second terms are called the dynamic phases since
they encapsulate information directly linked to the
Hamiltonian. The expression of the total phase as a
sum of a geometric and a dynamic phase is not
intrinsic and is connection dependent. It can even
happen that one of these summands vanishes. We
shall consider now two concrete examples: the free
rigid body and the heavy top.
2 I1
I2
I3
where I1 , I2 , I3 > 0 are the principal moments of
inertia of the body. Let I := diag(I1 , I2 , I3 ) denote the
moment of inertia tensor diagonalized in a principalaxis body frame. The LiePoisson bracket on R3 is
given by {f , g}() = (rf () rg()) and the
= , where 2 R3 is
equation of motions are
the body angular velocity given in terms of by
i := =Ii , for i = 1, 2, 3, that is, = I1 . The
trajectories of the these equations are found by
intersecting a family of homothetic energy ellipsoids
with the angular momentum concentric spheres. If
I1 > I2 > I3 , one immediately sees that all orbits are
periodic with the exception of four centers (the two
possible rotations about the long and the short
moment of inertia axis of the body), two saddles
(the two rotations about the middle moment of
inertia axis of the body), and four heteroclinic orbits
connecting the two saddles.
Suppose that (t) is a periodic orbit on the sphere
S2kk with period T. After time T, by how much has
the rigid body rotated in space? The answer to this
question follows directly from [3]. Taking
= =kk
and the potential v 0 we get
ZZ
2h T
kk
2kIsk2 s Istr I
kk3
Z
0
s Is2
T
ds
ds
s Is
2h T 2Mg
!
s ds
D
0
ZZ
2kIsk2 s Istr I
ds
s Is2
D
Z T
ds
0 s Is
where D is the spherical cap on the unit sphere
whose boundary is the closed curve (t) and D is a
two-dimensional submanifold of the orbit O
bounded by the closed integral curve ((t), (t)).
The first terms in each summand represent the
geometric phase and the second terms the dynamic
phase.
Gauged Poisson Structures
665
[q]
f
g
;
s;
s
s
D
E
v; Q=G Bq d SA~f s] ; d SA~gs]
where v = [q, ] 2 e
g . The Poisson bracket f , g 2
C1 (W) is given by
W
e f w] ; r
e W gw]
ff ; ggW w Q=G q r
A
A
f
g
;
w;
w
w
D
W
E
e f w] ; r
e W gw]
v; Q=G Bq r
A
A
667
Further Reading
Abraham R and Marsden JE (1978) Foundations of Mechanics,
2nd edn. Reading, MA: Addison-Wesley.
Guichardet A (1984) On rotation and vibration motions of
molecules. Annales de lInstitut Henri Poincare. Physique
Theorique 40: 329342.
Iwai T (1987) A geometric setting for classical molecular
dynamics. Annales de lInstitut Henri Poincare. Physique
Theorique. 47: 199219.
Kummer M (1981) On the construction of the reduced phase
space of a Hamiltonian system with symmetry. Indiana
University Mathematics Journal 30: 281291.
Lewis D, Marsden JE, Montgomery R, and Ratiu TS (1986) The
Hamiltonian structure for dynamic free boundary problems.
Physica D 18: 391404.
Marsden JE, Montgomery R, and Ratiu TS (1990) Reduction,
symmetry, and phases in mechanics. Memoirs of the American
Mathematical Society 88(436).
Marsden JE, Misioek G, Ortega J-P, Perlmutter M, and Ratiu TS
(2005) Hamiltonian Reduction by Stages, Lecture Notes in
Mathematics. Springer.
Marsden JE and Perlmutter M (2000) The orbit bundle picture of
cotangent bundle reduction. Comptes Rendus Mathematiques de
lAcademie des Sciences. La Societe Royale du Canada
22: 3354.
Marsden JE and Ratiu TS (2003) Introduction to Mechanics and
Symmetry, 2nd edn. second printing; 1st edn. (1994), Texts in
Applied Mathematics, vol. 17. New York: Springer.
Montgomery R (1984) Canonical formulations of a particle in a
YangMills field. Letters in Mathematical Physics 8: 5967.
Montgomery R (1991) How much does a rigid body rotate? A
Berrys phase from the eighteenth century. American Journal
of Physics 59: 394398.
Montgomery R, Marsden JE, and Ratiu TS (1984) Gauged Lie
Poisson structures. In: Marsden J (ed.) Fluids and Plasmas:
Geometry and Dynamics, Contemporary Mathematiques,
vol. 28, pp. 101114. Providence, RI: American Mathematical
Society.
Satzer WJ (1977) Canonical reduction of mechanical systems
invariant under abelian group actions with an application to
celestial mechanics. Indiana, University Mathematics Journal
26: 951976.
Simo JC, Lewis D, and Marsden JE (1991) The stability of relative
equilibria. Part I: The reduced energymomentum method.
Archive for Rational Mechanics and Analysis 115: 1559.
Smale S (1970) Topology and mechanics. Inventiones Mathematicae 10: 305331, 11: 4564.
Sternberg S (1977) Minimal coupling and the symplectic mechanics of
a classical particle in the presence of a YangMills field.
Proceedings of the National Academy of Sciences 74: 52535254.
Weinstein A (1978) A universal phase space for particles in Yang
Mills fields. Letters in Mathematical Physics 2: 417420.
Zaalani N (1999) Phase space reduction and Poisson structure.
Journal of Mathematical Physics 40(7): 34313438.
Introduction
Sufficiently dense concentrations of massenergy in
general relativity collapse irreversibly and form black
holes. More precisely, the singularity theorems state
that once a closed trapped surface has developed, some
world lines will only extend to a finite length in the
future they end in a spacetime singularity. Furthermore, the cosmic censorship hypothesis states that this
singularity is hidden away inside a black hole. One
can, therefore, classify initial data in general relativity
which describe an isolated system with no black hole
present into those which remain regular, and those
which form a black hole during their evolution.
Theorems on the stability of Minkowski spacetime,
and similar results for some types of matter coupled to
gravity, imply that sufficiently weak (in some technical
sense) initial data will remain regular. On the other
hand, no necessary or sufficient criterion for black hole
formation is known. For very strong data the existence
of a closed trapped surface implies black hole
formation, but although the data themselves may be
regular, the trapped surface must already be inside the
black hole. Between the very weak and very strong
regime, there is a middle regime of initial data for
which one cannot decide if they will or will not form a
black hole, other than evolving them in time.
The threshold between collapse and dispersion was
first explored systematically by Choptuik (1992). He
concentrated on the simple model of a spherically
symmetric massless scalar (matter) field (r, t). In this
model, the scalar-field matter must either form a black
hole, or disperse to infinity it cannot form stable
stars. Choptuik explored the space of initial data by
means of one-parameter families of initial data which
interpolate between strong data (say with large
parameter p) that form a black hole and weak data
(with small p) that disperse. The critical value p of the
parameter p can be found for each family by evolving
many data sets from that family. Near the black hole
threshold, Choptuik found the following phenomena:
1. Mass scaling. By fine-tuning the initial data to
the threshold along any one-parameter family,
one can make arbitrarily small black holes. Near
the threshold, the black hole mass scales as
M Cp p
for p p
1
4
669
6
threshold
Critical
point
One-parameter
family of
initial data
p<p
p=p *
p>p *
*
8
Z x; Z x;
9
with
10
11
1
X
671
18
i
Zi x
12
i0
15
r
r
Zr; tp Z
Z0
16
Lp
Lp
where
Lp Lep
17
20
eH
eH
21
microstates0
673
Outlook
Critical phenomena in gravitational collapse are now
well understood in spherical symmetry, both theoretically and in numerical simulations. In some matter
models, the phenomenology is quite complicated, but
it still fits into the basic picture outlined here.
The crucial question as to what happens beyond
spherical symmetry remains largely unanswered at
the time of writing. Perturbation theory around
spherical symmetry suggests that critical phenomena are not restricted to exactly spherical situations. This is also supported by simulations in
axisymmetric (highly nonspherical) vacuum gravity. Other simulations of nonspherical gravitational
collapse which cover the necessary range of spacetime scales required to see critical phenomena are
only just becoming available, and the results are
not yet clear-cut. For collapse with angular
momentum, no high-resolution calculations have
yet been carried out. As the necessary techniques
become available, one should be prepared for
numerical simulations to make dramatic extensions
or corrections to the picture of critical collapse
drawn up here.
See also: Computational Methods in General Relativity:
The Theory; Spacetime Topology, Causal Structure and
Singularities; Stability of Minkowski Space; Stationary
Black Holes.
Further Reading
Abrahams AM and Evans CR (1993) Critical behavior and
scaling in vacuum axisymmetric gravitational collapse. Physical Review Letters 70: 29802983.
Choptuik MW (1993) Universality and scaling in gravitational
collapse of a massless scalar field. Physical Review Letters
70: 912.
Choptuik MW (1999) Critical behavior in gravitational collapse.
Progress of Theoretical Physics 136 (suppl.): 353365.
Evans CR and Coleman JS (1994) Critical phenomena and selfsimilarity in the gravitational collapse of radiation fluid.
Physical Review Letters 72: 17821785.
Gundlach C (1999) Living Reviews in Relativity 2: 4 (published
electronically at http://www.livingreviews.org).
Gundlach C (2002) Critical gravitational collapse with angular
momentum: from critical exponents to universal scaling
functions. Physical Review D 65: 064019.
Current Algebra
G A Goldin, Rutgers University, Piscataway, NJ, USA
2006 Elsevier Ltd. All rights reserved.
Introduction
Certain commutation relations among the current
density operators in quantum field theories define
an infinite-dimensional Lie algebra. The original
current algebra of Gell-Mann described weak and
electromagnetic currents of the strongly interacting
particles (hadrons), leading to the AdlerWeisberger
formula and other important physical results. This
helped inspire mathematical and quantum-theoretic
developments such as the Sugawara model, light
cone currents, Virasoro algebra, the mathematical
theory of affine KacMoody algebras, and nonrelativistic current algebra in quantum and statistical physics. Lie algebras of local currents may be
the infinitesimal representations of loop groups,
local current groups or gauge groups, diffeomorphism groups, and their semidirect products or other
extensions. Broadly construed, current algebra thus
leads directly into the representation theory of
infinite-dimensional groups and algebras. Applications have ranged across conformally invariant
field theory, vertex operator algebras, exactly
solvable lattice and continuum models in statistical
physics, exotic particle statistics and q-commutation relations, hydrodynamics and quantized vortex
motion. This brief survey describes but a few
highlights.
0 0
0
F a x ; x; F 50
b y ; y x0 y0
X
0
i3 x y
cabd F 50
d x ; x
4
50 0
0
F a x ; x; F 50
b y ; y x0 y0
X
i3 x y
cabd F 0d x0 ; x
d
Current Algebra
X
0
F 50
F a fa ; F 50
b fb i
d cabd fa fb
7
6
675
Ja0 x; Jbk y
i
x y
cabd Jdk x
8
@
icab k 3 x yI
@x
Jak x; Jb y 0 k; 1; 2; 3
where Ja = (Ja0 , Jak ), k = 1, 2, 3, is again a 4-vector, c is a
finite constant, and I is the identity operator. The time
components in eqns [8] behave like the local currents in
eqns [4]. The Schwinger term is a c-number, while
setting the commutators of the space components to
zero is the simplest choice consistent with the Jacobi
identity. The Sugawara Hamiltonian is given in terms of
the local currents by the formal expression:
"
#
Z
3
X
1X
3
2
2
0
k
H
d x Ja x
Ja x
9
2c a R3
k1
where the pointwise products of the currents require
interpretation in the particular representation. This
Hamiltonian leads to current conservation equations
for the Ja .
11
m 1mm 1
m;n I 12
12
Current Algebra
13
Roughly speaking, the KacMoody generators correspond to Fourier transforms of charge densities on
S1 , whereas the Virasoro generators correspond to
Fourier transforms of infinitesimal motions in S1 .
The central extensions provide the finite, c-number
Schwinger terms. These structures have important
application to light cone current algebra, conformally invariant quantum field theories in (1 1)dimensional spacetime, the quantum theory of
strings, exactly solvable models in statistical
mechanics, and many other domains.
Of greatest physical importance, both in quantum
field theory and statistical mechanics, are those
irreducible, self-adjoint representations of the Virasoro
algebra known as highest weight representations,
where the spectrum of the operator L(m = 0) is bounded
below. In these applications, one represents a pair of
Virasoro algebras by mutually commuting sets of
(m) . In the quantum theory, for
operators L(m) and L
(0) L(0) ,
example, one takes the total energy H / L
(0)
(0)
and the total momentum P / L L . In a highest
weight representation, there is a unique eigenstate of
L(0) having the lowest eigenvalue h; for this vacuum
jhi, L(m) jhi = 0, m > 0.
Friedan, Qiu, and Shenker showed in 1984 that
highest weight representations are characterized by a
class of specific, non-negative values of the central
charge c and, correspondingly, of h: either c 1 (and
h 0) or c = 1 6( 2)1 ( 3)1 , = 1, 2, 3, . . .
(and h assumes a corresponding, specified set of values
for each value of ). In a beautiful application to the
study of the critical behavior of well-known statistical
systems, in which the generator of dilations is
(0) L(0) , they discovered a direct
proportional to L
correspondence with permitted values of the central
charge; thus, c = 1=2 for the Ising model, c = 7=10 for
the tricritical Ising model, c = 4=5 for the three-state
Potts model, and c = 6=7 for the tricritical three-state
Potts model.
677
@
3 x yJk x
@x
14
15
Jg1 ; Jg 2 i
hJg1 ; g2
Equations [15] are a representation by self-adjoint
operators of the semidirect sum of the abelian Lie
algebra D with vect0 (R3 ). The corresponding group
is the natural semidirect product of the space D
(regarded as an abelian topological group under
addition) with Diff 0 (R3 ).
The construction generalizes to a general manifold
M or manifold with boundary (in place of R3 ), and
to a general set of charge densities that generate the
local Lie algebra map0 (M, G). When M = S1 , we have
the KacMoody and Virasoro algebras with central
charge zero. However, L(0) in the nonrelativistic
(1 1)-dimensional quantum theories is proportional to the total momentum P, and thus is
unbounded above and below.
The continuous unitary representations of
Diff 0 (M), or its semidirect product with a local
current group at fixed time, thus describe nonrelativistic quantum systems (Albeverio et al. 1999,
Goldin 2004). The unitary representation V(), 2
Diff 0 (M), satisfies V(gr ) = exp [i(r=
h)J(g)], where
r 2 R and gr is the one-parameter flow in Diff 0 (M)
generated by the vector field g. Such a representation may be described very generally by means of a
measure on a configuration space , quasi-invariant
under a group action of Diff 0 (M) on , together
with a unitary 1-cocycle on Diff 0 (M) . The
Hilbert
space
for
the
representation
is
H = L2d (, W), which is the space of measurable
functions (
),
2 , taking values in an inner
product space W, and square integrable with respect
to . The unitary representation V is given by
s
d
V
16
d
where
denotes the group action Diff0 (M) !
; is the measure on transformed by (which,
by the quasi-invariance of , is absolutely continuous with respect to ); d =d is the Radon
Nikodym derivative; and (
) : W ! W is a system
of unitary operators in W obeying the cocycle
equation
1 2
1
2 1
17
and (
) 1 to obtain a unitary group representation on complex-valued wave functions; but inequivalent cocycles describe unitarily inequivalent
representations.
The configuration space (N) , N = 1, 2, 3, . . . ,
consists of N-point subsets of Rs , and (N) is the
(local) Lebesgue measure on (N) . The corresponding diffeomorphism group and local current algebra
representations describe N identical quantum particles in s-dimensional space. When 1, we have
bosonic exchange symmetry. Inequivalent cocycles
on (N) are obtained (for s 2) by inducing
(generalizing Mackeys method) from inequivalent
unitary representations of the fundamental group
1 [(N) ]. For s 3, this fundamental group is the
symmetric group SN of particle permutations; the
odd representation of SN , N 2, gives fermionic
exchange symmetry, while the higher-dimensional
representations are associated with particles satisfying the parastatistics of Greenberg and Messiah.
When s = 2, however, 1 [(N) ] is the braid group
BN . Goldin, Menikoff, and Sharp obtained induced
representations of the current algebra describing the
intermediate statistics proposed by Leinaas and
Myrheim for identical particles in 2-space. Such
excitations, subsequently termed anyons by Wilczek and characterized as charge-flux tube composites, are important constructs in the theory of
surface phenomena such as the quantum Hall effect,
and anyonic statistics has also been applied to the
study of high-Tc superconductivity. Current algebra
representations induced by higher-dimensional
representations of BN describe the statistics of
plektons. Similarly, current algebra in nonsimply
connected space describes the AharonovBohm
effect.
R
Let (h) = Rs ds x h(x) (x) denote the smeared
creation field. Let the indexed set of representations
N , JN , N = 0, 1, 2, . . . , satisfying the current algebra
[15], act in Hilbert spaces HN , where (h) : HN !
HN1 , (h) : HN1 ! HN , (h)jH0 = 0, so that
and
intertwine the N-particle diffeomorphism
group
representations.
Let (f ) and J(g) act on
L1
H
,
so
that
(f
)N = N (f )N , J(g)N =
N
N=0
JN (g)N . Then conditions for a Fock space hierarchy are specified by commutator brackets of the
fields with the currents:
f ;
h
N1 f h
Jg;
h
JN1 gh
18
Current Algebra
679
Further Reading
Adler SL and Dashen RF (1968) Current Algebras and Applications to Particle Physics. New York: Benjamin.
Albeverio S, Kondratiev YuG, and Rockner M (1999) Diffeomorphism groups and current algebras: configuration space
analysis in quantum theory. Reviews in Mathematical Physics
11: 123.
Arnold VI and Khesin BA (1998) Topological Methods in
Hydrodynamics. Applied Mathematical Sciences, vol. 125.
Berlin: Springer.
Gell-Mann M and Neeman Y (2000) The Eightfold Way (1964)
(reissued). Cambridge, MA: Perseus Publishing.
Goddard P and Olive D (1986) KacMoody and Virasoro
algebras in relation to quantum physics. International Journal
of Modern Physics A 1: 303414.
Goldin GA (1996) Quantum vortex configurations. Acta Physica
Polonica B 27: 23412355.
Goldin GA (2004) Lectures on diffeomorphism groups in
quantum physics. In: Govaerts J, Hounkonnou N, and
Msezane AZ (eds.) Contemporary Problems in Mathematical
Physics: Proceedings of the Third International Workshop,
pp. 393. Singapore: World Scientific.
Goldin GA and Sharp DH (1991) The diffeomorphism group
approach to anyons. International Journal of Modern Physics
B 5: 26252640.
Ismagilov RS (1996) Representations of Infinite-Dimensional
Groups. Translations of Mathematical Monographs, vol. 152.
Providence, RI: American Mathematical Society.
Kac V (1990) Infinite Dimensional Lie Algebras. Cambridge:
Cambridge University Press.
Marsden J and Weinstein A (1983) Coadjoint orbits, vortices, and
Clebsch variables for incompressible fluids. Physica D
7: 305323.
Mickelsson J (1989) Current Algebras and Groups. New York:
Plenum.
Ottesen JT (1995) Infinite Dimensional Groups and Algebras in
Quantum Physics. Berlin: Springer.
Pressley A and Segal G (1986) Loop Groups. Oxford: Oxford
University Press.
Renner B (1968) Current Algebras and Their Applications.
Oxford: Pergamon.
Sharp DH and Wightman AS (eds.) (1974) Local Currents and
Their Applications. New York: Elsevier.
Treiman SB, Jackiw R, Zumino B, and Witten E (1985) Current
Algebra and Anomalies. Singapore: World Scientific.