You are on page 1of 76

N E WS L E T T ER

OF THE EUROPEAN MATHEMATICAL SOCIETY

S E European
M M Mathematical
E S 25 years Society

December 2015
Issue 98
ISSN 1027-488X
Courtesy Sebsti Xamb-Descamps.
4 Presidents of the EMS.

Editorial Interviews
The EMS Jubilee Louis Nirenberg
25th EMS Manjul Bhargava
Anniversary History
5ECM in Amsterdam George Boole
6ECM in Krakw and Boolean Algebra
FREE
ACCESS
EXCEPTIONAL OFFER

Pure Mathematics
Research
COMPLIMENTARY ACCESS TO OVER 100
ARTICLES IN NOVEMBER AND DECEMBER
Authors in this collection include:
Toby Gee, Mark Kisin, Laura DeMarco, Terence Tao, Xander Faber, Gregg Musiker, Ralf Schiffler,
Lauren K. Williams, Peter Scholze, Jonathan Pila, Michael Atiyah, Claude LeBrun, Cristina Brndle,
Simon Donaldson, Evgeny Khukhro, Nicolas Bergeron, Akshay Venkatesh, David Conlon, Jacob Fox,
Benny Sudakov, Itay Neeman

Publishing in the highest quality journals:


Compositio Mathematica
u Journal
u of the Institute of Mathematics
Combinatorics, Probability
u and Computing of Jussieu
Bulletin of the Australian Mathematical
u Mathematical Proceedings of the Cambridge
u
Society Philosophical Society
Ergodic Theory and Dynamical Systems
u Mathematika
u
Forum of Mathematics, Pi
u Proceedings of the Edinburgh Mathematical Society
u
Forum of Mathematics, Sigma
u Proceedings of the Royal Society of Edinburgh,
u
Glasgow Mathematical Journal
u Section A: Mathematics
Journal of the Australian Mathematical
u The Bulletin of Symbolic Logic
u
Society The Journal of Symbolic Logic
u
The Review of Symbolic Logic
u

journals.cambridge.org/pm15
Contents

Editorial Team
Editor-in-Chief Vladimir R. Kostic
European
Mathematical
(Social Media)
Lucia Di Vizio Department of Mathematics
LMV, UVSQ and Informatics
45 avenue des tats-Unis University of Novi Sad

Society
78035 Versailles cedex, France 21000 Novi Sad, Serbia
e-mail: divizio@math.cnrs.fr e-mail: vladimir.slk@gmail.com

Copy Editor Eva Miranda


Departament de Matemtica
Chris Nunn Aplicada I, EPSEB, Edifici P
119 St Michaels Road, Universitat Politcnica
Aldershot, GU12 4JW, UK de Catalunya Newsletter No. 98, December 2015
e-mail: nunn2quick@gmail.com Av. del Dr Maranon 4450
08028 Barcelona, Spain
Editors e-mail: eva.miranda@upc.edu Editorial: The EMS Jubilee Challenges for the Next 25 Years
Ramla Abdellatif R. Elwes................................................................................. 3
Vladimir L. Popov
LAMFA UPJV Steklov Mathematical Institute 5ECM in Amsterdam A. Ran & H. te Riele................................. 9
80039 Amiens Cedex 1, France Russian Academy of Sciences
e-mail: Ramla.Abdellatif@u-picardie.fr 6ECM in Krakw. Organizers Reminiscences S. Jackowski....... 10
Gubkina 8
119991 Moscow, Russia Problems for Children 5 to 15 Years Old V. Arnold.................... 14
Jean-Paul Allouche e-mail: popovvl@mi.ras.ru
(Book Reviews) Additive Eigenvalue Problem S. Kumar..................................... 20
IMJ-PRG, UPMC Themistocles M. Rassias George Boole and Boolean Algebra S. Burris............................ 27
4, Place Jussieu, Case 247 (Problem Corner)
75252 Paris Cedex 05, France Interview with Abel Laureate Louis Nirenberg M. Raussen &
Department of Mathematics
e-mail: jean-paul.allouche@imj-prg.fr
National Technical University C. Skau.................................................................................. 33
of Athens, Zografou Campus Interview with Manjul Bhargava U. Persson............................... 39
Jorge Buescu GR-15780 Athens, Greece
(Societies) e-mail: trassias@math.ntua.gr Recollection of a Singular School S. Paycha.............................. 45
Dep. Matemtica, Faculdade
de Cincias, Edifcio C6, A Tour of the Exhibition MadeInItaly. Mathematicians in Search
Volker R. Remmert
Piso 2 Campo Grande (History of Mathematics) of the Future G. Bini........................................................... 48
1749-006 Lisboa, Portugal IZWT, Wuppertal University
e-mail: jbuescu@ptmat.fc.ul.pt Explain Your Thesis in Three Minutes M. Kreusch..................... 50
D-42119 Wuppertal, Germany
e-mail: remmert@uni-wuppertal.de Mathematical Sciences Research Institute H. Friedman............. 52
Jean-Luc Dorier
(Math. Education) Vladimir Salnikov
The Portuguese Mathematical Society (SPM) at 75
FPSE Universit de Genve University of Luxembourg F.P. da Costa......................................................................... 56
Bd du pont dArve, 40 Mathematics Research Unit
1211 Genve 4, Switzerland Campus Kirchberg
ICMI Column J.-L. Dorier......................................................... 58
Jean-Luc.Dorier@unige.ch
6, rue Richard Coudenhove- Professional Development Centres as Levers for Change in
Kalergi
Eva-Maria Feichtner L-1359 Luxembourg
Mathematics Education K. Maa et al...................................... 59
(Research Centres) vladimir.salnikov@uni.lu Connecting Old and New Information: zBMATH as a Hub
Department of Mathematics
University of Bremen Dierk Schleicher Connecting Digital Resources O. Teschke.............................. 61
28359 Bremen, Germany
e-mail: emf@math.uni-bremen.de
Research I Book Reviews............................................................................ 63
Jacobs University Bremen
Postfach 750 561 Letters to the Editor.................................................................... 71
Javier Fresn
(Young Mathematicians Column)
28725 Bremen, Germany Personal Column........................................................................ 72
dierk@jacobs-university.de
Departement Mathematik
ETH Zrich Olaf Teschke The views expressed in this Newsletter are those of the
8092 Zrich, Switzerland
e-mail: javier.fresan@math.ethz.ch
(Zentralblatt Column) authors and do not necessarily represent those of the
FIZ Karlsruhe EMS or the Editorial Team.
Franklinstrae 11
10587 Berlin, Germany ISSN 1027-488X
e-mail: teschke@zentralblatt-math.org
2015 European Mathematical Society
Jaap Top
Published by the
University of Groningen EMS Publishing House
Department of Mathematics ETH-Zentrum SEW A27
P.O. Box 407 CH-8092 Zrich, Switzerland.
9700 AK Groningen, homepage: www.ems-ph.org
Scan the QR code to go to the The Netherlands
Newsletter web page: e-mail: j.top@rug.nl For advertisements and reprint permission requests
http://euro-math-soc.eu/newsletter contact: newsletter@ems-ph.org

EMS Newsletter December 2015 1


EMS Agenda

EMS Executive Committee EMS Agenda


President Prof. Gert-Martin Greuel 2016
(20132016)
Prof. Pavel Exner Department of Mathematics 36 March
(20152018) University of Kaiserslautern Ethics Committee Meeting, Warsaw, Poland
Doppler Institute Erwin-Schroedinger Str. Contact: matarne@math.aau.dk
Czech Technical University D-67663 Kaiserslautern
Brehov 7 Germany 1820 March
CZ-11519 Prague 1 e-mail: greuel@mathematik.uni-kl.de Executive Committee Meeting, Institut Mittag-Leffler,
Czech Republic Prof. Laurence Halpern Djursholm, Sweden
e-mail: ems@ujf.cas.cz
(20132016)
Laboratoire Analyse, Gomtrie 23 April
& Applications Presidents Meeting, Budapest, Hungary
Vice-Presidents
UMR 7539 CNRS
Universit Paris 13 9 April
Prof. Franco Brezzi
F-93430 Villetaneuse Annual Meeting of the Committee for Developing Countries
(20132016)
France of the EMS, ICTP Trieste, Italy
Istituto di Matematica Applicata
e-mail: halpern@math.univ-paris13.fr http://euro-math-soc.eu/EMS-CDC/
e Tecnologie Informatiche del
C.N.R. Prof. Volker Mehrmann
via Ferrata 3 1516 April
(20112014)
I-27100 Pavia ERCOM meeting, St. Petersburg, Russia
Institut fr Mathematik
Italy TU Berlin MA 45
1617 July
e-mail: brezzi@imati.cnr.it Strasse des 17. Juni 136
EMS Council, Humboldt University, Berlin, Germany
Prof. Martin Raussen D-10623 Berlin
(20132016) Germany
e-mail: mehrmann@math.TU-Berlin.DE
Department of Mathematical
Sciences Prof. Armen Sergeev
Aalborg University
Fredrik Bajers Vej 7G
(20132016)
Steklov Mathematical Institute
EMS Scientific Events
DK-9220 Aalborg st Russian Academy of Sciences
Denmark Gubkina str. 8 2016
e-mail: raussen@math.aau.dk 119991 Moscow
Russia 15 March
Secretary e-mail: sergeev@mi.ras.ru Diderot Mathematical Forum 2016 Biomedical Applications
of Mathematics
Prof. Sjoerd Verduyn Lunel
EMS Secretariat Universidad Complutense de Madrid, Spain; Universit Paris
(20152018)
Descartes, France; Politecnico di Milano, Italy
Department of Mathematics
Utrecht University Ms Elvira Hyvnen and
Budapestlaan 6 Ms Erica Runolinna 1620 March
NL-3584 CD Utrecht Department of Mathematics 27th Nordic Congress of Mathematicians
The Netherlands and Statistics Stockholm, Sweden
e-mail: s.m.verduynlunel@uu.nl P.O. Box 68 Bernoulli Society-EMS Joint Lecture: Sara van de Geer (ETH
(Gustaf Hllstrmin katu 2b) Zurich)
Treasurer FIN-00014 University of Helsinki
Finland 30 March3 April
Tel: (+358)-9-191 51503 EUROMATH, Thessaloniki, Greece
Prof. Mats Gyllenberg Fax: (+358)-9-191 51400
(20152018) e-mail: ems-office@helsinki.fi 1115 July
Department of Mathematics Web site: http://www.euro-math-soc.eu EMS-IAMP Summer School in Mathematical Physics on
and Statistics Universality, Scaling Limits and Effective Theories
University of Helsinki EMS Publicity Officer Roma, Italy
P.O. Box 68 http://www.smp2016.cond-math.it/
FIN-00014 University of Helsinki Dr. Richard H. Elwes
Finland School of Mathematics 1822 July
e-mail: mats.gyllenberg@helsinki.fi University of Leeds 7th European Congress of Mathematics, Berlin, Germany
Leeds, LS2 9JT http://www.7ecm.de/
Ordinary Members UK
e-mail: R.H.Elwes@leeds.ac.uk 2526 August
Prof. Alice Fialowski Second Caucasian Mathematics Conference (CMC-II)
(20132016) Lake Van, Turkey
Institute of Mathematics Cover photograph:
Etvs Lornd University P. Exner, President, and 2018
Pzmny Pter stny 1/C J.-P. Bourguignon, M. Sanz-
H-1117 Budapest Sol, A. Laptev, former 19 August
Hungary presidents, at the celebration ICM 2018
e-mail: fialowsk@cs.elte.hu for the 25th anniversary of Rio Centro Convention Center, Rio de Janeiro, Brazil
the EMS.

2 EMS Newsletter December 2015


Editorial

The EMS Jubilee: Challenges for the


Next 25 Years
Richard Elwes (EMS Publicity Officer; University of Leeds, UK))

The roots of mathematics, like those of humanity it-


self, lie in Southern Africa. At the dawn of civilisation,
thinkers in Mesopotamia and Egypt made early break-
throughs in notation and technique. Indian and Chinese
mathematicians produced insights which remain with us
today, while the Persian and Arabic traditions developed
the subject over hundreds of years. Today, as in so many
areas of life, the USA is a modern powerhouse. But, in
our desire to give credit where it is due, and to honour
the contributions of cultures which are too often over-
looked, we should not get carried away. Our own conti-
nent of Europe has been home to a multitude of math-
ematical advancements since the time of Pythagoras. Just
occasionally, it is worth reflecting and celebrating this
glorious tradition.
Indeed, many European nations have their own illus-
trious mathematical histories. In the 19th and early 20th The guests for the EMS Jubilee at Institut Henri Poincar.
centuries, this led to the founding of plethora of national Photo courtesy of Elvira Hyvnen.
and regional mathematical societies, of which the oldest
surviving is the Dutch Koninklijk Wiskundig Genoot to both. He offered the opinion that mathematicians to-
schap, founded in 1778. The European Mathematical So- day stand in a similar position relative to Data, as they
ciety is thus a latecomer, not born until 1990 in the Polish did to Analysis in the time of Fourier: we already have
town of Mdralin. This year therefore, our society has the basic language and techniques, but a revolution is
reached 25 years of age, a youthful milestone which was surely imminent.
celebrated in magnificent style at the Institut Henri Poin- Why do mathematicians do mathematics? In truth,
car in Paris, on 22nd October. The day opened with an the answer is not usually because of its societal benefits
address from the Societys President, Pavel Exner, and or economic impact. At an individual level, we do it be-
comprised 4 plenary talks followed by a panel discussion cause we enjoy it. Depending on your perspective, we are
on the state of European mathematics, as we look to the either playful people who enjoy amusing ourselves with
challenges of the next 25 years and beyond. Several of puzzles, or deep-thinkers who provide answers to some
the themes from that conversation were well represented of the most profound questions our species can ask. (The
in the day as a whole, and perhaps it is worth drawing paradox of our subject is that this distinction is, in fact, no
them out. distinction at all.)
A major focus was the need for mathematics to be an The day saw two talks in this vein. The opening lec-
outward-facing discipline, in several senses. Every aspect ture by Hendrik Lenstra was an entertaining investiga-
of todays society is influenced by life-transforming tech- tion of profinite number theory, meaning the structure
nologies whose design and operation relies on sophisti- of Z, the profinite completion of the integers. One de-
cated mathematics. According to recent reports in UK, lightful discussion involved the profinite extension of
France, and Netherlands, the economic impact of Mathe- that staple of recreational mathematics, the Fibonacci
matics is enormous, to the tune of 9% of all jobs and 16% numbers. Generations of school-students and lay-people
of Gross National Product. The market thus presents an have been amused by this recursive sequence, and as
unprecedented and growing demand for mathematical Lenstra showed, there is plenty of enjoyment to be had
expertise, with data science in particular being an area of for professional mathematicians too. (See his essay in the
explosive growth. Newsletter of the European Mathematical Society61,
In his plenary talk, Andrew Stuart spoke eloquently September 2006, 1519.)
about one pressing challenge in this arena: the relation- In the afternoon session, we were treated to a talk from
ship between mathematical models and data. Taking the Lszl Lovsz on geometric representations of graphs.
example of numerical weather forecasting, a technology Here were beautiful problems and deep theorems, whose
as technically demanding as it is socially important, he origins lie in puzzles accessible to school-children. He be-
discussed the difficulty of incorporating observational gan with the theorem of Koebe that every planar graph
data into theoretical models, giving appropriate weight has a circle representation: a set of non-overlapping discs

EMS Newsletter December 2015 3


EMS News

in the plane, with two vertices joined by an edge if and


only if their corresponding discs touch. Extending this is
the Cage Theorem first proved by Andreev: that every
3-connected planar graph is isomorphic to the 1-skeleton
of a convex 3-polytope, where every edge of the polytope
touches a given sphere.
Another sense in which mathematics needs to out-
ward facing is in the need to engage with our colleagues
in other sciences. Mathematics has a long-standing rela-
tionship with physics, of course, but in the modern era bi-
ologists, chemists, medics, and others all have increasing
needs of our expertise.
Laure Saint-Raymond delivered her plenary lecture
on the subject of kinetic theory, a remarkable success
story which other subjects would surely love to emulate. Pavel Exner (EMS President), Maria Esteban (ICIAM President
Here, mathematics has wonderfully illuminated an ap- elect), Florence Berthou (Maire du 5me Arrondissement, Paris).
parently intractable question: how to predict or analyse
the (seemingly inherently unpredictable) behaviour of a and Ari Laptev, a fascinating and wide-ranging conversa-
gigantic system of interacting particles. Through the lan- tion touching on mathematical education, outreach, lob-
guage of entropy and chaos theory, these terms have been bying, publishing, ethics, and application. Afterwards, the
picked apart and robust results obtained. In particular, assembled company retired to the splendid surroundings
she spoke of recent breakthrough work of her own with of the Mairie du5meArrondissement for a welcome
Isabelle Gallagher and Benjamin Texier, wherein the from the Mayor, Mme Florence Berthou, and to drink
Boltzmann equation is rigorously derived as the limit of the health of the European Mathematical Society: Heres
a system of hard spheres. to another 25 years!
The formal part of the day was brought to an end with (Readers not able to attend the jubilee in Paris will
the panel discussion comprising Jean-Pierre Bourguig- find more coverage of the event in a special 100th edition
non, Maria Esteban, Roberto Natalini, Peter Bhlmann, of the Newsletter, next year.)

Announcement of the Next Meeting


of the EMS Council
Berlin, July 16 and 17, 2016
The EMS Council meets every second year. The next which was introduced by the 2008 Council. However, the
meeting will be held in Berlin, 16-17 July 2016, in the number of delegates for the 2016 Council is determined
Senate Meeting Room in the Main Building of the Hum- by the current membership class of the society.
boldt University at Unter den Linden 6. The council Each society is responsible for the election of its del-
meeting starts at 14:00 on 16 July and ends at lunchtime egates.
on 17 July. There is an online nomination form for delegates of
full members. The nomination deadline for delegates of
Delegates full members is 24 April 2016.
Delegates to the council are elected for a period of four
years. A delegate may be re-elected provided that con- (b) Associate Members
secutive service in the same capacity does not exceed Delegates representing associate members shall be elect-
eight years. Delegates will be elected by the following ed by a ballot organised by the Executive Committee
categories of members. from a list of candidates who have been nominated and
seconded by associate members, and who have agreed to
(a) Full Members serve. In October 2015, there were two associate mem-
Full members are national mathematical societies, which bers and, according to our statutes, these members may
elect 1, 2, 3 or 4 delegates according to their member- be represented by (up to) one delegate.
ship class. The membership class is decided by the coun- The delegate whose term includes 2016 is Mats Gyl-
cil, and societies are invited to apply for the new class 4, lenberg.

4 EMS Newsletter December 2015


EMS News

There is an online nomination form for delegates of as- There is an online nomination form for delegates of indi-
sociate members. The nomination deadline for delegates vidual members. The nomination deadline for delegates of
of associate members is 17 March 2016. individual members is 17 March 2016.

(c) Institutional Members Agenda


Delegates representing institutional members shall be The Executive Committee is responsible for preparing the
elected by a ballot organised by the Executive Commit- matters to be discussed at council meetings. Items for the
tee from a list of candidates who have been nominated agenda of this meeting of the council should be sent as
and seconded by institutional members, and who have soon as possible, and no later than 24 April 2016, to the
agreed to serve. In October 2015, there were 42 insti- EMS Secretariat in Helsinki.
tutional members and, according to our statutes, these
members may be represented by (up to) four delegates. Executive Committee
The delegates whose terms include 2016 are Joaquium The council is responsible for electing the President, Vice-
Bruna and Alberto Pinto. The delegate who can be re- Presidents, Secretary, Treasurer and other Members of the
elected is Sverre Olaf Smalo. Executive Committee. The present membership of the Ex-
There is an online nomination form for delegates of ecutive Committee, together with their individual terms of
institutional members. The nomination deadline for del- office, is as follows.
egates of institutional members is 17 March 2016.
President: Pavel Exner (20152018)
(d) Individual Members
Delegates representing individual members shall be Vice-Presidents: Franco Brezzi (20132016)
elected by a ballot organised by the Executive Commit- Martin Raussen (20132016)
tee from a list of candidates who have been nominated
and seconded, and who have agreed to serve. These del- Secretary: Sjoerd Verduyn Lunel (20152018)
egates must themselves be individual members of the
European Mathematical Society. Treasurer: Mats Gyllenberg (20152018)
In October 2015, there were 2650 individual members
and, according to our statutes, these members may be rep- Members: Alice Fialowski (20132016)
resented by (up to) 27 delegates. However, this number Gert-Martin Greuel (20132016)
may have increased by the time we call the election (if Laurence Halpern (20132016)
any) for individual members. Volker Mehrmann (20112018)
Here is a list of the current delegates of individual Armen Sergeev (20132016)
members whose terms include 2016:
Members of the Executive Committee are elected for a
Arne Ball period of four years. The president can only serve one
Vasile Berinde term. Committee members may be re-elected, provided
Maria Esteban that consecutive service does not exceed eight years.
Vincenzo Ferone The council may, at its meeting, add to the nominations
Christian Kassel received and set up a Nominations Committee, disjoint
Luis Narvaez Macarro from the Executive Committee, to consider all candidates.
Ji Rkosnk After hearing the report by the Chair of the Nominations
Oriol Serra Committee (if one has been set up), the council will pro-
ceed to the election of Executive Committee posts.
Here is a list of the delegates of individual members who All these arrangements are as required in the Statutes
could be re-elected to the 2016 Council: and By-Laws, which can be viewed on the webpage of the
council:
Peter Benner
Thierry Bouche http://www.euro-math-soc.eu/governance
Mireille Chaleyat-Maurel
Krzysztof Ciesielski The nomination forms for delegates can be found here
Mirna Damonja
Pavel Exner http://www.euro-math-soc.eu/nomination-forms-
Vincent Heuveline council-delegates
Arne Jensen
Paul C. Kettler
Ari Laptev Secretary: S  joerd Verduyn Lunel
Jos Francisco Rodrigues (s.m.verduynlunel@uu.nl)
Marie-Franoise Roy Secretariat: ems-office@helsinki.fi
Stepan Agop Tersian
Robin Wilson

EMS Newsletter December 2015 5


EMS News

LMS-EMS Joint Mathematical


Week-end in Birmingham, UK
Richard Elwes (EMS Publicity Officer; University of Leeds, UK)

The week-end 1820 September saw mathematicians ings at the Barber Institute of Fine Arts, next door to the
from around the world congregate at the University of School of Mathematics.
Birmingham, UK, for a conference in celebration of two Stefanie Petermichl (Toulouse) delivered the first
birthdays: the 150th of the venerable London Mathemat- plenary session of Saturday 19th, on Optimal control of
ical Society (LMS), and the 25th of the relatively youth- second order Riesz transforms on multiply-connected Lie
ful European Mathematical Society. groups, discussing progress on controlling the norms of
Under the watch of the Joseph Chamberlain Memo- certain classical operators on groups. She was followed
rial Clock-tower (or Old Joe, the worlds tallest free- by Bla Bollobs (Cambridge and Memphis) speaking
standing clock-tower), participants divided between par- on Percolation and random cellular automata. He paused
allel sessions on the themes of Algebra, Combinatorics, during his talk to pay tribute to two friends who had
and Analysis, and reunited for plenary talks from some recently passed away: IanCassels, Head of Mathemat-
of mathematics current leading lights. icsduringhisPhD at Cambridge, and Bollobss own
After warm greetings from Terry Lyons and Pavel graduate student Charles Read (Leeds). The days final
Exner, the two societies respective Presidents, and from plenary session was from Timothy Gowers on the subject
Andrew Schofield, Head of Birmingham Universitys of Interleaved products in highly non-Abelian groups, an
College of Engineering and Physical Sciences, the meet- algebraic problem motivated by a question in cryptog-
ing got underway with a plenary talk from Noga Alon raphy.
(Tel Aviv and Princeton), on the subject of Graphs, vec The conference dinner took place on Saturday even-
tors and integers. His focus was Cayley Sum Graphs of ing in Birmingham University Staff House, where deli-
finite Abelian groups, and the role they play in subjects cious food was consumed, and many glasses were raised
from Graph Theory to Information Theory. Aner Shalev in cheerful celebration of the two societies birthdays.
(Jerusalem) later delivered the days second plenary talk, Rounding off the meeting on Sunday 20th was Keith
on Groups in Interaction, discussing several instances of Ball (Warwick), with an entertaining plenary talk ex-
interplay between group theory and other subjects, in- ploring The probabilistic character of high-dimensional
cluding probability theory, algebraic geometry, and num- objects. Then with hearty thanks to the organisers, Chris
ber theory. Parker, Anton Evseev, Maria Carmen Reguera and An-
Away from the lecture theatres, mathematicians were drew Treglown, and with congratulations to EIisa Covato
spotted enjoying Balti curry (a famous Birmingham cre- (Bristol) and Robert Hancock (Birmingham) winners of
ation, along with the postage stamp and the pneumatic the graduate student poster competition, an excellent
tyre) and enjoying the outstanding collection of paint- celebratory weekend drew to a close.

European Mathematical Society Publishing House


Seminar for Applied Mathematics
ETH-Zentrum SEW A27, CH-8092 Zrich, Switzerland
orders@ems-ph.org / www.ems-ph.org

Jean-Pierre Bourguignon (IHS, Bures-sur-Yvette, France), Oussama Hijazi (Universit de Lorraine, Nancy, France), Jean-Louis
Milhorat (Universit de Nantes, France), Andrei Moroianu (Universit de Versailles-St Quentin, France), Sergiu Moroianu (Institu-
tul de Matematica al Academiei Romne, Bucuresti, Romania)
A Spinorial Approach to Riemannian and Conformal Geometry (EMS Monographs in Mathematics)
ISBN 978-3-03719-136-1. 2015. 462 pages. Hardcover. 16.5 x 23.5 cm. 78.00 Euro
The book gives an elementary and comprehensive introduction to Spin Geometry, with particular emphasis on the Dirac operator
which plays a fundamental role in differential geometry and mathematical physics.
After a self-contained presentation of the basic algebraic, geometrical, analytical and topological ingredients, a systematic study
of the spectral properties of the Dirac operator on compact spin manifolds is carried out. The classical estimates on eigenvalues
and their limiting cases are discussed next, and several applications of these ideas are presented. The special features of the
book include a unified treatment of Spinc and conformal spin geometry (with special emphasis on the conformal covariance of the
Dirac operator), an overview with proofs of the theory of elliptic differential operators on compact manifolds based on pseudodifferential calculus, a spinorial
characterization of special geometries, and a self-contained presentation of the representation-theoretical tools needed in order to apprehend spinors.
This book will help advanced graduate students and researchers to get more familiar with this beautiful domain of mathematics with great relevance to both
theoretical physics and geometry.

6 EMS Newsletter December 2015


EMS News

A Message from the Ethics Committee


Arne Jensen (Aalborg University, Denmark)

One of the tasks given to the Ethics Committee is to consider possible violations of the Code of Practice, in par-
ticular concerning plagiarism.
If you, as an author, believe that you are the victim of plagiarism and wish to bring a case to the Ethics Committee,
you should undertake the following steps:

1. Gather detailed information on the suspected plagiarism.


2. Give this material to a colleague whose opinions you respect and ask this person to evaluate your case.

Assuming that this colleague agrees with you:

3. Contact the author(s) and the editor(s) of the journal involved, presenting your evidence of plagiarism.

This step may resolve the case. If no replies or negative replies are received, you may consider bringing the case to
the Ethics Committee. You should start by contacting the Chair informally to ensure that your submission satisfies
the requirements of the Code of Practice. Then make a formal submission.

4. The Committee will acknowledge receipt and will decide whether there is a prima facie case. If so, the commit-
tee will consult experts as specified in the Code of Practice.
Arne Jensen
Chair (until end of 2015)
Ethics Committee
The Code of Practice can be found at
http://www.euro-math-soc.eu/committee/ethics

IMU Committee for Women in


Mathematics (CWM)
Funding Call for 2016
Marie-Franoise Roy (Universit de Rennes 1, France) and Caroline Series (University of Warwick, Coventry, UK)

The IMUs Committee for Women in Mathematics on how CWM money would be spent and other funding
(http://www.mathunion.org/cwm/) invites proposals for which may be available. There will be one call for appli-
funding of up to 3000 for activities or initiatives tak- cations regarding activities in 2016 with deadline of 15th
ing place in 2016, and aimed at establishing or support- January 2016. It is anticipated that further calls will be
ing networks for women in mathematics, preferably at made in subsequent years.
the continental or regional level, and with priority given Applications should be sent to info-for-cwm@math
to networks and individuals in developing or emerg- union.org
ing countries. CWMs help could include, for example, Successful applications will be informed no later than
funding meetings, travel for individuals for consultation February 29, 2016. Depending on demand, successful
purposes, or advice and support in creating websites. applications may not be funded in full. Successful appli-
Other ideas for researching and/or addressing problems cants will be asked to send before the end of 2016 a short
encountered by women in mathematics may also be con- report of the activity with details of how the budget was
sidered. spent.
Proposers should write a short account (no more
than two pages) explaining the nature of their activity IMU-CWM Committee
and how it fulfills the above aims, as well as indications October 2015

EMS Newsletter December 2015 7


A COMPREHENSIVE COURSE IN ANALYSIS
Barry Simon, California Institute of Technology
A Comprehensive Course in Analysis by Poincar Prize winner Barry Simon is a five-volume set that can serve as a graduate-level
analysis textbook with a lot of additional bonus information, including hundreds of problems and numerous notes that extend the
text and provide important historical background. Depth and breadth of exposition make this set a valuable reference source for
almost all areas of classical analysis.
Part 1: Real Analysis: 789pp 9781470410995 Hardback 105.00
Part 2A: Basic Complex Analysis: 641pp 9781470411008 Hardback 105.00
Part 2B: Advanced Complex Analysis: 321pp 9781470411015 Hardback 105.00
Part 3: Harmonic Analysis: 759pp 9781470411022 Hardback 105.00
Part 4: Operator Theory: 749pp 9781470411039 Hardback 105.00
5 Volume Set: 3259pp 9781470410988 Hardback 385.00

PARTIAL DIFFERENTIAL EQUATIONS


An Accessible Route through Theory and Applications
Andrs Vasy, Stanford University
Intended for readers who want to understand the theoretical underpinnings of modern PDEs in settings that are important
for the applications without using extensive analytic tools required by most advanced texts. The key goal of this book is to be
mathematically complete without overwhelming the reader, and to develop PDE theory in a manner that reflects how researchers
would think about the material.

Graduate Studies in Mathematics, Vol. 169


Jan 2016 280pp 9781470418816 Hardback 76.00

PERSISTENCE THEORY
From Quiver Representations to Data Analysis
Steve Y. Oudot, Inria Saclay
Persistence theory emerged in the early 2000s as a new theory in the area of applied and computational topology. This book
provides a broad and modern view of the subject, including its algebraic, topological, and algorithmic aspects. It also elaborates on
applications in data analysis. The level of detail of the exposition has been set so as to keep a survey style, while providing sufficient
insights into the proofs so the reader can understand the mechanisms at work.

Mathematical Surveys and Monographs, Vol. 209


Jan 2016 218pp 9781470425456 Hardback 125.00

RANDOM OPERATORS
Disorder Effects on Quantum Spectra and Dynamics
Michael Aizenman, Princeton University & Simone Warzel, Technische Universitt Mnchen
Provides an introduction to the mathematical theory of disorder effects on quantum spectra and dynamics. Topics covered range
from the basic theory of spectra and dynamics of self-adjoint operators through Anderson localization - presented here via the
fractional moment method, up to recent results on resonant delocalization.
The subjects multifaceted presentation is organized into seventeen chapters, each focused on either a specific mathematical topic
or on a demonstration of the theorys relevance to physics, e.g., its implications for the quantum Hall effect.

Graduate Studies in Mathematics, Vol. 168


Jan 2016 320pp 9781470419134 Hardback 87.00

Free delivery worldwide at www.eurospanbookstore.com/ams


AMS is distributed by Eurospan|group
CUSTOMER SERVICES: FURTHER INFORMATION:
Tel: +44 (0)1767 604972 Tel: +44 (0)20 7240 0856
Fax: +44 (0)1767 601640 Fax: +44 (0)20 7379 0609
Email: eurospan@turpin-distribution.com Email: info@eurospangroup.com
25th Anniversary of the EMS

5ECM in Amsterdam
Andr Ran (Vrije Universiteit Amsterdam, The Netherlands) and Herman te Riele (CWI, Amsterdam, The Netherlands)

In 2001, the two of us worked


together for the first time on
the organisation of the Dutch
Mathematical Conference,
which is an annual congress
of the Dutch Royal Math-
ematical Society (KWG).
Apparently, we did a nice job
because, later in the year, we
received phone calls from our
respective bosses telling us that we were the chosen vic-
tims to try to get the fifth European Congress of Mathe-
matics to Amsterdam. We said yes (reluctantly) and were
joined by Jan Wiegerinck from the Korteweg-de Vries
Institute of Mathematics of the University of Amster-
dam, making a triumvirate. Thus, it became a joint effort
of the three mathematics institutes in Amsterdam. By the time the congress was about to start, we still
In our innocence, we thought we would get it to Am- had some small items on our list to do but, on the week-
sterdam and then the big shots in the Dutch mathemati- end before the congress, these were finally put to rest. On
cal world would take over to actually organise it. So, we Monday 14 July 2008 at 8:45 in the morning, about 800
thought it would be a job spread over a few years, to write mathematicians from more than 60 European countries
a bid-book and negotiate with the selection committee of gathered at the RAI convention centre in Amsterdam
the EMS. No big deal. Boy, how wrong we were. for the opening ceremony.
Initially, everything seemed to go as planned. The bid- The opening ceremony was a spectacular affair. A
book was a piece of cake and we obtained promises of tableau vivante of Rembrandt van Rijns most famous
support from ministries, the municipality, scientific or- painting The Nightwatch was created on the stage by a
ganisations, etc. The site visit in 2003 was a very pleas- group of re-enactors. This was accompanied by drum rolls
ant experience and, in 2004, the decision was taken that from the drummers in the painting. In their midst was the
5ECM would be held in Amsterdam. To get an idea of painter himself, who was played by the first speaker of
what this would entail, the triumvirate participated in the opening session Robbert Dijkgraaf, currently Direc-
4ECM in Stockholm and we got a lot of good ideas. By tor of the Institute for Advanced Study in Princeton, at
then, it was already abundantly clear to us that now we the time a professor at the University of Amsterdam and
had secured 5ECM for the Netherlands, and for Amster- Chairman of the KNAW (Royal Dutch Academy of Sci-
dam in particular, it was up to us to actually organise it. ences). An excellent opening address by Dijkgraaf was
And then the hard work began. Obviously, many who followed by a warm welcome by Ari Laptev, then presi-
had pledged support notified us that that was meant as dent of the European Mathematical Society. After that
support in spirit and not in actual euros. Finding funding came the ceremony announcing the prize winners: 10
was a constant worry for us but Jan took it upon himself young mathematicians were on stage for the EMS prizes,
to organise this and managed to come up with interest- together with the winner of the Felix Klein prize. All in
ing sponsors. It became clear that support was a matter all, the conference was off to a good start.
of asking the right persons and we finally managed to get There were several highlights from the conference: of
a balanced budget, as well as a nice location. All kinds of course the 10 plenary lectures (including those of Richard
other organisational matters were solved for us by the Taylor and Lszl Lovsz) and 33 invited lectures, the lec-
people from the organising agency ICS International tures by the 10 EMS prize winners sponsored by the Dutch
Conference Services B.V. (now called MCI Amsterdam). Foundation Compositio Mathematica: Arthur Avila, Alexei
The matter of a suitable venue was actually solved Borodin, Ben Green, Olga Holtz, Boaz Klartag, Alexan-
during the site visit in 2003. We had several options but der Kuznetsov, Assaf Naor, Laure Saint-Raymond, Agata
the RAI convention centre in the southern part of town Smoktunowicz and Cdric Villani, and a lecture by the Fe-
turned out to be an excellent choice, with several hotels lix Klein prize winner Josselin Garnier. Also, there were
nearby and easy train connections to Schiphol Airport. three science lectures: one on quantum information theory
In the meantime, the scientific organising commit- by Ignacio Cirac, one on climate change by Tim Palmer
tee, headed by Lex Schrijver, and the prize committee, and one on mathematical biology by Jonathan Sherratt.
headed by Rob Tijdeman, had jointly put together a very The 5ECM conference also incorporated the an-
interesting programme. nual conference of the Royal Dutch Mathematical So-

EMS Newsletter December 2015 9


25th Anniversary of the EMS

ciety (Koninklijk Wiskundig Genootschap, KWG) and tion very much enjoyed the presidents dinner later that
two highlights were prizes awarded at those meetings. evening at the head office of ING, close to the conference
On Monday evening, there was the ceremony around venue.
the Brouwer prize. This is a prize that is awarded by On Wednesday, the congress dinner party was held in
the KWG every three years, and 2008 happened to be Hotel Arena. That turned out to be more than a bit mis-
one of those years. The recipient was Phillip Griffiths leading: the party was actually held in an old church and
(Princeton), who gave his lecture on Monday evening in was, like the rest of the congress, a memorable event.
the auditorium of the Vrije Universiteit. The ceremony For us as organisers, the week went by in a blur. There
was preceded by organ music, played by mathematician was always something to be done but our team got us
Jozef Steenbrink, and followed by the presentation of the through the week in one piece. Nevertheless, we were
Brouwer Medal and a welcome reception. In the after- quite relieved when the closing ceremony was over on
noon, Dirk van Dalen gave an historical lecture about Friday 18 July, around six in the afternoon. We survived
L.E.J. Brouwer. Also during 5ECM, the Beeger prize the experience and we are happy that we were able to
was awarded to Dan Bernstein. The Beeger prize is given give so many of our colleagues from around the world a
by CWI Amsterdam at every other meeting of the KWG nice conference in what we hope they found a very en-
and, again, 2008 happened to be a year for this. joyable and hospitable city.
During lunch breaks, movies were shown about Kurt
Gdel and Wolfgang Dblin. Two round table meetings
took place during the conference: one on mathemat- Herman te Riele joined CWI Amsterdam
ics and industry, and one on mathematics in develop- in 1970, until his retirement by the end of
ing countries. On Thursday afternoon, the Philips PhD 2011. He carried out research on numeri
prize lectures were presented and the Philips prize was cal mathematics and computational num
awarded to Erik Jan van Leeuwen of CWI Amsterdam. ber theory, often with help of large-scale
All in all, the conference programme had many prize lec- computers. Besides that he has been and
tures, special lectures and other features, like the selling still is active in organizational matters con
of Brouwer stamps and of Gaussian prime tablecloths by cerning mathematics (like member of the
KWG. A special issue of Nieuw Archief voor Wiskunde board of the Royal Dutch Mathematical Society (since
called Amsterdam Archive (http://www.nieuwarchief.nl/ 2003), the organization of the BeNeLux Mathematical
serie5/pdf/naw5-2008-09-2-091.pdf) was included in the Congress in Amsterdam (2223 March, 2016), the organi
conference bag. It presents a Dutch view of the world of zation of the fifth European Congress of Mathematics
mathematics. The proceedings of 5ECM were published (Amsterdam, 2008), secretary of ERCOM (European re
in 2010 by the EMS. search Centers on Mathematics (20062009) and secretary
No conference goes without its problems and up- of the Review Committee for mathematical research at six
heavals. For us, the most notable was the following. An Dutch Universities (20092010).
extremely well known mathematician was stopped when
entering the main lecture hall by a student: Sir where Andr Ran is the Desmond Tutu professor
is your badge? I dont have one, but my name is, in mathematics at Vrije Universiteit Am
to which the student answered: Well, sir, I do not care sterdam, where he has been since 1985.
who you are; we are under strict orders not to let anyone He also holds an extra-ordinary profes
in without a badge. We do apologise but we had to be sorship at North West University in South
strict; nobody got in without a badge, not even one of the Africa. His research interests are linear
most prominent mathematicians of the last half century. algebra, operator theory and systems and
Notably, the problem was solved and the person in ques- control theory.

6ECM in Krakw
Organizers Reminiscences
Stefan Jackowski (University of Warsaw, Poland)

6ECM, held in Krakw, 27 July 2012, was the most re- to my personal reminiscences related to the congress and
cent ECM, thus many readers may still remember it, if not to what happened behind the scenes.
as participants then as readers of the EMS Newsletter,
where a detailed report A Dozen Facts about the 6th Eu- A long time ago Mdralin 1990
ropean Congress of Mathematics was published in Sep- I think the first time I heard about the idea of the Euro-
tember 2012. This article will therefore be mainly devoted pean Congresses of Mathematics was in October 1990, at

10 EMS Newsletter December 2015


25th Anniversary of the EMS

a meeting in a residence of the Polish Academy of Sci- great local influence and the bid is strongly supported
ence in Mdralin near Warsaw, when the European Math- by other Universities. The team behind him was not so
ematical Society was founded. I must confess that, at the obvious to us, though. Another possible factor in Kra
time, the idea of having another big congress, similar to kow is the football competition [Euro 2012, organised
the ICM, in which prizes similar to Fields Medals were jointly by Poland and Ukraine SJ], which will almost
awarded, did not appeal to me. I was hoping to hear some certainly inhibit outreach activity.
new ideas of European collaboration.
I cant remember why I attended the meeting of The EMS Executive Council met in Utrecht just before
prominent representatives of the European mathematical the 5ECM in Amsterdam and chose Krakw for 6ECM.
community. At that time, I was less than 40 and I did not I think the decisive role was played by the strong deter-
have any formal position. I think I was invited because mination of Andrzej Pelczar (a figure well known to the
my wife and I had had the honour of receiving Michael EMS) to organise the congress, as well as the visiting
Atiyah at home when he came to Warsaw to discuss the teams very positive impression of the congress venue:
organisation of the ICM 1983. Sir Michael was also the the Auditorium Maximum of Jagiellonian University. The
key person in the process of founding the EMS and the other two cities presented very attractive outreach pro-
most distinguished participant of the Mdralin meeting. grammes but couldnt offer as convenient facilities.
His negotiating partner back in the 1980s and 1990
was Professor Bogdan Bojarski from the Polish Academy Back home
of Sciences. In some Soviet bloc countries, academies had When Polish representatives returned home from Utre-
a monopoly on representing the scientific community in- cht, difficult negotiations between the PTM and Jagiellon-
ternationally, thus the Polish mathematical community ian University began. Since both institutions signed the
couldnt be formally represented by the Polish Math- bid and accepted responsibility for the congress, it seemed
ematical Society a limited monopoly lasted long after natural to specify the contributions expected of each of
the fall of the system. The second host in Mdralin was them. Public universities in Poland are subject to detailed
Professor Andrzej Pelczar, then President of the Polish and restrictive regulations concerning spending money,
Mathematical Society (PTM), who later became Vice- whereas a society, as a private entity, can be much more
President of the EMS (19972000). flexible. At the time, Andrzej Pelczar was very active lo-
cally and at the national and international level trying to
Utrecht 2008 get support and attract interest in the congress. It was all
It was Andrzej Pelczars idea to invite the European Con- stopped by Andrzejs sudden death on 18 May 2010.
gress of Mathematics to Krakw, his beloved hometown, Some people couldnt imagine ECM without Andrzej.
and to Jagiellonian University, his alma mater. Having I even heard comments that it would be better to give up
this in mind, he promoted, in 2006, the upgrade of the Pol- organisation of the congress. For my PTM collaborators
ish Mathematical Society membership of the EMS and and me, this was unthinkable. My second term as PTM
then prepared a bid which was submitted to the Execu- president was ending on 31 Dec 2010 but, under these new
tive Committee of the EMS in 2007. I signed the bid as circumstances, I decided to stay for a third term, 201113.
President of the PTM, assuming that all the work would The unexpected absence of Professor Pelczar created
be done in Krakw under Andrzejs supervision and my many misunderstandings. We were not aware of various
role would just be formal representation of the society. traditional and oral commitments that had been made,
Organisation of the ECM was not on my list of priorities e.g. concerning funding. The misunderstandings became
when I became President of the PTM in 2005. Then I got clear when I was invited to report on preparations of the
involved in the organisation of a joint meeting with the congress at the Executive Committee meetings in Laus-
AMS in Warsaw in 2007 and, after its success, I planned anne, Firenze and Ljubljana. In Ljubljana, the Executive
similar meetings with Germany and Israel. I believed Committee decided that for the next congress (2016), a
that bilateral meetings, where both partners collaborate document would be signed by the EMS President and the
on the programme, served the Polish community better Chair of the Organising Committee, listing the commit-
then a big international meeting, where the local organ- ments agreed to by the Organising Committee.
isers are primarily responsible for financial support and After Andrzejs death, the Dean of Mathematics at
logistics, with little influence on the scientific programme. Jagiellonian University Roman Srzednicki (a former
Three cities wanted to organise 6ECM and submitted student of Andrzej) assumed his duties. In the Summer
their bids: Krakw, Prague and Vienna. Representatives of 2010, we both gave presentations on the preparations
of the EMS Executive Committee visited all three cities of the 6ECM at the Executive Council Meeting in So-
in late October 2007. After the visit they presented their fia. Later, President of the EMS Marta Sanz-Sol and I
evaluation, writing about Krakw as follows: made a presentation to the IMU General Assembly in
Hyderabad. In January 2011, Marta visited Krakw to
The conference venue in Krakow was extremely beau discuss progress in the organisation of the congress. Al-
tiful and practical. Everything needed for the confer though several representatives of the university and city
ence is available in one lovely building, close to the city authorities received her and assured her of their support,
centre. Also, this is owned by the University, so the price she probably realised that organisation of the congress
will be low. Andrzej Pelczar (the former Rector) has was at a crossroads.

EMS Newsletter December 2015 11


25th Anniversary of the EMS

ending in settling bills. She was involved in almost every


aspect of preparation of the 6ECM. The EOC was also
greatly helped by a large group of young Krakw math-
ematicians and students.
Registration of participants opened on the 6ECM
website in early November 2011. Based on the experience
of past ECMs, we were concerned that participation may
turn out to be low [see diagram below]. At the conference
on mathematical biology mentioned above, there were
950 participants, out of which over 600 presented their re-
sults. We heard comments that it was difficult to get sup-
port for coming to the 6ECM without the opportunity to
The lobby of the 6ECM. Photo by Ada Paka. speak, and complaints about an unbalanced scientific pro-
gramme. Some mini-symposia organisers needed more
In the Spring of 2011, a new Executive Organising time. As a response to these signals, the EOC proposed
Committee (EOC) was founded, consisting of three rep- holding Satellite Thematic Sessions (STSs) during the
resentatives of the PTM: Krystyna Jaworska (Secretary congress. After announcing it in the Second Announce-
and Treasurer of the PTM, Military University of Tech- ment, we received many applications from participants
nology, Warszawa), Wacaw Marzantowicz (Vice-Presi- who wanted to organise STSs. They submitted proposals
dent of the PTM, Adam Mickiewicz University, Pozna) of subjects and speakers. Making a complete list of speak-
as well as the author (Chair of the EOC) and three rep- ers, etc., was left to the STS organisers. Proposals were ac-
resentatives of Jagiellonian University (UJ). Shortly after, cepted by the EOC; in some cases, we suggested merging
Roman got sick and had to withdraw from organisational proposals into one session. Only registered 6ECM partic-
work. Representatives of UJ were changing; finally, from ipants were eligible to speak at an STS. I believe it helped
January 2012, they were Zbigniew Bocki (Director of the to increase participation in the congress. Some people
Mathematics Institute UJ, Vice-Chair of the EOC), Piotr told me that they came to Krakw mainly for STSs. There
Tworzewski (Vice-Rector of the UJ) and Robert Wolak were 15 STSs held during the 6ECM with over 150 talks.
(UJ). The EMS Executive Committee was seriously wor-
ried but, to support the smooth and more efficient organi-
sation of the congress, the change was accepted.

Last lap
Invitations to plenary and invited speakers were sent out
in May 2011. The President of the EMS, and representa-
tives of the PTM and UJ as co-organisers of the congress,
signed the invitations. We offered to cover all the usual
expenses. However, at that time, applications for support
were still pending. The First Announcement was distrib-
uted in July 2011, a year before the congress. Pre-registra-
tion was opened on the 6ECM website.
At the end of June 2011, the 8th European Confer-
Participants of the past ECMs.
ence on Mathematical and Theoretical Biology was held
at Auditorium Maximum, the future 6ECM venue. Mem-
bers of the EOC went there to monitor how a conference The congress
of a size comparable to the ECM would fit in the building. There are two gratifying moments in a conference organ-
Observations were carefully analysed and this helped us isers life: a smooth opening and a successful closing.
to arrange the space more comfortably. In particular, we The opening of the 6ECM was well-attended by state
noticed that there were not enough seats in the lobbies officials. The Minister of Science and Higher Education
for informal discussions and registration booths took up (Professor of Law Barbara Kudrycka) attended and gave
too much space. Thus, just before the 6ECM, we rented a a well-received speech, assuring the audience of the Gov-
number of sofas and comfortable chairs; instead of regis- ernments high regard for mathematics: Poland is very
tration booths, we just put tables at the side. aware of the importance of science education, which of
I distinctly remember the day at the beginning of July course includes mathematics. She went on to say: The
when, while on vacation, I worked on some 6ECM grant Government is aware of the special role of mathemati-
application from that day, preparations of the 6ECM cal abilities in the labour market. She mentioned an in-
became my full-time job, which continued until the last crease of funds for fundamental research, in particular
day of the congress a year later. My closest collabora- for mathematics. After the speeches, the President of the
tor was Krystyna Jaworska, who travelled with me many EMS, together with the chairs of the prize committees,
times from Warsaw to Krakw. She was in charge of fi- presented the EMS prizes to the recipients. They gave lec-
nances of the 6ECM, starting from grant applications and tures during the congress. In the booklet devoted to 2012

12 EMS Newsletter December 2015


25th Anniversary of the EMS

Finally, I invited all 6ECM organisers that were pre-


sent up to the podium. The President of the EMS ex-
pressed thanks to the speakers and the participants, as
well as the organisers, and declared 6ECM closed. [photo
in Newsletter Sept 2012]
Christian Baer, the President of Deutsche Mathema-
tiker-Vereinigung, invited 6ECM participants to the next
7ECM in Berlin.

After the congress was it worth it?


After the 6ECM experience, I like congresses more than
I did before and Im more convinced that the effort to
Students wearing 6ECM T-shirts at the receptionist desk. Photo by organise them is sensible. As Marta said at the opening
Ada Paka. speech, the congress is a feast of mathematics and it
helps to maintain unity of mathematics which is excep-
Prize Winners, a list of the past-winners of the EMS prize tional in contemporary science. It did serve this purpose!
for young mathematicians was included, with information While the function of congresses and conferences as a
on which of them had later received the Fields Medal (9 means of transmitting scientific information is diminish-
out of 50). ing, their role in maintaining personal relations between
Adrian Constantin gave the first plenary lecture with mathematicians may be growing. You could see it in the
the appealing title Some mathematical aspects of wa- lobby of the 6ECM. It is a great opportunity to meet
ter waves. The plenary speakers not only gave lectures people from different fields, whom you dont see at the
but also answered two questions asked by the editors specialised meetings. In todays world, it is particularly
of the special volume of the PTM journal Wiadomoci important to promote close contact between European
matematyczne. The first concerned the motivation of mathematicians.
interest in the subject of the lecture; the second con- During preparations for the congress, I profited a lot
cerned challenges in the field of the research. The an- from the experience of the organisers of the past ECMs
swers took up 30 pages of the volume distributed to all and ICMs. But these were informal conversations and
participants. The volume also contained over 20 survey passing on of various materials, etc. I think it would be
articles by foreign mathematicians on ideas and re- useful to create an ECM advisory committee which
sults of selected Polish mathematicians and their influ- would preserve know-how, experience and knowledge
ence on mathematics. Articles were devoted to Stefan about good practices. There is no reason to re-invent the
Banach, Karol Borsuk, Samuel Eilenberg, Jzef Maria wheel every time; it is enough to improve it.
Hoene-Wroski, Andrzej Lasota, Stanisaw ojasiewicz, A great reward for the organisers was receiving many
Wacaw Sierpiski, Hugo Steinhaus, Witold Wolibner, kind letters after the congress. Let me quote one from
Tadeusz Waewski and Antoni Zygmund, and many the founder of the first ECM in Paris in 1992:
other names were mentioned.
The next best attended event after the opening was cer- I take this opportunity to congratulate you again for
tainly the congress banquet, which was held at the cloister the organization of the 6th ecm. It was perfect. (). On
of the functioning Franciscan Monastery in Krakw Old the top, the statistics you wrote at the closing ceremony
Town. Before the banquet, there was a guided tour to a were quite interesting. I am happy that my idea called
nearby church of great historical and artistic value. foolish by influential French mathematicians in the
At the closing ceremony, prizes for the 10 best research 90s has finally taken a good shape thanks to Euro
posters were presented. The poster session turned out to pean mathematicians like you. (). Max Karoubi (or
be a great success. From over 300 submissions, the poster ganiser of the 1ECM in Paris).
committee selected 186 posters, which were displayed in
a suitably arranged basement (garage) in the auditorium.
A jury, consisting of ad hoc invited congress participants, Stefan Jackowski lectures at the University
awarded the prizes. Publishers who exhibited their publi- of Warsaw (UW). His research interests
cations during the congress funded prizes mostly books are in algebraic topology. He has been:
or electronic subscriptions. Dean of the Faculty of Mathematics, In
Marta told me that there is a tradition of present- formatics and Mechanics UW (199096,
ing a report on the congress at the closing ceremony. I 19992005); President of the Polish Math
wondered what kind of report might be interesting to the ematical Society (20062013); Member of
most faithful participants who had stayed to its very end. the EMS Executive Council (200814);
I asked my collaborators to prepare statistics concerning EMS lay auditor for 201516; and Chair of the Executive
the structure of participation of the congress and present- Organising Committee of 6ECM (201112). Since 2012,
ed them to the audience most are included in my article he has been a member of the Committee for Evaluation
report A Dozen Facts about the 6th European Congress of Research Units at the Polish Ministry of Science and
of Mathematics. Higher Education.

EMS Newsletter December 2015 13


Feature

Problems for
Problems for Children
Children
5 to 15 Years Old
Old
Vladimir Arnold
Vladimir Arnold

I wrote these problems in Paris in the spring of 2004. Some American high school students had been successfully
Russian residents of Paris had asked me to help cultivate a solving this problem for over a decade. But then some Rus-
culture of thought in their young children. This tradition in sian students arrived from Moscow, and none of them was
Russia far surpasses similar traditions in the West. able to solve it as their American peers had (by giving 30
I am deeply convinced that this culture is developed best square inches as the answer). Why not?
through early and independent reflection on simple, but not 7. Victor has 2 more sisters than he has brothers. How many
easy, questions, such as are given below. (I particularly rec- more daughters than sons do Victors parents have?
ommend Problems 1, 3, and 13.)
My long experience has shown that C-level students, lag- 8. There is a round lake in South America. Every year, on
ging in school, can solve these problems better than outstand- June 1, a Victoria Regia flower appears at its center. (Its stem
ing students, because the survival in their intellectual Kam- rises from the bottom, and its petals lie on the water like those
chatka at the back of the classroom demanded more abil- of a water lily). Every day the area of the flower doubles, and
ities than are requisite to govern Empires, as Figaro said on July 1, it finally covers the entire lake, drops its petals, and
of himself in the Beaumarchais play. A-level students, on the its seeds sink to the bottom. On what date is the area of the
other hand, cannot figure out what to multiply by what in flower half that of the lake?
these problems. I have even noticed that five year olds can 9. A peasant must take a wolf, a goat and a cabbage across a
solve problems like this better than can school-age children, river in his boat. However the boat is so small that he is able
who have been ruined by coaching, but who, in turn, find them to take only one of the three on board with him. How can he
easier than college students who are busy cramming at their transport all three across the river? (The wolf cannot be left
universities. (And Nobel prize or Fields Medal winners are alone with the goat, and the goat cannot be left alone with the
the worst at all in solving such problems.) cabbage.)
10. During the daytime a snail climbs 3 cm up a post. During
1. Masha was seven kopecks short of the price of an alphabet the night it falls asleep and slips down 2 cm. The post is 10 m
book, and Misha was one kopeck short. They combined their high, and a delicious sweet is waiting for the snail on its top.
money to buy one book to share, but even then they did not In how many days will the snail get the sweet?
have enough. How much did the book cost? 11. A hunter walked from his tent 10 km. south, then turned
east, walked straight eastward 10 more km, shot a bear, turned
2. A bottle with a cork costs $1.10, while the bottle alone
north and after another 10 km found himself by his tent. What
costs 10 cents more than the cork. How much does the cork
color was the bear and where did all this happen?
cost?
12. High tide occurred today at 12 noon. What time will it
3. A brick weighs one pound plus half a brick. How many occur (at the same place) tomorrow?
pounds does the brick weigh? 13. Two volumes of Pushkin, the first and the second, are
side-by-side on a bookshelf. The pages of each volume are
4. A spoonful of wine from a barrel of wine is put into a glass
2 cm thick, and the front and back covers are each 2 mm
of tea (which is not full). After that, an equal spoonful of the
thick. A bookworm has gnawed through (perpendicular to the
(non-homogeneous) mixture from the glass is put back into
pages) from the first page of volume 1 to the last page of vol-
the barrel. Now there is a certain volume of foreign liquid
ume 2. How long is the bookworms track? [This topological
in each vessel (wine in the glass and tea in the barrel). Is the
problem with an incredible answer 4 mm is totally impos-
volume of foreign liquid greater in the glass or in the barrel?
sible for academicians, but some preschoolers handle it with
5. Two elderly women left at dawn, one traveling from A ease.]
to B and the other from B to A. They were heading towards 14. Viewed from above and from the front, a certain ob-
one another (along the same road). They met at noon, but did ject (a polyhedron) gives the shapes shown. Draw its shape
not stop, and each of them kept walking at the same speed as as viewed from the side. (Hidden edges of the polyhedron are
before. The first woman arrived at B at 4 PM, and the second to be shown as dotted lines.)
arrived at A at 9 PM. At what time was dawn on that day?
...........................
6. The hypotenuse of a right-angled triangle (on an Ameri-
..........
.......

can standardized test) is 10 inches, and the altitude dropped ..........

Top view Front view


....................................................
1 .................................................................................... x ................................. A
to it is 6 inches. Find the area of the triangle.
To Problem 14 To Problem 16 To

14 EMS Newsletter December 2015


Feature
.................................
....................... ....
15. How many ways are there to break the number 64 up into ....... ...............
..................................................................................................
. ...
..
...... ... ....... .... ............................................
....... ....... ... ..............
the sum of ten natural numbers, none of which is greater than
. ...... ....
........................................................................................... ... ......... ..
...
... .. ...
. ..
. ...............................................
.
12? Sums which differ only in the order of the addends are
.. ... .
. ........................................
... .. ... .
.
..... .........
...
.. ... ...................................
... ...
not counted as different.
...
... .. ...
5
. .
.
.
.
... .. .
... .
....
16. We have a number of identical bars (say, dominoes). We
...
...
... .
..
.
...... ... ... ... ... ... ... ....... ... ... .............
...
.. ... 3
..
want to stack them so that the highest hangs out over the low-
... ... .......
... . ... ... ... .... . ..... .
.. .
. .
..
.... ... ..
.... ........ ...........................
................................................................................. ............................................... .

est by a length equal to x bar-lengths. What is the largest pos- To Problem 20 To Problem 21 To Problem 22
sible value of x? 22. You have two vessels of volumes 5 liters and 3 liters.
...................................
................................................................................................... Measure
...........................................
out one liter, leaving the liquid in one of the vessels.
............................................. ................................................................................................. ....................................................................................................................................
..................................................
............................................................ . . ........................................................
...................................... ...... .. ....... ......................
... ........
.......................................................................................... ... .... ..
.......... ... .......... ...
... ..... .
. ... .. ..............................................
x
.
1 A B
... .
. .... .......................................
..................................................... .................................................................................... ................................. .
Top view Front view ..........
... . ...
... ..
..
..
.
.
...
...
... ...................................
... ...
.....
To Problem 14 To Problem 16 To Problem 17
...
...
..
.. ...
...
.. 5
..
... .
.
..... ... ... ... ... ... ... ..... ... ... ... ...
.
..
.
.
.
.
.
. 3
17. The distance between towns A and B is 40km. Two cy-
... .
. ..
... . ... ..... ............
....
............ .... .....
..................................... .. .. ... .. .. ........
..........................................
clists leave from A and B simultaneously traveling towards
..................................................................................

one another, one at a speed of 10km/h To Problem


and the20other at a To23.
Problem 21 To Problem 22
There are five heads and fourteen legs in a family. How
speed of 15 km/h. A fly leaves A together with the first cy- many people and how many dogs are in the family?
clist, and flies towards the second at a speed of 100 km/h. The
24. Equilateral triangles are constructed externally on sides
fly reaches the second cyclist, touches his forehead, then flies
AB, BC, and CA of a triangle ABC. Prove that their centers
back to the first, touches his forehead, returns to the second,
(marked by asterisks on the diagram) form an equilateral tri-
and so on until the cyclists collide with their foreheads and
angle.
squash the fly. How many kilometers has the fly flown alto-
gether? ......
.. .....
.. .... . .....
... .....
... .....
.....
..................................................................................................................... ..... ..... ....................................................................................................
....... . .
................................................................................................... ... ..... ....... .......... ....
.............................................................................. . ..... ....... . .... ... ....... .
............................................................ ... ..... ................................................................................................... .. .....
C
. .....
........................................
................... .... .. ... ... ... .. ...
........... .......... ... ....................... ... ... .. .. ...
... .............. ... ... . ...
....... .. .... ... ..
... ....................................... ...
x A B
..
1
..
... ...... ... .. ..
..................................................................................... ..................................
A
..
....................................... .....................
... ................ ... .... ...
.. ... .... ....
... .............
............. .....
.. ....
. ...
... ...
...
...
...
...
... .................................................................. ..
...
...
...
...
.
. ...
To Problem 16 To Problem 17
...
B
... .
... .. .. .
... ...... ... ... .
.. ... ..
... ...... .. .. ........... ....
..... ....
18. Vanya solved a problem about two pre-school age chil-
... .............
... .......... ...
............................................................................
... ......
........
dren. He had to find their ages (which are integers), given the To Problem 24 To Problem 25
product of their ages.
25. What polygons may be obtained as sections of a cube
Vanya said that this problem could not be solved. The
cut off by a plane? Can we get a pentagon? A heptagon? A
teacher praised him for a correct answer, but added to the
regular
......
... ...... hexagon?
problem the condition that the name of the older child was .... .......
.. .....
.....
Petya. Then Vanya could solve the problem right away. Now
...
...
.....
.....
..... ..................................................................................................
....... . . .......
......................

..... ....... ........ .... ......................................
you solve it.
... ..... ....... .... ... ..........
..... ................................................................................................. .. ..... .
...
.
A
C
. . .
. ..
... ..... ... ... ... .. ......... .. ......
....
. ... ... .... ... ... ........
... ...
.......
. ........................... ...
...
...
... ... .
. . ...
... . ..
..................................................................................
.
..... .. .... ... .. .. ... ...............
... ........................................ ... ...........
19. Is the number 140 359 156 002 848 divisible by
..
... ...... ...
..... ..
B
... .. ..
A ...
..................... . ... .. ....
... ................ . ... ... ....
.... .... ... ... ........
... .... .. ... ... ....... .... ... ........
.............
............. .....
.
.
4 206 377 084?
... .. ... .... ... .
...... .... .... .......
... ................................................................ ..
...
..
...
... .
.
..
.
.. .. .......................... .. . . . ... .....
... .. ... . ... .....
B
.. .. ..
...
... ......
. ....
. ..
...
..
... .
. ... . .
.
..
.
..
.. ... .. .....
.. ....
... ..... .. .. ........... ...
. ... .
. ....
.....
20. One domino covers two squares of a chessboard. Cover
... .
.. ... . ..
... ... ........ .
... .......... ............................................................................. . ..
... ......
all the squares except for its two opposite corners (on the same
.......
To Problem 24 To Problem 25 To Problem 27
diagonal) with 31 dominoes. (A chessboard consists of 88 = 26. Draw a straight line through the center of a cube so that
64 squares.) the sum of the squares of the distances to it from the eight
vertices of the cube is (a) maximal, (b) minimal (as compared
................................... with other such lines).
....................... ....
.............. ..........................................................................................................
...... 27. A..........right ....................................circular cone is cut by a plane along a closed
. ...
.
... .
..... ..
........ ..... ..... ................
... . .
.. .... ....
curve. Two spheres inscribed in the cone are tangent to the
..
.
............................................................................... . .
. .... ..
... .
. ... ...
.
.............
....... .
. ..... .... .... ..
..
. ...... ..........
... ...
.. ..... .......................................
plane, one at point A and the ... other at point B. Find a point C
... ... ..........
... .. ... ... ... . ..
... .. .
. . .. . ... .
..
.... ..... .. . .
... . .
...
.....
on ... the cross-section such that the sum of the distances CA +
... ...
5
..
... .. ... ...
...
3
... .
.... ... ... ... ... ... ... ...... ... ... .CB is (a) maximal, (b) minimal.
... . .
... ..
........... ... .. . .. . ...
... ...... ... ... ... ......
.
. ..... ... ... ... ... ........... .... .....
..... ..... . .. .. .. .. ...... .....................................
..... .................................................................................. .............................................................................................................................................
.....
.
.......
.....
... ..... ......... ....
..... ... ...
To Problem 20 ..... To Problem ....... 21 .... ... ... .... ... .............To Problem 22
..... ... ... ....................
..... ..... .. .......... .... ....................................
.... ....... A .. ......
.... . ...
... C
. .
. . . . ..
..... .............................................................................. . .....
..
... ... .. .... .... ... .......
.....................
... ... ... ............................................................................
21. A caterpillar wants to slither ......from
..... .
. ..........................the ................... front ... ..left corner
...
...... . . ... . .
. .
.. .
. . ...
.
. .. . ... ... ... . . .
.....
...
..........
. . .. .........
... .... ... ... ............
B
.... ...
A .............................................. ... ... . ... ... ... ..... ..... ... .......
of the floor of a cubical room to the .....opposite .............. corner
.............................................................(the right
.
.. .
... .......
.. .. .
... .
..... . . . .
. . .. .......
.... ..... ...... .... .... ........
... ....... .....
......... . . ... .... ... ...... ....... ... .....
rear corner of the ceiling). Find the ......shortest .......... route for such ... a
... . . .
. .
. . .
. . . ....
.. ..... . . . ..
. ... .....
B
. ....... ... ... .
.
. .
..
.. .. .. .....
... .... .. .
... ....
. ... ....
. . .. .
journey along the walls of the room. ..... ...........
.. . . ....
...
. ..
.... . ..
. .
... .
... .......
..
.. ..
..
... ....
....
. ..
................................................................................... . .
... ....
...
To Problem 24 To Problem 25 To Problem 27

EMS Newsletter December 2015 15


Feature

28. The Earths surface is projected onto a cylinder formed


... .......
..
.................... ..........................................................
..... ....... .................... ..... ............ ... .......................................................................
.... ........ ...... ................... ................... .............. . .......................... ....................................................................................
... ............. ... ......................................................................................
by the lines tangent to the meridians at the points where .... .... they
........ .... ................
.
...
. . . . . .. .. . . .
... ... .......... . .... ... ... ... .. ...................
. . ........... .. .. .. ..... .... .............................. .. ..... ... ...... ...... ... ......... ......
... ...... .... .... ... .. . .. .. ................. ......................... .. .. ... ..... ... .... ......................... ........ ........
.... ..... .. .... . .. .. .................................. .. ...
intersect the equator. The projection is made along rays ... .... paral-
.... .... ... .... ..... .... ..... ....
... ... ... ... .... ..... . ........................................ ...................
. ... .... .............. ....... .. . .. . . .. ..
... .. ..... .... ..................... .... ......... ....
. .. .
... ... .... .... ... .... ... .. .. ...... ...... ...... ... ..................
. . ... .......... ....... .. .. ...
. . ... ... ..... .... ........................ ... ..... .................. ... ... ... ...... ... ...... .......... ..... ....
lel to the plane of the equator and passing through......the .. axis .....of
. .. .... .. . . . .. .. ... ..... ... ...... ..... .
.
.
.
.. .... .. ..
...
.... ....
.... .... . . . . .
.............. . . ... ..
.
. . ... .... ..... .....
... .......... .... .. ... ... ................ .... ........ .... .... .... ...... ....... ... ............................ ....
... .
................................... . .. ... ... ............. ... .... ... ............... ..... ...... . .. ............ ... .
.. ... .... .... .... .. . .. .. .............. ... .......... ...... .... ...... ........ ... ....................................................... ................. .....
the earth that connects its north and south poles. ....Will the area
. ... ...........................
.
.. .... . ... ... ... ................. ..... ..... ............................ .. ... .... .... ..................... ... ... ..... ... .................................. .. ... ... ............. ........ ...
..
.. ..... .
. ...
. . . . ..... .. .. ..... ..... .......................... . .. .. ..
..
............................. ... ..... ..... ... ............ ..... . .... .... .. ... ..
..
...... . ... ... . .. .. ..... ..... .......................... .. .. ........................... . . ... .... .. ... ............. ..... .... . . . .. .. ..... ......... .....
of the projection of France be greater or less than
....................................the area of
.. ... . .......... ... ... .... ..... ..... ..... .................... .. ........................ .. .... .. .............. .... ..... .. .. ... .... ... ........................
.... ..... . .... ......... ..... ..... . ....... ....... .. ... .. .. ..
... ... .. ...
... ... .. .. . ..... ..... ... .. ........................................... ... . .. . .... ...... ... .... ... .
. .
..... ....... .... ... ... .... .... . . . . . . ... ... .
.................
France itself?
............................ ... ... .. ... ..... ...... . . ... ...... .... ..... ............. ... ... ... ........ ... ... .. ........................
............................
............................ ... ..... .. ..... ..... . .. .. ...............
..... ....... .... ... .... ................ .. . ... .. ............... .................
........................ .. .... .. ............................. .. ....... ..... ..... ................ ....................................... .. ...........................................................................
.......................................................... .............................. ..... ................ ................................................ ............................ . ............
........................ ............ ...... ... ...................................................................
........... ....... ......................................
....... .....
.....
tetrahedron
tetrahedron octahedron
octahedron icosahedron
icosahedron
...
.....
.... ..
....... ....
.......... ......................................................... ................
..................... .............
..... ...................................................... ..........
..................................... ..................................
..
.............
.....
(tetra==4)4)
(tetra (octa==8)8)
(octa (icosa==20)
(icosa 20)
France
.......
....
... ....... ............ .. ..... ....
.............
... ........................................ ....
.......
.... .... ........ ......... ......... ........ ....... ..
.......
32. There is one more Platonic solid (there are 5 of them
... ....... ............ ........ . ....... ..... ...... ..........
.............
......... .. ...... ... ......... ...
...........
...
Image ..... .........
altogether): a dodecahedron. It is a convex polyhedron with
.... .. ........
........ ... .......
... ........ . ..... ......... ......
......
..... ...
twelve (regular) pentagonal faces, twenty vertices and thirty
... ............... .....
.... .. ............ .
... ................................................................. .... . .............
... ... ................
.... .... ...

edges (its vertices are the centers of the faces of an icosahe-


..... ...
......
. ... ..........
.......................... ..
.......... .............. ............................
dron).
...... ................................... ......
.... ...
...
.... .
Inscribe five cubes in a dodecahedron, whose vertices are
..... ...
........
............ . ...... .....
........................... ........ ..
. ..
........
To Problem 30
.................................. ..
To Problem 28 also vertices of the dodecahedron, and whose edges are diago-
29. Prove that the remainder upon division of the number nals of faces of the dodecahedron. (A cube has 12 edges, one
2 p1 by an odd prime p is 1 (for example: 22 = 3a + 1, 24 = for each face of the dodecahedron). [This construction was
5b + 1, 26 = 7c + 1, 210 1 = 1023 = 11 93). invented by Kepler to describe his model of the planets.]
30. A needle, 10 cm long, is thrown randomly onto ruled
....
........... .........
......... .... ...............
..........
paper. The distance between neighboring lines on the paper ....
.... .. .. .
. .. .. . .... ........
..
................................................................
.... ..........
. .. .. .... ....... .
is also 10 cm. This is repeated N (say, a million) times. How
.
. .. ..
.... ... .. ... .................................... . ..... .. .......
.. ....... ........ ....
... ................................................................... ...
. ...
many times (approximately, up to a few per cent error) will
.. .. . .. . ...
..
.. ... ....... .. .. .... ...
...... ... ... . ...
the needle fall so that it intersects a line on the paper?
.. ... ... . .
.
. ...
. .. .
.... .. ... .. ... . ... ... .
. ...
........................... .... .. .. .. .. .. ... .. ............................. ....
One can perform this experiment with N = 100 instead of
... .. ...
....... . . . .
..... ..
. . .
.. ... . .... .
.... ... ......... .... . ... ... . .. .. .. .. .. .. .. .. .. .. .........
a million throws. (I did this when I was 10 years old.) .... .
....... ...... ............ .... .......
......... ..... ...............
... .. .. . . .. .
. ..........
. .
..............................................................
.. . ...
[The answer to this problem is surprising: 2 N. Moreover,
........ ..........
....
even for a curved needle of length a 10 cm the number of To Problem 32 To Problem 33bis
intersections observed over N throws will be approximately 33. Two regular tetrahedra can be inscribed in a cube, so that
2a ....................................................................
.N. The number 355 22 their vertices are also vertices of the cube, and their edges
......
....
113 7 .]
..............
............
are diagonals of the cubes faces. Describe the intersection of
........
. . ......
..... ...
..... ..

these tetrahedra.
...
..... ...
........ ............ .....
................................................ ................................................
... . . ..
...... ............................................................... ...........
..................................... ..................................
. .. .. ................................... ........France.......
.. ................... ..
...
... ...
..
..
.
........... What fraction of the cubes volume is the volume of this
.......
intersection?
... ..... ........ ......... ......... ........ ........ ... .......
... ........... ........ ........ .... ...... ...
.................
..... ..... ....... .... ....... .. .
.........
....

33bis . Construct the section cut of a cube cut off by the plane
..... Image
. .. . ..... .......... ..........
. ..........
..... .. .....
.......... ...... ......
passing through three given points on its edges. [Draw the
.... .. . .....
... ........... ... .
...... .. ...
... .................. ...
... . ...
..
............
... ................................................................. .. .......... .
.. ..............
polygon along which the plane intersects the faces of the
.... ... .
.... ....
..... ....
...... ....
............................ .....
..cube.]
.. ............. .......................... .......
......... ....................................... ...... ... .........
. ....... .... . .......
..... ... ... .....
..... ...
. ... .....
.......
........... ........
..... ...
.
.....
.....
.................................................
.....

31.
.
Some polyhedra have only triangular faces. Some exam-
........................ . .....
...... .
.
................................................ .....
To Problem 28 To Problem 30 ..
.
...
.
.....
..... C
..... .
...
...
. A
....
ples are the Platonic solids: the (regular) tetrahedron (4 faces), ....... ............................................................................
... ...
....
... ...
.
the octahedron (8 faces), and the icosahedron (20 faces). The
... ........................ . . .
B
....
A .............................................
...
... ... ... .....
...
... .......
... ...
.....
faces of the icosahedron are all identical, it has 12 vertices, ......
... ............. . . ...
. .... .
............................................................ . .
... ...
. .
..... B
.
... ..... .....
and it has 30 edges.
.
..... .....
...
......
... ......
... .. ... . ........
Is it true that for any such solid (a bounded convex poly- .......................
. .
...
.. .
................................................................................ .....

hedron with triangular faces) the number of faces is equal to To Prob em 24 To Prob em 25 To Prob em 27
twice the number of vertices minus four? 34 How many symme r es does a e rahedron have? A cube?
........
..........
...... .. .....
An oc ahedron? ........ . ........An
........................................... cosahedron? A dodecahedron? A sym-
.
...... . .. ..........................................................................................
me ry of .a
. ....figure .... s a ..... ransforma on of h s figure preserv ng
. ... ........... ........ .... ........ .
..
....... ......................... .... ......... .
. .
.. ... ..... .
.
.. .. .. ....... ... .
...... ... ....
.. ... .... ..
..
.. . .. ..... ..
. ..
... ...... . . . .
.. .. ... .. ......
... ... ..... ......... ... .........
eng hs
.... ... .... ..
...... ..... .. ..... . . .. ..................... ... .. . . .
..
.
.. ... ....
. .
.. .. .
.. .. ... .. . ... ..... . ......
.
.
...........
. .. ...
. ... ....... . ... ....
.
... .... .
.... .. . .. ..... ... ..... .
... . ....... . ... ... ..... . ..
How .. .. many
......of hese symme r es are ro a ons and how
.... ... .... ..... ... . .. ...
... ... ........
. .. . ........ ... ... ... ...... ..... .....
.... .... ........ .. .. ...
... ... .... .......... .. .... ... ...... .. .. ............. ....
... ... ... ... ... .................................. ....
.... . ..
... ... ..
.....many are
... ... reflec ...........................ons ....................... n..... p ... ..anes ( n each of he five cases
.. ... .. ... . .
. .
. ... ..
..... .. . .
.. ... ... ..
... . ...
... .... ............. .. .. .
........
.. ..... ... ... ..... .. ........ . ... ..
.... ...
. .
.. ..
. ....
.
................
. ....
. ..
..
. .... ... . . .. .
.. . .
. . .
. ..... ... .
..... ........... ... .
..... ... . . .
... .. ... .......... . . . ..
s ed)?
... . .... ........
.
... ..... . .
............ ..... . ......
... ... ... .. ..... ....................... ..
... ......................
... .. ... ..... .... .
... .. ... ........
.
..... . ... ..
. ... ....
. ..
.. ...... ... ... .
. . .
.. . .
. .
.
. .
. .... ..
. ..
. .
. .
..
.. . ... .. ... . .. .. ....
..... . ................ ....
.. . ... ... .. ... ... .... . ... . .. ....
........................ ... ..... .. ..... ... .. ........
35 How
..
... .... many ... ways are
. .... .....here o pa n he s x faces of s m ar
.............. .....
..... .. .... ........ .... .... .. .... .. ... .... . ... . . ... . ... . .........
..............
..............
... ...
. . ..... .. ... ..... .. ......
. ...... ... .........
. . ..... . ...... ................
........................ ... ..... .. ........... .. .... ..........
cubes w h s x co ...ors ..........................(1 6) [one co or per face] so ha no
............... ..
............. ....... ..... . ........
........ ................................... .... .........
......... ...

tetrahedron octahedron wo of he co icosahedron ored cubes ob a ned are he same ( ha s no


(tetra = 4) (octa = 8) wo can be (icosa ransformed = 20) n o each o her by a ro a on)?

16 EMS Newsletter December 2015


Feature
...................................................................
...... ...... ...
......
..... 6 ...... .... 43. Suppose, in Problem 42, there turn out to be a lattice
6 .......................................................................... ...
..
..
...
....
...
... points inside the parallelogram, and b lattice points on its
.. ..
1 2 3 4 ..
..
...
.. 3 ....
. ...
sides. Find its area.
.. 2 ...
.. ..
..
..
...
... .
.. .. 44. Is the statement analogous to the result of problem 43
5 .. ... ..........
..
..................................................................
........ true for parallelepipeds in 3-space?
45. The Fibonacci (rabbit) numbers are the sequence (a1 =
36. How many different ways are there to permute n objects? 1), 1, 2, 3, 5, 8, 13, 21, 34, . . . , for which an+2 = an+1 + an for
For n = 3 there are six ways: (1, 2, 3), (1, 3, 2), (2, 1, 3), any n = 1, 2, . . . . Find the greatest common divisor of the
(2, 3, 1), (3, 1, 2), (3, 2, 1). What if the number of objects is numbers a100 and a99 .
n = 4? n = 5? n = 6? n = 10?
46. Find the number of ways to cut a convex n-gon into tri-
37. A cube has 4 major diagonals (that connect its opposite angles by cutting along non-intersecting diagonals. (These are
vertices). How many different permutations of these four ob- the Catalan numbers, c(n)). For example, c(4) = 2, c(5) = 5,
jects are obtained by rotations of a cube? c(6) = 14. How can one find c(10)?
.................................................. ...................................................
................................................................. .... ...
........................................................ ....
............................................................... .................................................................. .. .......... ...........
........................................................................... .... .......................................................................... ....
. ......... ............ .. ........... ..
... ... ...... ... .. .. .......... ..... ........................
.......... .......... ... ........... ...
....... .... ..... .... ................ ..............
. . .. ...
.
... ...
... ..... ....
...
...
....... ..... ....
.
. . ..
.. . . ... ..
.. ..
.... . .... .
.
.
..... ... ......... ........
....
... ... .... ...
........... .. .
.. .... ....
... .
. ... .
.
... ..... ..................................
. ..
..... ... ... ... ...
.......... ... ..... ..... ...
... ................................
... ...
..... ..... ..... .......
... ... ... .. ................................
.... ..............
... .. ..
....
. ..
... ...
. . .. .... . .. . ..
................................................... ......................................................................
.. .
...................................................... . .... . .....
.................................................................... ...... ................
. ..
..... ...... ........
..... ...........
........
...... ...........
.......
..... ...........
..
..
................ ...... ...... ..... ...... ...... ......
...... ........ ........ ........
38. The sum of the cubes of several integers is subtracted ..............
..... . . .
..
..... ..... ... ... ... ... ... ..
.. .. ...
from the cube of the sum of these numbers. Is this difference
... ..... ... ... ...
..... ..... ... .. ... .
. ... ..
............ ....... .
........................
.
........................ ........................
always divisible by 3? ......
.....
.....
.....
..... ..... ...... ......
...... ........... ...... ...........
39. Answer the same question for the fifth powers and divis-........ .....
..... ....
..... ...... ..... ......
... ........ ... ........ ...
ibility by 5, and for the seventh powers and divisibility by 7. ..................
..... ..
. . ... ...
.. ..
... ... ... ...
... . ... ..
40. Calculate the sum .......................... .........................

1 1 1 1 a = 2, b = 2
+ + + +
12 23 34 99 100 To Problems 42,47.43 There are n teamsToparticipating Problem 46in a tournament. After
(with an error of not more than 1% of the correct answer). each game, the losing team is knocked out of the tournament,
and after n 1 games the team left is the winner of the tour-
41. If two polygons have equal areas, then they can be cut nament.
into a finite number of polygonal parts which may then be A schedule for the tournament may be written symboli-
rearranged to obtain both the first and second polygons. Prove cally as (for example) ((a, (b, c)), d). This notation means that
this. [For spatial solids this is not the case: the cube and the there are four teams participating. First b plays c, then the
tetrahedron of equal volumes cannot be cut this way!] winner plays a, then the winner of this second game plays d.
........................................................ ....................................................
... ....
How many possible schedules are there if there are 10
teams in the tournament?
. .
.. .. ...
... ... ... .
..
.
For 2 teams, we have only (a, b), and there is only one
. . ..
.. .. ..................... .... ...
... ... ...
...
schedule.
. . ...
.. .. .....................................................
.....................................................
For 3 teams, the only possible schedules are ((a, b), c), or
42. Four lattice points on a piece of graph paper are the ver- ((a, c), b), or ((b, c), a), and are 3 possible schedules.
tices of a parallelogram. It turns out that there are no other For 4 teams we have 15 possible schedules:
lattice points either on the sides of the parallelogram or inside
it. Prove that the area of such a parallelogram is equal to that (((a, b), c), d); (((a, c), b), d); (((a, d), b), c); (((b, c), a), d);
of one of the squares of the graph paper. (((b, d), a), c); (((c, d), a), b); (((a, b), d), c); (((a, c), d), b);
(((a, d), c), b); (((b, c), d), a); (((b, d), c), a); (((c, d), b), a);
((a, b),..(c,
... d)); ((a, c), (b, d)); ((a, d), (b, c)).
..
.......... .........
............ ...
... .
............. ...
.............. ............ . ............ ..
. .
.... . .............. .. .............. ...
...............
....... .
.................
. ...
...
....
. 48. ... We connect
... .
.
. n points 1, 2, . . . , n with n 1 segments to
... ............................. . ........................
form.....a......tree. How many different trees can we get? (Even the
..
. ..
. ......
....
. .....
...... ......
.................
..........
... . .. ...... case
...... n = 5 is .interesting!) ......
............ ..... ........... ..... ........... .... ...........
.......... ...... ...... ...... ...... 1 2 .......... ......
...... .........
..... .......
...
..
..
.......
... n = 2: ............................. the number
. .
.. = 1;
..
..... .......... ... .. ... . .. ..
.... 1.. 2.. ....3.. 2.. 1...... 3.. 1.. 3.. 2..
...
.....
. ..... ... .... ...
.......... .....
....... .......................... .........n
......=
.......3:
. . ........... ........... . . ........... ........... . . ........... ........... .
... .. . ............................ . .. . . the number = 3;
..... ..... 1 2 3 4
2 1 1 1
..... ..... ...................................................
1................. ........... 2........... .. 3........... .. 4........... ..
..... ..... .
....
. ....
. .
. ...
. .. .. .. ..
...... ............
3....... ....... 3 ....... 2 ....... 2 .1...............3...............2...............4..... the number = 16.
.
..... ....... ...... n = ..4:
..... ... ...
.. ...
.. . .. . ... .
...... . .. ..
.. .........
.. .... ..... .. ..
.. .... ..... .. ..
..
. . ..
..... ........ ..... ..........
4 .... ... 4 4 3
... .. ..... .....
..... .....
..... ....
. ... ...
. ... ...........
... . ... ..
.... . .
........................ 49. .A .........permutation (x1 , x2 , . . . , xn ) of the numbers {1, 2, . . . ,
... .. ... ..
. .
..............
a = 2, b = 2 n} is called a snake (of length n) if x1 < x2 > x3 < x4 > .
To Problems 42, 43 To Problem 46

EMS Newsletter December 2015 17


Feature

Examples. n = 2, only 1 < 2 , the number = 1; 55. The sequence of Fibonacci numbers was defined in Prob-
lem 45. Find the limit of the ratio an+1 /an as n approaches

n = 3, 1<3>2
, the number = 2;
2<3>1 infinity:
n = 4, 1 < 3 > 2 < 4
an+1 3 5 8 13 21 34
= 2, , , , , , , . . . .
1 < 4 > 2 < 3 an 2 3 5 8 13 21


2 < 3 > 1 < 4 , the number = 5;
2 < 4 > 1 < 3

Answer: The golden ratio, 5+1 2 1.618.
This is the ratio of the sides of a postcard which stays sim-

3<4>1<2
ilar to itself if we snip off a square whose side is the smaller
Find the number of snakes of length 10. side of the postcard.
How is the golden ratio related to a regular pentagon and
50. Let sn denote the number of snakes of length n, so that a five-pointed star?
56. Calculate the value of the infinite continued fraction
s1 = 1, s2 = 1, s3 = 2, s4 = 5, s5 = 16, s6 = 61.
1 1
Prove that the Taylor series for the tangent function is 1+ = a0 + ,
1 1
2+ a1 +
x1 x3 x5  x2k1
1 1
tan x = 1 + 2 + 16 + = s2k1 . 1+ a2 +
1! 3! 5! (2k 1)! 1 1
k=1 2+ a3 +
1 ..
1+ .
..
51. Find the sum of the series .

x2 x4 x6  x2k
where a2k = 1, a2k+1 = 2.
1+1 + 5 + 61 + = s2k . That is, find the limit as n approaches infinity of
2! 4! 6! k=0
(2k)!
1
52. For s > 1, prove the identity a0 + .
1
a1 +

 1 
1 1
= . a2 +
1 n s ..
p=2 1 n+1 . 1
ps +
an
(The product is over all prime numbers p, and the summation
over all natural numbers n.)
57. Find the polynomials y = cos 3(arccos x), y = cos 4
(arccos x), y = cos n(arccos x), where |x| 1.
53. Find the sum of the series
 1
1 1 58. Calculate the sum of the kth powers of the n complex nth
1+ + + = . roots of unity.
4 9 n=1
n2

(Prove that it is equal to 2


or approximately 32 .) 59. On the (x, y)plane, draw the curves defined parametri-
6,
cally:
54. Find the probability that the fraction qp is in lowest terms.
{x = cos 2t, y = sin 3t}, {x = t3 3t, y = t4 2t2 }.
This probability is defined as follows: in the disk p2 +
q R2 , we count the number N(R) of points with integer
2

coordinates p and q not having a common divisor greater than 60. Calculate (with an error of not more than 10 % of the
1. Then we take the limit of the ratio N(R)/M(R), where M(R) answer)
 2
is the total number of integer points in the disk (M R2 ). sin100 x dx.
0
...............................................................
..........
....... ...............
61. Calculate (with an error of not more than 10 % of the
. .
. ...
......... ...........
answer)
...
....... .........
 10
.
.... .....
.
.
...... N (10) = 192
...
..
..
. ..... M (10) = 316 x x dx.
...
.. .... N/M = 192/316 1
... .....
.
0.6076
.... ...
. ..
...
... ....
. 62. Find the area of a spherical triangle with angles (, , )
...
... .. ..
.
. on a sphere of radius 1. (The sides of such a triangle are great
...
... .... circles; that is, cross-sections of the sphere formed by planes
... ....
...
... ..... passing through its center).
.... .......
.....
........... Answer: S = + + . (For example, for a triangle with
.....
......
......
....... ...........
..............
................................... .............. three right angles, S = /2, that is, one-eighth of the total
....................
area of the sphere).

18 EMS Newsletter December 2015


Feature

......
.........
.......................................
....... ......
.........
.......................................
....... B....
.. . .
. ...... ..
... ...... ..... ....
..... K
..... ..
... ..... .... .
. .... .. .... .
..
....
.............. ....
.. .
. .
.
.... .......................
. ... ... .
......... ... . .. L
.. ....... ........

... .. ...... . .... ..... ..... .........
1.....
......... ... ... ... ... ..
. ..
.. .. .. ... .. . ... .
.... ... .. ...
.....
.. .. .
.. ..... ...
...
... .. .
. ..... ... ....... ... .... ....
... .. ... .. . ... ... .. .....
..... ... .. ...... .. .....
..
...
.....
..... .. ..
.
... ........... ...................
. ..
.
. ........ ... ...
.....
...
..
.. ... .
... ............. ............. ... .
..
...... .. .... ............ ........ ................................................... ...........................
A C
... . .. . ... ..... ..
....... . .
M
... ...... .
. .. ... ... ........... .... ... ..
... .. ..
... ......
....... .. .. ... .... r....... .. ...
... ....... .... .... ... ... ....
......... .. ... ... ....
Hint: ...................The answer for non-acute angled triangles is not nearly
.... ....
..... .......... ..... ....... ....
...... .....
.....
...... .....
......
......................as ..........beautiful as the answer for acute angled triangles.
....... .
.........
................................. .
. ..... .......
........... ......

67.63 Calculate the average value of the function 1/r (where


To Problem To Problem 62
63. A circle of radius r rolls (without slipping) inside a circle r2 = x2 + y2 + z2 is the distance to the origin from the point
of radius 1. Draw the whole trajectory of a point on the rolling with coordinates (x, y, z)) on the sphere of radius R centred at
circle (this trajectory is called a hypocycloid) for r = 1/3, the point (X, Y, Z).
r = 1/4 for r = 1/n, for r = p/q, and for r = 1/2. Hint: The problem is related to Newtons law of gravitation
and Coulombs law in electricity. In the two-dimensional ver-
sion of the problem, the given function should be replaced by
.................................. .................................
........... ....... .......... .......
....... ...... ....... ......
ln r, and the sphere by a circle.
.... .. ..... ....
.. .....
.... .... ..
.. ....
.... .... ....
.... ... .............
68. The fact that 210 = 1024 103 implies that log10 2
.
.. .
. .
...... ....
... ...
.............. .
. ..

... .. . ... .....
..
1.......
.......
.. .
... .
.. ... ... .
...
....... .. ... ...
. ..
.....
...... ...
.. .....
0.3. Estimate by how much they differ, and calculate log10 2
.. ...
.....
...
. .. .... .
. .
....... ..
.
to three decimal places.
..... .. .. .
... . ..
...... .
.. . ... ....... .......... .
.....
...... .. ... ... . .
.. ................... ......... ....
. ........... ............ ...
...... .. .. ... ... .... ..... .
.
.
69. Find log10 4, log10 8, log10 5, log10 50, log10 32,
...... . ..
.. .. ... . ..... . .. ..
......
...... . .
... ....
... ..... r........ .. ..
........ .. .... .....
log10 128, log10 125, and log10 64 with the same precision.
... ....... . ... .. .
.... .......... .... .... ... .
..
..... ..... .
..... . ..........................
70. Using the fact that 72 50, find an approximate value
...... ...... ......
......
.......
.......... ........... ......
........ ......
.................................... ......................................
for log10 7.
To Problem 62 To Problem 63
71. Knowing the values of log10 64 and log10 7, find log10 9,
64. In a class of n students, estimate the probability that two log10 3, log10 6, log10 27, and log10 12.
students have the same birthday. Is this a high probability? Or
a low one? 72. Using the fact that ln(1 + x) x (where ln means loge ),
find log10 e and ln 10 from the relation1
Answer: (Very) high if the number of the pupils is (well)
above some number n0 , (very) low if it is (well) below n0 , ln a
log10 a =
and what this n0 actually is (when the probability p 1/2) is ln 10
what the problem is asking. and from the values of log10 a computed earlier (for example,
for a = 128/125, a = 1024/1000 and so on).
65. Snells law states that the angle made by a ray of light Solutions to Problems 6771 will give us, after a half hour
with the normal to layers of a stratified medium satisfies the of computation, a table of four-digit logarithms of any num-
equation n(y) sin = const, where n(y) is the index of refrac- bers using products of numbers whose logarithms have been
tion of the layer at height y. (The quantity n is inversely pro- already found as points of support and the formula
portional to the speed of light in the medium if we take its
speed in a vacuum to be 1. In water n = 4/3). x2 x 3 x4
ln(1 + x) x
+ +
Draw the rays forming the lights trajectories in the 2 3 4
medium air above a desert, where the index n(y) has a max- for corrections. (This is how Newton compiled a table of 40-
imum at a certain height. (See the diagram on the right.) digit logarithms!).
73. Consider the sequence of powers of two: 1, 2, 4, 8, 16,
... y ... y
32, 64, 128, 256, 512, 1024, 2048, . . . Among the first twelve
........
.. ...................... .......
... numbers, four have decimal numerals starting with 1, and
...........
none have decimal numerals starting with 7.
... .............. ..........
......... ... ......
... ... ....
Prove that in the limit as n each digit will be met
... ........ ...
....... ....
... ...... ..
with as the first digit of the numbers 2m , 0 m n, with a
... ...... .... ..
....
.....
.... .... ...
.. n(y)
..
certain average frequency: p1 30 %, p2 18 %, . . . , p9
... .
.... ....
..................................................................................................................................... ......................................................................
.
..........
4 %.
74. Verify the behavior of the first digits of powers of three:
(A solution to this problem explains the phenomenon of mi-
1, 3, 9, 2, 8, 2, 7, . . . Prove that, in the limit, here we also get
rages to those who understand how trajectories of rays ema-
nating from objects are related to their images).
1 Eulers
 nconstant e = 2.71828 is defined as the limit of the sequence
66. In an acute angled triangle ABC inscribe a triangle KLM 1 + 1n as n . It is equal to the sum of the series 1+ 1!1 + 2!1 + 3!1 + .
of minimal perimeter (with its vertex K on AB, L on BC, M It can also be defined by the given formula for ln(1 + x) : lim x0 ln(1+x)
=
x
on CA). 1.

EMS Newsletter December 2015 19


Feature

certain frequencies and that the frequencies are same as for 79. In Problems 76 and 78, prove that the sequence {gT (x)},
the powers of two. Find an exact formula for p1 , . . . , p9 . T = 1, 2, . . . is distributed over the torus uniformly: if a
Hint: The first digit of a number x is determined by the frac- domain A contains kn (A) points out of the n points with
tional part of the number log10 x. Therefore one has to con- T = 1, 2, . . . , n, then
sider the sequence of fractional parts of the numbers ma, kn (A) mes A
lim =
where a = log10 2. n n mes M
Prove that these fractional parts are uniformly distributed (for example, for a Jordan measurable domain A of measure
over the interval from 0 to 1: of the n fractional parts of the mes A).
numbers ma, 0 m < n, a subinterval A will contain the
quantity kn (A) such that as n , lim(kn (A)/n) = the length
of the subinterval A.
75. Let g : M M be a smooth map of a bounded do- Note to Problem 13. In posing this problem, I have tried to
main M onto itself which is one-to-one and preserves areas illustrate the difference in approaches to research by mathe-
(volumes in the multi-dimensional case) of domains. maticians and physicists in my invited paper in the journal
Prove that in any neighborhood U of any point of M and Advances in Physical Sciences for the 2000 Centennial issue.
for any N there exists a point x such that gT x is also in U for My success far surpassed the goal I had in mind: the editors,
a certain integer T > N (the Recurrence Theorem). unlike the preschool students on the experience with whom I
76. Let M be the surface of a torus (with coordinates based my plans, could not solve the problem. So they changed

(mod 2), (mod 2)), and let g(, ) = ( + 1, + 2). it to fit my answer of 4mm. in the following way: instead of
Prove that for every point x of M the sequence of points from the first page of the first volume to the last page of the
{gT (x)}, T = 1, 2, . . . is everywhere dense on the torus. second, they wrote from the last page of the first volume to
the first page of the second.
77. In the notation of problem 76, let
This true story is so implausible that I am including it
g(, ) = (2 + , + ) (mod 2). here: the proof is the editors version published by the journal.
Prove that there is an everywhere dense subset of the torus
consisting of periodic points x (that is, such that gT (x) = x for This excerpt was taken from the book V. I. Arnold, Lectures
some integer T (x) > 0). and Problems: A Gift to Young Mathematicians (AMS, 2015;
78. In the notation of Problem 77 prove that, for almost all ISBN-13: 978-1-4704-2259-2). To learn more about this book
points x of the torus, the sequence of points {gT (x)}, T = and to order a copy, visit www.ams.org/bookstore-getitem/
www.ams.org/ bookstore-getitem/
1, 2, . . . is everywhere dense on the torus (that is, the points x item=mcl-17.
item=mcl-17 The Newsletter thanks the AMS and the MSRI
without this property form a set of measure zero). for authorizing the reprint.

Additive Eigenvalue
Additive EigenvalueProblem
Problem
Shrawan Kumar(University
Shrawan Kumar (Universityofof North
North Carolina,
Carolina, Chapel
Chapel Hill,Hill,
USA) USA)

1 Introduction lem has been extended by Berenstein-Sjamaar (2000) and


Kapovich-Leeb-Millson (2009) for any semisimple complex
The classical Hermitian eigenvalue problem addresses the fol- algebraic group G. Their solution is again in terms of a sys-
lowing question. What are the possible eigenvalues of the tem of linear inequalities obtained from certain triples of
sum A + B of two Hermitian matrices A and B, provided Schubert classes in the singular cohomology of the partial
we fix the eigenvalues of A and B. A systematic study of flag varieties G/P (P being a maximal parabolic subgroup)
this problem was initiated by H. Weyl (1912). By virtue and the standard cup product. However, their solution is far
of contributions from a long list of mathematicians, notably from being optimal. In a joint piece of work with P. Bel-
Weyl (1912), Horn (1962), Klyachko (1998) and Knutson- kale, we defined a deformation of the cup product in the
Tao (1999), the problem was finally settled. The solution cohomology of G/P and used this new product to generate
asserts that the eigenvalues of A + B are given in terms a system of inequalities which solves the problem for any
of a certain system of linear inequalities in the eigenval- G optimally (as shown by Ressayre). This article is a brief
ues of A and B. These inequalities are given explicitly in survey of this additive eigenvalue problem. The eigenvalue
terms of certain triples of Schubert classes in the singular problem is equivalent to the saturated tensor product prob-
cohomology of Grassmannians and the standard cup prod- lem.
uct. Belkale (2001) gave an optimal set of inequalities for This note was written during my visit to the University
the problem in this case. The Hermitian eigenvalue prob- of Sydney and their hospitality is gratefully acknowledged.

20 EMS Newsletter December 2015


Feature

Partial support from NSF grant number DMS-1501094 is also : k/K h+ , where K acts on k by the adjoint representation
gratefully acknowledged. and h+ := {h h : i (h) 0 i} is the positive Weyl chamber
in h. The inverse map 1 takes any h h+ to the K-conjugacy

2 Main Results class of 1h.
For any positive integer s, define the eigencone
We now explain the classical Hermitian eigenvalue problem 
s
and its generalisation to an arbitrary connected semisimple s (g) :={(h1 , . . . , h s) h+s | (k1 , . . . , k s ) k s : kj
group more precisely. j=1
For any n n Hermitian matrix A, let A = (1 n ) = 0 and (k j ) = h j j}.
be its set of eigenvalues written in descending order. Re-
call the following classical problem, known as the Hermitian By virtue of the general convexity result in symplectic geome-
eigenvalue problem. Given two n-tuples of nonincreasing real try, the subset s (g) h+s is a convex rational polyhedral cone
numbers: = (1 n ) and = (1 n ), (defined by certain inequalities with rational coefficients). The
determine all possible = (1 n ) such that there aim of the general additive eigenvalue problem is to find the
exist Hermitian matrices A, B, C with A = , B = , C = inequalities describing s (g) explicitly. (The case g = sl(n)
and C = A + B. This problem has a long history, starting and s = 3 gives the Hermitian eigenvalue problem if we re-
with the work of Weyl (1912), followed by works of Fan place C by C.)
(1949), Lidskii (1950) and Wielandt (1955), and culminating Let = (H) denote the character group of H and let
in the following conjecture given by Horn (1962). (See also + := { : (i ) 0 simple coroots i } denote the
Thompson-Freede (1971).) set of all the dominant characters. Then, the set of isomor-
For any positive integer r < n, inductively define the set phism classes of irreducible (finite dimensional) representa-
S nr as the set of triples (I, J, K) of subsets of [n] := {1, . . . , n} tions of G is parametrised by + via the highest weights of
of cardinality r such that irreducible representations. For + , we denote by [] the
   corresponding irreducible representation (of highest weight
i+ j = r(r + 1)/2 + k (1) ).
iI jJ kK Similar to the eigencone s (g), one defines the saturated
and, for all 0 < p < r and (F, G, H) S rp , the following tensor semigroup:
inequality holds: s (G) ={(1 , . . . , s ) +s : ([N1 ] [N s ])G
  
if + jg p(p + 1)/2 + kh . (2)  0, for some N 1}.
f F gG hH
Then, under the identification : h h (via the Killing
Conjecture 1. A triple , , occurs as eigenvalues of Her-
form),
mitian n n matrices A, B, C respectively such that C = A + B
( s (g)) +s = s (G) (3)
if and only if
n n n (see Theorem 5).
i = i + i For any 1 j , define the element x j h by
i=1 i=1 i=1
and, for all 1 r < n and all triples (I, J, K) S nr , we have i (x j ) = i, j , 1 i . (4)
   Let P B be a standard parabolic subgroup with Lie alge-
k i + j.
bra p and let l be its unique Levi component containing the
kK iI jJ
Cartan subalgebra h. Let (P) be the set of simple roots
Horns conjecture above was settled in the affirmative contained in the set of roots of l. Let WP be the Weyl group of
(see Corollary 11) by combining the work of Klyachko [Kly] P (which is, by definition, the Weyl Group of the Levi com-
(1998) with the work of Knutson-Tao [KT] (1999) on the sat- ponent L) and let W P be the set of the minimal length repre-
uration problem. sentatives in the cosets of W/WP . For any w W P , define the
The above system of inequalities is overdetermined. Bel- Schubert variety:
kale (2001) proved that a certain subset of the set of inequal-
ities suffices. Subsequently, Knutson-Tao-Woodward (2004) XwP := BwP/P G/P.
proved that the subsystem of inequalities given by Belkale It is an irreducible (projective) subvariety of G/P of dimen-
forms an irredundant system of inequalities. sion (w). Let (XwP ) denote the fundamental class of XwP , con-
Now, we discuss a generalisation of the above Hermitian sidered as an element of the singular homology with integral
eigenvalue problem (which can be rephrased in terms of the coefficients H2(w) (G/P, Z) of G/P. Then, from the Bruhat
special unitary group SU(n) and its complexified Lie alge- decomposition, the elements {(XwP )}wW P form a Z-basis of
bra sl(n)) to an arbitrary complex semisimple group. Let G H (G/P, Z). Let {[XwP ]}wW P be the Poincar dual basis of the
be a connected, semisimple complex algebraic group. We fix singular cohomology H (G/P, Z). Thus,
a Borel subgroup B, a maximal torus H B and a maxi-
[XwP ] H 2(dim G/P(w)) (G/P, Z).
mal compact subgroup K. We denote their Lie algebras by
the corresponding Gothic
characters: g, b, h, k respectively. We Write the standard cup product in H (G/P, Z) in the {[XwP ]}
choose K such that 1 k h. Let R+ be the set of positive basis as follows:

roots (i.e. the set of roots of b) and let = {1 , . . . , } R+ [XuP ] [XvP ] = cwu,v [XwP ]. (5)
be the set of simple roots. There is a natural homeomorphism wW P

EMS Newsletter December 2015 21


Feature

Introduce the indeterminates i for each i \ (P) and cominuscule; hence, the deformed product 0 in H (G/P) co-
define a deformed cup product as follows: incides with the standard cup product and (c) and (d) are the
   1  same in this case.
u1 v1 )(xi ) w
[XuP ] [XvP ] = (w
i cu,v [XwP ], Because of the identification (3), the above theorem al-
wW P i \(P)
lows us to determine the saturated tensor semigroup s (G)
(6)
(see Theorem 18 for a precise statement).
where is the (usual) half sum of positive roots of g. By
The following result was proved by Ressayre [R] (2010).
Corollary 16 and the identity (13), whenever cwu,v is nonzero,
As mentioned above, for g = sl(n) it was proved by Knutson-
the exponent of i in the above is a nonnegative integer.
Tao-Wodward. Ressayres proof relies on the notion of well-
Moreover, the product is associative (and clearly commu-
covering pairs, which is equivalent to the notion of Levi-
tative). The cohomology algebra of G/P obtained by set-
movability with cup product 1.
ting each i = 0 in (H (G/P, Z) Z[i ], ) is denoted by
(H (G/P, Z), 0). Thus, as a Z-module, this is the same as Theorem 3. The inequalities given by (d) of the above the-
the singular cohomology H (G/P, Z) and, under the product orem form an irredundant system of inequalities determining
0 , it is associative (and commutative). Moreover, it contin- the cone s (g) (see Theorem 23 for a more precise statement).
ues to satisfy the Poincar duality (see [BK1 , Lemma 16(d)]).
As shown by Kumar-Leeb-Millson (2003), (c) of the
The definition of the deformed product 0 (now known as
above theorem gives rise to 126 inequalities for g of type B3
the Belkale-Kumar product) came from the crucial concept of
or C3 , whereas (d) gives only 93 inequalities.
Levi-movability in Definition 14. For a cominuscule maximal
We refer the reader to the survey article of Fulton [F] on
parabolic P, the product 0 coincides with the standard cup
the Hermitian eigenvalue problem and, for general G, the sur-
product (see Lemma 17).
vey articles by Brion [Br] and by Kumar [K3 ].
Now we are ready to state the main result on solution of
the eigenvalue problem for any connected semisimple G. For
a maximal parabolic P, let iP be the unique simple root not 3 Determination of the Eigencone
in the Levi of P and let P := iP be the corresponding fun- (A Weaker Result)
damental weight.
Below, we give an indication of the proof of the equivalence
Theorem 2. Let (h1 , . . . , h s ) h+s . Then, the following are of (a) and (b) in Theorem 2.
equivalent:
Definition 4. Let S be any (not necessarily reductive) alge-
(a) (h1 , . . . , h s ) s (g).
braic group acting on a (not necessarily projective) variety X
(b) For every standard maximal parabolic subgroup P in G
and let L be an S -equivariant line bundle on X. Any algebraic
and every choice of s-tuples (w1 , . . . , w s ) (W P ) s such
group morphism Gm S is called a one parameter subgroup
that
(OPS) in S . Let O(S ) be the set of all OPSs in S . Take any
[XwP1 ] [XwPs ] = d[XeP ] for some d  0, x X and O(S ) such that the limit limt0 (t)x exists in X
(i.e. the morphism x : Gm X given by t (t)x extends to
the following inequality holds:
a morphism  x : A1 X). Then, following Mumford, define

s  a number (x, ) as follows. Let xo X be the point 
L
x (0).
P
I(w 1 ,...,w s )
: P w1
j h j 0. Since xo is Gm -invariant via , the fiber of L over xo is a Gm -
j=1
module and, in particular, is given by a character of Gm . This
(c) For every standard maximal parabolic subgroup P in G integer is defined as L (x, ).
and every choice of s-tuples (w1 , . . . , w s ) (W P ) s such
that Under the identification : h h (via the Killing
form), s (G) corresponds to the set of integral points of s (g).
[XwP1 ] [XwPs ] = [XeP ],
Specifically, we have the following result, essentially follow-
P
the above inequality I(w 1 ,...,w s )
holds. ing from Mumford [N, Appendix] (see also [Sj, Theorem 7.6]
(d) For every standard maximal parabolic subgroup P in G and [Br, Thorme 1.3]).
and every choice of s-tuples (w1 , . . . , w s ) (W P ) s such
Theorem 5.
that  
s (g) +s = s (G).
[XwP1 ] 0 0 [XwPs ] = [XeP ],
P Let P be any standard parabolic subgroup of G acting on
the above inequality I(w 1 ,...,w s )
holds.
P/BL via the left multiplication, where L is the Levi subgroup
The equivalence of (a) and (b) in the above theorem for of P containing H, and BL := B L is a Borel subgroup
general G is due to Berenstein-Sjamaar (2000). Kapovich- of L. We call O(P) P-admissible if, for all x P/BL ,
Leeb-Millson (2009) showed the equivalence of (a) and (c). limt0 (t) x exists in P/BL . If P = G then P/BL = G/B
The equivalence of (a) and (d) is due to Belkale-Kumar and any O(G) is G-admissible since G/B is a projective
(2006). If we specialise the above theorem to G = SL(n) then, variety. For O(G) define the Kempfs parabolic subgroup
in view of Theorem 10, the equivalence of (a) and (b) is noth- of G by P() := {g G : limt0 (t)g(t)1 exists in G}.
ing but Horns conjecture (Corollary 11) solved by combining Observe that, BL being the semidirect product of its com-
the work of Klyachko (1998) with the work of Knutson-Tao mutator [BL , BL ] and H, any extends uniquely to a char-
(1999). In this case, the equivalence of (a) and (c) is due to acter of BL . Thus, for any , we have a P-equivariant line
Belkale (2001) and every maximal parabolic subgroup P is bundle LP () on P/BL associated to the principal BL -bundle

22 EMS Newsletter December 2015


Feature

P P/BL via the one dimensional BL -module 1 . We ab- Specialising Theorem 7 to G = SL(n), as seen below, we
breviate LG () by L(). We have taken the following lemma obtain the classical Horn inequalities.
from [BK1 , Lemma 14]. It is a generalisation of the corre- In this case, the partial flag varieties corresponding to the
sponding result in [BS, Section 4.2]. maximal parabolics Pr are precisely the Grassmannians of r-
planes in n-space G/Pr = Gr(r, n), for 0 < r < n. The Schu-
Lemma 6. Let O(H) be such that h+ . Then, is P-
bert cells in Gr(r, n) are parametrised by the subsets of cardi-
admissible and, moreover, for any and x = ulBL P/BL
nality r:
(for u U, l L), we have the following formula:
I = {i1 < . . . < ir } {1, . . . , n}.

LP () (x, ) = (w), The corresponding Weyl group element wI W Pr is nothing
where U is the unipotent radical of P, PL () := P() L and but the permutation
w WP /WPL () is any coset representative such that l1 1 i1 , 2 i2 , , r ir
BL wPL ().
and wI (r + 1), . . . , wI (n) are the elements in {1, . . . , n}\I ar-
Let = (1 , . . . , s ) +s and let L() denote the G- ranged in ascending order.
linearised line bundle L(1 ) L( s ) on (G/B) s (under Let I be the dual set
the diagonal action of G). Then, there exist unique standard I = {n + 1 i, i I}
parabolic subgroups P1 , . . . , P s such that the line bundle L()
descends as an ample line bundle L() on X() := G/P1 arranged in ascending order.
s
G/P s . We call a point x (G/B) , G-semistable (with respect Then, the Schubert class [XI := XwPIr ] is Poincar dual to
to not necessarily ample L()) if its image in X() under the the Schubert class [XI ] H (Gr(r, n), Z). Moreover,
canonical map : (G/B) s X() is semistable with respect    r(r + 1)
dim XI = codim XI = i . (8)
to the ample line bundle L(). Now, one has the following iI
2
fundamental theorem due to Klyachko [Kly] for G = SL(n),
For 0 < r < n, recall the definition of the set S nr of triples
extended to general G by Berenstein-Sjamaar [BS].
(I, J, K) of subsets of {1, . . . , n} of cardinality r from Sec-
Theorem 7. Let 1 , . . . , s + . Then, the following are tion 2. The following theorem follows from Theorem 7 for
equivalent: G = SL(n) (proved by Klyachko) and Theorem 9 (proved by
(a) (1 , . . . , s ) s (G). Knutson-Tao). Belkale [B3 ] gave another geometric proof of
(b) For every standard maximal parabolic subgroup P and ev- the theorem.
ery Weyl group elements w1 , . . . , w s W P W/WP such
Theorem 10. For subsets (I, J, K) of {1, . . . , n} of cardinality
that
r,
[XwP1 ] . . . [XwPs ] = d[XeP ], for some d  0, (7) [XI ] [X J ] [XK ] = d[XePr ], for some d  0
the following inequality is satisfied: (I, J, K) S nr .

s For an Hermitian n n matrix A, let A = (1 n )
P
I(w 1 ,...,w s )
: j (w j xP ) 0, be its set of eigenvalues (which are all real). Let a be the stan-
j=1 dard Cartan subalgebra of sl(n) consisting of traceless diago-
where iP is the unique simple root not in the Levi of P nal matrices and let b sl(n) be the standard Borel subalgebra
and xP := xiP . consisting of traceless upper triangular matrices (where sl(n)
is the Lie algebra of SL(n) consisting of traceless n n matri-
The equivalence of (a) and (b) in Theorem 2 follows easily by ces). Then, the Weyl chamber is:
combining Theorems 7 and 5.   
a+ = diag (e1 en ) : ei = 0 .
Remark 8. As proved by Belkale [B1 ] for G = SL(n) and ex- Define the Hermitian eigencone:
tended for an arbitrary G by Kapovich-Leeb-Millson [KLM], 
Theorem 7 remains true if we replace d by 1 in the identity
(n) = (a1 , a2 , a3 ) a3+ :
(7). A much sharper (and optimal) result for an arbitrary G is there exist n n Hermitian matrices A, B, C
obtained in Theorem 18. 
with A = a1 , B = a2 , C = a3 and A + B = C .
It is easy to see that (n) essentially coincides with the eigen-
4 Specialisation of Results to G = SL(n):
Horn Inequalities cone 3 (sl(n)). Specifically,
 
(a1 , a2 , a3 ) (n) (a1 , a2 , a ) 3 sl(n) ,
3
We first need to recall the Knutson-Tao saturation theorem
[KT], conjectured by Zelevinsky [Z]. Other proofs of their where (e1 en ) := (en e1 ).
result are given by Derksen-Weyman [DK], Belkale [B3 ] and Combining Theorems 7 and 5 for sl(n) with Theorem 10,
Kapovich-Millson [KM2 ]. we get the following Horn conjecture [Ho], established by the
works of Klyachko (equivalence of (a) and (b) in Theorem 2
Theorem 9. Let G = SL(n) and let (1 , . . . , s ) s (G) be for g = sl(n)) and Knutson-Tao (Theorem 9).
such that 1 + + s belongs to the root lattice. Then,
Corollary 11. For (a1 , a2 , a3 ) a3+ , the following are equiva-
 
[1 ] [ s ] G  0. lent:

EMS Newsletter December 2015 23


Feature

(a)
(a1 , a2 , a3 ) (n). By Proposition 13, if (w1 , . . . , w s ) is L-movable, then
[XwP1 ] . . . [XwPs ] = d[XeP ] in H (G/P) for some nonzero d.
(b) For all 0 < r < n and all (I, J, K) S nr , For w W P , define the character w by
|a3 (K)| |a1 (I)| + |a2 (J)|, 
w = , (12)
where, for a subset I = (i1 < < ir ) {1, . . . , n} (R+ \R+l )w1 R+
and a = (e1 en ) a+ , a(I) := (ei1
where R+l
is the set of positive roots of l.
eir ) and |a(I)| := ei1 + + eir .
Then, from [K1 , 1.3.22.3],
We have the following representation theory analogue of w = 2L + w1 , (13)
the above corollary, obtained by combining Theorems 7, 9 + L
where is half the sum of roots in R and is half the sum
and 10.
of roots in R+l .
Corollary 12. Let = (1 n 0), = (1
Proposition 15. Assume that (w1 , . . . , w s ) (W P ) s satisfies
n 0) and = (1 n 0) be three partitions such
equation (11). Then, the following are equivalent:
that || + || || nZ, where || := 1 + + n . Then, the
(a) (w1 , . . . , w s ) is L-movable.
following are equivalent:
(b) [XwP1 ] . . . [XwPs ] = d[XeP ] in H (G/P) for some nonzero d
(a) [] appears as an SL(n)-submodule of [] [].
and, for each i \ (P), we have
(b) For all 0 < r < n and all (I, J, K) S nr ,  
s 
r 
|(K)| |(I)| + |(J)| (|| + || ||), w j 1 (xi ) = 0.
n j=1
where, for a subset I = (i1 < < ir ) {1, . . . , n}, (I)
denotes (i1 ir ) and |(I)| := i1 + + ir . Corollary 16. For any u, v, w W P such that cwu,v  0 (see
Equation (5)), we have
5 Deformed Product (w u v )(xi ) 0, for each i \ (P). (14)
The above corollary, together with the identity (13), justi-
This section is based on the work [BK1 ] of Belkale-Kumar.
fies the definition of the deformed product 0 given in Section
Consider the shifted Bruhat cell:
2. This deformed product is used in determining the facets
wP := w1 BwP G/P. (codimension 1 faces) of s (g).
Let T P = T (G/P)e be the tangent space of G/P at e G/P. It Lemma 17. Let P be a cominuscule maximal standard para-
carries a canonical action of P. For w W P , define T wP to be bolic subgroup of G (i.e. the unique simple root P \ (P)
the tangent space of wP at e. We shall abbreviate T P and T wP appears with coefficient 1 in the highest root of R+ ). Then, the
by T and T w respectively when the reference to P is clear. It product coincides with the cup product in H (G/P).
is easy to see that BL stabilises wP keeping e fixed. Thus,
BL T w T w . (9) 6 Efficient determination of the eigencone
The following result follows from the Kleiman transversality
theorem by observing that gwP passes through e gwP = This section is again based on the work [BK1 ] of Belkale-
pwP for some p P. Kumar. The following theorem [BK1 , Theorem 22] deter-
mines the saturated tensor semigroup s (G) efficiently.
Proposition 13. Take any (w1 , . . . , w s ) (W P ) s such that Specifically, as proved by Ressayre (see Theorem 23), the set
s
of inequalities given by (b) of the following theorem is an
codim wP j dim G/P. (10) irredundant set of inequalities determining s (G).
j=1
For G = SL(n), each maximal parabolic P is cominuscule
Then, the following three conditions are equivalent: and, hence, by Lemma 17, 0 coincides with the standard cup
(a) [XwP1 ] . . . [XwPs ]  0 H (G/P). product in H (G/P). Thus, the following theorem in this case
(b) For general (p1 , . . . , p s ) P s , the intersection reduces to Theorem 7 with d = 1 in the identity (7).
p1 wP 1 p s wP s is transverse at e. It may be mentioned that replacing the product 0 in (b) of
(c) For general (p1 , . . . , p s ) P s , the following theorem by the standard cup product (i.e. Theo-

s rem 7 with d = 1 in the identity (7) see Remark 8), we get, in
dim(p1 T w1 p s T ws ) = dim G/P codim wP j . general, far more inequalities for simple groups other than
j=1 SL(n). For example, for G of type B3 (or C3 ), Theorem 7 with
The set of s-tuples in (b) as well as (c) is an open subset of d = 1 gives rise to 126 inequalities, whereas the following
Ps . theorem gives only 93 inequalities (see [KuLM]).

Definition 14. Let w1 , . . . , w s W P be such that Theorem 18. Let G be a connected semisimple group and let
 s (1 , . . . , s ) +s . Then, the following are equivalent:
codim wP j = dim G/P. (11) (a) = (1 , . . . , s ) s (G).
j=1 (b) For every standard maximal parabolic subgroup P in G
We then call the s-tuple (w1 , . . . , w s ) Levi-movable (or L- and every choice of s-tuples (w1 , . . . , w s ) (W P ) s such
movable) if, for general (l1 , . . . , l s ) L s , the intersection that
 
l1 wP 1 l s wP s is transverse at e. [XwP1 ] 0 0 [XwPs ] = [XeP ] H (G/P, Z), 0 ,

24 EMS Newsletter December 2015


Feature

the following inequality holds: Theorem 21. (i) The intersection sj=1 g j Bw j P G/P is

s the singleton { f P}.

P
I(w : j (w j xP ) 0, (ii) For the simple root iP \ (P), sj=1 j (w j xiP ) > 0.
1 ,...,w s )
j=1
The equivalence of (a) and (d) in Theorem 2 follows easily
where iP is the (unique) simple root in \(P) and xP := by combining Theorems 18 and 5.
xi P .
Remark 22. The cone 3 (g) h3+ is quite explicitly deter-
We briefly recall some of the main ingredients of the proof mined: for any simple g of rank 2 in [KLM, 7]; for any sim-
of the above theorem that are of independent interest. ple g of rank 3 in [KuLM]; and for g = so(8) in [KKM]. It has:
12 (6+6); 18 (9+9); 30 (15+15); 41 (10+21+10); 93 (18+48+
Definition 19. (Maximally destabilising one parameter sub- 27); 93 (18+48+27); 294 (36+186+36+36); 1290 (126+519+
groups.) Let X be a projective variety with the action of a 519+126); and 26661 (348+1662+4857+14589+4857+348)
connected reductive group S and let L be an S -linearised am- facets inside h3+ (intersecting the interior of h3+ ) for g of type
ple line bundle on X. Introduce the set M(S ) of fractional A2 , B2, G2 , A3 , B3, C3 , D4 , F4 and E6 respectively. The nota-
OPSs in S . This is the set consisting of the ordered pairs tion 30 (15+15) means that there are 15 (irredundant) inequal-
(, a), where O(S ) and a Z>0 , modulo the equiva- ities coming from G/P1 and there are 15 inequalities coming
lence relation (, a) (, b) if b = a . The equivalence class from G/P2 . (The indexing convention is as in [Bo, Planche
of (, a) is denoted by [, a]. An OPS of S can be thought IIX].)
of as the element [, 1] M(S ). The group S acts on M(S )
via conjugation: g [, a] = [gg1, a]. Choose an S -invariant The following result is due to Ressayre [R]. In the case
norm q : M(S ) R+ , where norm means that q|M(H) is the G = SL(n), the result was earlier proved by Knutson-Tao-
square root of a positive definite quadratic form on the Q- Woodward [KTW].
vector space M(H) for a maximal torus H of S . We can extend
Theorem 23. Let s 3. The set of inequalities provided by
the definition of L (x, ) to any element = [, a] M(S ) and
L
= (x,) . (b) of Theorem 18 is an irredundant system of inequalities de-
x X by setting L (x, ) a scribing the cone s (G)R generated by s (G) inside + (R) s ,
For any unstable (i.e. nonsemistable) point x X, define P
  i.e. the hyperplanes given by the equality in I(w 1 ,...,w s )
are pre-
q (x) = inf q() | L (x, )
1 cisely those facets of the cone s (G)R which intersect the in-

M(S ) terior of + (R) s , where + (R) := { h : (i ) R+ i }.
and the optimal class By Theorem 5, the same result is true for the cone s (g).

(x) = { M(S ) | L (x, ) 1, q() = q (x)}. Let g be a simple simply-laced Lie algebra and let :
g g be a diagram automorphism with fixed subalgebra k
Any (x) is called a Kempfs OPS associated to x. (which is necessarily a simple Lie algebra again). Let b be a
By a theorem of Kempf (see [Ki, Lemma 12.13]), (x) is Borel subalgebra and h be a Cartan subalgebra of g such that
nonempty and the parabolic P() := P() (for = [, a]) does
they are stable under . Then, bk := b is a Borel subalgebra
not depend upon the choice of (x). The parabolic P()
and hk := h is a Cartan subalgebra of k. Let h+ and hk+ be the

for (x) will be denoted by P(x) and called the Kempfs dominant chambers in h and hk respectively. Then,
parabolic associated to the unstable point x. Moreover, (x)
is a single conjugacy class under P(x). hk+ = h+ k.
We have the following result originally conjectured by
We recall the following theorem due to RamananRama- Belkale-Kumar [BK2 ] and proved by Belkale-Kumar [BK2 ],
nathan [RR, Proposition 1.9]. Braley [Bra] and Lee [Le] (case by case).
Theorem 20. For any unstable point x X and = [, a] Theorem 24. For any s 1,
(x), let
xo = limt0 (t) x X. s (k) = s (g) (hk+ ) s .

Then, xo is unstable and (xo ). 7 Saturation Problem


Indication of the Proof of Theorem 18: The implication
(a) (b) of Theorem 18 is, of course, a special case of The- In Section 2, we defined the saturated tensor semigroup s (G)
orem 7. (for any integer s 1) and determined it by describing its
To prove the implication (b) (a) in Theorem 18, we facets (see Theorems 18 and 23).
need to recall the following result due to Kapovich-Leeb- Define the tensor semigroup for G:
 
Millson [KLM]. Suppose that x = (g 1 , . . . , g s) (G/B) s is an s (G) = (1 , . . . , s ) +s : ([1 ] [ s ])G  0 .
unstable point and P(x) the Kempfs parabolic associated to
It is indeed a semigroup by [K2 , Lemma 3.9]. The saturation
(x), where : (G/B) s X() is the map defined above The-
problem aims to compare these two semigroups. We recall the
orem 7. Let = [, a] be a Kempfs OPS associated to (x).
following result (see [Br, Theorem 2.1]).
Express (t) = f (t) f 1 , where h+ . Then, the Kempfs
parabolic P() is a standard parabolic. Define w j W/WP() Lemma 25. There exists a uniform integer d > 0 (depend-
by f P() g j Bw j P() for j = 1, . . . , s. Let P be a maximal ing only upon s and G) such that, for any = (1 , . . . , s )
parabolic containing P(). s (G), d = (d1 , . . . , d s ) s (G).

EMS Newsletter December 2015 25


Feature

We now begin with the following definition. We take s = 3 Bibliography


as this is the most relevant case to the tensor product decom-
position. [B1 ] P. Belkale, Local systems on P1 S for S a finite set, Com-
positio Math. 129 (2001), 6786.
[B2 ] P. Belkale, Invariant theory of GL(n) and intersection the-
Definition 26. An integer d 1 is called a saturation factor ory of Grassmannians, IMRN 2004, no. 69, 37093721.
for G if, for any (, , ) 3 (G) such that + + Q, [B3 ] P. Belkale, Geometric proofs of Horn and saturation con-
we have (d, d, d) 3 (G), where Q is the root lattice of jectures, J. Alg. Geom. 15 (2006), 133173.
G. Of course, if d is a saturation factor then so is any of its [BK1 ] P. Belkale and S. Kumar, Eigenvalue problem and a new
product in cohomology of flag varieties, Inventiones Math.
multiples. If d = 1 is a saturation factor for G, we say that the
166 (2006), 185228.
saturation property holds for G. [BK2 ] P. Belkale and S. Kumar, Eigencone, saturation and Horn
problems for symplectic and odd orthogonal groups, J. of
The saturation theorem of Knutson-Tao (see Theorem 9) Algebraic Geom. 19 (2010), 199242.
asserts that the saturation property holds for G = SL(n). [BS] A. Berenstein and R. Sjamaar, Coadjoint orbits, mo-
ment polytopes, and the Hilbert-Mumford criterion, Journ.
The following general result (though not optimal) on sat-
Amer. Math. Soc. 13 (2000), 433466.
uration factors is obtained by Kapovich-Millson [KM2 ] by [Bo] N. Bourbaki, Groupes et algbres de Lie, Chap. 46, Mas-
using the geometry of geodesics in Euclidean buildings and son, Paris, 1981.
Littelmanns path model. A weaker form of the following the- [Bra] E. Braley, Eigencone Problems for Odd and Even Orthog-
orem was conjectured by Kumar in a private communication onal Groups, PhD Thesis (under the supervision of P. Bel-
to J. Millson (see also [KT, 7, Conjecture]). kale), University of North Carolina, 2012.
[Br] M. Brion, Restriction de represntations et projec-
Theorem 27. For any connected simple G, d = kg2 is a sat- tions dorbites coadjointes (daprs Belkale, Kumar et
uration factor, where kg is the least common multiple of the Ressayre), Sminaire Bourbaki, 64me anne no 1043,
coefficients of the highest root of the Lie algebra g of G 20112012.
written in terms of the simple roots {1 , . . . , }. [DW] H. Derksen and J. Weyman, Semi-invariants of quivers
Observe that the value of kg is 1 for g of type A ( 1); and saturation for Littlewood-Richardson coefficients, J.
Amer. Math. Soc. 13 (2000), 467479.
2 for g of type B ( 2), C ( 3), D ( 4); and 6, 12, 60,
[E] A. G. Elashvili, Invariant algebras, In: Lie Groups, their
12, 6 for g of type E6 , E7 , E8 , F4 , G2 respectively. Discrete Subgroups, and Invariant Theory (ed. E. B. Vin-
Kapovich-Millson determined 3 (G) explicitly for G = berg), Advances in Soviet Math. 8, Amer. Math. Soc.,
Providence, 1992, 5764.
Sp(4) and G2 (see [KM1 , Theorems 5.3, 6.1]). In particular,
[Fa] K. Fan, On a theorem of Weyl concerning eigenvalues of
from their description, the following theorem follows. linear transformations, Proc. Natl. Acad. Sci. USA, 35
Theorem 28. The saturation property does not hold for either (1949), 652655.
G = Sp(4) or G2 . Moreover, 2 is a saturation factor (and no [F] W. Fulton, Eigenvalues, invariant factors, highest
weights, and Schubert calculus, Bull. Amer. Math. Soc.
odd integer d is a saturation factor) for Sp(4), whereas both (N.S.) 37 (2000), 209249.
2 and 3 are saturation factors for G2 (and hence any integer [HS] J. Hong and L. Shen, Tensor invariants, saturation prob-
d > 1 is a saturation factor for G2 ). lems, and Dynkin automorphisms, Preprint (2014).
[Ho] A. Horn, Eigenvalues of sums of Hermitian matrices, Pa-
It was known earlier that the saturation property fails for
cific J. Math. 12 (1962), 225241.
G of type B (see [E]). [KKM] M. Kapovich, S. Kumar and J. J. Millson, The eigencone
Kapovich-Millson [KM1 ] made the following very inter- and saturation for Spin(8), Pure and Applied Math. Quar-
esting conjecture: terly 5 (2009), 755780.
[KLM] M. Kapovich, B. Leeb and J. J. Millson, Convex functions
Conjecture 29. If G is simply-laced then the saturation prop- on symmetric spaces, side lengths of polygons and the sta-
erty holds for G. bility inequalities for weighted configurations at infinity,
Journal of Differential Geometry 81 (2009), 297354.
Apart from G = SL(n), the only other simply-connected,
[KM1 ] M. Kapovich and J. J. Millson, Structure of the tensor
simple, simply-laced group G for which the above conjecture product semigroup, Asian J. of Math. 10 (2006), 493540.
is known so far is G = Spin(8), proved by Kapovich-Kumar- [KM2 ] M. Kapovich and J. J. Millson, A path model for geodesics
Millson [KKM, Theorem 5.3] by explicit calculation using in Euclidean buildings and its applications to representa-
Theorem 18. tion theory, Groups, Geometry and Dynamics 2 (2008),
Finally, we have the following improvement of Theo- 405480.
rem 27 for the classical groups SO(n) and Sp(2). It was [Ki] F. Kirwan, Cohomology of Quotients in Symplectic and
proved by Belkale-Kumar [BK2 , Theorems 25 and 26] for Algebraic Geometry, Princeton University Press, 1984.
[Kly] A. Klyachko, Stable bundles, representation theory and
the groups SO(2 + 1) and Sp(2) by using geometric tech-
Hermitian operators, Selecta Mathematica 4 (1998), 419
niques. Sam [S] proved it for SO(2) (and also for SO(2 + 1) 445.
and Sp(2)) via the quiver approach following the proof by [KT] A. Knutson and T. Tao, The honeycomb model of GLn (C)
Derksen-Weyman [DW] for G = SL(n). Further, it has been tensor products I: Proof of the saturation conjecture, J.
shown by Hong-Shen [HS] that the spin group Spin(2 + 1) Amer. Math. Soc. 12 (1999), 10551090.
has saturation factor 2. [KTW] A. Knutson, T. Tao and C. Woodward, The honeycomb
model of GLn (C) tensor products II: Puzzles determine
Theorem 30. For the groups SO(n) (n 7), Spin(2 + 1) and facets of the Littlewood-Richardson cone, J. Amer. Math.
Sp(2) ( 2), 2 is a saturation factor. Soc. 17 (2004), 1948.

26 EMS Newsletter December 2015


History

[K1 ] S. Kumar, Kac-Moody Groups, their Flag Varieties and [Sj] R. Sjamaar, Convexity properties of the moment mapping
Representation Theory, Progress in Mathematics, vol. re-examined, Adv. Math. 138 (1998), 4691.
204, Birkhuser, 2002. [S] S. Sam, Symmetric quivers, invariant theory, and satura-
[K2 ] S. Kumar, Tensor product decomposition, Proc. of the In- tion theorems for the classical groups, Adv. Math. 229
ternational Congress of Mathematicians, Hyderabad (In- (2012), 11041135.
dia), (2010), 12261261. [TF] R. C. Thompson and L. Freede, On the eigenvalues of
[K3 ] S. Kumar, A Survey of the Additive Eigenvalue Problem sums of Hermitian matrices, Linear Algebra Appl. 4
(with Appendix by M. Kapovich), Transformation Groups (1971), 369376.
19 (2014), 10511148. [W] H. Weyl, Das asymptotische Verteilungsgesetz der Eigen-
[KuLM] S. Kumar, B. Leeb and J. J. Millson, The generalized tri- werte linearer partieller Differentialgleichungen, Math.
angle inequalities for rank 3 symmetric spaces of noncom- Annalen 71 (1912), 441479.
pact type, Contemp. Math. 332 (2003), 171195. [Wi] H. Wielandt, An extremum property of sums of eigenval-
[Le] B. Lee, A Comparison of Eigencones Under Certain Dia- ues, Proc. Amer. Math. Soc. 6 (1955), 106110.
gram Automorphisms, PhD Thesis (under the supervision [Z] A. Zelevinsky, Littlewood-Richardson semigroups, In:
of S. Kumar), University of North Carolina, 2012. New Perspectives in Algebraic Combinatorics, MSRI
[Li] B.V. Lidskii, The proper values of the sum and product of Publ. 38 (1999), Cambridge Univ. Press, Cambridge, 337
symmetric matrices, Dokl. Acad. Nauk SSSR 74 (1950), 345.
769772.
[N] L. Ness, A stratification of the null cone via the moment
map (with an appendix by D. Mumford), Amer. J. Math. Shrawan Kumar [shrawan@email.unc.edu]
106 (1984), 12811329.
is the John R. and Louise S. Parker Dis-
[RR] S. Ramanan and A. Ramanathan, Some remarks on the
instability flag, Thoku Math. J. 36 (1984), 269291. tinguished Professor at the University of
[R] N. Ressayre, Geometric invariant theory and the gener- North Carolina, Chapel Hill, USA. He was
alized eigenvalue problem, Inventiones Math. 180 (2010), an invited speaker at ICM 2010. He is a
389441. Fellow of the American Mathematical So-
ciety.

George Boole and


George Boole andBoolean
BooleanAlgebra
Algebra
Stanley
Stanley Burris
Burris(University
(Universityofof
Waterloo, Waterloo,
Waterloo, Ontario,
Ontario, Canada)Canada)

George Boole (18151864) was responsible, in the years on which are founded the Mathematical Theories of Logic
18471854, for initiating the revolution in the subject of logic and Probabilities. We will refer to this book as LT. The first
by creating an algebra of logic for classes. This is all the more two-thirds of LT are on Booles algebra of logic for classes
remarkable because Boole was largely self-educated in math- (which will henceforth be called Booles algebra") and the
ematics (and several languages), having had to give up attend- last third on applications of this algebra to probability theory.
ing school at the age of 16 to start his career as a schoolteacher This article is only concerned with the logic portion of LT.
to provide financial support for his parents and siblings. He Booles algebra had equational laws, rules of inference for
started publishing mathematical papers, mainly on analysis, equational reasoning and a powerful Rule of 0 and 1, which
in 1841. Three years later, in 1844, at the age of 29, he won has only recently been deciphered. Indeed, it is remarkable
the first gold medal awarded in mathematics by the Royal So- how long it has taken to properly understand Booles algebra
ciety. the breakthrough came in 1976 with the publication of [15]
Boole struck up a friendly correspondence with Augustus by Theodore Hailperin (19162014).1
De Morgan (18061871). Subsequently, De Morgans noisy
feud (over a rather trivial matter in logic) with the respected 1 Using Ordinary Algebra
philosopher Sir William Hamilton (17881856) of Edinburgh
inspired Boole to write a booklet [2] in 1847 applying algebra One of the distinguishing features of Booles algebra of logic
to logic. In 1849, at the age of 34, Boole left school-teaching is the extent to which it looks like ordinary algebra. In-
in Lincolnshire, England, for a professorship at Queens Col- deed, this fact is beautifully summarised in the following two
lege in Cork, Ireland. For fascinating details on Booles life, quotes:
see the excellent biography [23] by Desmond MacHale
including such remarkable details as the fact that Booles That the symbolic processes of algebra, invented as tools of nu-
youngest daughter Ethel Lilian (18641960) wrote a novel merical calculation, should be competent to express every act of
called The Gadfly which essentially became the bible of the
Russian revolution.
1 For a modern introduction to Booles algebra of logic, see the authors
Although Boole was primarily an algebraist and analyst,
article [7] on Boole in the online Stanford Encyclopedia of Philosophy,
today he is best known for his work in logic, in particular as well as the texts [8], [9] of two recent talks by the author on Booles
for his 1854 book An Investigation of the Laws of Thought algebra of logic.

EMS Newsletter December 2015 27


History

thought, and to furnish the grammar and dictionary of an all- easily derives A B = 0, that is, A and B must be disjoint. This
containing system of logic, would not have been believed until it tells us that (1) addition must be a partial operation, and (2)
was proved. (Augustus De Morgan, in A Budget of Paradoxes.) the law x2 = x cannot apply to terms in general (in particular,
The algebra which Boole himself used was simply ordinary nu- not to x + y). An easy argument shows that if A + B is defined,
merical algebra as applied to a collection of quantities each it must be A B. Likewise, if A B is defined then one has
of which was assumed to be subject to the quadratic equation B A and A B = A  B.
x(1 x) = 0, and Boole showed how this hypothesis could be The price Boole paid for being so close to ordinary nu-
applied to the solution of many logical problems. (C. S. Peirce, merical algebra is that his models were partial algebras. This
from the Nachlass Notes on the list of postulates of Dr Hunting- might have discouraged most from continuing but not Boole.
tons Section 2, p. 264 in Vol. IV of Peirces collected papers.) In Chapter V of LT he stated his Principles of Symbol-
The equational theory of commutative rings with unity, with ical Reasoning, which essentially said that one can carry
a conveniently chosen axiom set CR1 , is a good place to out equational reasoning in his system as though the mod-
start when defining ordinary numerical algebra. This is of- els were total algebras.3 These Principles are in general false
ten thought of as the algebra of polynomials. The modern but, thanks to Hailperins work, we know they hold in Booles
way to build an algebra of logic on CR1 is to add the idem- system. Booles Principles allowed him to claim that an equa-
potent law (x)(x2 = x), giving the set BR of axioms for tional argument 1 (x), . . . , k (x) (x), with totally defined
Boolean rings. The relevant models of BR are the algebras equational premises and conclusion, could be justified by an
P(U), +, , , 0, 1, where P(U) is the power set of the uni- equational derivation that involved (partially) uninterpretable
verse U; addition and subtraction are symmetric difference, terms. He said it was just like using the uninterpretable 1
multiplication is intersection, 0 is the empty set and 1 is the to derive trigonometric identities.
universe. But this is, indeed, a modern approach and it did not
appear until the 20th century, most notably in the 1936 paper 2 Booles Rule of 0 and 1
[31] of Marshall Stone (19031989).
Booles approach was closer to ordinary numerical al- After introducing his laws, rules of inference and partial alge-
gebra than Boolean rings. He (implicitly) added the quasi- bra models, Boole stated a remarkable foundational principle
identities (x)(nx = 0 x = 0), for n = 1, 2, . . ., to CR1 , which we will call his Rule of 0 and 1 (LT, p. 37):
giving the set of axioms that will be called NCR1 .2 These ax- Let us conceive, then, of an Algebra in which the symbols x,
ioms are indeed true of the ordinary number systems, e.g. the y, z, etc. admit indifferently of the values 0 and 1, and of these
integers Z. values alone. The laws, the axioms, and the processes, of such
Boole also added one non-numerical law, namely x2 = x. an Algebra will be identical in their whole extent with the laws,
This was an unusual law in that it only applied to variables, the axioms, and the processes of an Algebra of Logic. Difference
not to compound terms like x + y. of interpretation will alone divide them. Upon this principle the
method of the following work is established.
Booles models P(U) := P(U), +, , , 0, 1 were given
by the definitions This foundation has been essentially ignored until the last
A B := A B, decade. What Boole meant is that an equational argument
 1 (x), . . . , k (x) (x) is correct in his system if and only
A B if A B =
A + B := if when the variables are restricted to the values 0 and 1
undefined otherwise,
 then it holds in the integers Z. We write this in modern no-
A  B if B A tation as follows, where Idemp(x) says the variables in the
A B :=
undefined otherwise, list x := x1 , . . . , xm are idempotent:
 
0 := , NCR1  Idemp(x) 1 (x) k (x) (x)
1 := U. iff
Since addition and subtraction were only partially defined, his  
Z |= Idemp(x) 1 (x) k (x) (x)
models were partial algebras. Booles partial algebras can be
seen to arise naturally as follows. His definition of multipli- iff  
cation as intersection was a consequence of using composi- Z |= 1 () k () () ,

tion of selection operators to determine multiplication here
a selection operator such as Red chooses the red elements where is a string of 0s and 1s of the same length as the list of
in a class. (The use of composition of selection operators is variables x. It would be nearly a century after the publication
clearly stated in his discussion of multiplication on p. 165 of of LT before the corresponding result for Boolean rings was
LT.) The definition of multiplication gave Boole his idempo- known, namely:
tent law A2 = A. Then, the numerical equations x 0 = 0 and BR  1 (x) k (x) (x)
x 1 = x led to his definitions of 0 and 1 as stated above.
iff
Now, suppose A + B is defined; then, it must satisfy (A +
Z2 |= 1 (x) k (x) (x),
B)2 = A+B. From NCR1 and the fact that A2 = A, B2 = B, one

3 To justify equational arguments 1 (x), . . . , k (x) (x) using purely


2 The N in NCR1 is supposed to suggest no additively nilpotent elements,
equational reasoning, one needs to treat the laws (x)(nx = 0 x = 0)
to use Hailperins terminology. Some might prefer to say that the addi- nx = 0
tive group is torsion-free". as equational rules of inference .
x=0

28 EMS Newsletter December 2015


History

where Z2 is the two-element Boolean ring. The same result Within a decade of the publication of LT, the move was
holds if we change BR to BA, a set of equational axioms for underway to replace Booles algebra with the modern ver-
Boolean algebras, and Z2 to 2BA , the two-element Boolean sion. Starting with the 1864 book Pure Logic by William
algebra. Stanley Jevons (18351882), there followed the 1867 paper
[24] of Charles S. Peirce (18391914), the 1872 book Die
3 Booles Four Main Theorems Formenlehre oder Mathematik by Robert Grassmann (1815
1901), the 1877 Operationskreis des Logikkalkuls of Ernst
This brief overview of Booles work will conclude with his Schrder (18411902) and the 1880 paper [25] of Peirce.
four main theorems. But first we need Booles notion of a Schrders Operationskreis was the first publication to absorb
constituent C (x), where is a string of 0s and 1s of the all of Booles main theorems into the modern setting.
same length as the string x of variables. A simple example Booles translations of propositions into equations also
will suffice: C101 (x1 , x2 , x3 ) := x1 (1 x2 )x3 . came under fire, and were replaced in the 1890s by equa-
Booles four main theorems (expressed in modern termi- tions and negated equations in Schrders monumental three-
nology) are: volume Algebra der Logik. This was possibly the last time

1. p(x) = p()C (x) (EXPANSION). anyone seriously read Booles LT as a fundamental source of

2. p1 (x) = = pk (x) = 0 iff 0 = i pi (x)2 (REDUC- information on the algebra of logic for classes; future research
TION). in this subject, aside from historical studies, used the modern

3. (x)p(x, y) = 0 iff 0 = p(, y) (ELIMINATION). version. However, as with many new subjects introduced into
4. q(x) y = p(x) iff mathematics, it took decades before mathematicians could
p(x) (p(x) q(x)) = 0 agree on standard nomenclature, symbols and axioms for the
modern version. The 1904 paper [17] of the Harvard mathe-
and matician E. V. Huntington (18741952) was one of the most
famous papers devoted to describing axioms for the modern
 
(v) y = C (y) + v C (y) , algebra of logic for classes it provided three sets of axioms.
J1 J2 In 1933, Huntington [18] returned to this topic, offering three
for Ji suitably determined (see [7]) by the equalities and more sets of axioms.
inequalities holding within the triples (p(), q(), 0) (SO-
LUTION). 5 Harvard and the name Boolean Algebra"
All of these results can be readily translated into theorems of
Boolean rings as well as of Boolean algebras, the latter having The name of the modern version fluctuated for decades, usu-
been carried out in 1877 by Schrder in [27], where the fourth ally being called the algebra of logic, and sometimes the
theorem has a considerably simpler statement. calculus of logic. Peirce was the only one who occasionally
In the following, for any set of axioms, let + be named the subject after Boole, calling it Boolian algebra or
{0  1}. Booles theorems, as stated above, are easily Boolian calculus. At times he referred to those who used
proved by strengthening Booles Rule of 0 and 1 to Horn sen- equational logic to develop an algebra of logic for classes,
tences : such as Jevons and Schrder, as Boolians. In 1898, Peirce
NCR+1  |Idemp iff Z |= |Idemp , gave a series of invited lectures at Harvard titled Reason-
ing and the Logic of Things". These lectures so inspired the
where |Idemp means the variables of are relativised to Harvard philosophy Professor Josiah Royce (18551916) that
idempotent elements (see [10]). The corresponding result for he decided, mid-career, to embark on a crash course to learn
Boolean rings is mathematics and modern logic. Royces PhD students who
BR+  iff Z2 |= . benefited from this included Henry M. Sheffer (18821964;
For Boolean algebras, one has the parallel result PhD 1908), C. I. Lewis (18831964; PhD 1910) and Norbert
Wiener (18941964; PhD 1913). In 1913, both Royce and
BA+  iff 2BA |= , Sheffer published papers [26], [30] using the phrase Boolean
stated in 1967 by Fred Galvin [13]. algebra" for the modern version of the subject. Wiener [36]
added a paper on Boolean algebra" in 1917. Sheffers paper
4 The Reaction to Booles Algebra of Logic became the more famous and it used Boolean algebra in the
title this was the paper with the Sheffer stroke as the sole
Booles early successors appreciated his main results but, with operation. Bertrand Russell (18721970) was so impressed
the exception of John Venn (18341923), they found Booles with Sheffers stroke that it was used in the second edition of
use of the algebra of numbers with partial algebra models at Principia Mathematica. The name Boolean algebra" for the
best unattractive, at worst completely unacceptable. Booles modern version became ever more standard; in 1933, Hunt-
work was so mysterious that the first attempt at a substantial ington [18] would credit Sheffer with introducing this name.
review [33] of LT did not occur until 1876, 22 years after the Another Harvard mathematician Marshall Stone (18031989)
publication of LT; it was written by Venn. It provided little would also use the name Boolean algebra" in his famous pa-
insight into why Booles algebra actually gives correct results pers of the 1930s. Up till 1940, the name Boolean algebra"
and would not be improved upon for the next 100 years. Fi- was used mainly in papers published in American journals;
nally, in 1976, Hailperin [15] showed that characteristic func- but by 1950, it was the worldwide standard for the modern
tions were the key to justifying Booles algebra. version of the algebra of logic for classes.

EMS Newsletter December 2015 29


History

6 Booles Translations circuits, to be used in the building of the Mark IV computer


the results were published in 1951 in [1]. He said that there
In order to apply his algebra of classes to Aristotelian logic, were those who found the language of propositional logic a
Boole first needed a translation of categorical propositions good choice, as well as those who liked the use of Boolean
into equations. Recall that there are four kinds of categorical algebra made by Claude Shannon (19162001) in his famous
propositions: (A) All x is y; (E) No x is y; (I) Some x is y; and 1938 paper [29]; but Aiken decided to use Booles arithmeti-
(O) Some x is not y. The following gives Booles translations cal algebra on the grounds that it would be more comfort-
of 1847 and 1854, followed by a recommended revision: able working with a familiar algebra. The first examples in
1847 1854 (see [11]) the book are the switching function for a triode tube being
t(x) = 1 x and the switching function for a pentode tube be-
(A) x = xy x = vy x = xy ing p(x, y) = 1 xy. So it seems that, even in the mid 1950s,
(E) xy = 0 x = v(1 y) xy = 0
there were people who knew that Booles algebra was based
(I) v = xy vx = vy v = vxy
on the algebra of numbers and was not to be confused with
(O) v = x(1 y) vx = v(1 y) v = vx(1 y)
modern Boolean algebra. However, it is likely that the major-
Boole viewed the syllogisms as simple applications of the ity of modern mathematicians and computer scientists believe
elimination theorem. Traditional logicians of the Aristotelian that Booles algebra is Boolean algebra. Hailperin tried to set
school argued that Boole offered nothing new since his gen- the record straight with his 1981 publication Booles algebra
eral eliminations could be achieved by a sequence of syllo- isnt Boolean algebra but it is doubtful that it has had signifi-
gisms. Boole eventually agreed (see LT, p. 240) that syllo- cant impact.
gisms were indeed sufficient to achieve his elimination results
but pointed out that the traditionalists lacked a description of 8 Justifying Booles Algebra of Classes
how the sequence of syllogisms was to be created. Then he
added that (with the Solution Theorem) his system went be- The reasons that Booles algebra of numbers worked so well
yond what the Aristotelians had achieved. in his algebra of logic remained a complete mystery until
Hailperins 1976/1986 book Booles Algebra and Probabil-
7 Was Boole misguided to have used the ity noted that Booles partial algebra P(U) was isomorphic
algebra of numbers? to the partial algebra consisting of the idempotent elements
of the ring ZU . The ring ZU satisfied NCR1 and the uninter-
Most of Booles successors found his use of ordinary numer- pretables of Booles system became interpretable as elements
ical algebra a totally unnecessary piece of baggage. However, of the subring of ZU generated by the idempotent elements.
an obvious advantage of being able to use the algebra of num- Still, some issues remained unresolved until recently, namely
bers was that its notation was standardised; practising math- the deciphering of Booles Rule of 0 and 1 (see [7], [10]) and
ematicians were expected to be quite fluent in its use. By us- a revision of Booles translations (see the table in Section 6)
ing the algebra of numbers, it is likely that Boole was able to to allow one to translate particular categorical propositions as
try out a large number of ideas and examples, quickly noting equations (see [11]). Finally, we can say that Booles algebra
which ones succeeded, thereby rapidly gaining a sense of the of logic for classes was essentially sound.
big picture. Perhaps it was because he was using a famil-
iar algebra that he was able to go so far in such short time Acknowledgements.
[2] was written in a few weeks in 1847. Those opting for Throughout 2015, University College Cork (UCC) has been
a modern algebra of classes spent decades trying to come to celebrating the bicentennial of the birth of Boole, its first and
agreement on notation and axioms; their best theorems were most famous mathematics professor. This article is the direct
the ones borrowed from Boole. result of being invited to give a talk in August 2015 A Primer
It is interesting to note that, in 1933, Hassler Whitney on Booles Algebra of Logic at the Irish Mathematical So-
(19071989) published a paper [35] showing how to convert ciety meeting in Cork. The financial support of the European
expressions in the modern Boolean algebra of classes into ex- Mathematical Society is gratefully acknowledged. Thanks are
pressions in the algebra of numbers, noting that this made the also due to Michel Schellekens of UCC for many discussions
verification of Boolean algebraic properties quite straightfor- on Booles contributions.
ward. His method was to map a class A to its characteristic
function A , for then one had A B  A + B A B , Bibliography
A B  A B and A  1 A . Unfortunately, he saw
little connection between his work and that of Boole he [1] Howard H. Aiken, Synthesis of Electronic Computing and
did not notice that the fragment of the ring ZU consisting of Control Circuits. Harvard University Press, Cambridge, Mass.
1951.
the characteristic functions was isomorphic to Booles par-
tial algebra P(U); otherwise, he would have been able to pro- [2] George Boole, The Mathematical Analysis of Logic, Being an
Essay Towards a Calculus of Deductive Reasoning, Originally
vide a justification of Booles main theorems 40 years before published in Cambridge by Macmillan, Barclay & Macmillan,
Hailperin. 1847. Reprinted in Oxford by Basil Blackwell, 1951.
A point of direct contact in the 20th century with Booles [3] , The Calculus of Logic, The Cambridge and Dublin Math-
approach came from the Harvard Computation Laboratory ematical Journal, 3 (1847), 183198.
under the direction of Howard Aiken (19001973). In 1947, [4] , An Investigation of The Laws of Thought on Which are
he set out to create a mathematics for electronic switching Founded the Mathematical Theories of Logic and Proba-

30 EMS Newsletter December 2015


History

bilities. Originally published by Macmillan, London, 1854. [29] Claude E. Shannon, A symbolic analysis of relay and switch-
Reprint by Dover, 1958. ing circuits. AIEE Transactions, Vol. 57 (1938), pp. 713723.
[5] , Selected Manuscripts on Logic and its Philosophy. eds. [30] Henry Maurice Sheffer, A set of five independent postulates for
Ivor Grattan-Guinness, Grard Bornet, Birkhuser, 1997. Boolean algebra, with application to logical constants. Trans-
[6] Frank Mark Brown, George Booles Deductive System. Notre actions of the American Mathematical Society, Vol. 14, No. 4
Dame J. Formal Logic 50 No. 3 (303330), 2009. (Oct 1913), pp. 481488.
[7] Stanley Burris, George Boole. The online Stanford Encyclope- [31] M. H. Stone, The theory of representation for Boolean al-
dia of Philosophy at http://plato.stanford.edu/entries/boole/. gebras. Transactions of the American Mathematical Society,
[8] , A primer on Booles algebra of logic. The text of an in- Vol. 40, No. 1 (Jul 1936), pp. 37111.
vited talk to the Irish Mathematical Society meeting, 27 Au- [32] , Applications of the theory of Boolean rings to general
gust 2015, in Cork, Ireland, available at http://www.math. topology. Transactions of the American Mathematical Society,
uwaterloo.ca/~snburris/. Vol. 41, No. 3 (May 1937), pp. 375481.
[9] , Justifying Booles algebra of logic. The text of an invited [33] J. Venn, Booles logical system. Mind, Vol. 1, No. 4 (Oct
talk on 28 August 2015 at the George Boole Mathematical 1876), pp. 479491.
Sciences Conference in Cork, Ireland, available at http://www. [34] , Symbolic Logic. London: Macmillan. pp. xl. 446. Second
math.uwaterloo.ca/~snburris/. ed. 1894.
[10] and H. P. Sankappanavar, The Horn Theory of Booles par- [35] Hassler Whitney, Characteristic functions and the algebra of
tial algebras. The Bulletin of Symbolic Logic 19 (2013), 97 logic. Annals of Mathematics 34 (1933), 405414.
105. [36] Norbert Wiener, Certain formal invariances in Boolean al-
[11] , Booles method I. A Modern Version. Preprint 2014, avail- gebras. Transactions of the American Mathematical Society,
able at arXiv:1404.0784. Vol. 18, No. 1 (Jan 1917), pp. 6572.
[12] , Booles Principles of Symbolical Reasoning, Preprint
2014, available from arXiv:1412.2953.
[13] Fred Galvin, Reduced products, Horn sentences, and decision Stanley Burris [snburris@math.uwaterloo.
problems. Bull. Amer. Math. Soc. 73, 5964. ca] joined the Pure Mathematics Depart-
[14] Robert Grassmann, Die Formenlehre oder Mathematik. Stet- ment at the University of Waterloo in 1968
tin, 1872. and has published over 70 research papers
[15] Theodore Hailperin, Booles Logic and Probability, Series: since then. He was a research associate of
Studies in Logic and the Foundations of Mathematics, 85,
Amsterdam, New York, Oxford: Elsevier North-Holland.
Alfred Tarski in 1971. He has carried out
1976. 2nd edition, Revised and enlarged, 1986. research in universal algebra for over two
[16] , Booles algebra isnt Boolean algebra, Mathematics Mag- decades, with a primary interest in Boolean
azine, Vol. 54, No. 4 (Sep 1981), pp. 172184. constructions and decidability, leading to the 1981 mem-
[17] E. V. Huntington, Sets of independent postulates for the al- oir Decidability and Boolean Constructions with Ralph
gebra of logic. Transactions of the American Mathematical McKenzie and the popular textbook A Course in Universal
Society, 5 (1904), 288309. Algebra with H. P. Sankappanavar. With increasing interest
[18] , New Sets of Independent Postulates for the Algebra of
in computer science, in 1996 he published Logic for Mathe-
Logic, With Special Reference to Whitehead and Russells
Principia Mathematica. Transactions of the American Mathe- matics and Computer Science. In the 1990s, he started work
matical Society, Vol. 35, No. 1 (Jan 1933), pp. 274304. on asymptotics and logic, leading to the 2001 book Number
[19] William Stanley Jevons, Pure Logic, or the Logic of Quality Theoretic Density and Logical Limit Laws. He has studied
apart from Quantity: with Remarks on Booles System and the history of logic (especially of the 19th century) for over
on the Relation of Logic and Mathematics. Edward Stanford, two decades, clarifying several aspects of Booles algebra of
London, 1864. Reprinted 1971 in Pure Logic and Other Mi- logic and writing the article George Boole in the online
nor Works, ed. by R. Adamson and H. A. Jevons, Lennox Hill
Stanford Encyclopedia of Philosophy. Being an avid fan of
Pub. & Dist. Co., NY.
[20] C. I. Lewis, A Survey of Symbolic Logic. Berkeley, University
Photoshop and of recognition of women in mathematics, he
of California Press, 1918. pp. 6 + 409. spearheaded the MAA Women of Mathematics poster project,
[21] Clarence Irving Lewis and Cooper Harold Langford, Symbolic strongly supported by Joseph Gallian, the MAA President at
Logic. The Century Philosophy Series. New York and London: the time (20072008).
The Century Co., 1932. pp. xi + 506.
[22] Alexander Macfarlane, Principles of the Algebra of Logic. Ed-
inburgh: Douglas, 1879.
[23] Desmond MacHale, George Boole, Boole Press Dublin, 1985.
The second edition appeared in 2014 from the Cork University
Press.
[24] C. S. Peirce, On an improvement in Booles calculus of logic.
Proceedings of the American Academy of Arts and Sciences,
Vol. 7 (May 1865 May 1868), pp. 249261.
[25] , On the algebra of logic. American Journal of Mathematics,
Vol. 3, No. 1 (Mar 1880), pp. 1557
[26] Josiah Royce, An extension of the algebra of logic. The Jour-
nal of Philosophy, Psychology and Scientific Methods, Vol. X.
No. 23, 6 November 1913.
[27] Ernst Schrder, Operationskreis des Logikkalkuls. Leipzig:
Teubner, 1877.
[28] , Algebra der Logik. Vols. IIII, 18901910; reprint by
Chelsea, 1966.

EMS Newsletter December 2015 31


Leonhard Euler
Mathematical Genius
in the Enlightenment
Ronald S. Calinger

Cloth $55.00

In this comprehensive and authoritative account,


Ronald Calinger connects the story of Eulers eventful
life to the astonishing achievements that place him in
the company of Archimedes, Newton, and Gauss.

Descent in
Buildings
The Princeton Companion to Applied Mathematics Bernhard Mhlherr,
Edited by Nicholas J. Higham Holger P. Petersson
& Richard M. Weiss
Mark R. Dennis, Paul Glendinning, Paul A. Martin,
Fadil Santosa & Jared Tanner, associate editors
Paper $75.00 | Cloth $165.00

Features nearly 200 entries organized thematically and written


by an international team of distinguished contributors
Presents the major ideas and branches of applied mathematics in a clear
and accessible way
Explains important mathematical concepts, methods, equations, and applications This is the third and final volume of a trilogy that
Gives a wide range of examples of mathematical modeling began with Richard Weiss The Structure of Spherical
Buildings and The Structure of Affine Buildings.
Covers continuum mechanics, dynamical systems, numerical analysis, discrete
and combinatorial mathematics, mathematical physics, and much more Annals of Mathematics Studies, 190
Phillip A. Griffiths, John N. Mather, and Elias M. Stein, Series Editors
Explores the connections between applied mathematics and other disciplines

Monumental and comprehensive, The Princeton Companion to Applied


Mathematics does a breathtaking job of conveying the richness, depth, and
vitality of todays applied mathematics. . . . An instant classic.
Classification of
Steven Strogatz, Cornell University and author of The Joy of x Pseudo-reductive
Groups
This book will be a landmark for decades ahead.
Brian Conrad & Gopal Prasad
Nick Trefethen, University of Oxford
Paper $75.00 | Cloth $165.00
The treasures [in the Princeton Companion to Applied Mathematics]
go on and on.
Lloyd N. Trefethen, SIAM Review

Cloth $99.50

The results and methods developed in Classification of


Also available:
Pseudo-reductive Groups will interest mathematicians
The Princeton Companion to Mathematics
and graduate students who work with algebraic
Edited by Timothy Gowers groups in number theory and algebraic geometry in
June Barrow-Green and Imre Leader, associate editors
positive characteristic.
Cloth $99.50
Annals of Mathematics Studies, 191
Phillip A. Griffiths, John N. Mather, and Elias M. Stein, Series Editors

See our E-Books at


press.princeton.edu
Interview

Interview with Abel Laureate


Louis Nirenberg
Martin Raussen (Aalborg University, Denmark) and Christian Skau (Norwegian University of Science and
Technology, Trondheim, Norway)

This interview took place in Oslo on 18 May 2015.

Partial differential equations (and geometry/


physics)

Firstly, we want to congratulate you on being awarded


(with Professor Nash) the Abel Prize for 2015. You will
receive the Abel Prize from His Majesty the King of
Norway in a ceremony tomorrow.
Your first important achievement in mathematics was
solving the so-called Weyl problem in your PhD thesis.
Could you tell us what the Weyl problem is about?
The problem was originally stated by Hermann Weyl.
You have a two-dimensional sphere with a metric (that
is, a way of measuring distance), and connected with the
metric is its curvature. If this curvature is positive, the
question is whether you can find a convex body in three
dimensional space with a mapping to the sphere so that
when you measure the distance in Euclidean space, it Louis Nirenberg is awarded the Abel Prize by King Harald V of Nor-
way; left: John F. Nash (photo: NTB/Scanpix)
agrees with the metric? Weyl went quite far toward solv-
ing this problem but there were some estimates missing.
My contribution was to fill in those missing estimates. ther has dimension zero or it has dimension one. But, no!
When you express the problem mathematically, it in- There are concepts of dimension of any non-negative
volves partial differential equations. The equations were number. We proved that the one-dimensional measure
so-called non-linear partial differential equations and had to be zero, so the set could not have dimension one.
the problem was proving the existence of solutions of The paper is very technical.
these equations. Much of my career has really been de-
voted to studying partial differential equations in general But it is very important in connection with the Navier
but also applying them to problems from geometry and Stokes equations?
complex analysis. I even wrote two papers with a friend Well, it is a useful result mathematically whether engi-
in economics involving partial differential equations. In neers use it or not. Aeroplanes fly whether we solve the
my mind, it is a wonderful field. A big part of the problem NavierStokes equations or not. But it is a big mathemat-
is proving that solutions exist because equations can be ical challenge to show that there are smooth solutions.
written down for which one knows there are no solutions.
Do you often think about the NavierStokes equations?
Many of these problems come from physics so solutions Once in a while. But I dont really have any fresh ideas. I
would be expected to exist? think it is up to younger people.
Yes. But, for instance, for equations in fluid dynamics (the
so-called Navier-Stokes equations that were introduced Start of a career in mathematics
150 years ago), mathematicians have not been able to
prove that smooth solutions exist for all time. So that is May we ask how your mathematical life started? We
still an open problem. were told that a certain teacher of Hebrew played an
instrumental role. Is that true?
Is it true that the best result in that direction is your My father tried to teach me Hebrew. I resisted, stupidly,
joint work with Caffarelli and Kohn from 1982? and now I know no Hebrew at all. He hired a friend to
That result is not about the existence of solutions but give me lessons in Hebrew. This friend happened to like
about the dimension of singularities if they do occur. mathematical puzzles and half the lessons then consisted
They cannot have a high dimension; for instance, they of these puzzles. I found them quite fascinating but, I
cannot fill a curve. They must fill a set of dimension less must say now, at my age, I am no longer fascinated by
than one. You may wonder what the hell that is? It ei- puzzles. They are for young people.

EMS Newsletter December 2015 33


Interview

That started my interest in mathematics. I also went ented students. Some of them became well known math-
to a very good high school. This was during the depres- ematicians. I was part of a very good body of students
sion and to be a high school teacher was considered a and it was an exciting time.
very good job. I had excellent teachers and I must say Usually, when you get a PhD at some university in
that the quality of the students was also very good. I par- America, you then leave. You go to another university
ticularly enjoyed the mathematics courses and especially for your first job. Courant was different. He kept the
geometry and physics. I then decided I would like to good people. If good people got PhDs, he simply offered
study physics. them jobs.

Were there already clear signs that you had an excep- Did it help being offered a job if you played an instru-
tional talent for mathematics? ment?
The teachers considered me good but I think it became I didnt play an instrument. But if I had, it may have
clearer in college that I had some talent in mathematics. helped even more. Of course, the rumour was that he
When I graduated from university, I actually received a hired people who played instruments (unless they played
gold medal for my work in mathematics and physics. the piano, which he played himself).

You graduated from McGill University in Montreal. Did you meet with him often?
Perhaps you could tell us about your experience study- Oh, yes. He often invited the students socially to his
ing mathematics and physics at university? home. His wife was completely devoted to music and
I finished high school and applied for a scholarship at played a number of instruments. She was the daughter
McGill, which I didnt get. The high school offered an ad- of the mathematician Carl Runge,2 by the way. They had
ditional year, equivalent to a first year at college. I did two daughters who were both very ardent musicians.
that, applied again to McGill and then got a scholarship. One of them became a professional musician and is now
So I was at McGill for three years rather than the usual married to Peter Lax. Courant was wonderful with young
four. This was during World War II and I graduated in the people very encouraging and really exceptional.
Spring of 1945, just at the end of the war in Europe.
It was a pleasure to study mathematics and physics. Mathematically speaking, your mentor was Kurt Frie-
However, that was the only thing I studied. Because I drichs?3
missed the first year, I didnt take any courses in other Yes. Friedrichs was the person I regard as my Sensei (as
interesting subjects. I am sorry I didnt. the Japanese say). I really was most influenced by him.
He worked mainly with partial differential equations but
How did you end up at the Courant Institute in New he also did other things. He wrote a book on quantum
York? theory and a book, together with Courant, on shock wave
By pure luck! When I finished at McGill, I had a Sum- theory, which was widely used and translated into many
mer job at the National Research Council where they did languages.
atomic research. A son of Courant1 had married a young
woman from Montreal, whom I knew. They both worked You mentioned that there was a special atmosphere at
there and one day she said they were going to New York the Courant Institute, in part because no distinction was
to visit Courant. I asked her to ask him to suggest some made between pure and applied mathematics
place I could apply to do graduate work in physics. Thats right. Courant insisted there was no difference be-
She came back and said that Courant suggested that I tween pure and applied mathematics. He did both and he
come and take a Masters degree in mathematics. I could encouraged people to do the same. It is just mathematics.
then go on to do physics, he said. I went down for an in- When New York University hired him, they asked him
terview and was offered an assistantship in mathematics. what a mathematics department needs and he said: A
I got a Masters degree and I just stayed on. I never left library and a coffee room. So we have a very nice lounge
New York University. that we use all the time.

Courant, Friedrichs and the CUNY It is remarkable that you are the fourth Abel Prize Lau-
reate associated with the Courant Institute (after Peter
Courant was head of a very famous institute in Gttin- Lax in 2005, Srinivasa Varadhan in 2007 and Mikhail
gen, Germany. He was kicked out when the Nazis came Gromov in 2009). What has made this institute so suc-
to power but he was offered a position at New York Uni- cessful?
versity a year later to set up a graduate programme in the Well, partly it is just the warm atmosphere. I think gradu-
mathematics department. They only had undergraduate ate students are very happy there and there is a lot of
training at that time. He came to New York to do that interaction between the students and the faculty. It is, of
but there were very few students in those first years. The course, much bigger now than it was when I was a student
number of students only increased after the war. When I there. But the warm atmosphere has prevailed.
came, just after the war, there were a number of very tal-
2 18561927.
1 18881972. 3 19011982.

34 EMS Newsletter December 2015


Interview

Who were your most important colleagues over your ca- Lets move on to your joint research with Shmuel Ag-
reer? mon and Avron Douglis.6 There were two very impor-
There is Friedrichs but also two other students of Cou- tant papers. Can you explain what they contained?
rant: Fritz John,4 a wonderfully talented mathematician What we did was to extend some classic work, by the Pol-
who later became a faculty member (I had the fortune ish mathematician Schauder, to higher order equations.7
of writing one paper with him) and Hans Lewy5 (I wrote There is a fundamental paper of Schauder for second-
several papers related to some of his work). Hans left order, so-called elliptic equations. We thought it would
Germany immediately after Hitler came to power. He be useful for people to be able to deal with higher order
came to the United States and had a career at Berkeley. equations and systems of equations so we proved the
analogues of those results. In the other paper, we proved
Partial differential equations and inequalities the results for systems and also for different norms, that
is, for different ways of measuring the size of the solu-
Your name, often with various co-authors, is attached tions. We published several different kinds of inequalities
to many fundamental concepts and theorems in PDEs. and they have been used by many people.
If you just look at the citation list, your work has had
a tremendous impact. Lets start with Fritz John, with You wrote a paper with Joseph Kohn introducing the
whom you authored a very influential paper about important notion of pseudo-differential operators. You
BMO functions (BMO standing for Bounded Mean are one of the fathers of that concept. Can you explain
Oscillation). why this concept is so important and how you came
That was his idea. He introduced BMO functions. It came upon it?
from some work he had done in elasticity theory. He ap- Joe Kohn had published a fundamental paper in com-
proached me saying: I have a class of functions and I plex analysis. It involved the regularity of solutions for
believe they should have such-and-such a property. I a certain class of systems up to the boundary a rather
worked on it and was able to prove that property. He difficult paper! He suggested we should try to generalise
then improved it so the final version is better than what I this to more general systems of equations. We started to
had done. It became a joint paper and I must say a lot of look at it and we had to consider so-called commutators
people have referred to it. of operators. You apply an operator and then you apply
a second one. Then you take the difference of that result
Absolutely! It became famous if we may say so be- with the operator obtained by applying the second one
cause of the many applications. For instance, Charles and then the first. We needed properties of the commu-
Fefferman got the Fields Medal in 1978 and one of his tator. We were using a certain space, called an Lp-space,
main contributions was to show that the BMO space is and a theory due to Caldern8 and Zygmund9 for certain
dual to the Hardy space H1. singular integral operators. We needed to extend their re-
Charles Fefferman did many things but in particular, he sult to commutators so we thought: How do we extend
proved the duality result that you refer to. these singular integral operators to make an algebra out
of them?
Your paper with Fritz John contains the JohnNiren- That led to what we call pseudo-differential opera-
berg inequality. You love inequalities? tors. The concept came from a very specific problem in
I love inequalities. And what we proved in the paper was systems of partial differential equations but it turned out
really an inequality. to be a useful thing in itself. It grew out of Caldern and
Zygmunds theory. By the way, Caldern was a wonder-
Would you explain why inequalities are so important in ful mathematician and he danced the tango, which I ad-
the theory of PDEs? mired enormously.
When you look at a partial differential equation, you may
ask whether a solution exists. Now, you cant write down You had a very bright student, August Newlander, with
the solution so you need to know some bounds. It cannot whom you wrote a very important joint paper in 1957.
be too big, it cannot be too negative, its derivatives cannot Can you tell us about the results you proved there?
be too big and so on. You try to get estimates of the size of It was a problem I first heard of from Andr Weil.10 He
the function and of its derivatives. All these estimates are said: Heres a problem in complex analysis. Why dont
inequalities. You are not saying that something is equal you people in partial differential equations work on this
to something but that something is less than some con- kind of problem? I thought: Why not? Lets try. I took
stant. Thus, inequalities play an essential role in proving a student who was very bright and I said: Lets look at
the existence of solutions. In addition, you want to prove the very simplest case, in the lowest dimension. The stu-
properties of solutions and, again, inequalities play a cen- dent, Newlander, had the initial idea, which worked fine
tral role. Hence, inequalities are absolutely fundamental
to studying partial differential equations; for that matter, 6 19181995.
so are they for ordinary differential equations. 7 18991943.
8 19201998.
4 19101994. 9 19001992.
5 19041988. 10 19061998.

EMS Newsletter December 2015 35


Interview

in low dimensions but, to our surprise and dismay, didnt You also went to the Soviet Union?
work in higher dimensions. We had to come up with a Yes. The first time I went was in 1963. It was a joint So-
completely different proof in higher dimensions. It led viet-American symposium on partial differential equa-
from a linear problem to a non-linear problem. It was tions, arranged by Courant on one side and the Soviet
kind of strange but the non-linear problem was in some mathematician Lavrentyev15 on the other. There were
ways more accessible. about two dozen American mathematicians and about
120 Soviet mathematicians from all over the Soviet Un-
What was Andr Weils reaction when you solved the ion. It is one of the best meetings I have ever attended.
problem? It was in Novosibirsk, Siberia, which was the academic
He was very happy and so were other people in complex city that Lavrentyev had helped create. It was like be-
analysis. Many people have used the result. Some years ing aboard a ship for two weeks with people you make
later, Hrmander11 found a linear proof of the same re- friendships with immediately. I made friends with Rus-
sult. It was very technical but it was purely linear. sians that are still friends today. Some have died, unfor-
tunately, but I have had very good friends in Russia since
Are there any outstanding problems in the enormous then. I have never collaborated with any of them but they
field of partial differential equations, apart from the Na- are still very warm friends; we would meet and talk about
vierStokes problem, that you would like to highlight? mathematics, politics and all kind of things.
Well, I think almost nothing has been done in so-called
over-determined systems, that is, where there are more How about China?
equations than unknowns. You may have two unknowns I have been to China a number of times. The first visit
and five equations so there have to be some compatibility was arranged by Chern,16 a Chinese mathematician who
relations. Theres almost no analytic theory of that. There had settled in America. This was in 1975 and the Cultural
is a theory developed by Cartan12 and Khler13 but that Revolution was still going on, though I didnt realise it at
assumes that everything is analytic. Outside analytic cat- the time. For instance, I was visiting the Chinese Acad-
egory, almost nothing is known about such systems. They emy of Science but I was taken to Beijing University. I
often come up in geometry so I feel that this is a big gap said I would like to meet the faculty but they said they
in the theory of partial differential equations. were busy teaching which was simply a lie. There was
no teaching going on. They showed me the library and
Mathematics and mathematicians all over the then they wanted to take me to some other university
world but I said: Theres no point. Either I meet the faculty or
I dont go.
May we ask you some questions about international They had me give many lectures but I said I also want-
mathematics? We know that you travelled to post-war ed to hear what some of the people there were doing. So
Europe very soon after your graduation. some young people spoke about some of their research.
Yes. I had a fellowship and came to Zurich during the I learned later that they had to get permission to attend
academic year 1951/52. I went mainly to be with Heinz my lectures. I didnt make close friends at that time. It
Hopf,14 who was a geometer and a topologist. Heinz was an interesting experience and, of course, things have
Hopf was a wonderful person a lovely and extremely changed enormously since then. I did make friends with
kind man. I also spent one month in Gttingen that year. some who subsequently came and spent a year or two at
That was arranged by Courant who felt I should go there. Courant.
During that year, I didnt actually carry out any research.
What I did was to write up the things I had done before. We should also mention that you were awarded the first
I had been very slow at writing them up for publication Chern Medal of the International Mathematical Union.
because I somehow had a block against writing. So dur- Yes. Thats true. That was in 2010.
ing that year I wrote several papers.
You were also awarded the first Crafoord Prize in 1982,
Did Courant ever return to Gttingen? together with Arnold.17
Yes. After the war, he went back to Germany many times. Perhaps it was a tongue-in-cheek comment but
He had many contacts and he wanted to help build up Arnold once said something like: Mathematics is the
German mathematics again. part of physics in which experiments are cheap.
It wasnt entirely tongue-in-cheek. He really felt that the
He must also have been very bitter? contact of mathematics with physics and the real world
Well, he was bitter but, at the same time, he had friends was important.
and he wanted to encourage and help to develop math- He didnt get permission to go and get the Crafoord
ematics in Germany. Prize. I visited Moscow just before I went to Sweden and
had dinner with him in his home. He was waiting until
11 19312012.
12 18691951. 15 19001980.
13 19062000. 16 19112004.
14 18941971. 17 19372010.

36 EMS Newsletter December 2015


Interview

the last minute to see if he would get permission, but he My ideas may lead to a better proof or may lead to some-
didnt. thing new. The student said hed never seen a proof he
When I went back to America, I got a call from a didnt like and I thought: He is hopeless!
woman claiming to be Arnolds sister. I thought: How
is that possible? I had just seen Arnold a few weeks be- May we ask you a question that we have asked several
fore and he never mentioned he had a sister in New Jer- previous laureates? How does one find the proof of a
sey. She came to my office and, indeed, it was Arnolds mathematical result?
sister. He never mentioned a word. Its incredible! Some people work with perseverance until a proof
is complete but others tell us that insight appears in a
Talking about Arnold, on some occasions he expressed sudden flash like lightning. Do you have experiences
frustration that results proved in the West had already of this sort?
been proved in Russia but, because of poor communica- Both may happen. But most of the time you are stuck.
tion during the Cold War, these results were not known. Maybe you make a breakthrough with some problem as
Did he express these feelings to you? you get some insight and see something you didnt see
He tended to do that. I remember once he was visit- before. But the perseverance and all the work you car-
ing New York. Someone was giving a seminar talk and ried out before seems to be necessary to have this insight.
he was attending the lecture. During the talk, Arnold You need perseverance or, as the Germans say, you need
said: Oh, that was already proved by such-and-such a Sitzfleisch.
Russian. But the person giving the seminar talk then
checked and the Russian had never proved it. So Arnold Are you the kind of person that gets so involved in try-
was not always correct. He tended to give more credit to ing to solve a problem that you are, so to speak, lost to
Russians than was due. the world?
You may have heard the joke where the Russian says: Not all the time but it can happen for many hours. Some-
What you proved, I proved first. And anyway, the result times, I wake in the middle of the night and start thinking
is trivial. about a problem for hours and cannot sleep. When you
do that, it is very hard to fall asleep again! If I have an
Problems, collaboration and Sitzfleisch idea, I just follow it. I see if it leads to something. I still
try to do that but in the last few years it has not led any-
It is striking that 90% of your published papers describe where. I havent had any success.
joint work. Can you explain why this is so?
It is just a pleasure! It is just an enormous pleasure talk- Communicating mathematics
ing mathematics with others and working with them. Of
course, much of the work you do yourself. I mean, you You have had 45 PhD students. That is an impressive
discuss ideas and work with others but then you go home number! Can you tell us what your philosophy is? How
and think about what you have done. You get some ideas do you come up with problems for your students?
and you get together again and talk about the new ideas. Its hard to say. Sometimes it is hard to think of a suitable
You get reactions to your ideas and you react to their problem. It is easier to think of problems that are too
ideas. It is a wonderful experience. hard, and just not practical, than to think of a problem
that is good and can be solved in reasonable time. I cant
Do you usually start out with a goal in mind? really answer that question. I dont know how I go about
Usually there is a goal. But somebody once used the ex- posing problems.
pression: There are those mathematicians who, when
they come to a fork in the road, they take it. Im that Were there occasions when you had to help students
kind of mathematician. So, I may be working on a prob- along?
lem with a colleague when we come to something that Oh, yes. I meet the students regularly, usually once a
looks interesting, and we explore that and leave the origi- week. We discuss their progress and I might make sug-
nal problem for a while. gestions. I may say: Look at this paper, this may lead to
something.
Are you more of a problem solver?
Yes, definitely. There are two kinds of mathematicians. How would you describe your love for mathematics?
There are those who develop theories and those who are What is it about mathematics that is so appealing to
primarily problem solvers. I am of the latter. you? Is it possible to communicate this love to people
outside the mathematical community? Does one have to
Do you come up with interesting problems through be a mathematician to appreciate the appeal of math-
discussions with other mathematicians? What kinds of ematics?
problems are you attracted to? Is there any pattern? Some people are very good at communicating to the gen-
Its hard to say. A graduate student once asked me how eral public. I am not so very good at that. But once you
I find good research problems. I said to him that I some- are in it, once you are hooked, its very exciting and fun. I
times see a result but dont like the proof. If the problem have used the word fun before. But it is really fun to do
appeals to me, I start to think if there is a better proof. mathematics. It is an enormous pleasure to think about

EMS Newsletter December 2015 37


Interview

mathematics even though you are stuck 90% of the time, ies: Italian movies, Russian movies, French movies. I went
perhaps even more. crazy. I went almost every night to the movies. Since then,
I have loved movies.
That is what people outside mathematics cannot com-
prehend. Have you seen A Beautiful Mind?
Yes, it is hard to comprehend. You have to be in it and I Of course, and I have read the book.
think it does take some talent to be able to do mathemat-
ics. But it also takes, as I said, Sitzfleisch. You need to What kind of music do you like?
be stubborn and have perseverance, and you cant give Mainly classical but I also listen to jazz. My grandson,
up. I have been stuck on some problems for years. who will be at the ceremony tomorrow, is a professional
jazz drummer. And I love Argentinian tango. I have a
But you do think its important to try to communicate large collection of records of Argentinian tango.
to the general audience?
Yes, I do think that is important: (a) for the development Not only on behalf of us but also on behalf of the Norwe-
of mathematics, and (b) to show them that it is a pleas- gian, Danish and European Mathematical Societies, we
ure to do mathematics. Courant and Robbins18 wrote would like to thank you for a very interesting interview.
a very nice book: What is Mathematics?. It is a lovely
book. There is also a recent book by Edward Frenkel, a
mathematician who came as an immigrant from Russia
as a young man. It is called Love and Math. He makes a
valiant attempt to get the general public interested in the
branch of mathematics in which he works (which is also
connected to physics). It is very hard to do. He tries but I
think it is too hard for the general public. But he makes a
real attempt to do it and I must say I admire him for that.
I just recently read his book.

Music and movies

We have one final question that we have asked several


laureates before. What are you interested in when you
are not doing mathematics? Louis Nirenberg (left) interviewed by Christian Skau and Martin
I love music. I love movies. You wont believe this but Raussen (photo: Eirik F. Baardsen, DNVA).
at the time when I lived in Montreal, in the province of
Qubec, you could not get into a movie before you were Martin Raussen is professor with special responsibilities
16. Incredible! Now its hard to believe. So when I was 16, (mathematics) at Aalborg University, Denmark. Chris
I went crazy and started to go to movies. When I moved tian Skau is professor of mathematics at the Norwegian
to New York, there were suddenly all these foreign mov- University of Science and Technology at Trondheim. They
have together taken interviews with all Abel laureates
18 19152001. since 2003.

European Mathematical Society Publishing House


Seminar for Applied Mathematics
ETH-Zentrum SEW A27, CH-8092 Zrich, Switzerland
orders@ems-ph.org / www.ems-ph.org

Fa di Bruno Hopf Algebras, DysonSchwinger Equations, and LieButcher Series


(IRMA Lectures in Mathematics and Theoretical Physics Vol. 21)
Kurusch Ebrahimi-Fard (Universidad Autnoma de Madrid, Spain) and Frdric Fauvet (Universit de Strasbourg, France), Editors
ISBN 978-3-03719-143-9. 2015. 466 pages. Softcover. 17 x 24 cm. 48.00 Euro
Since the early works of G.-C. Rota and his school, Hopf algebras have been instrumental in algebraic combinatorics. In a seminal
1998 paper, A. Connes and D. Kreimer presented a Hopf algebraic approach to renormalization in perturbative Quantum Field
Theory (QFT). This work triggered an abundance of new research on applications of Hopf algebraic techniques in QFT as well as
other areas of theoretical physics.
The present volume emanated from a conference hosted by IRMA at Strasbourg University in France. Researchers from different
scientific communities who share similar techniques and objectives gathered at this meeting to discuss new ideas and results on
Fa di Bruno algebras, DysonSchwinger equations, and Butcher series.
The purpose of this book is to present a coherent set of lectures reflecting the state of the art of research on combinatorial Hopf algebras relevant to high
energy physics, control theory, dynamical systems, and numerical integration methods. This volume is aimed at researchers and graduate students interested
in these topics.

38 EMS Newsletter December 2015


Interview

Interview with Manjul Bhargava


Ulf Persson (Chalmers University of Technology, Goteborg, Sweden)

Were you surprised to get the Fields Medal? a lot of time in India growing up, which is where my ex-
Well, I guess maybe it wasnt the hugest surprise, since tended family is from. Indians tend to have very large
people had been talking about it and asking me about it and supportive families and I always received a lot of en-
for so long. But, of course, when it actually happens, it is couragement from them.
a surprise, and a pleasant one. My greatest inspiration was my grandfather, who was
a Sanskrit scholar. I spent a lot of time learning from him,
There has been a lot of attention this year, more I think not just Sanskrit but also history, philosophy, literature
than four years ago. It was impossible to reach you by and other languages like Hindi and English.
email. I also learned a lot about these subjects from my
Indeed, within a couple days of the announcement, I had mother, who is a mathematician (so, of course, I learned
received around 18,000 emails! There was no way I could a lot of mathematics from her as well).
read through them all. Im so sorry that I was hard to I didnt like to go to school so much. I liked learning
reach. But I plan to respond to all of them you will on my own but then Id show up for exams and, of course,
eventually hear from me! also for after-class activities at my school, which was a
running joke among my teachers and friends.
I can understand. I tried to reach you after your talk
as well. It was impossible; the crowd around you was What is your earliest memory of mathematics?
incredible. I felt like the paparazzi. I always liked mathematics, as far back as I can remem-
Yes, I have never experienced anything like it. It was fun ber. I loved shapes. My favourite toy as a 2- or 3-year-
at first, giving autographs and taking photos. But then old (my earliest memories are from then) was a cube-
the crowd was getting too big and compressing towards shaped puzzle with various shapes cut out of it, and the
the centre, and people were getting crushed. Two security goal of the puzzle was to quickly identify the shapes and
guards came to try and disperse the crowd but they too insert them into the slots where they exactly fit. A bit
got crushed. Then they radioed a whole team of security, later, I remember trying to work out ways to add and
who made their way in and formed a ring around me and then multiply large numbers in my head using my fin-
took me away. gers.
I never thought I would experience something like I was always interested in numbers. I really liked big
that as a mathematician! I am extremely impressed that numbers and learned the names of bigger and bigger
mathematics in Korea is held in such high regard. The numbers in both Hindi and English (mahaashankh is
ceremony was broadcast live on national television and 1018 in Hindi; then I learned the word vigintillion in
so many from the public watched it, enough so that we English, which is 1063; then I learned from my grand-
would get stopped for pictures and autographs just walk- father that back before 500 BC, there was the Sanskrit
ing on the streets of Seoul. word dhvajaagranishamani for 10421, which I found
very exciting).
So you are experiencing the proverbial 15 minutes of
fame. Big numbers are much more mindboggling and condu-
Yes, or more precisely about a week of fame, until I leave cive to vertigo than mere infinity.
Korea and return to the US! Do you know of the number Googolplex?

Will this change your life significantly? Of course.


Well, I hope not. I know it will give me new opportunities That really used to intrigue me after coming to terms
to disseminate mathematics to the public around the world with mahaashankh, vigintillion and dhvajaagranishama-
that I didnt have before, and that is a responsibility that ni, I used to try to wrap my head around what that [Goo-
I look forward to and will, of course, not take lightly. At golplex] meant. Yes, the feeling of vertigo is quite accu-
the same time, I know that it is mathematics research and rate!
teaching that gave me this platform and I do not wish to
compromise too much on my research and teaching and So you were a math prodigy?
the ongoing work with my students and collaborators. So it No, I would not say that. I did not skip grades or anything
will all be about finding the right balance between the two. like that. But I did like to discover things for myself, en-
couraged by my mother.
To start from the beginning, what is your family back-
ground? I read that you made algebraic calculations at the age
I was born in Canada and moved to the United States of eight such as computing the number of oranges in a
early on. I grew up in a very Indian home and I also spent triangular pyramid, given the number of oranges on an

EMS Newsletter December 2015 39


Interview

edge of the pyramid. This is impressive. When did you ematics, it is not in words or in terms of things that can
encounter, say, complex analysis? be written down yet. That usually comes later when I try
I always enjoyed doing such things on my own for fun. to translate these thoughts into more usual mathematical
But, as I said, I never skipped any grades. For example, I language.
did not learn complex analysis until I was a sophomore in
college and took a course. So I was never that much more What about collaboration?
advanced than my peers, particularly when I reached col- I love working with others. Thinking and talking about
lege at Harvard, where so many students were really ad- and working on mathematics together with others is one
vanced. of the joys of mathematics. It can also be more produc-
tive. Collaboration forces you to think things through
Did you ever read a math book cover to cover? by explaining them, and the other person can then give
Yes, I often did in graduate school. Princeton did not of- invaluable input, feedback and new perspectives, as
fer many courses for beginning graduate students so this you bounce ideas off one another. Sometimes, different
was the way I learned much mathematics. collaborators can bring different areas of expertise to
a problem, which can be very valuable and help bring
You always were a good boy? some cross-pollination between areas.
I guess so. Im told I was quite hyper and naughty when
I was a 3- or 4- or 5-year-old but it seems I outgrew that! Does working with someone not interfere with your
thinking process? When you are thinking hard about
Did you have other interests? Reading? The classical something, you do not want to be interfered with.
Indian epics such as Ramayana and Mahabharata con- Sometimes. But, in collaboration, one generally decides
stitute a treasure trove which must be very fascinating which parts are more productive or fun when discussed
to a child. together and which parts each should go home and think
Yes, these epics, as well as various other works of litera- about alone. In the latter case, collaborators can go home,
ture, were frequently the topics of conversation over think separately and then bring their ideas the next time.
family dinners at home. In fact, my uncle often made So, even in collaboration, there is usually a lot of indi-
these conversations into contests, as to who knew the vidual thinking. Actually, sometimes youll be working
most details about various aspects of these works. These with a collaborator where no one speaks for over an hour
contests were a lot of fun and added for my cousins and individual thinking is going on! But still its fun to do
me a further motivation to read! it together, knowing that there is someone there in case
you are ready to share an idea, or problem, or general
Later on, did you consider any alternative careers? confusion.
I did often toy with the idea of becoming a musician and
tabla player as a profession. But at some point, I realized Nothing beats direct human conversation, or should
that if I became a mathematician, Id still find time to I say interaction. Are there problems you have solved
keep up music as well but if I became a full-time musi- in collaboration that you would not have been able to
cian, Id probably have a tough time keeping up math- solve by yourself?
ematics! Almost certainly! This is, of course, impossible to know
for sure. But I definitely think so, and I suspect this is
So you were all set to be a mathematician from an early likely the case in most of my collaborations!
age?
I always loved math and because my mother was a math- Do you understand everything you are doing in math-
ematician, I was always aware of that pleasant career op- ematics? I mean, what is important about a theorem
tion. But I did have brief periods when I changed and is not its precise formulation but the idea of the proof
thought maybe Id like to do computer science, or eco- which can be used to prove many different theorems.
nomics, or physics, or linguistics, as I liked these subjects a If you just use the exact formulation, you are actually
lot too. But I always realized that mathematics was what treating the theorem as a black box.
was bringing these subjects together, and mathematics I used to understand all the mathematics that I do, from
was what I liked about them. the very basics all the way up. There were no black boxes.
But in some recent collaborations just in the last year, I
How do you do mathematics? have used some theorems I did not understand from first
I usually like thinking about mathematics while walking, principles but took on trust, using them as black boxes
or jogging, or pacing, or working with someone at the as you say. I must admit that at first it made me quite un-
blackboard. I quite rarely sit down with a pen and paper comfortable. It did not seem to bother my collaborators
and do mathematics; that usually comes after something as much though, who were already used to it after many
has been worked out already. years of it! This is a modern phenomenon in doing math-
ematics, building on years and years and pages and
So you cannot think when you are standing still. pages of our predecessors work, to the point where it
I guess thats a good way of putting it! The reason I stay is not possible for any one person to know every detail.
away from paper is that often when I think about math- But one must get used to it. In some sense, it is also awe-

40 EMS Newsletter December 2015


Interview

inspiring: mathematics is bigger than any one person and,


as a mathematical community, we are able to go further
than any one person could alone.

I think that this is a bit sad. The Australian aborigi-


nes are supposed to be the best trackers in the world.
They are personally very connected to their environ-
ment; their technology may be primitive but they have
an intimate command of it. Modern man, on the other
hand, is more alienated; he is using gadgets all the time
of which he has not the faintest understanding, such as
mobile phones black boxes in other words. One would
think that the life of an aborigine would in some sense
be more satisfying, while we become more and more
Manjul Bhargava receiving the Fields medal from the Korean
consumers rather than producers. Would you care to
president Mrs Park Geun-hye. (Photo courtesy T. Gowers)
elaborate on that?
I do agree in a sense as I said, until the recent develop- ways willing to listen and inspire and you learn a lot just
ment, everything I did in mathematics till now I under- through conversation with such a figure. But, above all,
stood quite thoroughly from the bottom up and that was, he had a wonderful sense of what is important in math-
as you say, very satisfying. I felt like I was firmly touching ematics.
the ground at all times rather than floating in air. But John Conway and Peter Sarnak were also inspirational
one cannot deny some of the things in mathematics that figures to me during graduate school; I also talked to them
have been achieved by mathematicians being consumers quite regularly and learned so much from them as well.
of other mathematicians works rather than just being
producers. This leads me to another question. Is too much math-
ematics being published?
When you think of mathematics, do you do it systemati- What do you mean? In terms of too much paper being
cally, one problem at the time? wasted too many trees being cut in the forests?
No, generally not I usually like to think about many
problems at the same time. That way, if Im stuck on one, I meant that there is too much and that the important
I can always move to another for a little while, and may- things are being buried under a heap of garbage.
be, when I get back to the one I was stuck on, Ill have a Strong words! But is the peer review system not taking
fresh perspective. I like to have problems around of dif- care of that?
ferent levels some of them easy, in the sense that I have
at least some methods that can be used to make progress Not very effectively.
on them, while others are much harder, for which I often As to important things being buried, I think with modern
do not have any clear idea how to solve them. And then search engines and MathSciNet, we are in a position, as
there are long-range problems I keep in the back of my never before, to look and find what is important to us and
mind, hoping that one day I will have a moment of inspi- the things that we need in our work.
ration and find some possible inroad.
So this means that, in the future, everything should be
Do they range over a wide field? available electronically, a huge database which you can
Well, they definitely range over a slightly larger field eve- mine. Maybe that will change the forms of publishing
ry year, as I learn more mathematics and get interested dramatically; instead of writing traditionally structured
in more problems. But the core of most of the problems papers, results will be presented much less polished
I get interested in tend to be related to classical areas of not an entirely pleasant scenario.
number theory somehow. Actually, I do not think this to be so bad. More material
is available at our fingertips than ever before and avail-
The psychologist William James claimed that every in- able sooner to us than in times past. We have the option
terest in adult life can be traced back to one formed in to look at less polished material that is available much
your childhood or youth. sooner; of course, we can still wait for the peer-reviewed
This is very much so with my mathematical interests and more polished versions as we always did. So we have
much of what I do can be directly traced back to interests more options than we used to have, which I feel is a good
I formed when I was young. thing.

By the way, who was your advisor? The connection between mathematics and music is often
Andrew Wiles he was definitely a great inspiration pointed out. Do you think there is one and, if so, how do
to me! I didnt end up working directly in his area and you explain it?
tended to work more on my own problems but he was There definitely is. Many mathematicians are musical
just wonderful to discuss mathematics with. He was al- and I dont think that is a coincidence. What is interest-

EMS Newsletter December 2015 41


Interview

ing is that it doesnt quite work the other way round, at But some mathematicians are not musical at all. Does
least not to the same extent. that mean that they are different as mathematicians,
maybe even defective?
If you meet a musician and tell him that he must be No perhaps they have the potential but never had the
good at mathematics, chances are that he or she will opportunity to develop it. Not all excellent musicians are
look at you blankly. also excellent painters or sculptors. It is up to each per-
Precisely. son to decide how best to apply and develop ones artis-
tic sensibilities and interests.
It also means that musical talent is much more common
than mathematical talent. In order to appreciate mathematics, you have to be a
I guess so an interest in music certainly tends to be mathematician but you can appreciate music without
much more common than an interest in mathematics. being able to compose music. I suspect that the emo-
tional impact of music has very little to do with the
But we are not talking about just an interest but about composing of music, maybe not even with musicality
the active sense of playing an instrument. Do you play per se. Mathematics does not have this aspect. It does
an instrument? not give the same kind of emotional sustainment. It is
I did learn several instruments as a child: tabla, sitar, gui- very different listening to a math lecture and a piece of
tar, violin and a bit of piano later on. But tabla is the in- music.
strument that I always enjoyed the most and stayed most Every subject can be appreciated at different levels. I
in touch with over the years. definitely think it is possible for the general public to ap-
preciate the beauty of mathematics, if it is presented in
Do you actually perform to paying audiences? an accessible way. Maybe a general person will not ap-
Yes, I do but definitely less so than in the past. It takes so preciate it at the same level and in the same way that a
much time to prepare and practice for such concerts and mathematician would, just as a general person might not
Im getting less time every year, unfortunately. appreciate a technical piece of music in the same way
that a professional musician would. But I do agree that
But why is there such a connection? Have you ever given music tends to appeal slightly more to ones emotional
it a thought? side and mathematics slightly more to ones logical side.
I have given it a lot of thought. But is not the sensation of beauty listening to a lecture
by Serre similar to listening to a beautiful piece of music?
So what are your thoughts? Serres lectures are like musical improvisations he is
Well, I think that both mathematics and music are about sensitive to the reactions of the audience and changes his
patterns and how they fit together. Practitioners of both presentation as he goes along.
subjects are guided by beauty and elegance, with the aim
of telling stories and conveying thoughts that ordinary No there is, I think, a profound difference. The appre-
words cannot express. ciation of mathematical beauty is not tied to a definite
presentation; it can easily be paraphrased.
I would perhaps put it a bit differently. It is about Well, it depends what kind of music you are talking
themes that repeat themselves with subtle variations, not about. Not all music is tied to a definite presentation. I
mechanically or predictably. Music is not about perfect think elegant chalk-and-blackboard mathematics lec-
symmetry, but almost, just as two different mathemati- tures are more like the improvisations of classical Indian
cal fields can be very similar without being isomorphic music or jazz, where the speaker/musicians presentation
if the latter were the case, it would be boring, just as al- is affected by the atmosphere and mood of the day and
gorithmically generated music is supposed to be boring. the reactions and type of the audience. The main themes
I would say that part of the divergence of the two sub- are decided in advance but the exact presentation can
jects has to do with the difference between the types of vary, and can come out differently each time. That is part
thinking in the right and left hemispheres of the brain. of the excitement of it.
There is a formal, logical aspect to mathematics, which
is the only one people in general are aware of, but there Misspellings or mispronunciations have no impact on
is also a very artistic side, having to do with creating and the mathematical beauty but playing the wrong notes
discovering patterns and seeing how they can be con- has a jarring effect.
nected in a coherent and expressive way. The same is true Wrong notes, i.e. notes played outside the framework of
of music as well: there are formal constraints on compo- the musical piece, are similar to mathematical errors in
sition/improvisation dictated by the genre of the music a lecture and the latter may be equally (if not more!)
but, beyond that, what remains is the artistic expression jarring.
given those constraints of the genre. Id say the formal
and logical aspects are slightly more critical in math- But take a standard piece in the Western classical rep-
ematics than in music (in that a piece of mathematics is ertoire, say the Goldbach variations by Bach. They are
rejected if it is not completely logical) but the artistic and essentially played in the same way over and over again,
logical aspects play an important role in both. although of course I am aware of the subtle variations

42 EMS Newsletter December 2015


Interview

from one performance to another, just like differences Are you interested in mathematical philosophy the is-
in the translations of the same novel. You can take a sue of mathematical Platonism?
recording and listen to it again and again but if you A bit, but I cant say I give it a lot of thought.
played the same video of a Serre lecture over and over
again, you would get bored. So it means that you are a Platonist.
Well, thats why I was saying that a chalk-and-blackboard [Laughing] Sure I do feel that good and natural math-
math lecture is more similar to an improvisational form ematics is discovered and not invented.
of music, such as classical Indian music or jazz, where
the themes are predetermined but the exact rendition So you believe in an external mathematical truth, that
of the themes varies from performance to performance, the theorems we prove are not just figments of our im-
depending on the mood and the audience that particu- aginations or mere social constructs.
lar day that is part of the excitement of those forms of I do. I think many mathematicians would probably not
music. do mathematics otherwise.
For example, the raaga Darbaari is one of the most
beautiful raagas in Indian classical music but each time One of the wonders of mathematics is that some math-
one goes to hear it live, one is likely to hear a different ematical conjectures are actually being solved.
rendition (possibly even on a different instrument or in- I agree. That is rather amazing! But it is conceivable that
struments) and that is part of the beauty and appeal of some famous conjectures that people work on might be
this melodious raaga. One would likely get bored if one undecidable.
went to hear exactly the same performance of Darbaari
over and over again, unless it is one of the truly historical Maybe because they are, as they say in the field of PDEs,
and legendary renditions (such as that of Ustad Vilayat not well-posed. The natural questions seem to be ame-
Khan or Pandit Bhimsen Joshi). nable to solutions, such as the rather remarkable link
Incidentally, I am not so sure I would get bored view- between Fermats conjecture and work in Elliptic curves
ing the same lecture by Serre over and over again no that your advisor exploited.
doubt I would always find new insights each time! That does seem true so far. But there are also several
well known and seemingly well-posed problems that are
But if you want to understand a piece of mathematics, it still open it could be possible that some of these are
is better to read different presentations rather than read undecidable
the same one over and over again. Mathematical ideas
are not as tightly tied to their presentations as musical. Do you have any personal opinions on the status, say, of
Again, the very same is true for improvisational forms the Riemann Hypothesis or the BirchSwinnertonDyer
of music such as classical Indian music and jazz: listening Conjecture?
to several renditions of the same raaga or an improvisa- Id certainly love to see a solution to one or both of these
tional jazz piece by different expert musicians is the very in my lifetime. But its not clear to me in either case how
best way to understand it, absorb it and appreciate it. close we are. Ive certainly thought more about the latter
and feel like it could be within reach.
A further example to illustrate the difference between
composing and listening is that Haydn reportedly burst You have ideas of how to prove it, not just ways of vin-
into tears when listening to his own creation (actually dicating it, as in your talk, which was very nice by the
`The Creation), claiming that he could not have com- way.
posed it; it was too beautiful. Thanks. I think I have a few ideas of how to prove it, in
Actually, I truly feel this happens to mathematicians all the sense that I have some ideas that I have not yet tried
the time. Who hasnt looked back on ones prior works and so they have not failed yet!
and asked: How did anyone come up with that? It hap-
pens to me on occasion. I come across things that I have Which you are not going to expand upon?
written in the past and am amusingly impressed. How It would take us too much afar.
did I ever think of that? At the time, it might have been
obvious but, of course, not later. Looking back at ones We started our conversation about large numbers and
artistic thought or creation as a third person can be an how they exercise a fascination, especially on burgeon-
extremely different experience than the moment or ex- ing mathematicians. But there are different kinds of
perience of creating it. numbers; not all numbers seem directly related to cardi-
nality, and some basic cardinalities are very low. When
That is true. When you write something down, it is no it comes to mathematical abstraction, I believe that
longer a part of you; it has an independent objective there may only be three or four significant levels, just
existence. Popper would express it by saying it has been as we can only sense number, so called subitising, when
moved from World Two to World Three, from your own there are very few.
personal world to one which can be shared by all think- That is certainly true. There are very few levels of ab-
ing people. straction that mathematicians tend to employ. But that
Indeed. may be because mathematics is still in its infancy.

EMS Newsletter December 2015 43


Interview

It seems to me that when it comes to very large numbers, ics. Number theory inevitably touches on it and where it
they do not really have mathematical significance. The does, it gets hopeless.
celebrated Skewes number resulted from a crude esti- Could be. At the moment, though, there are far easier
mate, which has, I believe, been drastically improved. places where it gets hopeless! The fact that, as yet, there
Much of real mathematics only concerns the exponen- is no algorithm that provably determines whether an el-
tial; if you get double exponential bounds, your esti- liptic curve equation has a finite or an infinite number
mates are bound to be crude. of rational solutions is intriguing. You did go to my talk?
Well, that is true about Skewes number. But do you
know about the the ParisHarrington Theorem? It says Yes, I did, and this is why I brought this up. It was very
that the smallest number satisfying the strengthened fi- nice, perhaps because I knew a dense subset of it. It is
nite Ramsey theorem with given parameters grows well always the completion at infinity that you can bring
beyond exponentially in those parameters; it is not even home with you. If you do not know enough, you could
primitive recursive and is beyond what can be defined get lost.
using Peano arithmetic. Its a pretty natural example True, there is only so much information you can process
where such a huge function naturally arises and cannot during a talk.
be made smaller. I would not be surprised if more exam-
ples are discovered in the future in number theory and So what makes a good math talk?
beyond. Well, as you say, you should not strive to convey too much
information and you must try to build on what your audi-
But there are many places where logic and mathematics ence already knows. Doing less with clarity and purpose
part ways. In logic, you can formally have arbitrarily is much better than doing more. I also feel a talk should
long chains of quantifiers but, in practice, those chains try to impart a sense of the wonder, the motivation, the
are severely limited in length, just like the levels of ab- connections with other related problems or applications
stractions. Does anything beyond the cardinality of the and a glimpse of some of the key ideas. I think that goes
continuum enter real mathematics? for all talks, to the public as well as to ones professional
Its true that we rarely work with cardinalities beyond colleagues, although how one goes about it in each case
the continuum but it does happen, no? Many theorems in can be quite different.
topology and model theory (which has also been exten-
sively used in number theory in recent years) make use Thanks so much for your time during this very busy pe-
of ultrafilters, which even on Z have cardinality beyond riod for you.
that of the continuum Im sure there are other exam-
ples. It was a pleasure. Thanks very much for your interest,
and for the excellent questions!
There are no known effective bounds on the sizes of min-
imal rational solutions to algebraic equations over the
rational numbers but, if they existed, would they not be Ulf Persson [ulfp@chalmers.se] has been a
exponential with respect to the coefficients? member of the editorial board of the EMS
Well, given the ParisHarrington theorem, it seems to Newsletter since 2006 and a professor of
me that larger-than-exponential functions could poten- mathematics at Chalmers University of
tially arise in this scenario as well, no? I dont see why Technology in Gothenburg, Sweden, since
not. In particular, if there were such general exponential 1989, receiving his Ph.D. at Harvard in
bounds, would that not give an algorithm to determine 1975. He has in the past interviewed recent Fields medal
the solubility of Diophantine equations, which we know ists specifically for the EMS Newsletter, as well as other
cannot exist by work of Matiyasevich, etc.? mathematicians for alternate assignments, in a conversa
tional style, some published, others as of yet unpublished.
What I am aiming at is some formulation along the There is a plan to make a selection and collect them into a
lines of the unreasonable reasonableness of mathemat- forthcoming book.

44 EMS Newsletter December 2015


Discussion

Recollection of a Singular School


Sylvie Paycha (University of Potsdam, Germany)*

Some 40 participants are seated in the large seminar still is not confirmed but considering the potential danger
room; they have come from 12 different African coun- of the situation, we should all get back to the hotels imme-
tries1 to Ouagadougou, the capital of the landlocked diately. By the time I have digested the news and further
country of Burkina Faso2 to learn about Fourier integral questioned my colleagues, we realise that the participants
operators and their many concrete applications inside and have already left for the hotels with the hotel van. In the
outside the realm of mathematics. The school3, funded by car that is driving us back, I insist that we should go to
the Volkswagen Foundation, started three days ago and, the hotels to inform the participants and speakers of the
for the last talk of the day, I am about to explain to them situation. My colleagues hesitantly comply as I argue that
how to use the inverse Fourier transform to build pseudo- informing the participants of the situation is the best thing
differential operators from rational functions, when my to do at that stage, after which they leave for their respec-
colleague Bernard Bonzi, co-organiser of the school, steps tive homes on the other side of the city. Back in our hotel
into the room calling me to the door. after the sudden interruption of the talks, we gather with
I follow him along the corridor, where three other col- other guests staying at the hotel around the television set
leagues and co-organisers, Marie-Francoise Ouedraogo, in the reception hall to hear that the event was indeed a
Stanislas Ouaro and Hamidou Tour, are waiting with anx- putsch by General Gilbert Diendr.
ious looks; something serious has happened, they tell me. This was how the school was suddenly interrupted on
Is it that serious that I cannot have another 10 minutes to Wednesday 16 September 2015, a date that probably none
conclude my presentation? My colleagues seem reluctant of the participants of the school will ever forget.
to let me finish but finally nod approvingly, insisting that Diendr had served for three decades as former Presi-
I should conclude hastily. I still have no idea why there is dent Blaise Compaors Chief of Staff. This coup was sup-
this sudden tension and I am suspecting a problem with ported by the presidential guard known as the RSP, the
the premises we are using, a three-storey, medium size, 1,200 strong Regiment of Presidential Security, of which
cream-coloured building on the outskirts of Ouagadou- Diendr was also seen as the figurehead commander de-
gou, some 20 minute bus drive from the two hotels most spite having retired from the force in 2014. The RSP ar-
of us are staying in and not too far from the campus of the rested both President Michel Kafando of the transition
University of Ouagadougou (co-organising institution of government and Prime Minister Isaac Zida. The air and
the event). Some 10 participants are staying at the guest terrestrial borders were closed for some days4 and a curfew
house of the university, which is near the conference hall from 7 pm to 6 am (lasting some 10 days) was declared.5
but far from the two hotels in the city centre where most Little had I anticipated such a dramatic event during
of the participants are housed. The local organising com- the opening ceremony,6 which had taken place that very
mittee has rented the conference hall as well as a nearby morning in the presence of the President of the Univer-
room where we gather for lunch and coffee breaks from sity of Ouagadougou.
the private University of Saint Thomas dAquin. I had expressed my hopes that this school might help
Back in front of the audience patiently waiting for me, overcome the obstacles that set walls between our respec-
I mumble a few words explaining, with a touch of irony, tive nations and continents. A putsch was not among the
that we should conclude rapidly before we are required
to leave the room. I feel I am legitimately and dutifully
finishing the explanations I had started. The students who
had got actively involved in the discussion about the cor-
respondence between symbols and operators via the Fou-
rier transform seem somewhat disappointed when we stop
soon after the interruption. My colleagues, who still look
very preoccupied, take me away into a room, which adds
to the mystery of their sudden interruption, and answer my
questioning look: There seems to have been a putsch; it
Place de la Rvolution.
* On leave from the Universit Blaise Pascal in Clermont-Fer-
rand 4 The airport was briefly reopened later to let the presidential
1 Benin, Cameroon, Chad, Congo, Democratic Republic of
delegation of Macky Sall (President of Senegal) and Thomas
Congo (RDC), Ethiopia, Ivory Coast, Mali, Morocco, Niger, Boni Yayi (President of Benin) land to start the negotiations
Nigeria and Senegal. with General Diendr.
2 The Land of people of integrity (Le pays des hommes in- 5 The curfew restrictions were to be lessened a week later to

tgres), formerly Upper Volta. 11pm5am.


3 Summer school on Fourier integral operators and applica- 6 The ceremony had been postponed to the third day of the

tions, Ouagadougou, 1425 September 2015. meeting, due to my late arrival.

EMS Newsletter December 2015 45


Discussion

obstacles I had envisaged but now, the school seemed to be


doomed to end three days after it had started.
The Revolution square (Place de la Rvolution7)
where protests started that very Wednesday evening
nearly a year after the October 2014 protests that had
forced the former President Blaise Compaor to resign
following some 27 years in power was to separate the
local organisers and participants in their homes from us
foreign organisers, speakers and participants in our hotels
on the other side of the city. What was named the event
Hardly no one on the otherwise lively streets of Ouagadougou (left) and
for a while was to set up an imaginary yet tangible wall be-
participants on their way to the improvised conference venue (right).
tween us. Apart from a very brief visit of a couple of local
organisers to one of the hotels three days after the event, The hotel manager kindly lent us a small seminar room,
only some eight days later (a day after the Tabaski celebra- which seemed unused. We found a tiny, narrow whiteboard
tions8, seriously hampered by the lack of food in the city) to lean against the wall, a board we had to hold up straight
when the city seemed to start getting back to normal life on the table with one hand while writing with the other.
did the local organisers dare to venture back to the hotels. In that small room and on that board were held some 40
How frustrating and disappointing for them when they had lectures (four a day over 10 days), thanks to the speak-
put so much energy into organising the school! Little had ers10 who gave talks in such difficult conditions. We could
they anticipated the event, especially as, on the request only count on those who were fit enough to give a talk
of the Volkswagen Foundation, they had asked the Presi- for many fell ill over several days due to the preventative
dent of Ouagadougou University to confirm that organis- medication against Malaria and who were ready to take
ing such a school in Ouagadougou would be safe. the risk of walking ten minutes from their hotel to the new,
Gradually measuring the importance of what was hap- improvised conference room in the other hotel nearby.
pening, and the potential danger of violent confrontations Participants searched their pockets for a few marker pens
between the population and the RSP, as co-organiser of to give us and, when we ran out of pens on the fourth day,
the school, I began to worry about what to do in such cir- the hotel sent out an employee on a difficult mission (con-
cumstances. What were we to do on the morning after the sidering the circumstances) to drive through Ouagadou-
putsch, and the days to follow? One option was to declare gou, where most shops had remained closed since the coup,
the school over but then what would the participants and in search of a stationers who might sell markers. Thus, we
speakers now stuck in their hotels do all day; would they could go on covering the white board from top to bottom
not start panicking? An alternative was to try to adapt with semi-groups, distributions, wave front sets, Fourier
the organisation of the school to the circumstances, an a transforms, pseudo-differential operators, characteristics,
priori risky solution considering the instability of the situ- singular supports, Lagrangian submanifolds, Bergmann
ation. Indeed, the Balai Citoyen (the Civic Broom9) had transforms, Fourier integral operators, fundamental groups
grown in determination and efficiency, so a violent reac- and Maslov indices. With the event, singularities, which
tion could be expected from were the central theme of the school, had become a char-
the Burkinab people, who acteristic of this very