Some Trends in Modern Mathematics and the Fields Medal

by Michael Monastyrsky
Institute for Theoretical and Experimental Physics, Moscow

This article is based on a talk presented at the symposium The Legacy of John Charles Fields, held in Toronto, June 79, 2000. It was originally published in CMS NOTES de la SMC, March and April 2001, Volume 33, nos. 2 and 3, and is reproduced here with the permission of the CMS.

Introduction
The Fields Medal is now indisputably the best known and most inuential award in mathematics. Sometimes it is compared with the Nobel prize, since there is no Nobel prize for mathematics. Publishers and journalists especially like this comparison. It seems to me that such a comparison is not adequate. The Fields medal was established on dierent principles. Unlike the Nobel prize, which is mostly awarded to mature scientists to crown their careers, the Fields medal is awarded to young scientists, less than 40 years old. The prize is intended not only to recognize results already obtained, but also to stimulate further research. Besides this it is awarded only every four years, at the International Mathematical Congress. The rst Fields Medal was awarded in 1936 in Oslo and the second one 14 years later, in 1950, in Cambridge, Massachusetts. So mathematicians born during 1900-1910 were automatically excluded from the list of candidates, for example brilliant mathematicians like A. Kolmogorov, H. Car1

tan, A. Weil, J. Leray, L. Pontryagin, S. S. Chern, and H. Whitney. Nevertheless, if we look at the achievements of Fields laureates from the point of view of the development of mathematics in the 20th century, we see an impressive picture. The founder of the prize, John Charles Fields, considered two fundamental principles for the award:(a) the solution of a dicult problem and (b) the creation of a new theory enlarging the elds of applications of mathematics. Both these principles are important for the development of mathematics. It is quite clear that they are not independent. Very often the solution of a concrete dicult problem is based on the creation of a new mathematical theory and, conversely, the creation of a new theory may lead to the solution of an old classical problem. It is absolutely impossible to cover in a one-hour talk the results of Fields laureates even in a condensed form. In this talk I shall take a stroll through modern mathematics, giving a kaleidoscopic view of some exciting pictures. I shall try to explain the characteristic features of the mathematics of the 20th century, what kind of mathematics is considered important in this or that period, and how the results of the Fields medallists look from this point of view. The role of prizes, like the role of international recognition in general, is important for individual scholars. Despite Franz Neumanns beauti-

ful quote, The discovery of new truth is the greatest joy; recognition can add almost nothing to it, this wise idea is only partially true. According to Niels Bohr, the opposite conclusion is also valid. Recognition is especially important to young researchers. Selecting young mathematicians supports the continuing development of mathematics. The Fields Committees consist of outstanding mathematicians of the older generation, which makes their assessment of the creativity of the young all the more interesting. As I already mentioned, the rst Fields Medal was awarded in 1936, and the next one in 1950, so with one exception the medals are connected with the second half of the 20th century. The second world war greatly affected the development of society and science in general, mathematics especially. The development of mathematics is a good illustration of the more general thesis about the continuous but nondierentiable nature of the development of science. If we consider the graph of the development of mathematics, we evidently see the changes of interest in the periods of the world wars. It is natural for science to develop continuously, a fact based both on internal factors and the succession of generations. Also, science is characterized by some conservatism, which I consider in general as a robust phenomenon. Great ideas appear in the world by noiseless steps, as Nietsche said. The acceptance of new ideas proceeds against great obstacles and requires long testing. As Max Planck joked, a new scientic truth does not triumph by convincing its opponents and making them seeing the light, but rather because its opponents eventually die, and a new generation grows up 2

with it. That each tragic world war destroyed a whole generation of scientists accelerated in addition an apparently objective process to accept new points of view in mathematics. If we look at the prizes of 1936 and 1950 from this point of view we can see that new waves such as the explosion of interest in algebraic topology and geometry in the rst years after the Second World War are not yet reected in the rst postwar award. The 1950 prize was awarded to L. Schwartz (for the theory of distributions) and to A. Selberg for his remarkable achievements in number theory, namely, the distribution of zeros of Riemann function and anelementary proof of the asymptotic distribution of primes. But in 1954 the prize was awarded to K. Kodaira and J. P. Serre for postwar achievements. Hermann Weyl, who chaired the Fields committtee in 1954, delivered a speech on the papers of Kodaira and Serre. Curiously, Weyl had diculty distinguishing the areas of research of the two mathematicians. He said, The uninitiated may get the impression that our committee erred in awarding the Fields Medals to two men whose research runs on such closely neighboring lines. It is the task of the Committee to show that, despite some overlap in methods, they give the solutions of completely dierent, extremely dicult problems. In the subsequent awards, we see a denite balance between the two leading principles established by the founder of the prize. For example, in 1958 Klaus Roth was honoured for the proof of a delicate estimate that renes the Thue-Siegel theorem on the approximation of algebraic numbers by rational numbers. Roths theorem: If is any algebraic number, not it-

self rational, then for any > 2 the inequality p 1 | | < q q has only a nite number of solutions in rational p/q. The second medalist was Ren e Thom, who constructed a powerful method in topology known as the cobordism theory. In 1962 the prize winners were Lars Hrmander and John Milnor. o Hrmander developed the general theo ory of linear partial dierential equations, including hypoelliptic operators. The work of the other laureate was absolutely astonishing and has had great inuence on the future development of topology. It is very dicult to nd an analogous invention in the past to his beautiful construction of the dierent dierential structures on the seven-dimensional sphere. Later, the result became the cornerstone of a new branch of topology dierential topology. The original proof of Milnor was not very constructive but later E. Briscorn showed that these dierential structures can be described in an extremely explicit and beautiful form. Four medals were awarded in 1966. Among those honoured was Paul Cohen, who showed that if the ZermeloFraenkel axioms are consistent, then the negation of the axiom of choice or even the negation of the continuum hypothesis can be adjoined and the theory will remain consistent. It was the rst and the last time that the award was given to a specialist in mathematical logic. Alexander Grothendieck, one of the most original and puzzling mathematicians of our time, revolutionized algebraic geometry. The concept of schemes that 3

he introduced raised algebraic geometry to a new level of abstraction, beyond the reach of mathematicians with a traditional education. The theory of sheaves, spectral sequences, and other innovations in the late 1940s and earlier 1950s are subsumed by this complicated technique. But if certain mathematicians could console themselves for a time with the hope that all these complicated structures were abstract nonsense(in algebra, the term abstract nonsense has a denite meaning without any pejorative connotation), the later papers of Grothendieck and others showed that classical problems of algebraic geometry and the theory of numbers, the solutions of which had resisted eorts of several generations of talanted mathematicians, could be solved in terms of the Grothendieck K-functor, motives, l-adic cohomology, and other equally complicated concepts. Two remarkable mathematicians are present at this conference. The traditions of a scientic community are rather dierent from those of writers, movie stars, and fashion models. It is not an accepted practice to compliment a renowned scientist in his presence. So I really will not touch on the results of the mathematicians present here, but make some exception and say some words about the results of Steven Smale and Michael Atiyah, because they beautifully characterised the level of the prize and the realisation of its principles. The results of Smale are especially near to me, since I started my own career in mathematics as a student of the well-known Russian mathematician Dmitry Anosov, and his rst advice was to study the papers of Smale about dynamical systems.

S. Smale was honoured mostly for two of his achievements. The rst one is the solution of the Poincar e conjecture in higher dimensions. The Poincar conjecture is among the most e dicult problems in topology. It can be stated as follows in modern terms: Poincar conjecture A closed e smooth simply connected manifold M n with the homology groups of the sphere S n is homeomorphic to S n . Poincar stated his conjecture in e three dimensions. He believed that a stronger assertion was true, namely that M n is dieomorphic to S n . But as follows from the existence of Milnors exotic spheres, the conjecture is not true in this form. Smale proved a more general theorem on h-cobordism, from which it follows that Poincar conjece ture holds for dimensions n 5. In dimensions 5 and 6, a stronger conjecture is true: M n is dieomorphic to Sn. At rst sight it seems paradoxical that the proof of the Poincar cone jecture for higher-dimensional spaces is more accessible than for three- and four-dimensional manifolds. The reason is that a map of a surface into a manifold of fewer than ve dimensions cannot be approximated by an embedding. The situation is similar to the classication of manifolds. This indisputably classical result corresponds to the rst principle of the Fields award. The second achievement of Smale is connected with the theory of dynamical systems. This eld has its origin in classical mechanics and the theory of ordinary dierential equations. It was developed at the beginning of the twentieth century by H. Poincar, e G. D. Birkho, J. Hadamard, and I. Bendixson. In the middle 30s, remarkable results were obtained by E. Hopf, 4

G. Hedlund, M. Morse, A. Andronov, L. Pontryagin, and some others. But almost all of them were of a twodimensional nature. Smale substantially developed a multidimensional case. He showed that so-called structurally stable dynamical systems in higher dimensions have radically dierent properties. Unlike two-dimensional systems, studied by Andronov and Pontryagin, in a multidimensional situation structurally stable systems may have innite number of singular points, limit cycles, etc . His rst construction was the famous horseshoe, generated by discrete automorphisms of the torus. He proposed a very interesting hypothesis about the structural stability of geodesic ows on compact manifolds of negative curvature, later proved by Anosov. These results led to the creation of the theory of multidimensional dynamical systems, a new eld of mathematics still actively being developed. These results of Smale are an excellent illustration of the second Fields principle. The other laureate of this year, M. Atiyah, was recognised for his work in algebraic topology, especially for the proof of the index theorem which is known as the Atiyah-Singer Theorem. This theorem is remarkable from several points of view. Firstly, it generalized the long sequence of famous theorems begining with the Euler theorem on polyhedra and including the Riemann-Roch Theorem and the Poincar-H. Hopf Theorem about e the singularities of vector elds. The original proof of Atiyah and I. M. Singer was extremely complicated and used a wide spectrum of mathematical concepts developed in algebraic topology, geometry, and partial dierential equations in previous

years. Later, essential simplications were obtained and, especially remarkable, in recent years the relation between this theory and important problems in quantum eld theory, for example the problem of quantum anomalies, became clear. The work of Atiyah and Singer, Grothendieck, F. Hirzebruch, and many other mathematicians established a new eld of mathematics, where the ideas of algebraic topology and geometry and complex analysis are so interwined that traditional division is absolutely impossible now. Using a nice phrase Atiyah said, topologists used to study simple operators on complicated manifolds while analysts studied complicated operators on simple spaces. The time has arrived to study complicated operators on complicated spaces. These results not only raised mathematics to a very high level of abstraction, but proved the fruithfulness of these methods in the solution of long standing unsolved classical problems. One of the best examples is the solution by J. Adams of the famous problem of the existence of division algebras. From the time of Cayley, the following division algebras were known: real numbers, complex numbers, quaternions, and Cayley numbers. As the dimension grows we lose some properties, e.g. quaternions are

non-commutative. A natural question is: Are there other division algebras? The negative answer was obtained only in the 1960s and proved to be closely related to the following topological problem: nd all spheres on which the number of independent, continuous vector elds is equal to its dimension. There are only three such: S1, S3, S7. I hope that this gives at least a hint of how the two principles of Fields are linked in the work of M. Atiyah. Mathematics is a single subject, a fact that is not always obvious when you study the daily reality of research. It becomes clear, however, when you become acquainted with results of great mathematicians. This realization is one by-product resulting from an analysis of the works of the Fields medallists. Although honours went to authors of the greatest achievments obtained in the years immediately preceding each congress and sometimes in areas of mathematics widely separated from one another, truly wonderful connections between them were discovered with the passage of time. For that reason an -grid over the works of the Field medalists covers a signicant portion of the achievements of modern mathematics. Editors note: This article will be continued in the next issue.

Some Trends in Modern Mathematics and the Fields Medal

by Michael Monastyrsky
This is the second and concluding part of this article. The rst part appeared in last months issue. If we very quickly review the results of the Fields medallists, keeping in mind the fundamental principles of J. Fields, we can observe several interesting developments: 1. Allocation of stable elds of interest. 2. Succession of mathematics. 3. Zigzags of mathematical fashion. I will try to illustrate these theses with excerpts from the Fields medallists results. well known. If we consider the parallel development of mathematics and fundamental physics, we are astonished that the most revolutionary theories in 20th century physics are based on mathematics, which was especially developed for this purpose. It is enough to mention Einsteins special and general relativity based on the classical dierential geometry of Riemann spaces, quantum mechanics and Hilbert spaces and the theory of linear operators, the Schrdinger equao tion and spectral theory, and so on. This connection was broken, somewhere in the 30s, at the time of the solution of several more concrete problems in physics, when it seemed to physicists that most of their problems could be solved without the application of sophisticated and abstract modern mathematics. The development of pure mathematics in the period between the two world wars, and especially in the post-World War II period, was also characterized by weak connections with applied science, in particular with physics. This association was especially true of the areas of mathematics in which many Fields medalists worked. It was dicult to imagine that the concepts of sheaf, tale cohomole ogy, J-functor, and the like would ever be applied in physics. It was still more dicult to imagine that physics could assist algebraic topology and geometry.

1. Allocation of Interest

Indisputably, if we divide the mathematics of the second half of the century into two parts, the rst thirty years is mainly concentrated around problems of algebraic topology, algebraic geometry, and complex analysis. Here new concepts and methods appeared, and this evidently is reected in the list of Fields medallists. A denite change in this tendency, a return to the more classical topics, but of course on a new level, can be observed in mathematics from the end of the 70s. With some delay, this has been reected in the Fields medals awarded at the last two congresses. This point of view was widespread. It is important to note the new convergence between mathematics and The French mathematician Jean physics. The traditional contacts be- Dieudonn, one of the founders of e tween mathematics and physics are Bourbaki, expressed himself unam6

biguously on this subject in 1962. I would like to stress how little recent history has been willing to conform to the pious platitudes of the prophets of doom who regularly warn us of the dire consequences that mathematics is bound to incur by cutting itself o from applications to other sciences. I do not intend to say that close contact with other elds, such as theoretical physics, is not benecial to all parties concerned; but it is perfectly clear that of all the striking progress I have been talking about, not a single one, with the possible exception of distribution theory, had anything to do with physical applications. (Quoted from an address delivered at the University of Wisconsin in 1962, in which Dieudonn e gave a survey of the achievements of the preceding decade in pure mathematics. He emphasized algebraic topology, algebraic geometry, complex analysis, and algebraic number theory.) But as often happens with globally expressed opinions, the situation underwent a vast change ten years later. At the begining of the 70s, both in mathematics and physics, results were obtained that absolutely changed this point of view. Among the mathematicians who quickly understood the new opportunities and challenges hidden in the new physics were some Fields laureates. It is enough to mention S. Novikov, S. T. Yau, A. Connes, S. Donaldson, and E. Witten. Witten was the rst physicist to be awarded a Fields medal. Among the results of Fields laureates which were inspired by physical ideas, let us mention rst of all the work of Simon Donaldson. After the work of Milnor on dierential structures on S 7 , the paper of Donaldson appearing in 1983 had a 7

similar striking impact. Donaldson proved the existence of dierent dierential structures on simply-connected 4-dimensional manifolds. (Unfortunately the case of S 4 is not covered by his method and is still open.) Immediately after the work of Donaldson, in papers of R. Gompf and C. Taubes, the following remarkable result was proved: There exist an innite number of dierent dierential structures on R4 . This result, that the well-known space R4 hides such deep structures, is absolutely astonishing. It has very deep consequences for quantum gravity, where integrating over all metrics and so over different dierential structures is necessary. It is not less important than the proof, which is based on earlier discoveries in eld theory, mostly in the gauge theory of strong and weak interactions. Such interactions in the world of elementary particles are described by highly nonlinear equations with deep topological propertiesthe so called Yang-Mills equations. These equations were invented by the physicists C. N. Yang and R. Mills in 1954, but for many years were considered as nonphysical and attracted very little attention from physicists. Only the newest development of the theory of elementary particlesthe creation of the theory of weak and strong interactions based on the Yang-Mills equations led physicists to a deeper study of the structure of these equations. In the early 70s, the physicalmathematical union lessened the gap in the transmission of information, leading to the nal score in this striking mathematical achievement. These and other more recent results led to a new and deeper connection between mathematics and physics. The value

of this union for modern mathematics is indispensible and is based on a series of achievements of the rst rank. It is enough to mention the results of V. Drinfeld, M. Kontsevich, and many others.

ory of discrete groups and ergodic theory, he recently solved an old problem of the geometry of numbers: Oppenheims conjecture on the representation of numbers by indenite quadratic forms.

Pierre Deligne received the prize for a proof of a conjectures of A. Weil on zeta functions over nite elds. His reThe best conrmation of continuity sults are included as a special case of and fruitfulness in the development of the proof of the classical Ramanujan mathematics is the solution of deep conjecture. classical problems left by the previous generations of mathematicians. And Ramanujan Conjecture: Conhere the results of Fields medalists con- sider the parabolic form rm this idea nicely. The rst recipient of the Fields Medal was Jesse Douglas, who solved 2 12 (z) = x (1xn )24 = n xn n=1 the classical two-dimensional Plato n=1 problem. It is necessary to say that this problem was solved simultaneously by Tibor Rado, but Douglas so- where x = exp(2iz). Then |p | lution was considered as deeper and 2p11/2 for all primes p. could be applied to higher dimensions. The proof of Deligne is one of the Mathematicians of this mind-set inmost brilliant and striking examples of clude the famous number theorists like the unity and continuity of mathematA. Selberg, K. Roth, and A. Baker. In ics. It is striking in its beauty and comthe latest period, we see this tradition plexity, but required the application of in the works of Gregory Margulis and the wealth of techniques accumulated Pierre Deligne. in algebraic geometry over preceding The most important result of Maryears. gulis is his proof of Selbergs conjecThe last example which I give, but ture that a certain class of discrete subgroups of the group of motions of sym- only to mention in passing to supmetric spaces of higher rank with - port this thesis, is the proof of the nite volume is arithmetic. While the Moonshine hypothesis by Richard conjecture can be stated rather easily, Borcherds. Here the statement reits proof required a virtuoso mastery garding the relations between the coof the technique of the theory of alge- ecients of special modular forms, braic groups, use of the multiplicative dimensions of the representations of ergodic theorem, the theory of quasi- the Monster group and some inniteconformal mappings, and much more. dimensional Kac-Moody algebras led In recent years Margulis has examined to the proof by applying methods from the properties of discrete groups in dif- dierent elds of mathematics. It was ferent and sometimes unexpected ar- inspired by the recent development of eas. By combining ideas from the the- string theory.

2. Mathematical Succession

3. Zigzags in Mathematics
What I mean are the zigzags of mathematical fashion. I already talked about the domination of three mathematical diciplines in the list of Fields awards. Some reaction to this bias, even beside some objective background, appeared at the two last congresses. The awardees were mathematicians working in more classical elds. Let us mention here Jean Bourgain and Tim GowersBanach spaces, harmonic analysis, combinatorics; P-L. Lions partial dierential equations; JeanCristoph Yoccoz, Curtis McMullen dynamical systems, holomorphic dynamics; and the algebraist Em Zelmanov, who solved the classical restricted Burnside problem. This result capped o an extended period in group theory. J. Bougain and T. Gowers solved several classical problems in the theory of Banach spaces, discovered in very deep structures. J. C. Yoccoz and C. McMullen got important results in the so-called holomorphic dynamics. Here the study of sequences of mappings of complex sets led to the theory of dynamical systems. A typical problem of holomorphic dynamics is to describe the limiting sets of points of the mapping z R(z), where R(z) is a rational function and z is in C or C. Even the the study of sequences of iterations of such a seemingly simple map as fc (z) = z 2 + c conceals highly nontrivial results. This theory is placed at the meeting point of many beautiful mathematical theories, such as dynamical systems, Kleinian groups, FrickeTeichmller spaces and many others, u including computer graphics. This theory is very remarkable and

instructive if you look at it from a historical perspective. Created in the end of the 19th and the beginning of the 20th centuries in the works of the famous mathematicians P. Fatou, P. Montel, and G. Julia, it was seriously forgotten for more than forty years and was restored only in modern times. Now, besides being a very interesting theory, it has a wide eld of applications in physics. Let us mention the famous universality law of Feigenbaum which has important applications in turbulance. The unity of mathematics is shown best with these seemingly simple yet extraordinary complicated examples. To nish this very sketchy review of some of the achievements of modern mathematics in the light of Fields medals, let me say that the results honored by Fields medals substantially determined the development of mathematics in our time and its laureates are worthly representives of the mathematical community. Whether or not the Fields medal can be compared with Nobel prize, Fields idea of awarding it to the young has met with complete success. References 1. M. Monastyrsky, Modern Mathematics in the light of Fields medals, (with a forward by F. Dyson), A. K. Peters, Wellesley, 1998. 2. M. Monastyrsky, Modern Mathematics in the light of Fields medals (in Russian), Yanus-K, Moscow, 2000 (expanded and revised version of (1)).