B’OR HA’TORAH 9 (1995) Zvi Victor Saks completed an MS in computer science at the State University of N ew York at Buffalo and a PhD in mathematics in 1972 at Wesleyan Univer sity. Having taught on a professorial level at various universities, including the University of Costa Rica and SUNY, Saks now conducts research and development in applications of artificial intelligence to solve real world problems at Carnegie Group in Pittsburgh, Pennsylvania. He has just completed a program of automated logistics distribution planning of commodities through a distribution network for the US Army, and is now working on a program that automatically schedules medical evacuation of wounded people from a military situation or a natural disaster. An internationally reputed topologist , Saks's speciality in abstract infinite space prepared him well for learning and teaching Kabbala and hasidic philosophy. He learned two years at the Hadar Ha'Torah Yeshiva and one year at Yeshiva Derekh Hayyim, both in Brooklyn Frequently lecturing at Habad activities - including the Ivy League Torah Study Program - Saks is currently teaching two community classes, "Reintroduction to Judaism: An Adult Approach to Judaism for Beginners" and "l:!asidism." On the Nature of Truth in Mathematics by Professor Zvi Victor Saks This paper was originally delivered at the Second International B'Or Ha'Torah Conference. It appeared in Hebrew in B'Or Ha'Torah 8H and has been revised by the author for publication here . Introduction Given that the focus of this conference is "Truth and Bias in Scientific Research," I am sure many of you are wondering how these issues can relate to mathematics. After all, in mathematics, the results of one mathematician can be verified by anyone else who is competent in that field. What room is there for bias and controversy in mathematics? As I will discuss in this paper, there has in fact been a great deal of controversy in mathematics. In this century, some of the greatest mathematicians have disagreed about what is valid mathematics. In addition, over the centuries, mathematicians have dramatically changed their view of what mathematics is and how it relates to "reality." In the modem view, a distinction is made between pure mathematics and applied mathematics. 1 In pure mathematics, the axioms of a deductive system are arbitrary assumptions, subject to the sole restriction that they must be consistent with each other, that is, they must not lead to a contradiction. Theorems are logical deductions from the axioms, and all have an implicit "if...then" form; if the axioms are true, then the theorems are true. There is no concern about whether or not the axioms or the theorems are actually true. 1 Irving Adler. A New Look at Geometry (New York : John Day, 1966) page 78. 96 On the Nature of Truth in Mathematics This is the prevailing view, although not all mathematicians are comfortable with it. For example, Richard Courant and Herbert Robbins say that: A serious threat to the very life of science is implied by. the assertion that mathematics is nothing but a system of conclusions drawn from definitions and postulates that must be consistent but otherwise may be created by the free will of the mathematician. If this description were accurate , mathematics 2 could not attract any intelligent person. Actually, this description is accurate as far as many pure mathematicians are concerned . I would say that for the mathematics that I specialize in, and surely for other branches of mathematics, no relationship with reality is required, or even relevant. In applied mathematics, a pure mathematical system is used as a model for some "real world" system of objects, in which the axioms are true. Then the theorems are true, and the pure mathematical system can be used to make inferences and calculations about the real world system. In this paper I shall go back to Greek mathematics, specifically geometry, and discuss some of the highlights in the evolution of geometry to the present with respect to the issues of truth and validity. Then I shall visit the domain of mathematical infinity, and discuss its evolution and the divergent views of some of the greatest mathematicians of the twentieth century regarding truth and validity. Finally, I shall discuss the hasidic view and offer some personal insights into these issues . B'Or Ha'Torah metry according to the system of Euclid . In fact, Euclid has been called the most successful textbook writer in history, since Euclidean geometry is still studied in high schools, with rather minor variations from his original text. In Euclid's view, axioms are "self-evident truths," that is, a description of "reality." There was no real distinction as far as truth is concerned between pure mathematics and applied mathematics. Euclid had one controversial axiom , the so-called "parallel postulate ." John Playfair gives the form of this axiom as follows: Parallel Postulate : Through a point not on a given line, there is one, and only one, straight line that does not intersect the given line .3 (See Figure 1.) Figure 1 point parallel line given line For around 2000 years , mathematicians tried to prove that this axiom was a logical consequence of Euclid's other axioms and that Euclid should not have assumed it as an axiom . The outstanding fact about all of these efforts is that they failed. The major breakthrough came from Karl Friedrich Gauss in about 1800. Gauss, who was probably the greatest mathematician of all time, realized that no one had successfully proved the parallel pos2 Geometry Virtually all of us have studied plane geo- 3 Richard Courant and Herbert Robbins . What Is Mathematics ? (Oxford , England : Oxford University Press, 1969) page XVII . Adler, page 198. B'Or Ha'Torah tulate as a consequence of the other axioms because it was really independent of the other axioms. Gauss assumed a negation of the parallel postulate, that there exist infinitely many parallel lines through a given point that is not on a given line, which are parallel to the given . line. This was the first example of non-euclidean geometry. The results may appear strange, but it has been shown that his new kind of geometry is as valid as Euclid 's. Thus Euclid was both vindicated and dethroned.4 Euclid was correct in assuming the parallel postulate, since it was not a consequence of his other axioms. However, his view that axioms are "self-evident truths" was incorrect. Other alternative geometries are possible and equally valid to his. An interesting aspect of Gauss's discovery is that he kept it secret, informing only a few close friends by letter of his new geometry. Irving Adler explains that Gauss kept his discovery secret because the then dominant philosophy of Immanuel Kant stated that "properties of space supposedly expressed in Euclidean geometry were necessary properties imposed on reality by the process of perception. " Gauss was afraid to buck the tide and expose himself to ridicule by people who were his intellectual inferiors .5 Thirty or forty years later, other mathematicians, most notably John Bolyai, Nikolai Lobatchevski, and Georg Riemann, developed different non-euclidean geometries and publicized them. Riemann's geometry is probably the most important of these non-euclidean geometries , since it was used by Albert Einstein as his model of the universe in his theory of relativity. His geometry can probably most easily be described by the following example. In Riemannian geometry, there are no parallel lines whatsoever. The simplest model of On the Nature of Truth in Mathematics 97 this geometry is the surface of a sphere, for which lines are defined as arcs of great circles. In other words, given two points on the surface of a sphere, there is exactly one arc going around the sphere, which - if filled in to make a circular plane - would go through those two points and the center of the sphere. That arc is defined to be the line going through those two points . As can be seen in Figure 2, any two such lines intersect in two points , so there are no parallel lines. Also notice that any two of the lines in the picture intersect in a right angle .6 Figure 2 p oint point The plane is th e ar c "fill ed in " whi ch goes through the center . Thus each of the angles in the triangle is a right angle, and the sum of the angles in a triangle is 270 degrees. That may seem strange, but it is a valid geometry, and the correct geometry for the surface of a sphere. 4 5 6 Adler, page 196. Adler, page 212. Morris Kline. Mathematics in the Modern World: Readings from Scientific American, with introductions by Morris Kline (San Francisco : W. H . Freeman, 1968), page 129. 98 On the Nature of Truth in Mathematics As I mentioned before, Einstein used Riemannian geometry to formulate his theory of relativity, and this is certainly one of the greatest examples of applying pure mathematics to solve physical problems. There are a variety of interesting issues concerning the applicabili'ty of pure mathematics, which I will discuss in the context of this example. First of all, as Kline asks, Was Riemannian geometry just right for the theory of relativity? Almost certainly not. There is ample reason to believe that Einstein merely did the best he could with the mathematics he found available.7 In other words, for pure mathematics to be applicable to a given situation, it must only provide a good enough model. It need not be perfectly applicable. Kline goes on to argue that the reason why Riemann's pure mathematics turned out to be useful several generations later was that Riemann was rooted in physical problems for his research. 8 Since Riemann was trying to understand physical space, it is natural that Einstein found his work useful. But I don't agree with Kline's opinion that pure mathematics must be grounded in physical reality in order to possibly be useful in the future. For example, let's look at what Philip David and Reuben Hersch say about groups . A group is an abstract mathematical structure , one of the simplest and most pervasive in the whole of mathematics. The notion finds applications, for example, to the theory of equations ...to crystallography, to atomic and particle physics. The latter applications are particularly interesting in view of the fact that in 1910 a board of experts including B'Or Ha'Torah Oswald Veblen and Sir James Jeans, upon reviewing the mathematics curriculum at Princeton, concluded that group theory ought to be thrown out as useless. So much for the crystal ball of experts. 9 Although group theory is quite abstract and not grounded in physical reality, within a generation it turned out to be tremendously useful. I agree with the attitude just quoted that it is impossible to predict which pure mathematics some genius has produced will be applicable in the future. History teaches us that there is pure mathematics being produced today which appears totally useless but some day will become useful. Mathematical Infinit y For many mathematicians, including myself, modern mathematics begins with the revolutionary work of Georg Cantor, who formulated the theory of mathematical infinity in the 1870s. Before Cantor, mathematical infinity was "infinity as a potential." For example , the sequence of positive integers 1, 2, 3,... is infinite because there is no last number. But it was only potential infinity because there was no concept of reaching the end of the sequence. Cantor's basic contribution was that mathematically infinite objects can be considered to be actual well-defined objects that can be manipulated in many of the same ways as finite objects. For example , the set of positive integers {1, 2, 3,...} is an actual object. This can be formulated as an axiom : 7 8 9 Kline, page 233. Kline, page 234 . Philip J. Davi s and Reuben Hersh . The Mathematical Experienc e (Boston : Birkhauser, 1981) page 204-205 . B'Or Ha'Torah Axiom of Infinity: There exists an infinite set, or more precisely, {1, 2, 3,...} is a set, that is, an actual object. Cantor proved many remarkable results, including the notion of different levels of mathematical infinity, so that some mathematical infinities are greater than others. But paradoxes and contradictions were found in Cantor's theory and there was a "crisis in foundations" in mathematics. Some mathematicians wanted to throw the whole theory out the window. An extreme example is L .E.J. Brouwer's intuitionism, which in the words of Kurt Godel, "is utterly destructive in its results." 10 Brouwer's school holds that "difficulties arise only where the totality of integers is involved in some way," rejecting the notion of actual infinity. 11 Others felt that the basis of the theory was sound and with proper axioms; the core would be preserved and the contradictions removed . The great mathematician David Hilbert said: Where there is any hope of salvage, we will carefully investigate fruitful definitions and deductive methods. We will nurse them, strengthen them , and make them useful. No one shall drive us out of the paradise which Cantor has created for us. 12 I would now like to discuss several specific aspects of the development of theory of mathematical infinity. Is Mathematics Complete ? One of the important features of Hilbert's program for mathematics was that mathematics is complete, that is, every mathemati- On the Nature of Truth in Mathematics 99 cal problem is solvable. 13 This was shown false in 1930 by Godel, who produced perhaps the most remarkable theorems ever proven . Godel proved that: No axiomatic system containing arithmetic can demonstrate its own consistency, so we can never know for sure whether our system is consistent. Any such system must have true statements which are unprovable within the system. These results were extremely disappointing to many mathematicians. They are a tremendous expression of the limitation of mathematics. Thus mathematics is never going to contain objective truth. Moreover, for any finite system of axioms, there will always be undecidable statements within the system. Thus a mathematical system will always be describing only a limited portion of its domain of investigation. On the other hand, Godel's results are a wonderful expression of the vitality of mathematics to have that level of self-knowledge and to be able to pinpoint its limitations so precisely. Now let's discuss the attempts made to put Cantor's theory on solid foundation. In 1905, Ernst Zermelo offered a list of axioms for (infinite) set theory. He noticed that the following assumption was made implicitly by Cantor and others, and proved 1 ° Kurt Godel. "What is Cantor's Continuum Problem?" in P. Benacerraf and H. Putnam , eds. Philosophy of Mathematics : Selected Readings (Englewood Cliffs : Prentice-Hall, 1964) page 261. 11 A. Heyting . Intuitionism, An Introduction (North Holland, 1971) page 14. 12 David Hilbert . "On the Infinite" in Benacerraf and Putnam, page 141. 13 Hilbert, page 150. 100 On the Nature of Truth in Mathematics some remarkable lences. and unintuitive equiva- Axiom of Choice: Given any collection of nonempty sets, an element can ·be chosen from each set. Intuitively the axiom of choice should be true, because if a set is nonempty, then we can simply choose an element from it. If the collection of nonempty sets is finite, then we can simply pick a set, choose an element from it, pick another set and choose an element from it, and so on until we have chosen an element from each set in the collection. On the other hand, if the collection of sets is infinite, then the method I have outlined here would not work. Since we are finite beings, we can never perform an infinite number of operations and make an infinite number of choices. It is important to point out that the axiom of choice is only necessary in the case when no rule is available to choose an element from each set. If a rule is available , then the rule makes the choice. Bertrand Russell explained the axiom of choice as follows. Suppose that we have an infinite set of people. If we want to take one shoe from each person, then we can choose, say, the left shoe. This is a rule which applies to each pair of shoes and allows us to choose one shoe from each pair. Thus in this case we do not need the axiom of choice. However, if we want to take one sock from each person, then since the two socks are indistinguishable from each other, there is no rule that applies to each pair of socks that allows us to select one sock from each pair. Thus in this case we would need the axiom of choice to make the arbitrary choices for us. The axiom of choice is very controversial and is interesting and important enough to have a whole book written about it. 14 David B'Or Ha'Torah Hilbert wrote that the axiom of choice is "the most attacked up to15the present in the mathematical literature ." Moore summarizes the different degrees of opposition taken by various mathematicians who opposed the axiom of choice. 16 *** I would like now to discuss some basic questions: Is the mathematics of actual infinity consistent? From a philosophical point of view, what allows us to make these assumptions ? The consistency question was solved by Godel in 1938 when he proved the following result. If the regular axioms of set theory (without the axiom of choice) are consistent, then they remain consistent if the axiom of choice is assumed. In other words, once we assume the other axioms, from a consistency point of view, we lose nothing by also assuming the axiom of choice. Since Godel had already proven that an absolute proof of consistency from within the system is impossible, this result of relative consistency is the best result possible. From a philosophic point of view, there is wide divergence concerning the validity of Cantor's theory of mathematical infinity. As I mentioned before, Brouwer completely rejects it. At the other extreme , for Bertrand Russell and many others consistency is the only criterion. Since Godel showed that the theory is as consistent as possible , from this point of view, the system is valid . Godel 14 Gregory H . Moore . Zermelo's Axiom of Choice (New York : Springer-Verlag , 1982). 15 Moore, Prologue. 16 Moore, pages 139-141. B'Or Ha'Torah defends the theory and, I believe, requires more than just consistency. Godel writes: However, . this (Brouwer's) negative attitude toward Cantor's set theory, and toward classical mathematics , of which it is a natural generalization, is by no means a necessary outcome of a closer examination of their foundations, but only the result of a certain philosophical conception of the nature of mathematics, which admits mathematical objects only to the extent to which they are interpretable as our own constructions or, at least, can be completely given in mathematical intuition. For someone who considers mathematical objects to exist independently of our constructions and of our having an intuition of them individually, and who requires only that the general mathematical concepts must be sufficiently clear for us to recognize their soundness and the truth of the axioms concerning them , there exists, I believe, a satisfactory foundation of Cantor's set theory in its whole original extent and meaning , namely axiomatics of set theory. 17 On the Nature of Truth in Mathematics 101 a word, it requires the existence of G-d. This was my feeling many years ago, when I was a graduate student very far away from Jewish observance. Now that I have become an observant Jew and have studied hasidic philosophy for the last 18 years, my conception of G-d has expand d significantly, and I realize that G-d is much greater than a being who can make the axiom of choice work. Nonetheless, I would like to explore with you the significant relationship between the validity of actual mathematical infinity and G-d. Let's start with Cantor. Cantor was a deeply religious person who had an intimate relationship with G-d. He insisted that actual mathematical infinity (which he called "transfinite") exists because G-d exists . Cantor writes: That an "infinite creation " must be assumed to exist can be proved in many ways ... . One proof stems from the concept of G-d . Since G-d is of the highest perfection, one can conclude that it is possible for Him to create a transfinitum ordinatum. Therefore, in virtue of His goodness and majesty we can conclude that there is actually a created transfinitum. ...the transfinite not only expresses the extensive domain of the possible in G-d 's knowledge, but also presents a rich and continually increasing field of ideal discovery. Moreover, I am convinced that it also achieves reality and existence in the world of the created , so as to express more strongly than could have been the case with a mere "finite world" the majesty of the Creator following His Own free will. 18 Two important points in Godel's defense of Cantor 's theory are: l) Mathematical objects exist independently of humans. 2) We must be able to recognize the soundness of our mathematics. I would like to go one step further. What would it take to make Cantor's theory of actual infinity valid in some real sense? Specifically with reference to the axiom of choice, what would it take to make an infinite number of individual, arbitrary choices? It requires some type of Creative Infinite Intelligence . In 17 18 Godel, page 262. Michael Hallett . Cantorian Set Theory and Limitation of Size (Claredon , 1984) pages 23-24 . 102 On the Nature of Truth in Mathematics Cantor is saying that since G-d can create infinity, then He did. Moreover, Cantor believed in the absolute truth of his set theory because, as he once told Gosta Mittag-Leffler, it had been directly revealed to him by G-d. 19 Joseph Dauben writes extensively about the fundamental link between Cantor's deep religious convictions and his perception of mathematics. Dauben also expresses surprise that these religious convictions have received so little attention in discussions on Cantor's development of set theory.20 Hasidic Philosophy and Mathematical Infinity I would now like to discuss some concepts from hasidic philosophy that have a direct bearing on the existence of actual mathematical infinity. With great pleasure I thank Rabbi Simon Jacobson for helping me with the analysis and studying some of the referenced works with me. Several Lubavitch rabbis, including Rabbi Jacobson, have asked me about a certain passage in the works of the third Lubavitcher Rebbe, Menahem Mendel Schneerson, known as the Tsemah Tsedek . The passage seems to contradict the existence of actual mathematical infinity: B'Or Ha'Torah where each troop is limited . These sources would seem to contradict the Tsemah Tsedek 's statement, because each world or troop is limited but there are infinitely many of them. 24 To resolve the seeming contradiction, the present Lubavitcher Rebbe, Menahem Mendel Schneerson, writes that the fact that there are infinitely many worlds has been revealed to us by our sages, but there is no contradiction because G-d's power is above all limitations and contradictions. 25 There is a basic premise that, in general, G-d creates the world in ways that fit together with human logic. So the Tsemah Tsedek's statement is true according to logic and will apply to any normal situation to which logic applies. But since G-d is completely unlimited, He can choose to create in ways which contradict logic. In the case of infinitely many worlds of troops, He used His unlimited suprarational power to create infinitely many of them, and then logic no longer applies to these particular cases. Let's recall that I am discussing the real validity of actual mathematical infinity. My claim is that since G-d has created infinitely many worlds, then mathematics has the right and the ability to postulate the existence of actual infinity because it really does exist. In 19 .. .it is impossible that many finite individuals should join together to form 21 an actual infinity. This statement by the Tsemah Tsedek seems to contradict the axiom of infinity, in which the actual infinite set {I, 2, 3,...} is composed of (infinitely many) individual numbers. However, there are numerous references in hasidic literature which refer to G-d 22 23 having created infinitely many worlds. • It is also stated in the Talmud that G-d created infinitely many troops of hosts to serve Him, Joseph W. Dauben . Georg Cantor, His Mathematics and Ph ilosophy of the Infinite (Cambridge, MA: Harvard University Press, 1979) page 232 . 20 Dauben, page 232. 21 Menahem Mendel Schneerson (the Tsemah Tsedek). Derekh Mitsvotekha (Brooklyn, NY: Kehot , 1973) page 113 (in Hebrew) . 22 Yosef Yitshak Schneerson . Bosi L'Gani (Brooklyn , NY : Sichos in English, 1980) page 55. 23 Schneur Zalman Schneerson. Tanya (Brooklyn, NY: Kehot, 1981) page 243. 24 Talmud, !l.agiga 13b. (Available in English from the Traditional Press.) 25 Menahem Mendel Schneerson. Likutei Sichos , vol. X (Br-;;oklyn, NY: Kehot , 1975) pages 178-179 (in Hebrew) . On the Nature of Truth in Mathematics B'Or Ha'Torah other words, the axiom of infinity is true because the collection of worlds created by Gd is infinite . What about other axioms involving mathematical infinity, for example, the axiom of choice? Is the axiom of choice true? While Gd certainly could have performed constructions that would make the axiom of choice true , is there any evidence that He actually did so? In the same letter quoted above, the Rebbe emphasizes that the world was created by G-d to conform to logic, and that we can make exceptions only in cases where the Torah has revealed to us that G-d made such an exception. Thus G-d used His infinite power and overcame human logic when he created infinitely many worlds. But in other situations in which the Torah did not reveal to us that G-d made such an exception, we cannot assume that He did . Notice that this contradicts Cantor's belief that just because G-d could create actual infinity, then he did . Thus although G-d certainly could have performed constructions which would make the axiom of choice true, I really don't know at this point if He did or not, especially when the axiom of choice is expressed in its full generality. Of course, these are difficult questions, which I believe are worth more thought. For further discussion on the relationship between the mathematics of actual infinity and hasidic philosophy, see my paper given at the first B'Or Ha'Torah conference .26 In preparing this paper, I looked around for contemporary professional literature linking G-d with the theory of actual mathematical infinity. In a fascinating book entitled The Mathematical Experience, Philip Davis and Reuben Hersh do just that, as follows: Mathematics, then , asks us to believe in an infinite set. What does it mean that an infinite set exists? Why should one 103 believe it? In formal presentation this request is institutionalized by axiomatization . Thus, in Introduction to Set Theory by Hrbacek and Jech, we read on page 54: "Axiom of Infinity. An Infinite set exists." Compare this with the axiom of G-d as presented by Maimonides (Mishneh Torah, Book l, Chapter 1): The basic principles of all basic principles and the pillar of all the sciences is to realize that there is a First Being who brought every existing thing into being. Mathematical axioms have the reputation of being self-evident, but it might seem that the axioms of infinity and that of G-d have the same character as far as self-evidence is concerned. Which is mathematics and which is theology? Does this, then, lead us to the idea that an axiom is merely a dialectical position on which to base further argumentation, the opening move of a game without which the game cannot get started? 27 While I agree with this perspective, I take a positive stance. I believe, or rather I know, that G-d exists and that actual infinity exists because G-d created it. The main new idea in this paper is that although we finite humans could not construct actual infinity, G-d created an actual infinity of worlds, and that therefore when mathematics asserts the existence of actual infinity, this assertion is true . 26 Zvi Victor Saks. "Applications of Mathematical Infinity on Jewish Philosophy" in A. Gotfryd, H . Branover , and S. Lipskar , eds. Fusion : Absolute Standards in a World of Relativity (Jerusalem/New York: Feldheim, 1990) pages 123-142. 27 Davis and Hersh, pages 154-155.