معضل پیوسته کانتور-کورت گودل

What is Cantor's Continuum Problem? Author(s): Kurt Godel Source: The American Mathematical Monthly, Vol. 54, No. 9 (Nov., 1947), pp. 515-525 Published by: Mathematical Association of America Stable URL: http://www.jstor.org/stable/2304666 Accessed: 13/11/2008 11:28

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=maa.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.

Mathematical Association of America is collaborating with JSTOR to digitize, preserve and extend access to The American Mathematical Monthly.

http://www.jstor.org

WHAT IS CANTOR'S CONTINUUM PROBLEM?

KURT GÖDEL, Institute for Advanced Study 1. The concept of cardinal number. Cantor's continuum problem is simply the question: How many points are there on a straight line in Euclidean space? In other terms, the question is: How many different sets of integers do there exist?

This question, of course, could arise only after the concept of "number" had been extended to infinite sets; hence it might be doubted if this extension can be effected in a uniquely determined manner and if, therefore, the statement of the problem in the simple terms used above is justified. Closer examination, however, shows that Cantor's definition of infinite numbers really has this char acter of uniqueness, and that in a very striking manner. For whatever "number" as applied to infinite sets may mean, we certainly want it to have the property that the number of objects belonging to some class does not change if, leaving the objects the same, one changes in any way whatsoever their properties or mutual relations (e.g., their colors or their distribution in space). From this, how ever, it follows at once that two sets (at least two sets of changeable objects of the space-time world) will have the same cardinal number if their elements can be brought into a one-to-one correspondence, which is Cantor's definition of equality between numbers. For if there exists such a correspondence for two sets A and B it is possible (at least theoretically) to change the properties and rela tions of each element of A into those of the corresponding element of B, whereby A is transformed into a set completely indistinguishable from B, hence of the same cardinal number. For example, assuming a square and a line segment both completely filled with mass points (so that at each point of them exactly one mass point is situated), it follows owing to the demonstrable fact that there exists a one-to-one correspondence between the points of a square and of a line segment, and, therefore, also between the corresponding mass points, that the mass points of the square can be so rearranged as exactly to fill out the line seg ment, and vice versa. Such considerations, it is true, apply directly only to physical objects, but a definition of the concept of "number” which would de pend on the kind of objects that are numbered could hardly be considered as satisfactory.

So there is hardly any choice left but to accept Cantor's definition of equality between numbers, which can easily be extended to a definition of "greater" and "less” for infinite numbers by stipulating that the cardinal number M of a set A is to be called less than the cardinal number N of a set B if M is different from N but equal to the cardinal number of some subset of B. On the basis of these definitions it becomes possible to prove that there exist infinitely many different infinite cardinal numbers or "powers," and that, in particular, the number of subsets of a set is always greater than the number of its elements; furthermore it becomes possible to extend (again without any arbitrariness) the arithmetical operations to infinite numbers (including sums and products with any infinite

515

516

WHAT IS CANTOR'S CONTINUUM PROBLEM?

(November,

number of terms or factors) and to prove practically all ordinary rules of computation.

But, even after that, the problem to determine the cardinal number of an individual set, such as the linear continuum, would not be well defined if there did not exist some "natural" representation of the infinite cardinal numbers, comparable to the decimal or some other systematic denotation of the integers. This systematic representation, however, does exist owing to the theorem that for each cardinal number and each set of cardinal numbers- there exists exactly one cardinal number immediately succeeding in magnitude and that the cardinal number of every set occurs in the series thus obtained. This theorem makes it possible to denote the cardinal number immediately succeeding the set of finite numbers by N. (which is the power of the "denumerably infinite" sets), the next one by N1, etc.; the one immediately succeeding all Ni where i is an integer, by Nw, the next one by Xwt1, etc., and the theory of ordinal numbers furnishes the means to extend this series farther and farther.

2. The continuum problem, the continuum hypothesis and the partila results concerning its truth obtained so far. So the analysis of the phrase "how many" leads unambiguously to quite a definite meaning for the question stated in the second line of this paper, namely, to find out which one of the X's is the number of points on a straight line or (which is the same) on any other contin uum in Euclidean space. Cantor, after having proved that this number is cer tainly greater than 8o, conjectured that it is 81, or (which is an equivalent proposition) that every infinite subset of the continuum has either the power of the set of integers or of the whole continuum. This is Cantor's continuum hy pothesis.

But, although Cantor's set theory has now had a development of more than sixty years and the problem is evidently of great importance for it, nothing has been proved so far relative to the question what the power of the continuum is or whether its subsets satisfy the condition just stated, except (1) that the power of the continuum is not a cardinal number of a certain very special kind, namely, not a limit of denumerably many smaller cardinal numbers, and (2) that the proposition just mentioned about the subsets of the continuum is

1 As to the question why there does not exist a set of all cardinal numbers, see footnote 12.

2 In order to prove this theorem the axiom of choice (see:A. Fraenkel, Einleitung in die Mengen lehre, 3rd ed. Berlin, 1928, p. 288 ff.) is necessary, but it may be said that this axiom is, in the present state of knowledge, exactly as well founded as the system of the other axioms. It has been proved consistent, provided the other axioms are so. (See my paper quoted in footnote 13.) It is exactly as evident as the other axioms for sets in the sense of arbitrary multitudes and, as for sets in the sense of extensions of definable properties, it also is demonstrable for those concepts of definability for which, in the present state of knowledge, it is possible to prove the other axioms, namely, those explained in footnotes 17 and 21.

3 See F. Hausdorff, Mengenlehre, 1st ed. (1914), p. 68. The discoverer of this theorem. J. König, asserted more than he had actually proved (see Math. Ann. 60 (1904), p. 177).

1947]

WHAT IS CANTOR'S CONTINUUM PROBLEM?

517

true for a certain infinitesimal fraction of these subsets, the analytical4 sets.5 Not even an upper bound, however high, can be assigned for the power of the continuum. Nor is there any more known about the quality than about the quantity of the cardinal number of the continuum. It is undecided whether this number is regular or singular, accessible or inaccessible, and (except for König's negative result) what its character of confinality* is. The only thing one knows, in addition to the results just mentioned, is a great number of consequences of, and some propositions equivalent to, Cantor's conjecture.

This pronounced failure becomes still more striking if the problem is con sidered in its connection with general questions of cardinal arithmetic. It is easily proved that the power of the continuum is equal to 240. So the con tinuum problem turns out to be a question from the “multiplication table" of cardinal numbers, namely, the problem to evaluate a certain infinite product (in fact the simplest non-trivial one that can be formed). There is, however, not one infinite product (of factors >1) for which only as much as an upper bound for its value can be assigned. All one knows about the evaluation of infinite products are two lower bounds due to Cantor and König (the latter of which implies a generalization of the aforementioned negative theorem on the power of the con tinuum), and some theorems concerning the reduction of products with different factors to exponentiations and of exponentiations to exponentiations with smaller bases or exponents. These theorems reduce the whole problem of com puting infinite products to the evaluation of XF (Ned and the performance of certain fundamental operations on ordinal numbers, such as determining the limit of a series of them. X.Sa), and therewith all products and powers, can easily be computed if the "generalized continuum hypothesis" is assumed, i.e., if it is assumed that 2a=Na+1 for every a, or, in other terms, that the number of subsets of a set of power Xa is X2+1. But, without making any undemon strated assumption, it is not even known whether or not m
3. Restatement of the problem on the basis of an analysis of the foundations of set theory and results obtained along these lines. This scarcity of results, even as to the most fundamental questions in this field, may be due to some ex tent to purely mathematical difficulties; it seems, however (see Section 4 below), that there are also deeper reasons behind it and that a complete solution of

ended (1935), p. 32. Even for compleave lif they are infinite)

4 See the list of definitions at the end of this paper.

5 See F. Hausdorff, Mengenlehre, 3rd ed. (1935), p. 32. Even for complements of analytical sets the question is undecided at present, and it can be proved only that they have (if they are infinite) either the power X, or Xi or continuum (see; C. Kuratowski, Topologie I, Warszawa-Lwow, 1933, p. 246.)

6 See W. Sierpinski, Hypothese du Continu, Warsaw, 1934.

* This reduction can be effected owing to the results and methods of a paper by A. Tarski published in Fund. Math. 7 (1925), p. 1.

8 For regular numbers na one obtains immediately:

Marina) = Maka=2* = 40+1.

518

WHAT IS CANTOR'S CONTINUUM PROBLEM?

(November,

these problems can be obtained only by a more profound analysis (than mathe matics is accustomed to give) of the meanings of the terms occurring in them (such as "set," "one-to-one correspondence," etc.) and of the axioms underlying their use. Several such analyses have been proposed already. Let us see then what they give for our problem.

First of all there is Brouwer's intuitionism, which is utterly destructive in its results. The whole theory of the N's greater than Xi is rejected as meaning less." Cantor's conjecture itself receives several different meanings, all of which, though very interesting in themselves, are quite different from the original prob lem, and which lead partly to affirmative, partly to negative answers;10 not everything in this field, however, has been clarified sufficiently. The "half intuitionistic" standpoint along the lines of H. Poincare and H. Weylli would hardly preserve substantially more of set theory.

This negative attitude towards Cantor's set theory, however, is by no means a necessary outcome of a closer examination of its foundations, but only the result of certain philosophical conceptions of the nature of mathematics, which admit mathematical objects only to the extent in which they are (or are be lieved to be) interpretable as acts and constructions of our own mind, or at least completely penetrable by our intuition. For someone who does not share these views there exists a satisfactory foundation of Cantor's set theory in its whole original extent, namely, axiomatics of set theory, under which the logical system of Principia Mathematica (in a suitable interpretation) may be subsumed.

It might at first seem that the set theoretical paradoxes would stand in the way of such an undertaking, but closer examination shows that they cause no trouble at all. They are a very serious problem, but not for Cantor's set theory. As far as sets occur and are necessary in mathematics (at least in the mathe matics of today, including all of Cantor's set theory), they are sets of integers, or of rational numbers (i.e., of pairs of integers), or of real numbers (i.e., of sets of rational numbers), or of functions of real numbers (i.e., of sets of pairs of real numbers), etc.; when theorems about all sets (or the existence of sets) in general are asserted, they can always be interpreted without any difficulty to mean that they hold for sets of integers as well as for sets of real numbers, etc. (respectively, that there exist either sets of integers, or sets of real numbers, or ... etc., which have the asserted property). This concept of set, however, according to which a set is anything obtainable from the integers (or some other well defined

See L. E. J. Brouwer, Atti del IV Congresso Internazionale dei Matematici (Roma 1908), p. 569.

10 See L. E. J. Brouwer, Over de gondslagen der wiskunde (Amsterdam and Leipzig, 1907) I, 9; III, 2.

11 See H. Weyl, Das Kontinuum, 2nd ed. 1932. If the procedure of construction of sets de scribed there (p. 20) is iterated a sufficiently large (transfinite) number of times, one gets exactly the real numbers of the model for set theory spoken of below in Section 4, in which the continuum hypothesis is true. But this iteration would hardly be possible within the limits of the half intuition

istic standpoint.

1947]

WHAT IS CANTOR'S CONTINUUM PROBLEM?

519

objects) by iterated application12 of the operation "set of,"18 and not some thing obtained by dividing the totality of all existing things into two categories, has never led to any antinomy whatsoever; that is, the perfectly "naive" and uncritical working with this concept of set has so far proved completely self consistent.14

But, furthermore, the axioms underlying the unrestricted use of this concept of set, or, at least, a portion of them which suffice for all mathematical proofs ever produced up to now, have been so precisely formulated in axiomatic set theoryis that the question whether some given proposition follows from them can be transformed, by means of logistic symbolism, into a purely combinatorial problem concerning the manipulation of symbols which even the most radical intuitionist must acknowledge as meaningful. So Cantor's continuum problem, no matter what philosophical standpoint one takes, undeniably retains at least this meaning: to ascertain whether an answer, and if so what answer, can be derived from the axioms of set theory as formulated in the systems quoted.

Of course, if it is interpreted in this way, there are (assuming the consistency of the axioms) a priori three possibilities for Cantor's conjecture: It may be either demonstrable or disprovable or undecidable.16 The third alternative (which is only a precise formulation of the conjecture stated above that the diffi culties of the problem are perhaps not purely mathematical) is the most likely, and to seek a proof for it is at present one of the most promising ways of attack ing the problem. One result along these lines has been obtained already, namely, that Cantor's conjecture is not disprovable from the axioms of set theory, pro

vided that these axioms are consistent (see Section 4).

It is to be noted, however, that even if one should succeed in proving its undemonstrability as well, this would (in contradistinction, for example, to the proof for the transcendency of 7) by no means settle the question definitively.

12 This phrase is to be understood so as to include also transfinite iteration, the totality of sets obtained by finite iteration forming again a set and a basis for a further application of the operation "set of."

13 The operation "set of x's" cannot be defined satisfactorily (at least in the present state of knowledge), but only be paraphrased by other expressions involving again the concept of set, such as: "multitude of x's," "combination of any number of x's," "part of the totality of x's"; but as opposed to the concept of set in general (if considered as primitive) we have a clear notion of this operation.

14 It follows at once from this explanation of the term "set" that a set of all sets or other sets of a similar extension cannot exist, since every set obtained in this way immediately gives rise to further application of the operation "set of" and, therefore, to the existence of larger sets.

15 See, e.g., P. Bernays, A system of axiomatic set theory, J. Symb. Log. 2 (1937), p. 65; 6 (1941), p. 1; 7 (1942), p. 65; p. 133; 8 (1943), p. 89. J. von Neumann, Eine Axiomatisierung der Mengenlehre, J. reine u. angew. Math. 154 (1925), p. 219; cf also: ibid., 160 (1929), p. 227; Math. Zs 27 (1928), p. 669. K. Gödel, The Consistency of the Continuum Hypothesis (Ann. Math. Studies No. 3), 1940.

16 In case of the inconsistency of the axioms the last one of the four a priori possible alterna tives for Cantor's conjecture would occur, namely, it would then be both demonstrable and dis provable by the axioms of set theory.

520

WHAT IS CANTOR'S CONTINUUM PROBLEM?

(November,

Only someone who like the intuitionist) denies that the concepts and axioms of classical set theory have any meaning (or any well-defined meaning) could be satisfied with such a solution, not someone who believes them to describe some well-determined reality. For in this reality Cantor's conjecture must be either true or false, and its undecidability from the axioms as known today can only mean that these axioms do not contain a complete description of this reality; and such a belief is by no means chimerical, since it is possible to point out ways in which a decision of the question, even if it is undecidable from the axioms in their present form, might nevertheless be obtained.

For first of all the axioms of set theory by no means form a system closed in itself, but, quite on the contrary, the very concept of set17 on which they are based suggests their extension by new axioms which assert the existence of still further iterations of the operation "set of." These axioms can also be formulated as propositions asserting the existence of very great cardinal numbers or (which is the same) of sets having these cardinal numbers. The simplest of these strong "axioms of infinity" assert the existence of inaccessible numbers (and of num bers inaccessible in the stronger sense) > 0. The latter axiom, roughly speak ing, means nothing else but that the totality of sets obtainable by exclusive use of the processes of formation of sets expressed in the other axioms forms again a set (and, therefore, a new basis for a further application of these processes).18 Other axioms of infinity have been formulated by P. Mahlo.19 Very little is known about this section of set theory, but at any rate these axioms show clearly, not only that the axiomatic system of set theory as known today is incomplete, but also that it can be supplemented without arbitrariness by new axioms which are only the natural continuation of the series of those set up so far.

That these axioms have consequences also far outside the domain of very great transfinite numbers, which are their immediate object, can be proved; each of them (as far as they are known) can, under the assumption of consistency, be shown to increase the number of decidable propositions even in the field of Diophantine equations. As for the continuum problem, there is little hope of solving it by means of those axioms of infinity which can be set up on the basis of principles known today (the above-mentioned proof for the undisprovability of the continuum hypothesis, e.g., goes through for all of them without any change). But probably there exist others based on hitherto unknown principles; also there may exist, besides the ordinary axioms, the axioms of infinity and

17 Similarly also the concept "property of set" (the second of the primitive terms of set theory) can constantly be enlarged, and furthermore concepts of "property of property of set" etc. be in troduced whereby new axioms are obtained, which, however, as to their consequences for proposi tions referring to limited domains of sets (such as the continuum hypothesis) are contained in the axioms depending on the concept of set.

18 See E. Zermelo, Fund. Math. 16 (1930), p. 29.

19 See: Ber. Verh. Sächs. Ges. Wiss. 63 (1911), pp. 190–200; 65 (1913), pp. 269–276. FromMahlo's presentation of the subject, however, it does not appear that the numbers he defines actually exist.

1947)

WHAT IS CANTOR'S CONTINUUM PROBLEM?

521

the axioms mentioned in footnote 14, other (hitherto unknown) axioms of set theory which a more profound understanding of the concepts underlying logic and mathematics would enable us to recognize as implied by these concepts.

Furthermore, however, even disregarding the intrinsic necessity of some new axiom, and even in case it had no intrinsic necessity at all, a decision about its truth is possible also in another way, namely, inductively by studying its "success," that is, its fruitfulness in consequences and in particular in "veri fiable" consequences, i.e., consequences demonstrable without the new axiom, whose proofs by means of the new axiom, however, are considerably simpler and easier to discover, and make it possible to condense into one proof many different proofs. The axioms for the system of real numbers, rejected by the intuitionists, have in this sense been verified to some extent owing to the fact that analytical number theory frequently allows us to prove number theoretical theorems which can subsequently be verified by elementary methods. A much higher degree of verification than that, however, is conceiv able. There might exist axioms so abundant in their verifiable consequences, shedding so much light upon a whole discipline, and furnishing such powerful methods for solving given problems (and even solving them, as far as that is possible, in a constructivistic way) that quite irrespective of their intrinsic neces sity they would have to be assumed at least in the same sense as any well established physical theory.

4. Some observations about the question: In what sense and in which direc tion may a solution of the continuum problem be expected? But are such consid erations appropriate for the continuum problem? Are there really any strong indications for its unsolubility by the known axioms? I think there are at least two.

The first one is furnished by the fact that there are two quite differently de fined classes of objects which both satisfy all axioms of set theory written down so far. One class consists of the sets definable in a certain manner by properties of their elements, 20 the other of the sets in the sense of arbitrary multitudes, irre spective of, if, or how they can be defined. Now, before it is settled what objects are to be numbered, and on the basis of what one-to-one correspondences, one could hardly expect to be able to determine their number (except perhaps in case of some fortunate coincidence). If, however, someone believes that it is meaningless to speak of sets except in the sense of extensions of definable prop erties, or, at least, that no other sets exist, then, too, he can hardly expect more than a small fraction of the problems of set theory to be solvable without making use of this, in his opinion essential, characteristic of sets, namely, that they are

20 Namely, definable "in terms of ordinal numbers" (i.e., roughly speaking, under the assump tion that for each ordinal number a symbol denoting it is given) by means of transfinite recursions, the primitive terms of logic, and the e-relation, admitting, however, as elements of sets and of ranges of quantifiers only previously defined sets. See my papers quoted in footnotes 13 and 19, where an exactly equivalent, although in its definition slightly different, concept of definability (under the name of "constructibility") is used. The paradox of Richard, of course, does not apply to this kind of definability, since the totality of ordinals is certainly not denumerable.

522

WHAT IS CANTOR'S CONTINUUM PROBLEM?

[November,

all derived from or in a sense even identical with) definable properties. This characteristic of sets, however, is neither formulated explicitly nor contained implicitly in the accepted axioms of set theory. So from either point of view, if in addition one has regard to what was said above in Section 2, it is plausible that the continuum problem will not be solvable by the axioms set up so far, but, on the other hand, may be solvable by means of a new axiom which would state or at least imply something about the definability of sets.21

The latter half of this conjecture has already been verified; namely, the concept of definability just mentioned (which is itself definable in terms of the primitive notions of set theory) makes it possible to derive the generalized continuum hypothesis from the axiom that every set is definable in this sense 22 Since this axiom (let us call it“A”) turns out to be demonstrably consistent with the other axioms, under the assumption of the consistency of these axioms, this result (irrespective of any philosophical opinion) shows the consistency of the continuum hypothesis with the axioms of set theory, provided that these axioms themselves are consistent.23 This proof in its structure is analogous to the con sistency-proof for non-Euclidean geometry by means of a model within Eu clidean geometry, insofar as it follows from the axioms of set theory that the sets definable in the above sense form a model for set theory in which furthermore the proposition A and, therefore, the generalized con tinuum hypothesis is true. But the definition of "definability" can also be so formulated that it becomes a definition of a concept of "set" and a relation of "element of” (satisfying the axioms of set theory) in terms of entirely different concepts, namely, the concept of "ordinal numbers,” in the sense of elements ordered by some relation of "greater" and "less," this ordering relation itself, and the notion of “recursively defined function of ordinals," which can be taken as primitive and be described axiomatically by way of an extension of Peano's axioms.24 (Note that this does not apply to my original formulation presented in the papers quoted above, because there the general concept of "set" with its element relation occurs in the definition of "definable set," although the definable sets remain the same if, afterwards, in the defini tion of "definability” the term “set” is replaced by "definable set."]

21 D. Hilbert's attempt at a solution of the continuum problem (see Math. Ann. 95 (1926), p. 161), which, however, has never been carried through, also was based on a consideration of all possible definitions of real numbers.

22 On the other hand, from an axiom in some sense directly opposite to this one the negation of Cantor's conjecture could perhaps be derived.

23 See my paper quoted in footnote 13 and note Proc. Nat. Ac. Sci. 25 (1939), p. 220. I take this opportunity to correct a mistake in the notation and a misprint which occurred in the latter paper: in the lines 25 to 29 of page 221, 4 to 6 and 10 of page 222, 11 to 19 of page 223, the letter a should be replaced (in all places where it occurs) by . Also, in Theorem 6 on page 222 the symbol "=" should be inserted between pa(*) and da(*'). For a carrying through of the proof in all de tails the paper quoted in footnote 13 is to be consulted.

24 For such an extension see A. Tarski, Ann. Soc. Pol. Math. 3 (1924), p. 148, where, however, the general concept of "set of ordinal numbers” is used in the axioms; this could be avoided, with out any loss in demonstrable theorems, by confining oneself from the beginning to recursively de finable sets of ordinals.

1947]

WHAT IS CANTOR'S CONTINUUM PROBLEM?

523

A second argument in favor of the unsolubility of the continuum problem on the basis of the ordinary axioms can be based on certain facts (not known or not existing at Cantor's time) which seem to indicate that Cantor's conjecture will turn out to be wrong,25 for a negative decision the question is (as just explained) demonstrably impossible on the basis of the axioms as known today.

There exists a considerable number of facts of this kind which, of course, at the same time make it likely that not all sets are definable in the above sense. 28 One such fact, for example, is the existence of certain properties of point sets (asserting an extreme rareness of the sets concerned) for which one has suc ceeded in proving the existence of undenumerable sets having these properties, but no way is apparent by means of which one could expect to prove the exist ence of examples of the power of the continuum. Properties of this type of sub sets of a straight line) are: (1) being of the first category on every perfect set, 27 (2) being carried into a zero set by every continuous one-to-one mapping of the line on itself.28 Another property of a similar nature is that of being coverable by infinitely many intervals of any given lengths. But in this latter case one has so far not even succeeded in proving the existence of undenumerable examples. From the continuum hypothesis, however, it follows that there exist in all three cases not only examples of the power of the continuum,29 but even such as are carried into themselves (up to denumerably many points) by every translation of the straight line.30

And this is not the only paradoxical consequence of the continuum hy pothesis. Others, for example, are that there exist: (1) subsets of a straight line of the power of the continuum which are covered (up to denumerably many points) by every dense set of intervals, or in other terms) which contain no undenumerable subset nowhere dense on the straight line, 31 (2) subsets of a straight line of the power of the continuum which contain no undenumerable zero set, 32 (3) subsets of Hilbert space of the power of the continuum which con tain no undenumerable subset of finite dimension,33 (4) an infinite sequence A' of decompositions of any set M of the power of the continuum into continuum

25 Views tending in this direction have been expressed also by N. Lusin in Fund. Math. 25 (1935), p. 129 ff. See also: W. Sierpinski, ibid., p. 132.

26 That all sets are "definable in terms of ordinals" if all procedures of definition, i.e. also quantification and the operation f with respect to all sets, irrespective of whether they have or can be defined, are admitted could be expected with more reason, but still it would not at all be justi fied to assume this as an axiom. It is worth noting that the proof that the continuum hypothesis holds for the definable sets or follows from the assumption that all sets are definable, does not go through for this kind of definability, although the assumption that these two concepts of definabil ity are equivalent is, of course, demonstrably consistent with the axioms.

27 See W. Sierpinski, Fund. Math. 22 (1934), p. 270 and C. Kuratowski, Topologie I, p. 269 ff.

28 See N. Lusin and W. Sierpinski, Bull. Internat. Ac. Sci. Cracovie 1918, p. 35, and W. Sier pinski, Fund Math. 22 (1934), p. 270.

29 For the 3rd case see: 1.c. 6, p. 39, Th. 1. 30 See W. Sierpinski, Ann. Scuol. Norm. Sup. Pisa 4 (1935), p. 43. 81 See N. Lusin, C. R. Paris 158 (1914), p. 1259. 32 See W. Sierpinski, Fund. Math. 5 (1924), p. 184. 33 See W. Hurewicz, Fund. Math. 19 (1932), p. 8.

524

WHAT IS CANTOR'S CONTINUUM PROBLEM?

[November,

many mutually exclusive sets A such that, in whichever way a set As, is chosen

for each i, II;(M-A) is always denumerable.34 Even if in (1)-(4) "power of the continuum" is replaced by "X" these propositions are very implausible; the proposition obtained from (3) in this way is even equivalent with (3).

One may say that many of the results of point-set theory obtained without using the continuum hypothesis also are highly unexpected and implausible.35 But, true as that may be, still the situation is different there, insofar as in those instances (such as, e.g., Peano's curves) the appearance to the contrary can in general be explained by a lack of agreement between our intuitive geometrical concepts and the set-theoretical ones occurring in the theorems. Also, it is very suspicious that, as against the numerous plausible propositions which imply the negation of the continuum hypothesis, not one plausible proposition is known which would imply the continuum hypothesis. Therefore one may on good reason suspect that the role of the continuum problem in set theory will be this, that it will finally lead to the discovery of new axioms which will make it possible to disprove Cantor's conjecture.

Definitions of some of the technical terms Definitions 4-12 refer to subsets of a straight line, but can be literally transferred to subsets of Euclidean spaces of any number of dimensions; definitions 13-14 refer to subsets of Euclidean spaces,

1. I call "character of confinality" of a cardinal number m (abbreviated by "of(m)") the smallest

number n such that m is the sum of n numbers
(i.e., if, although it is a limit of numbers
to finite numbers.] 4. A set of intervals is dense if every interval has points in common with some interval of the set.

The end-points of an interval are not considered as points of the interval.) 5. A zero-set is a set which can be covered by infinite sets of intervals with arbitrarily small

lengths-sum. 6. A neighborhood of a point P is an interval containing P. 7. A subset A of B is dense in B if every neighborhood of any point of B contains points of A. 8. A point P is in the exterior of A if it has a neighborhood containing no point of A. 9. A subset A of B is nowhere dense on B if those points of B which are in the exterior of A are

dense in B. Such sets A are exactly the subsets of the borders of the open sets in B, but the

term "border-set" is unfortunately used in a different sense. 10. A subset A of B is of the first category in B if it is the sum of denumerably many sets nowhere

dense in B.

34 See S. Braun and W. Sierpinski, Fund. Math. 19, (1932), p. 1, proposition (C). This proposi tion and the one stated under (3) in the text are equivalent with the continuum hypothesis.

35 See, e.g., L. Blumenthal, Am. Math. Monthly 47 (1940), p. 346.

19471

PERSONNEL AND TRAINING PROBLEMS IN STATISTICS

525

11. Set A is of the first category on B if the intersection A B is of the first category in B. 12. A set is perfect if it is closed and has no isolated point (1.e., no point with a neighborhood con

taining no other point of the set). 13. Borel-sets are defined as the smallest system of sets satisfying the postulates:

(1) The closed sets at Borel-sets. (2) The complement of a Borel-set is a Borel-set.

(3) The sum of denumerably many Borel-sets is a Borel-set. 14. A set is analytic if it is the orthogonal projection of some Borel-set of a space of next higher

dimension. [Every Borel-set therefore is, of course, analytic.) 15. Quantifiers are the logistic symbols standing for the phrases: "for all objects *" and "there

exist objects x." The totality of objects x to which they refer is called their range. 16. The symbol "e" means "the set of those objects x for which ..."

PERSONNEL AND TRAINING PROBLEMS IN STATISTICS

S. S. WILKS, Princeton University 1. Introduction. A report* has been prepared by the Committee** on Applied Mathematical Statistics of the National Research Council to: (1) analyze some factors contributing to the recent extraordinary growth of interest in the use of statistical methods, (2) present some information on the current and future needs of statistically trained personnel, (3) examine the impact of these needs on present training facilities, and (4) indicate some steps which might estab lish a training program adequate to meet these growing needs. This is a summary of the report.

2. Statistical organizations. As a simple indication of growth of interest in statistical methods, the Committee describes the formation and recent growth of statistical organizations. The American Statistical Association, founded more than 100 years ago, had a membership of 1700 in 1935. By the end of 1946 it had nearly 4000 members. The Institute of Mathematical Statistics, formed in

* "Personnel and Training Problems Created by the Recent Growth of Applied Statistics in the United States," a report by the Committee on Applied Mathematical Statistics, National Research Council, Washington, D. C., NRC Reprint and Circular Series No. 128, May, 1947, 17 pp.

** Luther P. Eisenhart, Chairman, Former Chairman Division of Physical Sciences, National Research Council

Samuel S. Wilks, Secretary, Professor of Mathematical Statistics, Princeton University Chester I. Bliss, Biometrician, Connecticut Agricultural Experiment Station Edward U. Condon, Director, National Bureau of Standards Harold O. Gulliksen, Professor of Psychology, Princeton University

Lowell J. Reed, Vice-President of the University, and Professor of Biostatistics, School of Hygiene and Public Health, The Johns Hopkins University

Charles F. Roos, President, The Econometric Institute, Inc. Walter A. Shewhart, Research Engineer, Bell Telephone Laboratories

Hugh M. Smallwood, Director, Department of Physical Research, General Laboratories, U.S. Rubber Company

Frederick F. Stephan, Professor of Sociology and Statistics, Cornell University