servation and experience.

The "intui-
tionism" of Brouwer and Poincare, in-
sofar as it conceived of the natural
The Role of Intuition numbers as "intutitively given," seems
to proceed from this native intuition
(4).
Intuition plays basic and indispensable role in
a I do not believe that my teachers
had in mind anything like this native
mathematical research and in modern teaching methods. intuition. Moreover, I have always
doubted whether they ever tried to
R. L. Wilder analyze ju;st exactly what they meant
by "intuition." But I believe that they
associated it, in some way, with experi-
ence-mathematical experience, to be
more precise-and that the more ex-
I can recall that when I was a doc- problem of defining exactly what they perienced the mathematician became,
toral student, I was admonished again meant by "intelligence." Subsequently, the more reliable did his "intuition"
and again by my advisers, "Don't let these testers came up with a number, become. That is, mathematical intui-
your intuition fool you." I cannot, how- the "intelligence quotient" or I.Q., and tion, like intelligence, is a psychologi-
ever, remember just what I took this to it was demonstrated that the student cal quality which stems possibly from
mean; I probably thought it meant, with the higher I.Q. would, generally a hereditarily derived faculty, but
"Don't let your imagination lead you speaking, do better in his studies than which is, at any given time, principal-
astray; what you think is true may the one with lower I.Q. But the con- ly an accumulation of attitudes derived
very possibly turn out to be false." cept of the I.Q. as a measure of some- from one's mathematical experience.
One of my favorite articles in this thing called "native intelligence," that This should not be taken to mean
connection is a transcription of a lec- is, the intelligence bequeathed to the that mathematical intuition is some-
ture by Hans Hahn, entitled "The individual by his heredity, had to be thing which already contains one's at-
crisis in intuition," in the anthology abandoned. After numerous experi- titude toward a mathematical situation
edited by J. R. Newman entitled The ments, especially with -inductees dur- which one has never faced before. In-
World of Mathematics (1). This article ing World War I, it became clear that deed, in this day of widely diversified
echoes the warnings of my early teach- the cultural environment so modifies branches of mathematics, a mathema-
ers, and especially the admonition that this native intelligence as to render the tician may be expected to have little
"what you think is true may very pos- I.Q., at most, a measure of the com- or no intuition regarding a branch of
sibly turn out to be false." In fact, one bined effect of heredity and environ- mathematics in which he has never
can easily get the impression from ment on the individual's capacity for worked; his intuition is of use chiefly
Hahn's article that "intuition" is a learning, his perception, and his de- in those areas with which he has had
thoroughly unreliable guide and that gree of conformity to cultural direc- some experience. There is some agree-
one should regard it with suspicion tives. And an I.Q. could, over a period ment between this assertion, I think,
even when its every suggestion has been of time, be lowered or raised by the and the one with which Hahn con-
rigorously checked. environmental factors active during the cluded his article (1), namely, that in-
Now insofar as checking carefully period (2). tuition "is force of habit rooted in
the suggestions of one's intuition is Coming back to mathematical intui- psychological inertia." Like intelli-
concerned, no one would quarrel with tiion, we might expect to find an anal- gence, and I refer here to the kind
this, I believe. But as for intuition be- ogous situation. I believe that the in- that I.Q. testers have in mind, intui-
ing thoroughly unreliable, I am of the tuition about which some philosophers tion is greatly influenced, possibly whol-
opinion that this mental quality, what- speak is-if not wholly, at least partial- ly formed, by the cultural environ-
ever it is, has been too much maligned. ly-a "native intuition." Thus Des- ment-probably even more so than is
Indeed, I would go so far as to say cartes stated (3): "By intuition I un- intelligence. For I believe, in particular,
that without it, mathematical creation derstand, not the fluctuating testimony that the average nonmathematician has
would well-nigh cease, and modern of the senses, nor the misleading judg- no mathematical intuition at all, ex-
methods of teaching would be difficult ment that proceeds from the blunder- cept that nebulous quality of the mind
to justify. ing constructions of imagination, but which, if nourished by experience with
the conception which an unclouded mathematics, would develop into what
and attentive mind gives us so readily we call mathematical intuition.
Nature of Mathematical Intuition and distinctly that we are wholly freed
from doubt about that which we un-
In order to support these conten- derstand." And Kant, as I interpret Individual versus Collective Intuition
tions, it must be made clear just what him, conceived of the concepts of both
is meant, in mathematics, by "intui- time and space as deriving from an I have used the word "psychologi-
tion." a priori intuition which is independent cal" with reference to intuition (from
Not so long ago, those who were of experience. Among the more mod- here on, "intuition" will mean "mathe-
trying to test intelligence got into a ern philosophers, especially those of a matical intuition"). I wish to emphasize
predicament because they ignored the mystical bent, knowledge imparted by that my ultimate concern is with the
The author is research professor of mathematics this native intuition may be considered intuition of the individual mathema-
at theUniversity of Michigan, Ann Arbor. more valid than that gained from oib- tician. I am not unaware of the fact
5 MAY 1967 605
that concerning certain questions there of intuition-it is advisable, I believe, of research, the effects of which caused
is essentially what might be called a to look at some specific examples first. a virtual revolution in philosophical
collective or cultural intuition. For in- And since the manner in which intui- and mathematical thought.
stance, before Weierstrass gave his ex- tion exerts its influence varies accord- I have already mentioned the intui-
ample of a real continuous function ing to whether it is collective (cultur- tion, also false, which underlay the
having no derivative at any point of al) or individual, and whether it is conviction that every continuous func-
its interval of definition, probably al- true or false, I shall separate my ex- tion must have a derivative at some
most every mathematician felt intui- amples along those lines. point of its interval of definition. I
tively that such a function could not am confident that, if a study were
exist; this intuition had become a cul- made of the historical background pre-
tural attitude, a common belief. But Examples ceding publication of Weierstrass's ex-
consider the four-color map problem ample (6), it would be found that the
(5): I doubt if the average mathemati- Let us first consider the intuition influence of this false intuition had
cian today has any intuitive feeling apparently possessed by the Greeks, had its beneficial aspects. I can im-
regarding whether there exists or does and certainly by their medieval suc- mediately recall Lagrange's proposed
not exist a map that cannot be colored cessors, that the parallel axiom was method for calculating derivatives by
with at most four colors-simply be- true. I use the word "true" in the expanding functions in Taylor's series;
cause he has never worked on the absolute sense in which they seem to thereby making a start on the theory
problem. And an analogous statement have used it. This was an intuitive be- of analytic functions. If he had known,
can be made about the so-called "last lief possessed by all mathematicians, as we now know, that most continu-
theorem of Fermat," as well as a host since during the period involved every- ous function ("most" in the sense
of other problems. Before one can one who professd to be expert in of the Baire category) have no deriva-
have a really intuitive feeling about mathematics was expected to be famil- tives anywhere in the interval of defi-
such problems, one must have worked iar with Euclid's Elements. It was an nition, miight he not have been de-
on them. But everyone who has gone instance in which the collective intui- terred from proposing a method which
very far in mathematics will have tion was a false guide-a case typical he considered applicable to all contin-
worked with functions of a real vari- of those which Hahn cited in his arti- uous functions?
able and can be expected to have de- cle. It is interesting, however, to try To take a more recent case, con-
veloped an intuition for them. A simi- to assess the overall influence which sider the general "closed curve"; more
lar remark holds for the structure of this intuition had on mathematics. That specifically, a curve which is a com-
the real-number continuum. So far as an intuition was false is not sufficient mon boundary of two domains in the
those mathematical concepts that form reason to conclude that it was bad. plane. There was more general interest
part of the equipment of every mathe- And in this case, I believe that the in- in this topological configuration 65
matician are concerned, there can be fluence was highly beneficial. For if it years ago than now, since both the
expected to exist a kind of intuition had not been for the conviction that Jordan curve theorem and Peano's
that is common to most members of the parallel axiom could be proved space-filling curve had stimulated in-
the mathematical community. But as from the other axioms of Euclid-and terest in plane curves. Although we
soon as one goes beyond these con- this conviction was a direct result of now know quite simple examples of
cepts to mathematical specialties-par- the common intuition concerning its closed curves which have complemen-
ticularly to their frontiers-then the in- truth-then possibly the appearance of tary domains other than the two of
tuition becomes a quite individual af- the non-Euclidean geometries together which they are the common boundary
fair; and it is this intuition that is of with their effect on all mathematics (7), apparently around the turn of the
immediate importance in creative and philosophy might have been de- century the common intuition was that
work. layed. Of course the non-Euclidean ge- there could be only two such domains,
But this is entirely in accord with ometry would most certainly have been an "inside" and an "outside." Just how
the concept of mathematical intuition discovered sooner or later, if for no much influence the proving of the
as an accumulation of attitudes de- other reason than that the axiomatic Jordan curve theorem had on this in-
rived from one's experience. Regard- method, as Hilbert, for example, con- tuition we can only surmise. At any
ing matters of common knowledge, ceived it in his Foundations of rate, Schoenflies, who had recently
such as function theory, the attitudes Geometry, was already beginning to given such a proof and who could be
we acquire are determined by our emerge in the work of such mathemati- considered an expert on the topology
teachers, and the relation to the gen- cians as Boole, Hamilton, and others. of the plane, as well as one of the
eral mathematical culture of the time And someone would no doubt ulti- principal founders of the topology of
is apparent. But when one cultivates mately have experimented with alter- Euclidean spaces, published a number
a special area of interest, and especial- natives for the parallel axiom, just as of results in which he took it for
ly as he becomes involved in research Hamilton and Grassmann experimented granted, as intuitively clear, that a
on its frontiers, then one develops his with denials of the commutative laws closed curve could have only two
own attitudes in the light of his own of algebra. But because of the special complementary domains. Now this was
personal experiencs. Only he can make position held by Euclidean geometry, bad, of course; but was its influence
the educated guess,qsince he has devel- not only in philosophy, buit as part of on the development of mathematics
oped his own intuition. And although the general mathematical curriculum, bad? I think not. For example, it evi-
the connection with the current cul- the impact of the eventual realization dently came to the attention of L. E.
tural atmosphere is still traceable, it of the independence of the parallel J. Brouwer, the "father" of modern
is much less direct. axiom on the mathematical and phil- Intuitionism, and inspired him to look
Coming to my main topic-the role osophical community started a chain into the validity of the assumption.
606 SCIENCE, VOL. 156
I surmise that this helped arouse Brou- burst forth through the medium of sev- nately. Consequently I would like to
wer's interest in topology (although eral mathematical leaders, all working cite some cases where the intuition
that interest possibly had other stim- independently. In the case of the dis- was correct; but at the same time I
tili too), and that his classical work covery of incommensurability, some would like to consider how these cases
in this regard (giving counterexam- have attributed it to Pythagoras him- dovetail, so to speak, with the former
ples which included closed curves self, others to Hippasus (a student of cases in forcing the formulation of
which are the common boundaries of Pythagoras); but the truth is that no- new concepts.
an arbitrary countable number of do- body really knows. However, since Let me begin with the most basic
mains) influenced his continuing inter- the Pythagorean theorenm had become concept of all, namely number; more
est in topology. In particular, it led known quite generally at that time, specifically, the "counting" or "nat-
to his interest in the topological in- the incommensurability between the ural" numbers, 1, 2, 3, .... The
variance of closed curves. In the proof diagonal and side of a square could Intuitionist philosophy regards the ori-
of this, which he was the first to give, not have been long concealed, no mat- gin of these to be in man's intuition
he started a chain of ideas which led ter who first detected it. In the case of "fundamental series" of mental acts,
to the extension of homology theory of the parallel axiom, Gauss, Bolyai, consisting of a first act, a second act,
to general spaces. For several years and Lobachewski all discovered the a third act, and so on. I presume this
thereafter he was quite active in this facts at about the same time. We must have been an intuition which was
branch of mathematics, finding a num- now know that Bolzano had an exam- derived from the physical and cultural
ber of results which have become clas- ple similar to Weierstrass's. And about environment. More specifically-and
sical (such as his fixed point theorems the same time that Brouwer found his this can be inferred from a study of
and work on mappings of locally Eu- example of a "pathological" closed the forms of primitive number-words,
clidean manifolds), and which were curve, a Japanese mathematician, as well as of the practice of tallying
not fully appreciated in the mainstream Wada, also apparently produced one. --the use of one-to-one correspondence
of topology until over a decade later. And no one knows how many other to compare collections of physical ob-
individual mathematicians were either jects, along with the repetitive charac-
working on, or had produced exam- ter of the actual determination of such
Common Features of the Examples ples, proving the faulty character of correspondences, built up a set of at-
each of these collective intuitions. The titudes which formed, ultimately, the
All three of the examples of collec- discovery of the space-filling curve, intuition of fundamental series. And I
tive intuition that I have mentioned which I mentioned above only inci- presume this was an intuition on the
were false, yet it is difficult to believe dentally, was evidently another typi- cultural level, shared by virtually all
that their influence was entirely bad. cal case. Parametric representation of who found it necessary to engage in
It is curious how much good mathe- plane curves had proved extremely use- the primitive forms of counting. Prob-
matics can be done even when the col- ful, although evidently its introduction ably an analogous kind of intuition
lective intuition concerning basic mat- (possibly by Cauchy) was made under was involved in the genesis of geome-
ters is false. This is most striking dur- the influence of the intuitive belief try, where it became necessary to com-
ing the period preceding Weierstrass's that such curves would always be pare lengths and areas. All of this is
example. For during that period the curves of the intuitively accepted kind very conjectural, of course, but it seems
collective intuition concerning conti- -that is, having no "breadth" or "thick- fairly representative of what occurred
nuity, existence of derivatives, infinite ness." The usual pattern of events fol- prior to those periods for which the
series, the real-number system, and a lowed. After much good research based historical records are more complete.
host of other fundamental concepts on the concept, almost simultaneously And it is our earliest example of how
was at best faulty and usually full of Peano, E. H. Moore, and Hilbert came correct intuition on the collective level
error. Yet on such a basis much of up with examples showing that the in- serves to build the mathematical edifice.
classical analysis was built up. It would tuition underlying the concept of para- However, it was an intuition which
be quite as apt to speak of the "mod- metric representation was false (8). finally led to concepts that produced
ern miracle" as we do of the "Greek There followed a period of 40 years the "Greek crisis"; and it was neces-
miracle." or so of research in plane topology sary for Eudoxus and his contempo-
Speaking of the "Greek miracle" re- and problems related to it. The pat- raries to create a new conceptual frame-
calls the classical crisis regarding com- tern is quite typical. work which, while containing the ma-
mensurability. Here again, the collec- jor part of the old, rejected the parts
tive intuition regarding number and that had been found false. There fol-
magnitude, according to which all Role of Intuition in Evolution of lowed that flowering of activity that
magnitudes were commensurable, Concepts we call the "Greek miracle," based
though false, was able to support the on a new intuition of the number
creation of much good mathematics. What do these case histories indi- concept-the so-called geometric "mag-
Moreover, the ultimate discovery of cate concerning the manner in which nitudes"-which permitted a further
their true character led, in all these mathematical concepts evolve on the construction of mathematical theory
cases, to very fruitful periods of math- cultural level? I am asking this ques- atop the old of the Pythagoreans. Al-
ematical activity. It is my individual tion with a twofold purpose in mind. though couched in the language of
opinion that they all represent natural Thus far I have recounted only cases geometry, this intuition comprised vir-
phenomena in the evolution of mathe- in which the collective intuition was tually a complete theory of the real-
matics. In each case, the evidence is false-I have not given any cases number system. Unfortunately, the
strong that the discovery of the error where it was true-and mathematical course taken by Western culture pre-
in the basic intuition was about to intuition is not always wrong, fortu- cluded further development of the
5 MAY 1967 607
Greek intuition. And it was not until clusion that we can draw from all this, for many years and become "classi-
early times that activity in mathemati- namely, that so far as mathematics be- cal." The experienced individuals may
cal analysis, based on the foundation ing ultimately based on intuition is have stopped working on them, having
laid by the Greeks and their successors concerned, the Intuitionists are correct. found their efforts at solution frus-
(who added new symbolic representa- But the mathematical intuition, as I trated, and therefore have gone on to
tions for number), brought to light the have used the notion, is not precisely problems promising quicker results. I
inadequacy of the intuition created by that of Intuitionism; and, moreover, believe that what happens here is that
the work of Eudoxus. By the latter half the methods which the majority of the collective intuition in the field of
of the 19th century, real analysis had mathematicians use are not those of a particular problem continues to grow,
reached a more precise formulation of the Intuitionistic doctrine. being passed on by the older workers
the real-number continuum with the But to summarize the role of the to the younger. Ultimately, due to a
notion of set. The so-called "arithmeti- mathematical intuition in the evolution combination of a more mature collec-
zation of analysis" by Weierstrass and of mathematical concepts-our collec- tive intuition (which has been growing
others provided a new conception of tive intuition of basic concepts has unnoticed), new methods, and individ-
the real continuum and made possible grown by a series of discoveries of ual genius, someone (usually a young-
the theory of measure and the brilliant faulty features in the current concepts, er mathematician, relatively new in the
researches of the first half of the with ultimate replacement by new con- field, and possessing a fresh individual
present century in both analysis and cepts which not only clear up the intuition) is able to solve the problem.
topology. faults, but lead to feverish activity on That feeling of awe, which I am sure
But this new conception of the real the new foundation with consequent many older creative mathematicians
continuum gave birth to a new intui- creation of much good mathematics. must get regarding the powers of the
tion-that of the theory of sets. The Ultimately, the new concepts begin to younger generation of creative workers,
work of Cantor was the classical for- reveal faults; in particular, we dis- has a firm basis. The younger man
mulation of this new intuition. Some of cover that they have brought in with has not only come into the particular
its faults were discovered early, in the them new intuitions which have to be field without having to clutter up his
guise of the set-theoretic contradic- made more conceptually precise. And brain with concepts and methods which
tions. By now, the mathematical world the cycle goes on. served their purpose and are now dis-
had developed new standards of rigor, carded, but using new concepts and
and it was realized that the remedy methods he has built up an individual
must be sought in a more precise for- Role of Intuition in Research intuition which forms a platform from
mulation of the theory of sets. The axio- which he can regard his field of re-
matic method, used by the Greeks to I come now to the role of intuition search with an eye undimmed by the
avoid the Zeno paradoxes and the in research. The biographical com- recollection of earlier and faulty in-
commensurability assumption, was ap- ments in Poincare's writings, and the tuitions. The director of his first re-
proaching a new maturity and again more complete work of Hadamard search has no stronger responsibility
offered a method for attaining the de- (10), embody a good account of how than that of guiding and steering this
sired precision. For most ordinary pur- intuition works on the individual level young intuition into the most up-to-
poses, axiomatic systems for set theory in creative work. As I remarked be- date conceptual channels. It is almost
provided quite a satisfactory basis. But fore, this is intuition which is of a a truism that without intuition, there
so far as a unique formulation of gener- highly specialized nature. It relates to is no creativity in mathematics.
al set theory is concerned, we are today the particular problem on which only
in little better position than were the the individual, or a few individuals, are
Pythagoreans with respect to geometry, working. It is true, of course, that in Role of Intuition in Teaching
or the early analysts with respect to the their background is collective intuition,
real continuum. Our knowledge of the and they are certainly influenced by it. Like collective intuition, individual
axiom of choice, for instance, is pure- In particular, their choice of the prob- mathematical intuition is not a static
ly intuitive. We have an accumulation lems on which they work is guided by but a growing thing. It starts develop-
of good mathematics based upon its what the collective intuition deems the ing when we are children, during the
use, but we feel uneasy about its para- most fruitful direction for research. But time when we learn to distin-
doxical consequences, such as the once having selected the particular guish shapes and sizes (geometric in-
Banach-Tarski theorem (9). The same problem, the individual begins to build tuition) and to count (arithmetic in-
holds for the continuum hypothesis, new concepts and their resultant intui- tuition). We are not born with it, for
although this is perhaps not so serious tions. In a way, he repeats the experi- without a cultural basis for its develop-
for most of us. It does serve as a re- ence of the general mathematical cul- ment, there can apparently be no
minder, however, that our intuition of ture, but on a different level and at mathematical intuition. By the time
the real continuum was not thoroughly greater rates of change. His false intu- the child starts school in our culture,
clarified by the work of Weierstrass itions are usually recognized to be however, he usually has some basis
and his contemporaries. They, of neces- such in a relatively short time ("rela- to build on-his parents have probably
sity, brought into being a new intui- tively short" can be as much as sev- taught him to count, for example-
tion-the theory of sets-and so long eral years, of course), and they are and the continuing development of
as this theory has only an intuitive patched up by correct conceptual this basis undoubtedly forms one of
base, so must all the mathematics de- material. the central responsibilities of primary
pendent upon it. These remarks apply, too, in the teachers.
I think that there is only one con- case of problems that remain unsolved By the time the student reaches high
608 SCIENCE, VOL. 156
school he should have a fairly substan- his intuition. Probably the worst exam- of teaching-which exemplifies much
tial intuitive base from which to work. ple of this kind of thing is the writ- of what I have been saying-Moise
Presumably his teachers have used his ing of a definition on the board, then commented that "sheer knowledge does
arithmetic intuition to develop both explaining what it means and how it not play the crucial role in mathema-
higher arithmetic and algebra, and-at is used. tical development that most people sup-
least under new curricular ideas-his For example, consider the mathe- pose" (11). And a propos of the time
geometric intuition, not only to devel- matical induction principle. One can lost in using the Moore method, he
op elementary geometric facts but to proceed by first writing it on the black- stated: "The resulting ignorance ought
aid in solving arithmetic and algebraic board in the form in which it is usual- to be a hopeless handicap, but in fact
problems. And in this process, the ly stated, as an axiom; secondly, by it isn't; and the only way that I can
teacher should have added to the in- explaining what it means; and thirdly, see to resolve this paradox is to con-
tuitive base. In short, as the student by showing how to use it in proof clude that mathematics is capable of
comes to the high school teachers, his of simple arithmetic formulas. This is being learned as an activity, and that
mathematical equipment should have followed by homework in which the knowledge which is acquired in this
two main components-the intuitive student applies the process much as way has a power which is out of all
component and the knowledge com- a proof algorithm, imitating what the proportion to its quantity." And in the
ponent. These are difficult to separate, teacher has done. The brighter students second volume of Polya's recent book
particularly since the intuitive compo- will not have any trouble with this, Mathematical Discovery (12), there is
nent is dependent for its growth on perhaps, but the average ones will be the quotation from the 18th-century
the knowledge component. beset by minor questions such as: German physicist Lichtenberg: "What
Perhaps I can make this clearer by "How do I find the (n + 1) st term?" you have been obliged to discover by
stating my conception of what the -questions which are largely due to yourself leaves a path in your mind
new curricula being developed today the algorithmic character of what they which you can use again when the
should accomplish in contrast to the have been taught. need arises." This is an expressive way
old, standard, mathematical curricu- Now this kind of teaching is certain- of saying that you have added to the
lum. The old curriculum was designed ly not going to help the student rec- accumulation of your mathematical in-
chiefly for the knowledge component; ognize, when he later comes to a prob- tuition.
the student was taught how to per- lem in which mathematical induction The second matter that worries me
form arithmetic and algebraic opera- is a natural mode of proof or defini- is related to the use of the axiomatic
tions and how to prove theorems. But tion, that the mathematical induction method in secondary school teaching,
little conscious development of mathe- principle may be called upon. For particularly where the function is that
matical intuition took place; what there while he may "know" mathematical of definition. What I said before re-
was of this seemed to find expression induction he has not acquired any in- garding mathematical induction applies
chiefly in the problems that were tuitive feeling for it. If, on the other here. The student should not be intro-
given to be solved. But insofar as these hand, his teacher had given him credit duced to a theory by means of axioms.
were mechanical repetition of the opera- for knowing how to count and having Consider the arithmetic of integers.
tions or modes of proof that had been an intuition of "fundamental series," Here is a theory with which the stu-
taught, they added little or nothing to and if the teacher had proceeded from dent is already familiar, a circum-
the intuitive component. In contrast to there to guide him to the discovery of stance which makes it an excellent
this, the new curricula should try to the mathematical induction principle, subject for a proper introduction to
turn teaching of the knowledge compo- then the student would have acquired axiomatics. But before stating a single
nent into a process whereby the not only a knowledge of the principle, axiom, the teacher ought to respect
student's intuition is actually used and but also an intuitive base for later the student's imagination enough to tell
developed further in acquiring the new recognizing instances when the prin- him something about the purpose of
knowledge. ciple could be applied. In this way, axiomatics. In particular, he should be
For example, while under the old the intuition would be permitted to told that one wishes to seek out cer-
system the student was told the formu- play its proper role in creative teach- tain specific aspects of the arithmetic
la for carrying out a process, under ing. of integers from which the other as-
the new he should be invited to do a Perhaps most experienced teachers pects can be derived; for having done
little guessing as to what form the proc- already use such creative teaching meth- so, then not only can one test the ac-
ess should take. This guessing and ods, and they would not think of pre- curacy of an operation against the
the accompanying experimentation, re- senting a definition without first call- axiomatic base, but one can also try
sulting in a decision as to the final ing upon the student's intuitive powers out one's imagination by finding models
result, develops and strengthens his to help formulate the definition. How- for all or some of the axioms other
mathematical intuition. In an embryon- ever, two matters worry me: First, that than that of the arithmetic of integers
ic way, this procedure is precisely the they may find, under the pressure of -as, for example, the arithmetic of
same as that pursued by the research crowding a certain amount of material rationals, elementary algebra, and the
mathematician, and in my opinion the into a given amount of time, that it is like.
teacher who cultivates it is doing crea- necessary to resort to the old mode of After having decided to try to list
tive teaching. And I believe that all teaching which consists of (i) state- axioms, the student should be encour-
concepts should be introduced in this ment of the definition, (ii) explanation aged to discover suitable axioms him-
way. To explain a concept to a stu- of it, and (iii) application of the con- self-under the guiding hand of the
dent adds to his knowledge compo- cept to a particular problem. In dis- teacher, of course. It should go with-
nent, perhaps, but does not strengthen cussing the so-called "Moore method" out saying, however, that if these mat-
S MAY 1967 609
ters are too advanced to be compre- Summary appeals to the intuitive background al-
hended by the student, then the axio- ready developed. It is in this way that
matic method should not be introduced "Intuition," as used by the modern understanding and appreciation of new
at all. Nor should axioms be sneaked mathematician, means an accumulation mathematical knowledge may be prop-
in under the guise of so-called "laws" of attitudes (including beliefs and opin- erly instilled in the student.
presumably handed down by some ob- ions) derived from experience, both in-
References and Notes
scure mathematical Moses. dividual and cultural. It is closely asso-
Most of this applies, I believe, to ciated with mathematical knowledge, 1. J. R. Newman, The World of Mathematics
(Simon and Schuster, New York, 1956), pp.
college teaching-certainly up to the which forms the basis of intuition. This 1956-1976.
knowledge contributes to the growth 2. C. S. Yoakum, personal communication.
end of the first two years of college. As 3. R. Descartes, in The Philosophical Works of
the student goes on to more advanced of intuition and is in turn increased Descartes, E. S. Haldane and G. R. T.
Ross, Trans. (Cambridge Univ. Press, Cam-
work, the intuitive component of his by new conceptual materials suggested bridge, 1911).
training begins to assume more impor- by intuition. 4. R. L. Wilder, Introduction to the Foundations
of Mathematics (Wiley, New York, ed. 2,
tance. At this stage of his career it The major role of intuition is to 1965), pp. 211-212 and chap. 10. The Intui-
may be assumed that he is possibly provide a conceptual foundation that tionist philosophy holds that all mathematics
should be founded on the "intuitively given"
going on to do some kind of creative suggests the directions which new re- notion of sequence (such as a sequence of
acts) having a first element, a second ele-
work, if not in mathematics, then in search should take. The opinion of the ment, and so on. The sequence of natural
some other science. And it is desirable individual mathematician regarding ex- numbers, 1, 2, 3, ..., is the mathematical
model for such a sequence, and only con-
that his teachers have had some exper- istence of mathematical concepts (num- cepts that can be built on this sequence by
ience with creative work. This does ber, geometric notions, and the like) constructive methods can be accepted as
mathematically valid, according to Intuition-
not mean that the teacher must have are provided by this intuition; these istic tenets. Acceptance of these tenets would
a Ph.D. degree; this is a fetish I wish opinions are frequently so firmly held severely limit modem mathematical concep-
tualization, as well as rule out certain methods
we could get rid of. I would much as to merit the appellation "Platonic." of proof; proof of existence by contradiction
(contradictio ad absurdum) would not be
prefer a teacher without a Ph.D. who The role of intuition in research is acceptable, for example. In an even more
is excited about mathematics and can to provide the "educated guess," which restricted form, the tenets of Intuitionism are
to be found in the doctrines of the famous
teach creatively, than a teacher with may prove to be true or false; but in 19th-century mathematician, L. Kronecker.
a Ph.D. who is neither enthusilastic either case, progress cannot be made 5. See, for instance, P. Franklin, in Galols Lec-
tures (Scripta Mathematica Library, New
about mathematics nor capable of in- without it and even a false guess may York, 1941).
lead to progress. Thus intuition also 6. A simple example may be found in T. H.
spiring his students. Naturally, as the Hildebrandt, Amer. Math. Mon. 40, 547 (1933).
student progresses into graduate work, plays a major role in the evolution 7. Examples are described in R. L. Wilder,
Mathematics Teacher 55, 462 (1947).
most of his teachers will, as a matter of mathematical concepts. The advance 8. See J. W. Young, Fundamnental Concepts of
of course, have Ph.D.'s, since they of mathematical knowledge periodical- Algebra and Geometry (Macmillan, New
York, 1916), pp. 167-170. Young's statement
should themselves either be doing ly reveals flaws in cultural intuition; that this is a one-to-one correspondence be-
creative work, or at least have done these result in "crises," the solution tween the points of the line interval and the
square is false, however.
sufficient work to realize the role of of which result i;n a more mature in- 9. "A solid unit sphere can be decomposed into
a finite number of pieces which can be reas-
intuition in such work and the im- tuition. sembled to form two solid unit spheres." A
portance of using methods that will The ultimate basis of modern math- brief discussion may be found in E. Kasner
and S. Newman, Mathematics and the Imagi-
develop it. The student in the graduate ematics is thus mathematical intuition. nation (Simon and Schuster, New York, 1940),
stage should be capable of adding to and it is in this sense that the Intui- pp. 206-207; for technical details see S.
Banach and A. Tarski, Fund. Math. 6, 244
his knowledge component on his own; tionistic doctrine of Brouwer and his (1924); R. M. Robinson, (ibid. 34, 246 (1947)]
his mentor's responsibility is chiefly followers is correct. Modern instruc- shows that the minimal number may be 5.
10. J. Hadamard, The Psychology of Invention in
to nourish his mathematical intuition, tional methods recognize this role of the Mathematical Field, (Princeton Univ.
Press, Princeton, N.J., 1949).
for it is this that is going to be of intuition by replacing the "do this, do 11. E. E. Molse, Amer. Math. Mon. 72, 407
greater importance in his career as a that" mode of teaching by a "what (1965).
12. G. Polya, Mathematical Discovery (Wiley,
mathematician. should be done next?" attitude which New York, 1965), vol. 2, p. 103.

610 SCIENCE, VOL. 156