8 PART 1 I Science and Nonscience: Defining the Boundary

metaphysical pseudo-proposition as "the Absolute

enters into, but is itself incapable of, evolution and
bility as our criterion of significance, we are logically 2
obliged to treat these general propositions of law in
progress,"4 is not even in principle verifiable. For one the same fashion as we treat the statements of the
cannot conceive of an observation which would en- metaphysician.
able· one to determine whether the Absolute did, or In face of this difficulty, some positivists6 have KARL R. POPPER
did not, enter into evolution and progress. Of course adopted the heroic course of saying that these general
it is possible that the author of such a remark is using propositions are indeed pieces of nonsense, albeit an
English words in a way in which they are not com- essentially important type of nonsense. But here the Science: Conjectures and Refutations
monly used by English-speaking people, and that he introduction of the term "important" is simply an at-
does, in fact, intend to assert something which could tempt to hedge. It serves only to mark the authors'
be empirically verified. But until he makes us under- recognition that their view is somewhat too paradoxi- emplified by astrology, with its stupendous mass of
When I received the list of participants in this
stand how the proposition that he wishes to express cal, without in any way removing the paradox. Be- empirical evidence based on observation-on horo-
course and realized that I had been asked to speak
would be verified, he fails to communicate anything sides, the difficulty is not confined to the case of scopes and on biographies.
to philosophical colleagues I thought, after some
to us. And if he admits, as I think the author of the general propositions of law, though it is there revealed But as it was not the example of astrology which
hesitation and consultation, that you would proba-
remark in question would have admitted, that his most plainly. It is hardly less obvious in the case of led me to my problem I should perhaps briefly de-
bly prefer me to speak about those problems which
words were not intended to express either a tautol- propositions about the remote past. For it must surely scribe the atmosphere in which my problem arose
interest me most, and about those developments
ogy or a proposition which was capable, at least in be admitted that, however strong the evidence in favor and the examples by which it was stimulated. After
with which I am most intimately acquainted. I there-
principle, of being verified, then it follows that he has of historical statements may be, their truth can never the collapse of the Austrian Empire there had been
fore decided to do what I have never done before: to
made an utterance which has no literal significance become more than highly probable. And to maintain a revolution in Austria: the air was full of revolu-
give you a report on my own work in the philosophy
even for himself. that they also constituted an important, or unimpor- of science, since the autumn of 1919 when I first tionary slogans and ideas, and new and often wild
A further distinction which we must make is the tant, type of nonsense would be unplausible, to say the theories. Among the theories which interested me
began to grapple with the problem, 'W'hen should a
distinction between the "strong" and the "weak" very least. Indeed, it will be our contention that no theory be ranked as scientific?' or 'Is there a criterion Einstein's theory of relativity was no doubt by far
sense of the term "verifiable." A proposition is said proposition, other than a tautology, can possibly be for the scientific character or status of a theory?' the most important. Three others were Marx's the-
to be verifiable, in the strong sense of the term, if, anything more than a probable hypothesis. And if this The problem which troubled me at the time was ory of history, Freud's psycho-analysis, and Alfred
and only if, its truth could be conclusively estab- is correct, the principle that a sentence can be factu- neither, 'When is a theory true?' nor, 'When is a the- Adler's so-called 'individual psychology'.
lished in experience. But it is verifiable, in the weak ally significant only if it expresses what is conclusively ory acceptable?' My problem was different. I wished There was a lot of popular nonsense talked about
sense, if it is possible for experience to render it verifiable is self-stultifying as a criterion of signifi- to distinguish between science and pseudo-science; know- these theories, and especially about relativity (as still
probable. In which sense are we using the term cance. For it leads to the conclusion that it is impossi- ing very well that science often errs, and that pseudo- happens even today), but I was fortunate in those who
when we say that a putative proposition is genuine ble to make a significant statement of fact at all. science may happen to stumble on the truth. introduced me to the study of this theory. We all-the
only if it is verifiable?
I knew, of course, the most widely accepted an- small circle of students to which I belonged-were
It seems to me that if we adopt conclusive verifi-
swer to my problem: that science is distinguished thrilled with the result of Eddington's eclipse observa-
ability as our criterion of significance, as some pos-
NOTES from pseudo-science--or from 'metaphysics'-by its tions which in 1919 brought the first important con-
itivists have proposed, 5 our argument will prove too
empirical method, which is essentially inductive, pro- frrmation of Einstein's theory of gravitation. It was a
much. Consider, for example, the case of general 1. Tractatus l.og£co-Ph£losoph£cus, Preface. ceeding from observation or experiment. But this great experience for us, and one which had a lasting
propositions of law-such propositions, namely, as 2. Bradley, Appearance and Real£ty, 2nd ed., p. 1.
did not satisfy me. On the contrary, I often formu- influence on my intellectual development.
"arsenic is poisonous"; "all men are mortal"; "a 3. This example has been used by Professor Schlick to
lated my problem as one of distinguishing between a The three other theories I have mentioned were
body tends to expand when it is heated." It is of the illustrate the same point.
4. A remark taken at random from Appearance andRe- genuinely empirical method and a non-empirical or also widely discussed among students at that time. I
very nature of these propositions that their truth
al£ty, by F. H. Bradley. even a pseudo-empirical method-that is to say, a myself happened to come into personal contact with
cannot be established with certainty by any finite se-
5. E.g., M. Schlick, "Positivismus und Realismus," method which, although it appeals to observation Alfred Adler, and even to cooperate with him in his
ries of observations. But if it is recognized that such
Erkenntnis, Vol. I, 1930. F. Waismann, "Logische and experiment, nevertheless does not come up to social work among the children and young people in
general propositions of law are designed to cover an
Analyse des Warscheinlichkeitsbegriffs," Erkenntn£s, scientific standards. The latter method may be ex- the working-class districts of Vienna where he had
infinite number of cases, then it must be admitted Vol. I, 1930.
that they cannot, even in principle, be verified con- established social guidance clinics.
6. E.g., M. Schlick, "Die Kausalitiit in der gegenwiirti- It was during the summer of 1919 that I began to
clusively. And then, if we adopt conclusive verifia- gen Physik," Naturw£ssenschaft, Vol. 19, 1931. Conjectures and Refutations: The Growth of Scientific Knowledge
(London: Routledge and Kegan Paul, 1963), pp. 33-39. feel more and more dissatisfied with these three the-
Reprinted with permission. ories-the Marxist theory of history, psycho-analysis,
 10 PART 1 I Science and Nonscience: Defining the Boundary POPPER I Science: Conjectures and Refutations 11

and individual psychology; and I began to feel dubi- news, but also in its presentation-which revealed prediction, just then confirmed by the findings of 4. A theory which is not refutable by any con-
ous about their claims to scientific status. My prob- the class bias of the paper-and especially of course Eddington's expedition. Einstein's gravitational the- ceivable event is non-scientific. Irrefutability
lem perhaps fJrSt took the simple form, 'What is in what the paper did not say. The Freudian analysts ory had led to the result that light must be attracted is not a virtue of a theory (as people often
wrong with Marxism, psycho-analysis, and individual emphasized that their theories were constantly veri- by heavy bodies (such as the sun), precisely as ma- think) but a vice.
psychology? Why are they so different from physical fied by their 'clinical observations'. As for Adler, I terial bodies were attracted. As a consequence it 5. Every genuine test of a theory is an attempt
theories, from Newton's theory, and especially from much impressed by a personal experience. could be calculated that light from a distant fixed to falsify it, or to refute it. Testability is falsi-
the theory of relativity?' Once, in 1919, I reported to him a case which to me star whose apparent position was close to the sun fiability; but there are degrees of testability:
To make this contrast clear I should explain that did not seem particularly Adlerian, but which he would reach the earth from such a direction that the some theories are more testable, more ex-
few of us at the time would have said that we be- found no difficulty in analyzing in terms of his the- star would seem to be slightly shifted away from the posed to refutation, than others; they take, as
lieved in the truth of Einstein's theory of gravita- ory of inferiority feelings, although he had not even sun; or, in other words, that stars close to the sun it were, greater risks.
tion. This shows that it was not my doubting the seen the child. Slightly shocked, I asked him how he would look as if they had moved a little away from
truth of those other three theories which bothered 6. Confirming evidence should not count except
could be so sure. 'Because of my thousandfold expe- the sun, and from one another. This is a thing which
me, but something else. Yet neither was it that I when it is the result of a genuine test of the
rience,' he replied; whereupon I could not help say- cannot normally be observed since such stars are
merely felt mathematical physics to be more exact theory; and this means that it can be pre-
ing: 'And with this new case, I suppose, your rendered invisible in daytime by the sun's over-
than the sociological or psychological type of the- sented as a serious but unsuccessful attempt
experience has become thousand-and-one-fold.' whelming brightness; but during an eclipse it is pos-
ory. Thus what worried me was neither the problem to falsify the theory. (I now speak in such
What I had in mind was that his previous obser- sible to take photographs of them. If the same
of truth, at that stage at least, nor the problem of ex- cases of 'corroborating evidence'.)
vations may not have been much sounder than this constellation is photographed at night one can
actness or measurability. It was rather that I felt that new one; that each in its turn had been interpreted in measure the distances on the two photographs, and 7. Some genuinely testable theories, when
these other three theories, though posing as sci- the light of 'previous experience', and at the same check the predicted effect. found to be false, are still upheld by their
ences, had in fact more in common with primitive time counted as additional confirmation. What, I Now the impressive thing about this case is the admirers-for example by introducing ad hoc
myths than with science; that they resembled as- asked myself, did it confrrm? No more than that a risk involved in a prediction of this kind. If observa- some auxiliary assumption, or by reinterpret-
trology rather than astronomy. case could be interpreted in the light of the theory. tion shows that the predicted effect is definitely ab- ing the theory ad hoc in such a way that it
I found that those of my friends who were ad- But this meant very little, I reflected, since every sent, then the theory is simply refuted. The theory is escapes refutation. Such a procedure is al-
mirers of Marx, Freud, and Adler, were impressed conceivable case could be interpreted in the light of incompatible with certain possible results of ways possible, but it rescues the theory from
by a number of points common to these theories, Adler's theory, or equally of Freud's. I may illustrate observation-in fact with results which everybody be- refutation only at the price of destroying, or
and especially by their apparent explanatory power. this by two very different examples of human behav- fore Einstein would have expected. 1 This is quite dif- at least lowering, its scientific status. (I later
These theories appeared to be able to explain prac- ior: that of a man who pushes a child into the water ferent from the situation I have previously described, described such a rescuing operation as a
tically everything that happened within the fields with the intention of drowning it; and that of a man when it turned out that the theories in question were 'conventionalist twist' or a 'conventionalist
to which they referred. The study of any of them who sacrifices his life in an attempt to save the child. compatible with the most divergent human behavior, stratagem'.)
seemed to have the effect of an intellectual conver- Each of these two cases can be explained with equal so that it was practically impossible to describe any One can sum up all this by saying that the crite-
sion or revelation, opening your eyes to a new truth ease in Freudian and in Adlerian terms. According to human behavior that might not be claimed to be a rion of the scientific status of a theory is its falsifiability,
hidden from those not yet initiated. Once your eyes Freud the first man suffered from repression (say, of verification of these theories. or refutability, or testability.
were thus opened you saw confirming instances some component of his Oedipus complex), while the These considerations led me in the winter of
everywhere: the world was full of verifications of second man had achieved sublimation. According to 1919-20 to conclusions which I may now reformu-
the theory. Whatever happened always confirmed Adler the first man suffered from feelings of inferi- late as follows. II
it. Thus its truth appeared manifest; and unbeliev- ority (producing perhaps the need to prove to him-
ers were clearly people who did not want to see the 1. It is easy to obtain confirmations, or verifica-
self that he dared to commit some crime), and so did I may perhaps exemplify this with the help of the
manifest truth; who refused to see it, either be- tions, for nearly every theory-if we look for
the second man (whose need was to prove to himself various theories so far mentioned. Einstein's theory
cause it was against their class interest, or because confirmations.
that he dared to rescue the child). I could not think of gravitation clearly satisfied the criterion of falsifi-
of their repressions which were still 'un-analyzed' of any human behavior which could not be inter- 2. Confirmations should count only if they are ability. Even if our measuring instruments at the
and crying aloud for treatment. preted in terms of either theory. It was precisely this the result of risky predictions; that is to say, if, time did not allow us to pronounce on the results of
The most characteristic element in this situation fact-that they always fitted, that they were always unenlightened by the theory in question, we the tests with complete assurance, there was clearly
seemed to me the incessant stream of confirmations, confirmed-which in the eyes of their admirers con- should have expected an event which was a possibility of refuting the theory.
of observations which 'verified' the theories in ques- stituted the strongest argument in favor of these the- incompatible with the theory-an event Astrology did not pass the test. Astrologers were
tion; and this point was constantly emphasized by ories. It began to dawn on me that this apparent which would have refuted the theory. greatly impressed, and misled, by what they believed
their adherents. A Marxist could not open a newspa- strength was in fact their weakness. 3. Every 'good' scientific theory is a prohibi- to be confirming evidence-so much so that they
per without finding on every page confirming evi- With Einstein's theory the situation was strik- tion: it forbids certain things to happen. The were quite unimpressed by any unfavorable evi-
dence for his interpretation of history; not only in the ingly different. Take one typical instance-Einstein's more a theory forbids, the better it is. dence. Moreover, by making their interpretations
 KUHN I Logic of Discovery or Psychology of Research? 13
12 PART 1 I Science and Nonscience: Defining the Boundary

reason alone they are apt to seem to support those of this event. This is a characteristic and recurrent
and prophecies sufficiently vague they were able theories. Examples are Empedocles' theory of evo-
theories in the light of which they were interpreted. theme of such myths, but one which seems to have
to explain away anything that might have been a lution by trial and error, or Parmenides' myth of
But real support can be obtained only from obser- failed to attract the interest of the analysts, perhaps
refutation of the theory had the theory and the the unchanging block universe in which nothing not accidentally. (The problem of confirmatory
vations undertaken as tests (by 'attempted refuta-
prophecies been more precise. In order to escape ever happens and which, if we add another dimen- dreams suggested by the analyst is discussed by
tions'); and for this purpose criteria of refutation have
falsification they destroyed the testability of their sion, becomes Einstein's block universe (in which, Freud, for example in Gesammelte Schriften, ill, 1925,
to be laid down beforehand: it must be agreed which
theory. It is a typical soothsayer's trick to predict too, nothing ever happens, since everything is, observable situations, if actually observed, mean that where he says on p. 315: 'If anybody asserts that most
things so vaguely that the predictions can hardly fail: four-dimensionally speaking, determined and laid the theory is refuted. But what kind of clinical re- of the dreams which can be utilized in an analysis ...
that they become irrefutable. down from the beginning). I thus felt that if a the- sponses would refute to the satisfaction of the analyst owe their origin to [the analyst's] suggestion, then no
The Marxist theory of history, in spite of the se- ory is found to be non-scientific, or 'metaphysical' not merely a particular analytic diagnosis but objection can be made from the point of view of an-
rious efforts of some of its founders and followers, (as we might say), it is not thereby found to be psycho-analysis itself? And have such criteria ever alytic theory. Yet there is nothing in this fact', he sur-
ultimately adopted this soothsaying practice. In unimportant, or insignificant, or 'meaningless', or been discussed or agreed upon by analysts? Is there prisingly adds, 'which would detract from the
not, on the contrary, a whole family of analytic con- reliability of our results.')
some of its earlier formulations (for example in 'nonsensical'. 4 But it cannot claim to be backed by
cepts, such as 'ambivalence' (I do not suggest that 4. The case of astrology, nowadays a typical pseudo-
Marx's analysis of the character of the 'coming so- empirical evidence in the scientific sense-al- science, may illustrate this point. It was attacked, by
there is no such thing as ambivalence), which would
cial revolution') their predictions were testable, and though it may easily be, in some genetic sense, the Aristotelians and other rationalists, down to Newton's
make it difficult, if not impossible, to agree upon such
in fact falsified. 2 Yet instead of accepting the refuta- 'result of observation'. day, for the wrong reason-for its now accepted as-
criteria? Moreover, how much headway has been
tions the followers of Marx reinterpreted both the (There were a great many other theories of this made in investigating the question of the extent to sertion that the planets had an 'influence' upon ter-
theory and the evidence in order to make them pre-scientific or pseudo-scientific character, some which the (conscious or unconscious) expectations restrial ('sublunar') events. In fact Newton's theory of
agree. In this way they rescued the theory from of them, unfortunately, as influential as the Marxist and theories held by the analyst influence the 'clinical gravity, and especially the lunar theory of the tides,
refutation; but they did so at the price of adopting interpretation of history; for example, the racialist responses' of the patient? (To say nothing about the was historically speaking an offspring of astrological
a device which made it irrefutable. They thus gave interpretation of history-another of those impres- conscious attempts to influence the patient by pro- lore. Newton, it seems, was most reluctant to adopt a
a 'conventionalist twist' to the theory; and by this sive and all-explanatory theories which act upon posing interpretations to him, etc.) Years ago I in- theory which came from the same stable as for exam-
troduced the term 'Oedipus effect' to describe the ple the theory that 'influenza' epidemics are due to an
stratagem they destroyed its much advertised claim weak minds like revelations.)
influence of a theory or expectation or prediction astral 'influence'. And Galileo, no doubt for the same
to scientific status. Thus the problem which I tried to solve by pro-
upon the event which it predicts or describes: it will be reason, actually rejected the lunar theory of the tides;
The two psycho-analytic theories were in a dif- posing the criterion of falsifiability was neither a and his misgivings about Kepler may easily be ex-
remembered that the causal chain leading to Oedi-
ferent class. They were simply non-testable, ir- problem of meaningfulness or significance, nor a plained by his misgivings about astrology.
pus' parricide was started by the oracle's prediction
refutable. There was no conceivable human problem of truth or acceptability. It was the problem
behavior which could contradict them. This does of drawing a line (as well as this can be done) be-
not mean that Freud and Adler were not seeing tween the statements, or systems of statements, of
certain things correctly: I personally do not doubt the empirical sciences, and all other statements-
that much of what they say is of considerable im- whether they are of a religious or of a metaphysical
portance, and may well play its part one day in a character, or simply pseudo-scientific. Years later-
psychological science which is testable. But it does
mean that those 'clinical observations' which ana-
it must have been in 1928 or 1929-I called this first
problem of mine the 'problem of demarcation'. The
3
lysts naively believe confirm their theory cannot criterion of falsifiability is a solution to this problem
do this any more than the daily confirmations of demarcation, for it says that statements or sys-
which astrologers find in their practice. 3 And as tems of statements, in order to be ranked as scien- THOMAS S. KUHN
for Freud's epic of the Ego, the Super-ego, and the tific, must be capable of conflicting with possible, or
Id, no substantially stronger claim to scientific sta- conceivable, observations.
tus can be made for it than for Homer's collected Logic of Discovery or Psychology of Research?
stories from Olympus. These theories describe
some facts, but in the manner of myths. They con- NOTES known views of our chairman, Sir Karl Popper. 1 Or-
tain most interesting psychological suggestions, My object in these pages is to juxtapose the view of
1. This is a slight oversimplification, for about half of scientific development outlined in my book, The dinarily I should decline such an undertaking, for I
but not in a testable form. the Einstein effect may be derived from the classical Structure of Scientific Revolutions, with the better- am not so sanguine as Sir Karl about the utility of
At the same time I realized that such myths may theory, provided we assume a ballistic theory of light. confrontations. Besides, I have admired his work for
be developed, and become testable; that histori- 2. See, for example, my Open Society and Its Enemies, too long to turn critic easily at this date. Neverthe-
cally speaking all-or very nearly all-scientific ch. 15, section iii, and notes 13-14. Criticism and the Growth of Knowledge, Imre Lakatos and Alan less, I am persuaded that for this occasion the at-
theories originate from myths, and that a myth 3. 'Clinical observations', like all other observations, Musgrave, eds. (New York: Cambridge University Press,
1970), pp. 1-23. Reprinted with permission of the publisher. tempt must be made. Even before my book was
may contain important anticipations of scientific are interpretations in the light of theories; and for this