Now let us begin our course. The course is in scientific method—it is called ‘Introduction to Scientific Method’—and the first thing that I want to say as an introduction is that this is a subject that does not exist. Now the fact that this subject does not exist may strike some of you as strange. You are, after all, supposed to be taking a course in it. So this may require an explanation of its own. But the fact that the subject does not exist is something that I should know, since I happen to be the only professor of scientific method in the entire British Empire—a fact that I mentioned to you last time, and that may now suggest to you something about the nature of the subject itself.
There are, in any event, several senses in which the subject does not exist, two of which are important to understand.
First of all, it is important to understand that academic subjects in general do not exist. This is a very important point. Academic subjects do not exist, but are instituted by universities. They are
instituted by universities because universities have to pay their professors. And in order to pay their professors they have to appoint them. And so there has to be a certain fiction that one
appoints a person who is an expert in a subject. And since he is supposed to be an expert in a subject, the subject itself must somehow exist. But all of this is a fiction. What really exist are
problems—not subjects but problems. And when someone is interested in a problem and wants to solve it, then there is something that is really serious. Subjects, like history or
economics, are merely conveniences for university administrators: they are really more or less arbitrarily chosen collections of problems that have been ordered merely for the purpose of
administering universities and examining students!a function that I think a university should not fulfil, though that is another matter.
But scientific method exists even less than other subjects. The usual idea of scientific method is that there is a certain technique or procedure for making scientific discoveries, and that you can learn this technique and then apply it in science to make discoveries. According to the usual view, a man either knows scientific method and makes scientific discoveries, or he doesn’t know scientific method and makes no discoveries!unless, of course, he should
happen to make some by chance. Now a person may make discoveries by chance. But a scientist, according to the usual view, does not make discoveries by chance. That is what makes him a scientist: his mastery of the scientific method.
This, again, is the usual view. But now I want to say something in criticism of it. If this is what is meant by ‘scientific method’, then the subject certainly does not exist. For there simply are no
techniques or procedures for making discoveries in science. Contrary to the usual view, there are many examples of scientists —and many very great scientists—who have only once in their
lives made a major discovery. These are scientists who—even though they continued to be in the forefront of science and were able to criticize other people’s discoveries in a most excellent way, and even though they had problems that they would have loved to solve, and surely tried to solve—were never again able to make a major discovery.
Now this would be inexplicable if being a great scientist consisted in being an expert in scientific method.
So what about these people? If the usual view were true, then the only explanation would be that they had become too old, or had forgotten the scientific method, or something like that. But in many cases this obviously just isn’t true.
Consider Max Planck. Max Planck was surely one of the greatest German physicists, and probably one of the greatest physicists of all time. But Planck made only one great discovery in theoretical physics. In 1900, or thereabouts, he discovered what is called ‘the quantum of action’, which is the basis of quantum theory and of all atomic theory. Planck lived for nearly 50 years after he made this discovery. But he never made another one. He made very interesting contributions to scientific discussions, and he wrote very interesting books. But he made no further scientific discovery, although he surely tried. Now either Planck had forgotten his scientific method after he made his discovery, or he made his discovery by chance and without using scientific method at all. But in either case, the example of Planck just does not fit the usual idea of scientific method. And Planck is just one example among many.
Albert Einstein, who was certainly one of the greatest physicists of all time, is another. Einstein made several important discoveries. He discovered the theories of both general and special relativity. And he won the Nobel Prize in 1921 for his discovery of the photoelectric law. But Einstein did not solve the problem that was closest to his heart—the so-called problem of unified field theory —even though he worked on it for forty years. If there were something like a scientific method, then Einstein would surely have mastered it and would surely have solved the problem of unified field theory. So the thing isn’t really like that.
We have, in fact, quite a few examples of people who made a great discovery in science and then failed to add anything to it. You may, perhaps, say that this is due to a hardening in the arteries. But I do not think that this is a very good explanation. A better explanation is that there simply is no such thing as scientific method. So let me say to those of you who are taking this course in order to learn the method of science and how to use it to make scientific discoveries: You are all excused and may now leave the room. For there simply is no scientific method that yields scientific discoveries in this sense.
Now in so far as our subject does not exist, we need to explain how it is that there should be an entire course devoted to it. This course must have been introduced into the London School of Economics under a false assumption. It was, in fact, introduced by Sidney and Beatrice Webb who, as you may know, founded the LSE. The Webbs believed in scientific method, and they actually wrote a book on the methods of social science. They introduced the subject because they believed that the methods in the natural sciences were further advanced than those in the social sciences, and that social scientists could thus learn something from studying the methods of the natural scientists. In this they followed John Stuart Mill, who had published very similar ideas in his Logic. And I want to say from the beginning that John Stuart Mill was a very great man. He was a great economist and also a very great political philosopher. As many of you know, he wrote a book called On Liberty. This is a very great book, and one that everyone should read. It is really a book of lasting importance. But Mill was not terribly good, even for his time, as a logician and methodologist. He had a highly optimistic view of scientific method. And the Webbs would have done better not to have followed him.
But the interesting thing is that it is not only Mill and the Webbs. The majority of philosophers and working scientists also believe that there is something like a method of scientific discovery, and also that this method is better developed in the natural sciences than it is in the social sciences. This is, as I’ve said, the usual view, and I am, in fact, nearly alone in saying that it is not true. Nevertheless, they too are quite mistaken about it. But, of course, by this time many of you may be asking yourselves why you should now believe what I have to say about it? My answer is that you shouldn’t. I warned you last time against believing me. If there is something that I am now saying that you don’t understand, or if you think that what I am now saying is false, then you should challenge me, and force me to clarify my views.
What do most scientists believe about scientific method? I am sorry to say that most of them, fortunately not all, believe the following story. They believe that science begins somehow or other with a collection of material. This material is, in the main, observational—if not actual observations, then perhaps statistics collected from questionnaires. They believe that we collect this material, and that we then work on it—and working on it is, again, partly statistical—in order to discover some of its general characteristics.
Now this is a procedure that can be organised on a very large scale. People do. You get together a lot of money, maybe a grant or something, and you hire people. You pay them so much, and they go about collecting material. You then organise this material, and you feed it into a computer, and somehow out come the results. This, I am afraid, is especially the way in which social scientists view the method of the natural sciences. And the main idea behind it is that we can somehow or other get from the observations to some general results.
But all of this is just nonsense. And we can all make an experiment here and now to show that it is nonsense. I think that most of you have brought along a pencil and paper. I want you to take your pencil and your paper, and I want you to observe. I want you to do it carefully, and I want you to write down your observations, for, say, the next three minutes, beginning now.
What’s the matter? Is there a bit of a problem here? Oh yes, I forgot to tell you what to observe. But you see, that's just my point. In order to observe, you have to know what to observe. You must, in other words, have a theory, or at least a problem. So either I will have to tell you what to observe, or you will tell yourself what to observe without, perhaps, even being aware that you are doing so.
You see, my point is very simple: the idea that science can start with observations is in itself already impossible. You can’t start with observations. You have to start with problems. You have to start with problems, because only the problem will tell you what to observe. Only the problem will tell you what observations will be interesting or relevant. So what I said to you before—that only problems and not subjects exist—I now want to repeat again. There exist problems, there exist problem situations, situations in which we are forced to consider certain problems—and when someone becomes interested in a problem and tries to solve it, then we have something that is really serious.
Now, you can rightly ask, ‘Where do the problems come from? But I would say that there are all sorts of possibilities. Our problems may arise simply out of our need to live: they may be problems of finding food and shelter, or of improving the food and shelter that we have already found. Or they may arise out of more theoretical discussions: you may have read something in a book, or you may have talked to somebody, and a difficulty arose. Certainly our problems do not arise from nothing. They arise out of our various situations in life and, especially, from our contact with other people. In any event, I would suggest that observation is normally preceded by some problem, and that our observations are then relevant to this problem. It is quite true that certain problems may be the result of an observation—if, for example, you have a theory that led you to expect something else!but I am quite sure that no problem ever starts with collecting observations, as opposed to a chance observation, or something like that.
But most philosophers and scientists want to say that science is empirical, or observational, rather than philosophical, or speculative. ‘Science’, they say, ‘does not start with books—books are the result of science, not the beginning.’ In philosophy you are supposed to make books out of other books. But in science you are supposed to make experiments, and to observe, and to collect these observations—and only then are you supposed to make books.
According to these people, the main source of scientific observation is not what other people have thought but what you have observed. Now I am, in this regard, a great admirer and friend
of the physicist Erwin Schrodinger. But he too has said that science begins with sense-data, which are supposed to be observation reports like ‘Here’s a red patch’ and ‘There‘s a red patch’. He says, in fact, that science doesn’t really do anything else but find correlations between the observations that we make—statistical correlations and the like. And he says, to my great surprise, that nobody ever contests this view.
So there are some very good people who believe this story. But I would answer to all of them that it is, in my opinion, a complete mistake, a myth—a full-blown myth, very widely believed, but, strangely enough, it does not resemble the truth at all. I would say that the truth is that science always starts with problems and goes on with problems and that the problems are in the main very often found in books.
Now here you may say, ‘But what about the first books‘? Surely the problems in them did not come from books.’ And this, of course, is true. But I would suggest that the first books found their problems in some unwritten myths—not in books, but in myths and stories nevertheless. They may, for example, be creation stories. And they may be very primitive, like the beautiful story that the Maoris tell about the creation of the world. The Maoris, as you probably know, are the inhabitants of New Zealand. And New Zealand, as you probably know, is a longish island, or rather two islands, in the Pacific. So the Maoris have the wonderful story that one of their gods went out fishing one day, and he caught a fish, and he began to pull, and it was a terribly heavy fish, and he pulled and he pulled and he pulled, and in this way he finally pulled this
longish fish that we call New Zealand out of the Pacific.
The story, of course, is very primitive. But it doesn’t matter. The beginning of science is really the beginning of cosmology, and the beginning of cosmology was very primitive. So one can begin to ask some awkward questions!because there was, after all, a boat, and a fishing line, and the god, and the Pacific. And all of that was already there. So this cannot actually be the story of the creation of the world. At best, it is the story of the creation of New Zealand, if it is anything. So all sorts of criticism can be brought against it. But the important thing is that science begins with stories like this—or, I should say, it begins when we start to ask awkward
questions about them. ‘If the world were fished out of the sea, then what about the hook?’ That is a scientific problem. The moment you ask about the hook, you become a scientist.
My thesis is that science starts with problems and not with observations, and that the difference between science and unscientific myth consists primarily in the fact that some people begin to ask awkward questions about the myths. Someone, in other words, begins to criticise the myth, and out of this criticism there arise new stories, and the result of that is further criticism, and further stories, and so on. And so we get a certain history of views about the world, and the latest of these views is the so-called accepted scientific opinion.
Now, you will say, perhaps, that this is a very low view of science to take. But I happen to think that it is a very high view to take. Science, in this sense, is the result of a lot of thought and a lot of criticism. Observation does play a very important role. But it does not play the role that most people think. Our theories do not begin with observations. But we use our observations wherever we can to check them. It is, in fact, part of our criticism to check our theories against our observations. And I can check a theory against observations only if I first have a theory.
So why do so many philosophers and scientists say that science is empirical instead of speculative? The answer to this question is very complicated, and I cannot give all of it today. But at least part of the answer lies in the fact that we tend to write textbooks and to teach the scientific opinions that we accept. And when we do this, we tend to forget all about the myths and the problems that led us Now let me just ramble a bit about the fact that these problems are forgotten. It is really one of the most important facts in the history of science that scientific problems are forgotten due to our method of teaching. This fact not only leads to the idea that there are subjects instead of problems, it can also reinforce the idea that science begins with observation. Let me give you an example.
Someone may be teaching you geometry. And if he is a good teacher, he may tell you that the problem behind geometry had something to do with measuring the fields in Egypt. The usual
story—l don’t know whether you have heard it, but it is the story that I was taught—goes like this:
The word ‘geometry’ means measuring the earth, and by this is meant measuring the fields. The science of geometry was first developed in Egypt, due to some peculiar facts of nature. It seems that the river Nile overflowed its banks nearly every spring and in so doing destroyed the property borders. These fields were very valuable, so that when the Nile retreated the fields had to be remeasured and re-laid in order to insure that no injustice would be done to their various owners. By remeasuring and re-laying their fields, the Egyptians gradually developed some practical rules, and in so doing invented geometry.
I don’t know whether you have heard a story like that. Anyway, this is how the story is usually told:
And then, since geometry is both an art and a science, there arose one day the need to put it all together in a textbook. And the man who did it was the famous geometrician Euclid. So Euclid wrote his Elements, which was actually the first textbook in the history of mankind, and then for some thousands of years geometry was taught by teaching Euclid’s textbook and its results.
That is roughly how the story goes—and then only quite recently some very important deviations have been invented, the so-called non-Euclidean geometries, which are really very interesting and important.
Now let me tell you how I see the story. I may, of course, be wrong. I assure you that I may be as wrong as anybody. But from my studies of the thing I think the story was completely different. The beginning in Egypt may or may not be so—I don’t think it was in Egypt for reasons that are connected with the later part of my story. But the later part of my story is this. Some people began to speculate about the world. I don’t know whether ants speculate about the world and think that the world is bounded by certain trees and a blue sky—or what. But men at any rate do speculate about the world, and they do so especially after they have done some considerable travelling and shipping. Now the Greeks of that time had done some travelling. And they knew something not only about the Mediterranean, but even about India and the Sudan. So they began to speculate about the world and about how it was all connected, and about the great sky above us, and they soon developed cosmological theories.
These cosmological theories were very primitive. As I’ve already aid—I am not straying as far away from my subject as you may think, because we are speaking about scientific method and the beginning of science is really the beginning of cosmology—the beginning of cosmology was very primitive. I have already told you about the Maoris’ creation story and about how it was criticised. This sort of criticism can be brought against practically every cosmology, even, on a somewhat higher level, against Einstein's cosmology, and it is under the influence of such
criticism that our cosmologies change and our science develops. But to continue, it happened that one great Greek philosopher summed up all the beautiful cosmologies that had gone before him. This was the philosopher Plato, and he founded a school, which was really the first university, that he called the Academy. He dreamt of becoming a king, but instead of this he became the first professor of philosophy.
Now Plato raised quite a number of questions about the shape of the universe, and there are some very great and interesting difficulties in these questions. He also raised questions about the shape of crystals. And he had a theory, a very beautiful theory, according to which matter, all matter, earth and air and fire, actually consisted of crystals—liquid crystals or hard crystals, but crystals. These crystals were later called ‘polyhedra’ by his students in the Academy—a polyhedron is a body with many edges and many corners; a cube, for example, is a polyhedron. So Plato had the idea that the whole world was a huge globe, and that in the
world were the stars and all sorts of things, including the earth. The earth was in the centre of the world, and the earth, of course, consisted of crystals, just as all of the other things in the world consisted of crystals. And some of these crystals, the regular ones, were even called ‘Platonic bodies’ from this story. But the point is that Plato’s cosmology worked with geometrical notions.
Plato, as you can see, had lots and lots of problems that someone today could call either cosmological or geometrical problems. The two went together because he had a geometrical picture of the world—geometrical, because the world consisted fundamentally of crystals of various different shapes. So his cosmology was fundamentally a geometrical cosmology. Now the important thing is that Plato left quite a number of unsolved problems in his cosmology. Some of the most urgent of these were solved by his pupils—Eudoxus, for example, was one of them. But what Euclid did was to bring their solutions, and some of the solutions that they
did not succeed in making as well as some important solutions of his own, all together into one great system, which he really intended as a system of cosmology, of how the world is made.
Now Euclid did solve many of Plato’s problems. But he also did two things that he should not have done. The first thing that he should not have done, though it is understandable enough, is that he did not say that this is what he was doing. He did not have to say this, because everybody at the time already knew it. Everybody knew that he was a Platonic cosmologist. So he didn’t mention in his book that these were his problems. Secondly, he put into his
book only those problems that he could solve. And this, in a certain sense, was a crime. Plato did not do it. And none of the philosophers before Plato did it. The first philosopher who tried to
confine himself to what he thought he could solve was really Aristotle. Anyway, the result was that here was a book that contained geometrical problems and the geometrical solutions to
them. The fact that these geometrical problems were also cosmological problems was soon forgotten—though everyone in Euclid’s time knew about it, and we can still read in a commentary in Proclus, written seven hundred years later, a sentence that runs more or less as follows: “Some people say that Euclid’s are not a textbook of geometry but are actually a book on cosmology.”
The reason why people say this is that Euclid ends his Elements with the construction of the Platonic bodies, which is a good indication that what he really wanted was to solve some of Plato’s cosmological problems. Plato had tried to indicate how these bodies could be constructed. But it wasn’t very satisfactory. And Plato knew that it wasn’t satisfactory, because he stressed in the Republic that what is now called solid geometry, the geometry of bodies, did not yet exist. So Plato at least was clear about open problems. But Euclid did not stress the open problems, and so the first textbook of geometry was created. And this had truly devastating results. The main thing is that it practically put an end to geometrical research. There were some exceptions, there was, for example, Archimedes who developed the thing a bit further. And I am, of course, over-simplifying. But the main result is that a book that was intended to solve cosmological problems suddenly became a textbook of geometry!and with that, a new subject was invented, along with experts in that subject, since a geometer was now a man who knew his Euclid, instead of a man who could invent new problems and bring the science further. This became the criterion for being an expert in geometry—whether or not a
man knew his Euclid—and if a man did know his Euclid, then he could be regarded as an expert, and he could then be appointed professor by a university where he could teach and get paid. And this is really the way in which the subject was invented.
Now I have to close in a minute or two, but I promised you last time that I would reveal all of my secrets today. So let me give you a very brief statement of my secret. My main thesis in these
lectures is that there is, of course, something that we can call ‘scientific method’, but that it only very rarely and happily and occasionally and accidentally leads to results. Scientific method,
like many other of our activities, is problem-solving. And it is nothing very peculiar. It is the method of criticism that is, fundamentally, the method of finding problems, thinking up possible solutions, and then criticizing them.
You start with a problem, and then you try to solve it. And as a rule you will find that you haven’t even quite understood your problem. So there is suddenly a new problem: how can we understand our problem? But, just for a minute, never mind. We tart with a problem, and we try to solve it. How do we try to solve it? We try to solve it by proposing solutions, the best solutions that we can find. But we know that they won’t be very good, and we can call them tentative solutions. So we propose a tentative solution, and then we try to criticize it. We try to show that it isn’t really a solution of the problem. And when we have criticized it, and eliminated our errors, then we are faced once again with our problem, or, perhaps, with a slightly different problem, since we now know something new about it that we didn’t know before.
I have invented a very simple schema to express this method. It is:
P1 -> TT -> EE -> P2
And you will probably see it quite a lot if you stay in this course. P1 is your initial problem, and TT is your tentative solution to it. I write TT, instead of TS, because your tentative solution in science is almost always a tentative theory. By EE, I mean criticism. But I write it like this to indicate that the aim of criticism is to eliminate our errors. And this elimination of errors then leads to another problem, which is, namely, P2.
So it seems that this is a circular sort of progress, and it may even seem as if we haven’t made any progress at all. We start with a problem, we give a tentative solution, we make some criticisms, and then we are back again at our problem. But even though this problem brings us back to where we started even so—I assert that we have still made progress. Even though we are moving in a circle, we are making progress, and not merely circles. The progress we make is that we understand our problem better. We understand our problem better, because we now realize why the proposed tentative solution is not really a solution. And this means that we have learned something. We would not, for example, propose the same tentative solution again—not unless we thought that there was something wrong in the way we criticized it; not,
that is, without first criticizing our criticism.
But perhaps—and this sometimes happens after we have gone this way in a circle several times—perhaps we do think up a solution that really leads somewhere, that really produces something, and that has some real interest in it. Even in this case we will, as a rule, still find ourselves confronted with problems. The tentative solutions that seem to lead somewhere in fact lead us to the discovery of new problems that we may not have been able to see before. So while we do make progress, the progress that we make hardly ever means that we have really solved our problem. We may shift the problem and we may advance the problem, but we
very rarely ever really solve the problem. That may happen once or twice in the history of mankind. But in the main you can’t expect it. What usually happens is that we only shift the problem without really making it advance at all. And the best you can hope is to advance the problem just a little bit.
I told you that I would give my secret away to you at the very beginning. My secret is that this process of problem, tentative solution, criticism, problem, whether new or old—it’s always a bit
new if our tentative solution and the criticism was a real effort—my secret is that this is more or less all there is to scientific method, and that we can now, as it were, all go home.
To put it in another way, my secret is that science grows in the main by starting with problems and ending with problems, and whether or not it has grown can be measured by the difference
between the problems with which we start and the problems with which we end.
If the process leads us back to the same problem, or to one that is only slightly different, then science has grown very little. But if the difference between the problem with which we start and the problems with which we end is very marked, then science has grown quite a bit.
Let me just give you one illustration of this. You have all heard about chemical elements—about hydrogen, oxygen, and the others—and you probably all know a little bit about them. Now one problem in science was to find out whether or not there is anything like a natural order to these elements. That was problem one. A Russian scientist named Dmitri Mendeleyev actually found
such an order in the last century. It has since been superseded in many ways. But he discovered the beautiful order of the periodic table, with which many of you are probably familiar. This discovery advanced his problem very far. But now there was a new problem. If Mendeleyev’s system is correct, then certain elements seem to be missing. So is his system correct? That might be problem two.
Now if we say that the system is not correct, then we are immediately back once again with problem one. But if we assume for a moment that the system is more or less correct, then we have problem three: Where are the missing elements? And now a new search for elements arises. Chemists try to find these elements, which, according to the periodic system, ought to exist, though perhaps they really don’t. If these new elements don’t exist, then the periodic system is mistaken and we are, again, back at problem one.
But this search for the elements raises a fourth problem, namely, how do we search for elements that we do not yet know? Should we search the whole universe for these elements? But we can search for these elements only if we have some idea about where and how we should expect to find them.
So a fifth problem arises, and this leads us to search the periodic system itself for clues that might tell us enough about the properties of these unknown elements so that we will at least have some idea about what we are looking for, and how we should look for it.
Now as things turn out, the unknown elements are eventually found and we complete our periodic table of elements. But in this now revised table there are certain heavy elements at the very end—the heaviest elements are put at the very end—and someone now asks the question, ‘Why does the table end here? Are there, perhaps, some elements beyond this table?’ And in this way, one problem always and inevitably leads to another.
We constantly have before us questions, and we constantly try to answer these questions, and our answers constantly raise new questions. But let me continue, just to show you how these
problems may gain in depth.
We now have a beautiful table, our periodic table of elements. It’s a very interesting structure, and all the known elements fit into place. These elements, by the way, are also classified in this table according to their crystal structure, but that is another story. Anyway, someone now asks, ‘Why are the elements arranged in this queer table?’ We see that there is a certain method to the table, and we can explain how the table is constructed. But we don’t know why the elements should occur in just the way in which they do. So a new and deeper theoretical problem arises, namely, the problem of explaining this arrangement.
This particular problem was actually solved, and not so very long ago. It was solved in the main by Niels Bohr and Wolfgang Pauli at the beginning of the century. These two together explained the periodic table with the help of a certain theory about the structure of atoms—a theory that explains why atoms are built up in such a way as to make just those atoms that correspond to these elements. So this leads us into the depths of atomic theory, and from atomic theory new problems lead us to even further depths.
Let me give you an example. When Bohr explained the periodic table, he believed that atoms consist of only two particles, electrons and protons—the electrons being negatively electrically
charged particles and protons being positively electrically charged particles. This was a beautiful theory according to which all the matter in the universe is built out of only two kinds of particles. It was so beautiful that Bohr and many others believed that it was actually the final solution to the problem. But further searching led to new problems, and to problems of a greater depth. And since then, we have acquired lots and lots of new particles, so that we are
now in about the same situation as we were when we had our periodic table of elements. Instead of a periodic table of elements, we now have something like a periodic table of elementary particles out of which the atoms are built, and what once looked like a real solution now appears to have been only tentative.
Now this may give you just some idea of how one can look at science and say, ‘We start from problems and we end with problems, and the new problems, compared with the old problems,
give us some sort of measure of the great progress we have made. The new problems are a measure of the growth of our knowledge in science’.
Here, the very finding of a problem is in the main a form of criticism, since it often involves the criticism of what appeared to other people to be a solution. So you find a problem by criticizing
some alleged solution, then you try to produce a solution of your own and criticize that, and then you try to find a new solution and criticize it too—and the scheme looks circular. But it only looks circular. We make real progress in this way by discovering which solutions do not work. And sometimes we make great progress in this way. And when we do, it is often astonishing how different the problems that you end with are from the problems with which you started. This difference in the nature of your problems means that you have added something to science.
But these other ideas—that science is a body of knowledge, that you can add to it by learning the tools of the trade, that the primary tool of the trade is scientific method, and that the great scientist is a practitioner of scientific method—these ideas are, in my opinion, all wrong. In this sense, scientific method exists even less thanother subjects exist. It exists only in the sense that we can study the ways of criticism. And we can do this only up to a point. We can study how criticism works in actual cases in science—how itsometimes progresses science and shifts a problem forward instead of just sideways or backwards.