In Our Time - The Scientific Method

Episode Date: January 26, 2012

Melvyn Bragg and his guests discuss the evolution of the Scientific Method, the systematic and analytical approach to scientific thought. In 1620 the great philosopher and scientist Francis Bacon published the Novum Organum, a work outlining a new system of thought which he believed should inform all enquiry into the laws of nature. Philosophers before him had given their attention to the reasoning that underlies scientific enquiry; but Bacon's emphasis on observation and experience is often seen today as giving rise to a new phenomenon: the scientific method.The scientific method, and the logical processes on which it is based, became a topic of intense debate in the seventeenth century, and thinkers including Isaac Newton, Thomas Huxley and Karl Popper all made important contributions. Some of the greatest discoveries of the modern age were informed by their work, although even today the term 'scientific method' remains difficult to define.With: Simon SchafferProfessor of the History of Science at the University of CambridgeJohn WorrallProfessor of the Philosophy of Science at the London School of Economics and Political ScienceMichela MassimiSenior Lecturer in the Philosophy of Science at University College London.Producer: Thomas Morris.

This BBC podcast is supported by ads outside the UK. Every week, we cover the week's tech news on this week in tech. Hi, this is Leo LePort inviting you to join me and my panelists this week. Jason Heiner, Doc Rock, and Mike Elgin will talk about Anthropics' new AI. They say it's too dangerous to release Sam Altman in response to the firebombing of his house. And Samsung jumps profits eightfold thanks to AI. Hi, you'll find Twitter at twit.tv or wherever you get your podcasts. Thanks for downloading the In Our Time podcast.

For more details about In Our Time and for our terms of use, please go to BBC.co.com.uk forward slash radio four. I hope you enjoy the program. Hello, in 1620, the Lord Chancellor of England was the distinguished scholar Lord Verilum. He was not just a lawyer, born Francis Bacon. He was also a philosopher and scientist. And that year, he published a book that has come to be

seen as his masterpiece. The Novum, Organum, proposes a new approach to the investigation of the laws of nature, a scientific method based on experience and observation. My method, wrote Bacon, though hard to practice, is easy to explain. I open and lay out a new and certain path for the mind to proceed in, starting directly from the simple sensuous perception. Bacon's new and certain path is often seen as the beginning of the modern scientific method, a set of rules guiding science, scientific inquiry which have been the subject of intense debate over the intervening 500 years. Some scholars have seen the scientific method as an essential part of modern scholarship. Others, like the 19th century biologist Thomas Huxley, regarded as an extension of common sense.

With me to discuss the scientific method are Simon Schaffer, Professor of the History of Science at the University of Cambridge, John Warrel, Professor of the Philosophy of Science at the London School of economics, and Michaela Massimi, senior lecturer of the philosophy of science at University College London. Simon Schaffer, would you begin by giving us a slightly more precise idea of what is meant by the phrase scientific method? Well, we would like our stories about the world to be true and as simple as they possibly can be and to have as large an extent as possible to be maybe even universal. But the world, as finance ministers, for example, keep on discovering is messy and complicated and hard to explain or predict. And scientific method has come to be seen as a way

of reconciling those two. Our stories about the world with the way the world works. We'd like there, perhaps, to be a series of methods or recipes or rules that would as far as possible guarantee that our accounts of the world are what we want them to be. These rules might include principles of observation, experiment, reasoning. They're also very often going to specify who we might trust to produce such stories and indeed where they might be, what kinds of places,

what kinds of tools they might use. So scientific method in its widest sense is, I think, an essential part of perhaps a human desire to understand the way in which the world around us works. Where do we look for the first examples of the scientific method? Where else would we look? Where else would we look?

Presumably all great literate cultures have had these concerns with adequate accounts of the world and people and methods in which trust should be vested. For the Western tradition, the tradition to which Francis Bacon belongs, one looks back to the Greeks, to Plato and Aristotle, to the transmission of their debates through, Islamic and Latin commentaries in the Middle Ages. Plato proposes a rather pithy version of the problem of method that if we're engaged in an inquiry, says Plato, and we already know the topic of inquiry, then the inquiry is pointless. And if we don't know the point of inquiry,

then the inquiry is useless. To which Aristotle develops a very powerful answer, which is presumably observing the world, would give us an account of what, what problems there are that inquiry might address. And for Aristotle, there had always been these two kinds of issues, facts, things that happen in the world. We'd like to know why. We search for causes. And things, we'd like to know what there is in the world, and we'd like to know their

essences. And there are methods that Aristotle and his commentators discuss, which might allow us to do that. Now, for Aristotle and for the great medieval tradition, there was a moment. model of how that kind of reasoning might proceed, and that model came from mathematics. And a lot of the Western tradition prior to Bacon had focused on exactly how mathematical models of reasoning and natural philosophical models of reasoning, that's to say, reasoning about what there is in the world, could possibly relate to each other. And it's in that debate, I think,

that it's interesting to put Bacon's intervention. Given that mathematics through Euclid was thought to be an instance, perfect truth. Yes, if the premises of a mathematical demonstration are true and the demonstration is logically adequate, then its consequences are true. As we know, and John Warrul, Aristotle flowed through, translated, developed in the Arabic world, into the Renaissance in Western Europe. And so we're going to move on just as swiftly as I've moved on. But the scientific method relies heavily

on reasoning, and scientists adopted one of two or even three approaches. Can you tell us about the inductive method? I think it's probably better if we start with deduction and then compare it with induction if you don't mind. I don't mind. I'd just switch the question. So question four becomes questions three.

We start with deduction. Yes, thank you. So in a deduction, as Simon's just mentioned, so both methods of inference involve starting with certain premises and inferring certain conclusions from them. In a deduction or deductively valid inference, the conclusion is completely guaranteed, guaranteed to be true if the premises are.

Can you give us an example? So the old and famous example that people use all the time in elementary logic is all men are mortal, Socrates is a man, therefore Socrates is mortal. If it's true that all men are mortal and that Socrates is a man, then of course it must be true that Socrates is mortal. Maybe some of his young followers would like to have believed that Socrates was immortal, but if they accepted that he was only a man and that all men are mortal, they, in all consistency, not accept that. So in a deduction, really, we're teasing out the consequences of what are already

in the premises. The process may surprise us, but we're basically only going, we're getting out of the premises what's already in them. So can we just dig in there a little bit more? Sure. We've got plenty of time. Can you give us some more instances of these are sort of having an idea, having a notion, and then trying to prove it. Is that too basic? No, it, uh, it, uh, How you get your starting point is another matter, but we start in mathematics with certain axioms. We can talk about how to arrive at those actions, but we start, Euclid gives us certain axioms for simple relationships between points and lines and spaces and planes and so on. And then we, in mathematics, deduce from those axioms ineluctible consequences like of interesting kind, like the sum of all the internal angles. of any triangle add up to 180 degrees.

The main role of deduction in science is in the testing process where we take a theory, there's a whole story to be told, obviously, how we arrive at interesting theories, but let's put that to one side. We've arrived at some theory, and we want to test it,

and we test it by deducing from it, in this ineluctible sense, often very surprisingly, certain consequences from it, together with some background, assumptions and extra principles, but fundamentally from the theory. So, for example, from Einstein's general theory of relativity, you can deduce that if you look at the motion of mercury over time, you'll see that the point where it reaches its closest

point to the sun, it's so called periolian, moves in a certain distinctive way, and that turns out to be true, this consequence of Einstein. It has to be true if Einstein's theory is true, and it turns out that it is. again, you can do some Einstein's general theory of relativity that two stars that appear to be a certain distance apart in the night sky will appear to be a slightly different distance apart in the daytime sky when the sun's around. Now, of course, that's not normally possible to test,

but you can test it in a solar eclipse, which Eddington did. And again, what had to be true at the empirically checkable level, so you can't check Einstein's general theory of relativity directly. can't look at curved space time or anything. You have to deduce consequences from it that you can check. That again turned out to be true. Of course, had any of them turned out to be false, then by the flip side of, in a deduction,

if the premises are true, the conclusions got to be true. If the conclusion was actually false, then something in the premises has to be false. Can we move to induction now? Sure. Okay, well, induction's much trickier and more controversial issue. what you're looking for in an inductive inference is to do something that you cannot, by definition,

do in a deductive inference, namely become entitled to a conclusion that goes beyond the premises that tells you something more than is actually implicit in the premises. Can you give us an example? Yeah. Whenever you go from any statement about, or set of statements about the past or about so far tested cases, to a generalisation that covers not just past cases, future cases or untested cases, then you're making an inductive inference.

So the classic, though I think in science, unrepresentative cases, so-called simple inhumative induction, so 18th century Europeans saw thousands and thousands of swans, all of which were white, and they were tempted to infer that all swans are white. Well, of course, this may be a psychologically compelling inference if you've seen enough swans, but it can't by definition be deductive because Captain Cook, covered on behalf of the Europeans, of course, the Aborigines knew it already when he landed in Australia, that not all swans are white. So the premises about the observations that the Europeans have made were true, but the conclusion was fault. But I think in general,

there are many cases where in science we do want to go beyond what we strictly know. So, for example, if a theory has been confirmed enough, as the general theory of relativity has, we're inclined to say, well, it must be true. It must have at least approximately true, at least have latched onto the way things are in the world, and therefore to rationalise applications in future cases. Well, then we're going from past tests to the future, and of course there's a big, big issue about what the justification for that sort of inductive inference is.

Michael Masami, some people suggest that the starting point of the modern scientific method was how I started this programme, with Francis Bacon's work Novo Organ. Why was that so important? Francis Beacon was probably really one of the first philosophers in the crucial period of transition from the Renaissance to the early modern period that wrote a text about the scientific method. So it was one of the very first philosophers that very expressly thought that there is a method in science and he laid out what the method is.

and as you have already anticipated, Bacon identified the method of science with induction, with the inductive method, which was bound to remain very influential throughout the history and philosophy of science for centuries to come. How did he arrive at the inductive method, Michaela? Because he gives some vivid instances, doesn't he? It does.

His book was called the new organon, the novel organon in Latin, and was part of a much bigger project that Bacon had at the time. called the Instauratio Mania, which remained largely incomplete. Bacon wasn't just a philosopher. He was, as you said, the politician, eventually Lord Chancellor. It was also an experimenter.

So it's against the background of the emerging experimental sciences that Bacon's reflection on the method really took place. And I think we have to bear in mind something really important about the cultural context in which is reflection on the method that took place. Bacon didn't think that science was first and foremost pure theoretical formal knowledge. He believed that science was also applied science, experimental sciences. He was himself an experimenter. He died, I think, of pneumonia by playing with Heise, doing experiment with Heise.

And of course, when we say experimental sciences, at the time, we are at the beginning of the 17th century. There was a very thin line between experimental sciences and alchemy and magic. So it's a very complex cultural context. But it's not all that conflict in the sense that he's experimenting, his idea of observational experiment became fundamental to a group of very influential men a few years later who took him as a tutor de facto, he'd been a few decades dead, and they became the broader society.

So he's an observational experiment was a key in the Novo Mogarnum, and it became key to an extremely intelligent group. for Ben. Exactly. He called the book a new organon because he was reacting against Aristotle organon so I thought that we shouldn't start with first principle and deducing from first principle our knowledge but as you said we should start

with observation and data and from data try to arrive at universal laws or universal generalization. So for example a classical example that you find is the following if you want to investigate

what's the nature of what we will now call thermal energy or heat. You would proceed in the following way. You would compile a series of tables, what you call tables of instances, a table of presence, table of absence, and table of degrees. So in the table of presence,

you would enlist all phenomena in which heat seems to be present, say the sun, volcanic lava, liquid boil, and then you would have a table of absence where you would have all the phenomena that seem to have some relation with heat, but effectively there is no clear sign of it. And then you would have a table of degrees where you would rank those phenomena according to the higher or lower degrees of heat present in them.

And by comparing in a stepwise process, those different tables, you would arrive to the conclusion that heat is an expansive motion of the parts of the bodies, which is in surprising similarity with what the much later kinetic theory would say about thermal energy. Another founder of modern methods was Galileo. Can you briefly tell us what he brought to the table? Yes. Galileo, well obviously, key figure of the scientific revolution of the time, by contrast with Bacon, Galileo didn't write a book on the scientific method.

So his reflection on the method are part and parcel of his scientific. research and scientific endeavor. Galileo mark a turning point, I think, in the history of the scientific method. Because for the first time, science was no longer an enterprise of coming up with

hypotheses that can save the phenomena, save the appearances. But he was the first scientist to say that the aim of science is to tell us the truth about NACHA. which is the reason why the old Galileo affair with the religious authority began. So, for instance, the French physicist and philosopher Pierre Duem in 1980 brought a beautiful book called To Save the Phenomena,

where he gives his own a reconstruction of the history of scientific method from ancient Greek science through the Middle Ages. And he clearly said that Galileo was the person that put to an end the method of the astronomer that had been so overwhelming for, century. And the method of the astronomer consists in coming up with a hypothesis conjectures about planetary motion, testing your hypotheses, but never expecting your hypothesis to tell you really the truth

about the heaven. It's only with Galileo that we have that. And once we have that picture in place, then the method become essential, because you want to know what rules are going to lead us to the truth. And the connection with, sorry, interrupt, and the connection with Bacon is that it's observation again. Simon Schupper, can you

take us on to Isaac Newton He gave a lot of thought to this subject, and in Principia in 1687, he includes four rules of scientific reasoning. Is he gathering together, Bacon and Galileo, and developing it? Are we on a line here? Yes, I think to a large extent that's exactly what Newton was after. There are a couple of features of the rules of reasoning that Newton develops in the 1680s, which I think are fascinating.

One is the very first set of rules that Newton ever wrote down for reasoning, and not about natural philosophy, but about how to make sense of the apocalypse. So in the early 1670s, Newton was very keen on making sense of the revelation of St. John. And he wrote down a long series of rules for a book he was writing but never completely published on how to make sense of a very complicated set of arguments that he found in this piece of revelation. I think it's very interesting that Newton then shifts those rules to natural. philosophy. Within natural philosophy, the four rules set out a kind of program. Newton began with nine, and then in subsequent additions of Principia, changed them to three, and then four, and his disciples and admirers treated these four rules as rather sacred writ. The first two rules set out an aim for natural philosophy, which is that it must be about causes, and that these causes must be true. and that for the same effects, we should use the same causes.

The third rule, perhaps the most complicated rule, is that there will be some properties of the world that we think are universal in the world. And those will be properties that, as he says, don't suffer intention or remission. That's to say they don't, roughly speaking, suffer changes of intensity, like the fact that bodies have shape, that they move.

these are universal properties of bodies whatsoever. The final rule is Newton's account of induction, that we must derive our principles, he says, from as wide a possible induction as possible, and we should not tolerate hypotheses, which are simply concocted to explain away phenomena and principles that we've derived by induction. These are ambitious rules.

Newton doesn't always obey his own. own rules, and he doesn't use them explicitly very often, but because of the triumph of Newtonian natural philosophy, mathematics and optics, those rules become something of a program in the Enlightenment.

And John Warrell, can we take this on as if all sorts of ideas are coming in, and I think a supplementary and almost contradictory and parallel, but let's pretend, as it were, for the second of his conversation, that there's a driving argument, and the next stage I want to get to is the 19th century,

the debate between John Stuart Mill and William Hewell on the subject of scientific reasoning. Can you give us an idea of what that debate was about and why it's so important? Well, it's a very wide-ranging debate, actually, and covers interestingly not only scientific how to arrive at scientific truths, but also to arrive at moral truths, political truths and so on, and they're interestingly interconnected. You might think those were three quite separate things, but for Mill and Hewle, they were interconnected.

But even if we narrowly focus on science, their differences were wide-ranging and rather complex. I think just for the present purpose, you just pick on what I think is the most fundamental and it touches on what we've talked about before, namely that George Stuart Mill, without, of course, being a naive empiricist who believed that you could derive all scientific theories completely from the data, from observational and empirical data. It was much closer to that end of the empirical spectrum, of the epistemological spectrum, I beg your pardon, there was Huell. Conversely, Hewle, without quite believing the sort of naive

intuitionist or rationalist view that somehow the human mind could capture reality independently of any observation, was much closer to that end of the spectrum. And most of their detailed differences were because they came from those two separate viewpoints. Would it be too simplistic to say, in a way, can be described as a conflict between the, the power of observation and the power of inspiration.

Yes. I mean, He will, for example, believed that you couldn't, that, and I think he was right, that observation could not be done in a way that was revealing a 10-year-rate without bringing in ideas. He talked about the idea-ladeness of observation. So not meaning, of course, that you can see what you like in the world, depending on what your theoretical conceptions are, but the way that you would group phenomena together. So, you know, every two events, strictly.

speaking are distinct, but we nonetheless in science regard them often as a repetition of the quote, same event. Well, that could only be under a description that comes from some set of ideas that we have about the world. So the idea that we could somehow start with bare observation was incorrect for Huell. And there I think he was right. On the other hand, he does go too far for me towards the intuitionist end of the spectrum by claiming that certain principles, for example, Newton's first law of motion that says that everybody continues in its state of rest or uniform motion in a straight line unless acted on by some external force. Although you actually need observation to discover it, once you discover it, you see that it's an a priori truth. It's true independently of experience.

Our reason could have, in some weird sense of could, I think, have known that in advance in any observation. and then I think he's giving too much weight to inspiration, as you would put it, or intuition or the power of the natural light of reason, and too little to observation. One of the key questions of that debate that John's been talking about concerned the nature of scientific hypothesis. Can you tell us about, first of all, what you think a hypothesis is? And why is it significant?

A scientific hypothesis is a proposition an assertion that is taken as the starting point of a reasoning and which is truth has not been established so if you want it's a supposition it's a conjecture

and in the history of the scientific method we have already mentioned cases of people like Newton who absolutely rejected the idea that we start with hypotheses in our scientific procedure. But we had also other philosophers, precisely people like

William Ewell, that defended a very sophisticated form of hypothetical deductivism. John mentioned this already at the beginning of this program. So this is the view that says, really, science starts with hypotheses. No matter how we get those hypotheses. We can get them by experience, by analogies with other cases, no matter how we get them. And we then try to deduce empirical consequences from those hypotheses that we can

test. Now, the new thing about Ewell, I think, and the reason why the Ewell meal debate is very interesting for the story of the method, is that you will introduce the series of criteria for testing scientific hypothesis. One of those criteria is novelty. So a good hypothesis should be able to predict novel phenomena, shouldn't just accommodate the phenomena for which it was originally introduced. The other criterion is that the hypothesis

should be able to explain a variety of phenomena, not just one kind of phenomena. And the third criterion is, is coherence. So we expect the hypothesis to be part of a growing scientific theory, where eventually the hypothesis will be part of a simple, unified and coherent body of knowledge. And for instance, they all debated it took place in the 8040s between a UL and Mille on whether we should really go for induction as the method of science, or we should go for hypothetical deduction, was very relevant for the sciences of the time, or was very relevant for

assessing the validity of the wave theory of light. In the first half of the 19th century, people believed that light was a wave in a medium called the ether, and the Ewell thought that his method, the hypothetical method clearly vindicated the validity of the wave theory of light because the theory was able to predict a novel phenomena. So it was a hypothesis, okay? It's a hypothesis that light is a wave in a medium like the ether.

But it's able to predict novel phenomena. There was this famous experiment that was run in the 1820s called the Poisson experiment. It was able to explain different kinds of phenomena. It was originally devised to explain interference and diffraction, but then was able to explain also polarisation and was part of an increasing, growing, coherent body of knowledge from Koshy to Arago to B.O. and so forth. And another novelty that you'll introduce was the word scientist in 1832,

but onto Charles Darwin, Simon Chaffer, 1859, on the origin of a species. Now, why did that, that seem to pose a problem for people, those who were interested in the scientific method? It certainly did. I mean, Darwin's project to his more expert Victorian readers seemed to many of them either to have violated the principles of scientific method or to be using methods with which they were completely unfamiliar. So both Huell and Mill reacted quite directly to,

Darwin's work. Darwin's book, when it came out, was treated by many of the leaders of Victorian sciences horror. So Adam Sedgwick, one of the most important geologists of the time, said that Darwin's reasoning was like a pyramid

upside down. From one geometrical point, Darwin had seemed to be able to construct an entire account of evolution and natural selection, and Sedgwick thought that was horrific. For Huell, The huge methodological problem with Darwin's story is that we do not now see evolution happening.

So the pressure on Darwin from Huell was to show the processes he was describing actually happening around us. Mill, on the other hand, notoriously, while in many ways rather in favor of Darwin's enterprise, offers him a poison chalice and says, Darwin's theory is a magnificent hypothesis. Mr Darwin does not think that it's yet proved. And of course Darwin thought exactly that it had been proved and that it was much more than a hypothesis. And it was much more than a hypothesis because it hadn't simply been concocted

on the basis of the phenomena that it was also designed to explain. So Darwin's programme has huge effects on science, on culture, on religion and on politics, but one of its most interesting effects, perhaps, is on what counts as an adequate method in the sciences. And that debate, I think, stays live right down to the present. John Wall, can we talk, go and move forward now to Einstein,

and the effect he had on the notion of the scientific method, particularly in a relationship to the sort of godlike, if I might use that word, position that Newton's theories had assumed, Newton and his theories had assumed. Yeah, well, I think the impact of both the Einsteinian theory and quantum mechanics, the principal one, was that prior to the turn of the 20th century, to a good approximation, at least people who regard science as whatever them finicky problems

that may be with induction and so on that philosophers can point out, science has revealed the truth. Nature and nature's laws lay hidden night, God said let Newton be an always light and that stuff. Newton's theory of dynamics and gravitation was true. Science then built on this in a cumulative way, kinetic theory of heat, uses Newton's theory at fundamentals and develops it. Maxwell's theory of electromagnetism extends physics

without needing any correction from anything that went before. So there was this cumulative view of whatever the philosophical issues about justifying it, science had developed truth, and it had produced more and more truths as history of science have progressed. But then everybody saw that Einstein's theory was revolutionary, at least in certain ways. I mean, after all, Newton's theory says that there's action at a distance in the form of gravity. It says that space is absolute and infinite. It says that time is absolute, so that two events either are simultaneous or they aren't.

And Einstein, I beg a pardon, contradicts directly all of those things. There's no action to distance. Space is finite, though unbounded. Two events can be simultaneous in one frame of reference and not in another. They can be simultaneous in time and not in space. They can be simultaneous with respect to one observer and entirely legitimate. Other observer can see that they're not simultaneous in his frame of reference. So the issue that that really raised, together with quantum mechanics,

which is indeterministic, introduces genuine probabilities into the world, whereas previous physics had been deterministic, give me a complete description of the state at any given time the laws of physics will dictate exactly which state the system will be in later. That doesn't happen in quantum mechanics. Made people face up to the fallibility of science, at least at the highest theoretical levels.

And I think that posed the central question for science studies in the 20th and 21st centuries, namely how, if at all, to reconcile the traditional claim that science special epistemic warrant with its demonstrated fallibility at the highest theoretical levels, demonstrated from within science itself by the relativistic and quantum revolutions. Michaela Massimmy, did Carl Popper in 1934 in his book, The Logic of Scientific Discovery, did he take on that argument?

Yes. Popper in a way summarizes all the strands that we have already mentioned, so far from really bacon and Newton and so forth. And it took the debate on the scientific method really in a completely different direction. Because for the first time, Popper realized that we cannot take observation and experiments as the basis of the scientific method. Observations and experiments are always

what they call the theory-laden. They come as part and parcel of our scientific theory. So he suggested that, a better way of looking at the method of science is not to go from observation and experiments via induction to universal generalization, but instead to replace induction with a new criterion which is called falsification.

Falseification, yes. Yes. So the key idea is the following. No matter how many positive instances we find of a hypothesis, the hypothesis may still turn out to be wrong tomorrow, So we can accumulate as many examples as we like of white swans. The chances are we go to Australia tomorrow and we find a black swan that falsify our hypothesis. So Popper thought that a better, a more secured way of proceeding in science would be to look for that single counter-istence that could falsify hypotheses.

So its starting point is, again, it's a form of hypothetical deduptivism, a kind of modern version. sophisticated version of UL's method. We start with a hypothesis, we start with bold conjectures and then we try to test those conjecture as severely as we can. We try to look for potential falsifiers

of the theory, statements that if proved true, could falsify your theory. And he thought that this was a much more secured method than the other method that started with observation and proceeded by induction. Simon Chappah, what influenced it, what influenced did Popper's work have on the thinking of philosophers and scientists about the scientific method?

I think Popper is, to reveal my prejudice, surprisingly influential. He won a fellowship of the Royal Society. He is much discussed, was much discussed from the 1950s onwards in this country, where he came just after the Second World War. and I think Popper's project raised two absolutely fascinating questions, both of which we've already discussed. One is, if the aim is strenuously to try to falsify one's theories, what should the attitude to the theories one believes in in fact be?

Is it that they've so far survived appalling tests? can we say that they're true? That's a very big question. That's the question, as John said earlier, raised by the apparent defeat of Newtonian mechanics by Einstein. And a second question, which is, is the function of method to describe the way very good science proceeds, or to recommend how very good science should proceed? Good example. It's not at all clear. that scientists in fact, whether in the past or in the present, set out strenuously to falsify their favourite theories. Maybe they should, but many of them don't.

Do you want to work, Popper makes an important distinction between science and pseudoscience? Is it possible for you briefly to say what distinction that was and how important it was? Sure. Basically, it's that scientific theories are falsifiable. there are consequences that they have, deductive consequences that they have, that may turn out to be false. A pseudo-scientific theories, i.e. a theory that claims to be scientific but isn't really, is not truly falsifiable. The main targets he had in mind as pseudo-scientific theories were Marx's theory of history and Freud's theory of the psyche, both of which he felt could accommodate any phenomenon and therefore were strictly speaking unfalsifiable.

because of being unfalsifiable they were untrue. That's an interesting question. They weren't rationally believable whether they're true or not is a question that you would only decide at the pearly gates, I guess. But I think the easiest thing is to take a better example named Scientific Creationism, the theory that the world was created in 4004 BC.

It looks like it's in trouble with the existence of fossils and the existence of all sorts of evidence from carbon dating and other sources that the universe is a lot older than 6,000 years. Philip Goss in this wonderful book called Onfilos, which is Greek for naval, and was focused on the issue of why Adam had a naval, given that he obviously didn't need to have, came up with the obvious solution that God created the world with lots and lots of apparent fossils

and lots of apparently aged things in there right from the inception. Well, you can see that if you allow yourself that sort of ad hoc manoeuvre, within a theoretical framework, then it is genuinely immune to any testing, and I think rightly is characterized as pseudo-scientific. Well, in this brief history of method, Kayla, in 1962 we have Thomas Kuhn, who offers a new picture, because the scientific method keeps changing, we have to keep saying that to this, and there are disputes about which is a scientific method, whether it is a scientific method, whether it is just an extension of common sense,

therefore what is common sense, and so it goes. So we're following one line, but there are many. unfollowed lines for programs to do in years to come. Anyway, what did Kuhn bring to the feast? With Kuhn, philosophers of science stopped thinking that science was some sort of accumulation of theories which are more and more likely to be true. You see, the relevance of the scientific method emerged, as I said,

with Galileo in the history of science, when for the first time Galileo dare to say that the goal of physics is to give us the true story about the heaven. Now, when Kuhn came along in 1962 with the structure of scientific revolution and portrayed a completely new picture of science that said science goes through cycles, periods of what it called normal science, crisis, and scientific revolutions.

So you have a series of paradigms that alternate. it. We went from the Ptolemaic paradigm to the Copernican paradigm, from the Newtonian mechanics to Einstein Relativity Theory, but it's not the case that those paradigms are more and more likely to be true. It's not the case that Copernicus is more likely to be true than Ptolemy or that Einstein is more likely to be true than Newton. Once that picture is gone, of course, all sorts of implications arise for the scientific method, which doesn't mean that Kuhn dismiss the method. He thinks the methodology is an essential part of a paradigm.

So the method become internal to each of those different scientific paradigms. So Kuhn thought that during periods of normal science, periods that extends over centuries, students learn, say, Newtonian mechanics over Newton's Principia and they learn how to solve a standard problem using a standard. standard set of solution. So they have a method. That method is part and parcel of the Newtonian

paradigm, but it's not the method that is going to give us the ultimate truth about the world. So to summarize Simon, starting with you, where are we now? I think the question of whether there is a single scientific

method is still extraordinarily controversial. I mean, there is an enormous amount of evidence that different sciences use different methods and maybe the project to find a single scientific method that would embrace all the sciences might not be the best way to go. Furthermore, the ever more pressing presence of the sciences in public life, in public debate, in questions of politics, welfare, economy and progress seem to me to make questions of what the appropriate methods are and where trust and authority should be vested,

perhaps some of the most important questions in public life at the moment. So over the millennia that we've discussed in this programme, the question of how we make stories about the world that compel seem to me those issues remain absolutely crucial for us. And finally, John Warrel. Yeah, well, Simon and I would have lots of disagree about, and also Michaela, I think that one can restore a view of science as accumulating, truths or approximate truths at any rate.

And I think that the question of whether there's one scientific method depends on how specific you want to be. I think at root, you know, that we somehow or other get to hypotheses and test them is still true. And that, of course, the way that we do that, we may learn how to do that better and better as science itself progresses. Thank you very much. John Warrles, Simon Schaffer, and Michaela Massimi.

Next week we will be discussing something written 1800 years ago, the Kama Sutra. Thanks for listening. If you've enjoyed this BBC podcast, why not try others such as The Forum, the discussion programme about global ideas. To find out more, visit BBCworldservice.com slash forum.