Alice: Hello, and welcome to Simon's quick introduction to the philosophy of science.
Bob: Oh, hi, Alice. Now, why are we doing this as a dialogue?
Alice: Oh, that's traditional - philosophers of science have often used dialogues to get their point across, Galileo and Lewis Carroll to name but two. But of course there's always a problem with the use of dialogues.
Bob: What's that, then?
Alice: Well, suppose we have a discussion and I come out looking more reasonable. Does that mean that I have the more reasonable viewpoint?
Bob: I suppose that it just means that the person who wrote our dialogue wanted you to come out with the better points, and gave me worse lines to reply with. I remember Hofstadter wrote something about this, talking about Galileo's dialogues - he brought in a "neutral" third party to judge between the two people debating, but of course this third party was also written in by Galileo.
Alice: Exactly, and that's going to be a running theme through these dialogues - "who says?" The reason I might come out with a better argument is because the argument is constructed that way. You have to look behind the argument and think about it for yourself rather than be taken in by what the author wanted to achieve.
Bob: Well, that makes me more comfortable about what I say, then - if I lose the argument, it isn't really me losing at all.
Alice: But this applies to the philosophy of science, too. When we stop and think why we believe our perceptions of science, we have to ask where we got those perceptions from. And much of the time the answers will be "common sense", or "scientists".
Bob: I can understand that scientists are hardly impartial observers of their own actions, but common sense is, by definition, a reasonable way to look at the world.
Alice: And again we have to ask - who says? Unfortunately when dealing with the philosophy of science, you will often find yourself having to delve a lot more deeply into the reasons why you believe certain things to be "obvious" or "common sense", and the results can be quite uncomfortable.
Bob: Isn't this all a bit postmodern - question authority, everything's relative, we can't really be sure of anything and all that sort of thing? We don't really work like that - scientists certainly don't work like that, do they? They trust common sense in making their observations and theories, instead of this postmodern scepticism.
Alice: Well, this dialogue is certainly being presented in quite a postmodern way, and postmodernity has certainly picked up on some of the ideas found in the philosophy of science, but the scepticism you'll find in many quarters of philosophy is much older than postmodernism. In fact, we'll see it goes all the way back to the start of the philosophy of science itself.
Bob: But people aren't really sceptical about science - we read things in the newspaper about new scientific discoveries, and we fund science through our taxes, and we generally trust that when scientists say things, they're probably right.
Alice: Sure, but why is that? What's informing our view of the objectivity of science? Who says?
Bob: I suppose you're going to tell me that it's "common sense" and "scientists" again.
Alice: Well, that's not exactly fair. Many scientists have done some study of the philosophy of science - and even if they haven't, they can compare the ideal of how science is supposed to work with the day-to-day experience of how scientists really operate - and so there's a range of opinions within the scientific community about the reliability and objectivity of science. But - and this is a big generalisation - the media, including the specialist scientific media, tend to be both staunchly modernist and also pretty unaware of the debates about the scientific process. And we, on the whole, tend to still believe the media.
Bob: Fair enough. One more question, though; why are we in the preacher/ directory?
Alice: I suppose it's because there's a big debate about the role of science and religion at the moment, and it would be useful to have some background to the theories of science and what it can tell us, to help inform that debate a bit.
Bob: And I suppose also our author believes that people need to understand how knowledge works before they engage in Christian apologetics.
Alice: There's probably an element of that in there too. But sometimes I have a feeling that our author believes that people have to understand how everything works before they engage in apologetics...
Bob: Well, let's get back to that later. You were saying something earlier about scepticism being part of the philosophy of science from the get-go.
Deduction and abduction
Alice: OK. Let's start with basics. How do we know stuff, scientifically?
Bob: That's such a vague question, but let me try to answer it. There are lots of ways - we could make experiments, we draw up theories and test them, we work out the implications of other theories...
Alice: Well, let's start with that last one. Working out implications is part of the scientific method of deduction. I know that Socrates is a man, and that all men are mortal, and so I deduce that Socrates is a mortal.
Bob: Exactly.
Alice: The thing about deduction is that it doesn't tell you anything you didn't already know. It's nice to be able to prove that Socrates is mortal, but we could have proved that by the knowledge that he's dead. Deduction takes you from two things you know to give you a logical and reasoned path to a third thing, but it doesn't actually help you decide whether or not the premises were true.
Bob: So, Socrates is a duck, all ducks have bills, therefore Socrates has a bill.
Alice: A logical argument, well deduced, but unfortunately starting from false premises. Deduction is a useful tool for slicing and dicing information - pretty much an infallible tool, if you're happy to accept that the laws of logic are infallible. So we'll want to justify as much as we can by deduction. Unfortunately, we'll find that, because it doesn't help us test whether or not things are true, it's not actually all that great for doing science.
Bob: OK. What other tools do we have?
Alice: Another one is called abduction. Basically, you take the best approximation to the truth that you can find; you think up an explanation for an event, and you hang around until someone comes up with a better one.
Bob: That seems like the way we expect scientific theories to develop.
Alice: Well, it does, but abduction has two problems. The first is that it doesn't give you a qualitative measure - what makes one explanation "better" than another?
Bob: I see where you're going - that we can't say one theory is better than another because it's tidier or we like it more. But surely you can start by saying that one theory is better than another if it explains more of the facts.
Alice: But what if two theories explain the same facts equally well, just in different ways?
Bob: Well, it looked like the Copernican theory and the geocentric theory were coming up with the same explanations, until the Copernican theory edged ahead when more evidence came about.
Alice: But until the new evidence came about, you couldn't tell them apart - and they were seriously different in their view of the world! The problem is, you can always come up with instances of another theory which is equivalent to the one you've got, but differs in some important way. Let's take evolution - or intelligent design - versus the theory that the whole universe was created five minutes ago, but with all our circumstances and memories intact. They both explain the same facts perfectly well. Take your pick...
Bob: I can see that the second theory does explain all the evidence we have, and I can't think of a reason why we should choose that one over another, other than "common sense" again. So do we just choose the theory we like the best?
Alice: A guy called Feyerabend would say that you can't actually compare scientific theories, any more than you can compare chalk and cheese. Each theory constructs its own view of the world, and you can't compare worlds. Or another way of looking at it is that the view of the world is a social construction of the consensus of the scientific community, and "true" or "false" doesn't come into it.
Bob: And I suppose that you could also say that even if one theory is better than another, your "best" theory might well be the best of a bad lot. There's no guarantee that the "right" theory is in the mix somewhere.
Bob: Right. That's the second problem: abduction always comes with an element of doubt. There's no way you can know that the universe really is the way your theory predicts, just that it's probably more likely to be that way than someone else's theory says. Thankfully, Francis Bacon had a few words to say about these problems.
Induction : Bacon and Hume
Bob: What was Bacon's solution?
Alice: Well, he had two ideas - the first one stops you from getting things wrong; the second tries to help you get things right. Bacon said you get things wrong by following various erroneous patterns of thought, which he called "Idols". The Idols of the Tribe are what happen when we try to make the world conform to neat patterns, even when it doesn't want to.
Bob: What do you mean?
Alice: Well, for a long time, we thought that planets moved in circles. Once we thought about that kind of motion, we could have thought of circles or ellipses, but we initially went with circles. Why was that? Purely because circles are neater than ellipses. We want the universe to be neat, but that's not a very good reason for thinking that the universe is neat.
Bob: Fair enough. What are the others?
Alice: The Idols of the Cave are what happens when we draw the conclusion that we wanted to draw all along, when we let our personal prejudices inform our scientific thinking.
Bob: Like the way that all those scientists in favour of the superior IQ of the white over the negro all happened to be white?
Alice: Curious coincidence, wasn't it? And it all happened a long time after Bacon warned about these things...
Bob: But of course we know better now. So what else?
Alice: The Idols of the Marketplace are problems caused by the way we use language. For instance, we could say that things fall to earth because of gravity; but this is only true because we use the word "gravity" to mean "the reason things fall to earth". And the Idols of the Theatre are what happen when we substitute philosophical thought, or conventional thought processes, for proper scientific thought; maybe you could say that this is when "common sense" gets in the way...
Bob: How on earth are we supposed to get around these things?
Alice: Bacon's solution was really simple: we just try really hard. We gather a lot of evidence, and we record instances of things affecting other things. So you cool water down to less than zero degrees, and it freezes. Then you do it again, and it freezes again. By now you can be pretty convinced that water freezes at zero degrees.
Bob: Well, that's obvious.
Alice: It's obvious to us because we "know" that water freezes at zero degrees. But suppose you didn't know this. Let's propose that water only freezes at zero degrees during a full moon.
Bob: Well, that's an arbitrary requirement, and the principle of Occam's Razor says we should choose the simpler theory, the one without arbitrary requirements.
Alice: Again we need to ask, who says? Why should we agree with Occam? Maybe his Razor is just an Idol of the Tribe - us trying to make the world neat and patterned, and it might not be. Occam's Razor makes an aesthetic judgement in favour of the "simpler" theory, but not necessarily a scientific judgement.
Bob: Well, so what if it is an Idol of the Tribe? You're only concerned with Idols of the Tribe because you agree with Bacon, and there's no reason to do so...
Alice: Now you're getting the idea!
Bob: .. but if you want to be perverse about it, we can try the experiment again during a new moon, and see if the water still freezes.
Alice: Exactly - we try the experiment in all kinds of conditions until we're convinced that the only thing that causes the freezing is the move to zero degrees. But even if we're convinced, we don't know.
Bob: Why not?
Alice: Well, this is the process called induction - we take water to zero degrees, and it freezes; we try it again, it freezes again. Therefore water freezes at zero degrees. But actually this doesn't tell us anything about water - it just tells us about the history of our experiments so far. All we know is that every time we've tried getting water to zero degrees, it's frozen. That's telling us things about the past, but not about the future. It's equivalent to saying that since the sun rose yesterday, and the sun rose today, the sun will rise tomorrow. We might say there's a good chance of it, but we can't know it.
Bob: But that seems like a perfectly good way of making observations and predictions; we do it all the time.
Alice: Yes, we do do it all the time, but - and this is the important part - that's no reason for thinking it's reliable. We have no a priori reason for thinking that the future will always be like the past.
Bob: But that's the only reasonable way to make progress.
Alice: We're not looking at whether or not it's reasonable, we're looking at whether or not it's reliable, or whether it's true. This is what David Hume wanted to look at. If we only think about what we believe to be reasonable, we fall victim to the Idols of the Theatre again. Hume said that you've got to find a way of reasoning about the inductive process itself, and there's a huge problem there - the only reason we have to trust induction is that it's worked so far, and that's a huge circular argument.
Bob: I don't see what's circular about it.
Alice: Sure you don't, because the inductive method is so pervasive we don't notice it. But what we've said is that the inductive method worked yesterday, it worked today, and so it will work tomorrow. The fact that it's worked so far is only a guarantee that it will work in the future if you already believe the principle of induction - we actually have no reason to believe that the future will be like the past, except that it has been so far. So the only way we can prove the reliability of the inductive method is to apply it to itself; and if we don't think the inductive method is reliable, then that's not a reliable way to reason about it.
Bob: But surely the method itself stands to reason - it's a rational way of conducting experiments and learning about the universe.
Alice: Of course it is - but that's again due to a circular argument.
Bob: Now I really don't see how.
Alice: Because of the word "rational". Induction is rational reasoning because what we consider to be "rational" includes such processes as induction.
Bob: So what are you saying? That "rationality" is bogus and induction can't be trusted? Should we just throw the whole scientific method out the window?
Alice: No, no, no. Listen, nobody is going to be giving up induction just because it's not rationally justifiable - most scientists are pragmatists, and if it seems to work, they'll carry on using induction. All I'm saying is that we have no a priori reason to be sure that what the scientific method is giving us is actually the truth.
Bob: Well, inductive reasoning has given us a great deal of scientific advances, and without this way of doing science we wouldn't have any of the modern inventions we use today.
Alice: That's a big hypothetical, but I'll grant it's probably true. But nobody is denying that induction is useful; it certainly is useful. It's just not provable. And it has other big problems as well.
Bob: Such as?
Alice: Well, for starters, you could just as well say that induction is abduction - it only gives us the best guess - with all of the attendant problems that that has. When you're wanting to defend induction that way, you call it Induction to the Best Explanation.
Bob: Which ought to flag up warning lights about who says what's "Best"...
Alice: Another issue is that induction relies on us picking up a number of instances of an event in order to come up with an inductive proof. We have to make sure that we collect all the data, and don't ignore the results we don't like, but that's a slight side issue - that's a problem with the practise of the scientific method, not the theory behind it. But there is one problem with collecting instances of an event...
Falsification
Bob: Let me think - we have to decided how many instances are required before we believe something inductively?
Alice: Exactly. Or, in fact, the better question, how do we know how many instances are required before we believe something inductively? Again, who says?
Bob: OK. So what's the answer?
Alice: Of course, it depends who you believe. Some people will say that you can draw up a table of probabilities, and the more instances you have, the higher the probability that you're right.
Bob: That sounds reasonable.
Alice: Well, it seems reasonable, but it's actually argument in a circle again. If you don't know how many instances are required, how do you draw up your probabilities? This is one of the reasons why a guy called Popper came up with the idea of falsification. Really his work came out of wanting to know what made, say, physics a science, and astrology a pseudoscience.
Bob: But everyone knows that astrology is bunk.
Alice: That's the "everyone knows" argument again - who says? You're arguing from this vague thing called conventional wisdom instead of anything solid. I'm not suggesting that astrology isn't bunk, but you'll have to do a bit better than that to say why.
Bob: Well, first, it's unscientific.
Alice: And who's telling you that? The scientists...
Bob: No, what I mean by unscientific is that it doesn't turn cause into effect.
Alice: But it certainly does - it takes the alignment of heavenly bodies and describes their influence on human behaviour. Deductively, no less! It's only your preconditioning that tells you this is an invalid mode of inference. You've fallen victim to Bacon's Idols of the Tribe.
Bob: Yes, but the descriptions astrology produces are too vague to be tested - if my horoscope says "you will have a pleasant surprise", then I'm sure I will - eventually. That's equivalent to saying "the rain will stop", because if you wait long enough, it's bound to be true.
Alice: Right, and that's where falsification comes in. Popper said basically that science advances when theories are wrong, not when they are right. If you have a theory that all swans are white, it doesn't matter how many white swans you see; that doesn't advance your theory one bit. What matters is whether or not you see a black swan, because once you see a black swan, the theory is toast. That answers the problem of how many X's you need - the answer is, it doesn't matter how many X's you have, what matters is one ~X.
Bob: So it's like, scientific theories should have a built-in self-destruct clause?
Alice: That's exactly what Popper said - that any scientific theory should be able to come right out and say "This theory of X is wrong if an instance of ~X is ever discovered."
Bob: OK, but doesn't that mean that science doesn't actually give us any knowledge - it can only tell us things that haven't been proved wrong yet?
Alice: Pretty much. In that respect, all of current scientific "knowledge" is "just a theory".
Bob: Hold on, isn't this just replacing induction with abduction - "the best explanation we've got so far"?
Alice: Not exactly, because if you remember, for science to advance, you have to prove a theory wrong. That means that the more a theory attains the status quo, the more you have to try to find ways of disproving it. This is different from abduction because Popper's view of science challenges you to try to disprove your own theory. However, it does mean that our understanding of the universe is provisional. Popper calls this fallibilism - everything that we think we know could be overturned overnight if we came up with a bunch of critical experiments which falsified everything.
Bob: Well, that seems fine if you're prepared to believe that scientists have enough integrity to attempt to prove themselves wrong...
Alice: Hold that thought, because we're dealing with the philosophy of the scientific method in principle, not the dealings of scientists in practise; that's another area altogether. Anyway, according to Popper, pseudoscience comes about when people are so wedded to their theories that they cannot or will not accept any attempt to disprove them.
Bob: Like those conspiracy theorists who say that every time the government does something to dispel their theories, this is more and more conclusive evidence of a cover-up?
Alice: Exactly - even contradicting evidence is portrayed as "proving" the theory. And Popper said that this was often because the theories involved in pseudo-science are over-reaching - that they try to give you certainty about everything. For Popper, it's the element of doubt, not certainty, that makes science science.
Bob: So are all scientists really trying to prove themselves wrong?
Alice: No, not all of them, of course, because you've forgotten to ask "who says?" The only scientists who are actively trying to prove their own ideas wrong are the ones that agree with Popper... Besides, there are other problems with falsification.
Bob: Well, I can see one: if we can attack induction by saying that it can only be proved inductively, how about falsification - can it be falsified itself?
Alice: That's a good argument, but Popper saw that one coming. He said, basically, that falsification is just a tool, a philosophical process for thinking about science, not a part of doing science itself.
Bob: That doesn't seem fair - couldn't you say that about induction too?
Alice: Not really, because induction is bound up in the whole process of how we derive theories, whereas falsification just gives us a tool for determining whether or not to reject a theory.
Bob: I'm not convinced, but I'll let that slide for a second - what are the other problems, then?
Alice: Well, there are certain things you can't falsify easily. For instance, if I say that all swans are white, you just have to find a black swan to falsify my theory. But if I say that there is a green swan, you can only falsify that statement by completing a thorough search of the entire universe and not finding any green swans.
Bob: But I guess one way around that would be to say that statements like "there is a green swan" aren't scientific for the very reason that they're not falsifiable. It depends whether you're more wedded to the scientific theory or the falsification theory. That's a tricky one... And I suppose you can't falsify theories about things that have already happened? So according to Popper, evolution isn't scientific?
Alice: Not just evolution, but Intelligent Design too; let's be even-handed here. Strict adherents of Popper's views would say that neither of these are science, but historiography, since they don't have much in the way of predictive power. Or perhaps that either both of them are science or neither is. Remember that Popper was not trying to distinguish good science from bad science - just science from non-science. But anyway, Popper himself didn't go that far. He did express some concerns about how both of these theories produce such dogmatically entrenched positions, providing unfalsifiable certainty instead of the element of doubt, that they tend to smell like pseudo-science.
Bob: Are you saying that the more certain we get about scientific principles, the more they look like religion instead of science?
Alice: No - not yet, although we will get to that. We haven't finished with the problems of falsification yet.
Bob: What else is wrong with it?
Alice: You sort of hinted at it earlier: with "there is a green swan", is it the statement or the falsifiability that's the problem? Similarly, if we perform a critical experiment, and it looks like our theory has been falsified, we might want to ask what exactly has been falsified. Scientific theories depend on so many other scientific theories - and other background assumptions - that it's sometimes not possible to say clearly what's to blame.
Bob: So if I have a theory that a certain substance doubles in mass when burnt, and someone burns it and doesn't observe a doubling in mass, I could say this isn't a problem with my theory, but of the theory of combustion.
Alice: Or the theory of gravity, since weighing something involves the relation between mass and weight. Or even mathematics, since you might have found a falsification to multiplication. Or you could just say that they didn't perform the experiment properly, and it's a falsification of their scientific capabilities.
Bob: I see that all these are possible, but intuitively, it's more likely to be the theory that's at fault, particularly if other people try the same experiment and get the same results. But that's just going back to the problem of induction, which was what we were trying to get away from.
Alice: Right. And also, it reflects on what you're prepared to give up, and what you're prepared to keep. You want to keep hold of mathematics, and you're more prepared to sacrifice the new theory to keep them. So the odds are stacked against the new theory and in favour of "common sense". Actually, this exposes a big problem with both induction and falsification, and in fact much of the scientific method in general.
Bob: What's that, then?
Alice: Let's suppose we're trying to find out the temperature at which water freezes, again. We make the hypothesis that water freezes at zero, and we set out to do our experiments. But not once do we find that water freezes at zero. The first time, there's salt in the water, and it freezes at a lower temperature. The second time, there's something wrong with our thermometer. We want to throw these results out - but we want to throw them out because they disagree with our hypothesis. Eventually we find a few samples which do freeze at zero, so we keep those. Isn't that kind of selectivity of evidence guaranteed to prove our theory in the end?
Bob: Surely in those two cases there are reasonable grounds for discarding the results.
Alice: Certainly, but somewhere you have to draw the line at what component of the experiment you think matters. Can I throw a result away because the water was in a green beaker, not a red one? The line should be somewhere in the middle - well, who says it should? - but like I said, in reality we're prepared to give up very little when it comes to accepting a novel theory. On the other hand, we have philosophers like Quine who think you should be able to falsify absolutely anything. Quine said that even the "laws" of logic are human constructions, and you shouldn't necessarily throw out a theory if it leads to a paradox, because it might be that logic itself has been falsified.
Bob: But you can't just rewrite logic to suit any new theory that comes along - that's completely unacceptable! That way lies relativism...
Alice: OK, now we're getting somewhere.
Bob: We are? So you're happy to agree there are some things we can take as read?
Alice: Well - only the belief that there are some things we can take as read!
Bob: What do you mean?
Kuhn and Revolutions
Alice: Well, the next philosopher we need to look at is a man called Kuhn. We can explain his main idea in simple terms by saying that the way science is done is completely determined by the culture in which it operates.
Bob: I know that some Russian scientists work in ways that Western scientists don't accept, and that Chinese medicine and Western medicine don't always get on, but surely the scientific method is always more or less the same wherever you go.
Alice: More or less, I'll grant you. But Kuhn was talking about the culture of the scientific community itself. At any point, there are a series of unwritten rules about what you can and can't investigate. We tend to think that science is an intellectual free-for-all, and if you can come up with an idea and prove it then that becomes the scientific position, but it's not like that.
Bob: Because of factors like funding and status, and the personality of individual scientists?
Alice: These are factors relating to the sociology of science, and we don't have time to look into those today. The main point is something more fundamental. It's more that the ways of thinking are constrained by the prevailing consensus. For instance, in Aristotle's day, it simply wouldn't make sense to talk about "splitting the atom", because the atom was something that was, by definition, unsplittable. So splitting the atom was simply something you couldn't investigate without throwing out all the understanding of the way the world worked, which, as you demonstrated by your reaction to the idea of re-thinking logic or mathematics, is not something that scientists are very keen on doing.
Bob: That makes sense for Aristotle, but what about these days?
Alice: It's still the same. There are certain guiding principles that we use as signposts to tell us whether a theory is likely to be wrong. These are the basic assumptions that you just don't question, because they're established principles and any theory which seems to contradict them is, by definition, not established.
Bob: Such as?
Alice: We've smashed a few recently - for instance, we used to think that action at a distance is impossible. This is the principle of locality - it means that remote effects have local causes. If I talk to you across the room and you hear my voice, the sound didn't just "turn up" on the other side of the room, but moved steadily from place to place through the medium of air until it reached your ears. If it looks like some behaviour just "turns up", it's because an effect has moved steadily through some medium that we don't know about yet.
Bob: That makes sense.
Alice: Of course it does, and you'll probably say that what I'm about to tell you doesn't make sense: that messing with one atom in a quantum pairing caused a change in another atom miles away, and when they observed this, they would also observe that the change happened faster than the speed of light, so could not have happened by a steady local progression. This was predicted by quantum theories, but people didn't really like the theories because they broke locality.
Bob: Or they broke the idea that things can't happen faster than the speed of light.
Alice: That's right - that's another of the signposts we use. So we have to ask which of those two guiding principles to give up. Whenever we do give up one of those guiding principles, we tend to find that most of the other assumptions we use in science need rethinking too, and we get what Kuhn calls a "revolution" - science starts all over again.
Bob: Surely it doesn't start from scratch? After all, we're always building on the knowledge of what's gone before, even if our knowledge now is that the knowledge then was incorrect. And we certainly don't have a total revolution - even though we have quantum theory, we still use Newton's mechanics for real-world applications.
Alice: Yes, we do; we're often caught between two world-views for the sake of convenience. Even people who accept quantum theory, which tells us that space-time is Reimannian, will use Aristotelian geometry because it makes your head hurt less. Accepting a theory doesn't commit you to a particular worldview - we can accept Newton's theories and Einstein's theories at the same time without saying that the universe really is the way one or the other says.
Bob: But what about revolutions?
Alice: In revolutions, the process of science is not as cumulative as people tend to make out it is, because you have to throw out a lot of the progress you thought you had made. In fact, it's inevitable that you do so because it's simply impossible to move from one worldview to another by a cumulative process. There's no steady stream of progress to get you from Newton to Einstein - something big has to give. That's what Kuhn calls a "paradigm shift", and it was one in the eye for the logical positivists.
Logical Positivism and Scientism
Bob: OK, so who are the logical positivists?
Alice: "Were", although you still tend to find one or two around. Logical positivism was a viewpoint in science that wrapped up a number of different philosophical tenets. One was developmentalism.
Bob: What's that one?
Alice: You mentioned it earlier - that we're always building on what went before. In science, that all human knowledge is gradually increasing, and we're just getting smarter and smarter.
Bob: Kind of like evolution.
Alice: Very much like evolution, but developmentalism was a philosophical idea a long time before it was a biological theory. But as in biology, so in science; maybe meta-ontology recapitulates meta-phylogeny?
Bob: Hmm?
Alice: Sorry, meta-biologist's joke. Anyway, another big thing the logical positivists were into was empiricism.
Bob: It doesn't exist unless I can measure it?
Alice: That's the one. But logical positivists also took empiricism a lot further, and said that, basically, the only things worth talking about are statements about things that can be measured, tested and "proved" by science - although as we've seen, the nature of proof is a little bit ropey. They knew the problems with induction, but also believed in it very strongly, so fought hard philosophically trying to find ways around those problems. But they would say that anything other than scientifically provable statements are just meaningless.
Bob: So far this sounds like what most people believe about science.
Alice: Indeed it does, because logical positivism heavily influenced modernism, and most people out there are modernists. Now think about it - how would you defend the idea of logical positivism without resorting to "common sense"? On what basis would you argue?
Bob: It seems obvious that you can't argue logical positivism on its own terms, because it's not empirically verifiable. But can't you just use Popper's trick and say that it's just a philosophical tool, not a scientific tool?
Alice: No, not if you claiming that only empirically verifiable statements are worth talking about...
Bob: So you're saying that logical positivism is just another belief system, based on axioms that you just have to accept on their own merits?
Alice: Pretty much.
Bob: But most people don't have a problem with these issues - they don't care about the fundamental logical basis of the scientific ideal.
Alice: Or they don't think about it. Strictly speaking, most people aren't subscribers to full-blown logical positivism, but something a bit softer, called scientism. Scientism inherits the idea from logical positivism that scientific statements are of a different calibre of belief than other statements: say, theological, ethical, or aesthetic. Scientific statements are set apart because they're innately more trustworthy, because they've been "proven", and therefore concludes that science really does tell us solid facts about the way the world works. This, of course, requires unshakable belief in the scientific process.
Bob: Surely scientific statements are of a different calibre - they do tell us something verifiable about the universe.
Alice: Again, we have to ask "who says?" People like Kuhn and Popper and Feyerabend would say no, scientific statements are only of a different calibre because the experts in our tribe, the ones we've appointed to be the judges of reality for us, appeal to their own reliability - despite continually telling us that they've found something new and what they believed all along was wrong.
Bob: But we've been saying all this time that scientists don't necessarily "believe it all along"; they might be just holding onto some theory until it's falsified, or using abduction to take the best guess at the world's workings, or working with Newtonian mechanics one day and relativistic physics another day depending on which is more useful for solving the problem.
Alice: Exactly, but those taken in by scientism don't see the internal machinations; they see this machine called science that every so often expels some new wonderful theory to "expand our knowledge about the universe", and most of them don't, won't or can't think about the fundamental concepts behind the whole thing. Unfortunately, they think that the logical positivists are the only people playing the game - usually precisely because the logical positivists tell them that scientific scepticism is scientifically meaningless. But they don't see that...
Bob: I can't help thinking that you're caricaturing scientism to leave it in a bad light - but maybe that's the author getting in the way again. Anyway, what do you suppose you call someone who subscribes to scientism?
Alice: Anything you like, if you're a philosopher of science; they can't hear you.
Bob: Very funny.
Scientific Explanation and Demarcation
Alice: Let's begin to tie it all together. We looked at how science works; now, what would you say that science is for?
Bob: Well, it's for giving us knowledge and explanation about how the world works.
Alice: What kind of explanation should science give us?
Bob: I would say "a true one", but I suspect you're going to try to convince me that I should say "a useful one" or "a likely one", or even "a consensus one".
Alice: No, I'm talking about the different classes of explanation. There are many different ways of explaining things, with lots of different explanations which may all be true. Take, for instance, this question: "Why are John and Jane married?"
Bob: OK.
Alice: One possible answer would be "Because their personalities suit each other really well." That would be a psychological explanation, and it may be true. Or maybe you could say "Jill introduced them a few years ago, and they've been seeing each other ever since." That's more of a teleological explanation, it describes the whole chain of events. A more prosaic answer might be "Because they performed a certain ritual in front of a designated official, and therefore the law conveys upon them the title of a married couple." That's called a "nomic" explanation - it deals with laws and the consequences of actions. All of these are correct explanations, but they're explanations at very different levels.
Bob: The last one is more of a "how" than a "why".
Alice: That's right, but that's the kind of explanation that science tends to provide - that's why we talk about the "laws of science", or the "laws of physics". And the logical positivists would say "good, because that's the only kind of explanation that makes any sense."
Bob: But in a sense, it doesn't answer the question.
Alice: Right. So we could say with the logical positivists that yes, scientific statements are distinct from other statements of belief, and yes, we can say that the scientific method does require a particular standard of proof - although we may disagree first with the way real scientists doing their jobs realistically apply that standard, and also with the theoretical basis of the method at all. But scientific statements are distinct because their way of explaining the universe is incomplete. Yes, the positivists are right that is a demarcation between science and other forms of explanation, and that very fact opens the door for other forms of explanation to come from psychology or religion or aesthetics or ethics or anywhere else.
Bob: Now, wait. The logical positivists would say that the other possible levels of explanation are meaningless.
Alice: And we could equally well say that logical positivism is meaningless. What they mean is that the other levels are scientifically meaningless, and that's fine, because we've seen that there's a need for explanations that are not purely nomic.
Bob: But have we answered the question about whether or not science is different from any other belief system?
Alice: I think you ought to know the answer to that by now...
Bob: "It depends on who you ask."
Alice: Exactly.