Page 3.14

Sir Karl Raimund Popper (1902-1994) was a professor at the London School of Economics and among the most influential philsophers of science of the 20th century. Among his other projects, Popper dealt with the question of what is, and what is not, science.

Popper proposed that what separates scientific theory from non-science is empirical falsifiability: that is, whether an idea can be disproven by observation and experimental testing. God’s existence, for example, cannot be called a scientific theory because no observation can falsify the existence of God. A logical corollary to this idea is that scientific theories can never be proven, only disproven. Popper advanced the notion that no amount of positive evidence can confirm a scientific theory, but a single counterexample can prove it to be false.

Sounds bullet-proof, right? Not exactly.

A large part of science’s authority is the idea that experimental results are reproducible and can be verified by anyone. Increasingly, however, as science becomes more technolgically advanced and more secretive, we rely exclusively on the testimony of the experimenters to decide which theories do and do not hold water. Falsifying experimental results is becoming a lot less practical. Furthermore, some scientific theories are too complex to be falsified by just a few negative results. After all, a strict Popperian interpretation of science demands that global warming be dismissed as ascientific the moment contrary evidence comes to light.

Consequently, to contend that all science is as impartial, and as willing to change tacks, as Popper suggests, is to situate science in the public consciousness in a somewhat misleading way. In this conversation, ScienceBloggers Janet Stemwedel, a philosopher of science, and Ben Cohen, a science historian, get together on Instant Messenger to discuss the problems that accompany a strict Popperian interpretation of science–above all, that in his idealism, Popper is a little out-of-touch with how science is really done.

Writes Ben: “Karl Popper keeps coming up. Especially among groups of scientists. And people trying to demonstrate what is and what is not science. And others who want to hold out for a clear answer to what counts and what does not. But why? Cohen and Stemwedel wanted to talk about it some more. Both have some training in philosophy of science, and both have an abiding interest, and professional careers, in studying the practice and meaning of science in society.”

Janet adds: “I think the question came up in passing of why scientists’ knowledge of the philosophy of science seems to start and end with Karl Popper, especially given that those of us in philosophy of science/science studies don’t stop there (or even necessarily grant Popper a central role). As I mentioned in a comment on my recent post on the demarcation problem, if scientists had posters in their lockers, instead of Che Guevara they’d have posters of Sir Karl.

The part of Popper’s work scientists know about is his effort to characterize the line between science and pseudo-science (aka “the demarcation problem”) and, to a lesser extent, his attempt to show how the problem of induction (from David Hume in the 1700s — as far as the empirical evidence goes, we DON’T know that the sun will rise tomorrow) *isn’t* a problem for science.

Ben and I chatted to toss around ideas about why Popper is such a fave of the scientists.”

…And here, the transcript of their IM sesh begins:

janetdstemwedel: Want to talk Popper?

benjaminrcohen: Who doesn’t?

janetdstemwedel: Should I start with my sense of things, as someone who teaches Popper’s greatest hits every semester?

benjaminrcohen: So, now, yes, about Sir Karl Popper, Giant of 20th century philosophy. By all means, please do so

janetdstemwedel: The thing about Popper is that he’s the philosopher of science that scientists are likely to know about AND think isn’t dead wrong. By the scientists’ lights, he really captures something important about the *spirit* of the scientific endeavor.

benjaminrcohen: and as such that speaks to some ideal image of science, no?

janetdstemwedel: Yup. That being the hard-headed testing of even your most dearly held hypotheses. If the data says your favorite hypothesis is wrong, you’ve got to kiss it goodbye. No crying in science and all that.

benjaminrcohen: So do scientists believe that this ideal vision of how science works will some day come true? Or that they want to regain some past sense of it that once existed?

janetdstemwedel: Well, maybe it’s a regulative ideal. When they’re being hard core, they’re just that tough minded. But perhaps, given the fierce struggle for scarce funding and jobs and all that, you need to make allowances.

Also, Popper probably comes up with the only easy “solution” to the problem of induction: don’t rest your beliefs on it. The only thing you can be sure of is the falseness of the hypotheses the data have (deductively) shown to be false.

benjaminrcohen: But science produces data that doesn’t jive with prevailing theory quite a bit

benjaminrcohen: and we don’t just throw out the theory lickety split

janetdstemwedel: Yeah, but maybe that’s a case where the swashbuckling Popperian scientists can swing in on a vine and overtake the crusty old dogmatic hypothesis-lovers.

benjaminrcohen: that’s good, I can picture that. Philosophers go to wardrobe, get into costume, then save the science day. All Johnny Depp there.

janetdstemwedel: The individual scientists, to a certain extent, see themselves as at least capable of that kind of iconoclasm.

benjaminrcohen: Nevertheless, to be beholden to Popper suggests that the only thing that is science is theory

benjaminrcohen: Better put maybe: for Popper’s philosophy of science, theory is all there is

janetdstemwedel: It’s true, there’s very little about practice.

benjaminrcohen: And so any conversation about how science works or should work refers to what theory is or should be.

janetdstemwedel: But, he was inheriting the methodology of the logical positivists/empiricists, who were all about analyzing the logical connections between observation reports and theoretical claims.

benjaminrcohen: Yeah, he seems to have made some key fixes to the verificationist ideals of how theory progresses, or, how knowledge grows. Which reminds me: Do you know John Zammito’s book?

janetdstemwedel: I don’t know Zammito’s book. What’s it about?

benjaminrcohen: Called A Nice Derangement of Epistemes: Post-Positivism in the study of Science from Quine to Latour. It’s pretty much a history of philosophy of science in the 20th century, from responses to the positivists through the science wars.

benjaminrcohen: Do your students read Popper for his contribution to the demarcation problem? That he tells us what is and what is not science?

janetdstemwedel: Yeah, we talk about the science/pseudo-science (on “non-science”) sorting problem.

janetdstemwedel: Anyhoo, the line between science and non-science *does* seem to have something important to do with a commitment that things be testable against observables. Popper is definitely on the right track on that one.

benjaminrcohen: I agree, it’s a solid point that testing against observables is key. Though that doesn’t really explain what makes science science. Plus, that demarcation relies upon critical tests that can set things up one against another, in unproblematic ways. Which is why I’m always pressing to talk about practice and science-as-it-actually-is. sometimes annoyingly.

janetdstemwedel: Well, the reality-based knowledge thing might not be UNIQUE to science, but it’s definitely central. What else do we want to add to specify science vs. everything else? Explain natural phenomena by reference to natural causes. Uh … What else? The thing my students find most frustrating about Popper is that he only grants us knowledge of what we don’t know (i.e., the hypotheses we’ve demonstrated to be false), and of what has been the case in the past but won’t necessarily be again.

benjaminrcohen: Okay, okay, so I’m starting to realize that I’m looking in another direction in the Popper conversations and thus talking past a lot of the bloggers when I get replies on my comments about him

benjaminrcohen: It’s been a while for this, but it seems like the scientists’ view of Popper is one , not surprisingly, from the inside, looking inward, at how they could or should operate

benjaminrcohen: Whereas I’m thinking about science as part of society, as outward, as contributing and built from social mechanisms…of course, in relation to physical reality

janetdstemwedel: There’s Popper the inspirational speaker vs. Popper the guy who’s gonna check your lab notebooks. And I agree with you, that second Popper doesn’t pay nearly enough attention to practice. It does seem like if you were a hard core Popperian in the lab you’d never get anywhere, that some amount of seeing if a hypothesis pulls through after initial disappointment is warranted.

benjaminrcohen: But doesn’t that make it hairy, when Popper is referred to as the answer to some public debate about science’s status or value?

janetdstemwedel: Yeah, in public debates, there seems to be some hesitance to let on that the operation of science is actually complicated, and that the individual human beings doing it are not perfectly rational or disinterested or whatever. The Popperian hardheadedness may actually be more plausible if we look at the action of social groups rather than individuals, though — I think we can bring in practice and sociology of science without cutting Sir Karl loose altogether.

benjaminrcohen: I think you’re right, but that gets into a bigger debate, right?. (and one that Zammito is trying to show was *not* won by sociologists) — that being whether or not sociology of science has trumped or rendered less useful a traditional phil of science

benjaminrcohen: But even that assumes that all phil of sci is just analytical, whereas I’ve learned quite a bit from praxis-based philosophers of science

benjaminrcohen: And even so, I agree with you that the two are not necessarily mutually exclusive

janetdstemwedel: It seems important not to let go of the question, “Why does scientific knowledge seem so powerful compared to other kinds of knowledge?” — what kind of contact between scientific knowledge and what we can observe in the world explains that. I take it that’s really what philosophy of science is after. But ultimately, looking at what real scientists do to solve real problems (as well as what communities of scientists judge to be “good” or “bad” ways to practice science) tells us lots of important stuff about particular (rather than abstracted and idealized) ways scientific knowledge is built. In other words, I’m one of those practice-oriented philosophers of science.

benjaminrcohen: It certainly does have significant political implications, that science should be clearly visible as something distinct

janetdstemwedel: For sure, a philosophy of science that wants to hold forth on the basis of some idealized science is wrongheaded.

janetdstemwedel: I worry that “othering” science too much creates societal alienation from it. Kids decide they can’t understand it unless they have gigantic brains.

benjaminrcohen: Now, this reminds me too…Popper’s working in a scientific moment when physics is the ascendant form of inquiry amongst all others

benjaminrcohen: and that isn’t the case anymore, at least not in the same way it was in the first half of the century

janetdstemwedel: The Vienna Circle also thought physics was the bees’ knees.

benjaminrcohen: Hempel and Carnap and the lot?

janetdstemwedel: Also, Bertrand Russell circa 1912 when he wrote that mature science (like physics) had no use for a concept of cause.

benjaminrcohen: So we shouldn’t throw in Duhem, being French and all, since we know what’s what with the French

janetdstemwedel: Duhem was kind of into chemistry though, so I cut him slack

benjaminrcohen: But beyond that I’m starting to get in over my head, and too far into territory that you’re on far firmer ground to speak about.

janetdstemwedel: But honestly, would any scientist tolerate generalizing on the basis of one data point? So how did philosophers of science get away with saying, “All good science should look like physics?” The sweeping normative proclamation — which gets us back to Sir Karl.

janetdstemwedel: At the same time, it seems reasonable that you need *some* standards. There’s a reason Feyerabend freaks scientists out — he crazy!

janetdstemwedel: Where Popper wins the crowd is the claim that, if we’re trying to build a picture of reality, you have to be able to cut some stuff loose. Competing scientists, competing research groups, competing “schools of thought” on a particular scientific question, might actually be able to challenge each other (“How does what you’re saying fit with these observations?”) more effectively than individual scientists can challenge their own hunches. But it probably can’t be all Popper all the time.

janetdstemwedel: I don’t think the power of science comes down to more powerful instruments. I think the answer, if there is one, has to come down to methodology — and that probably includes attitude, and how scientists see what they’re doing, too.

benjaminrcohen: Do I take it this means they dislike Kuhn?

janetdstemwedel: I think Kuhn ruffles fewer feathers than he used to.

janetdstemwedel: The no-guarantee-you’re-making-progress-toward-truth thing is a bit of an issue, but the paradigms and normal science as puzzle-solving sell pretty well.

benjaminrcohen: yet this all seems so very odd to me — that all these ideas about science and how science works are really, really old, and that scientists don’t want to gain from the “advances” (the progress, the improvement) in studies in phil of science over the past three or 4 decades?

benjaminrcohen: this gets us back to the opening: no-guarantee-you’re-making-progress-toward-truth is premised on some ideal version of what science is, not what science actually is

janetdstemwedel: A few issues here: Lots of phil of science looks at what scientists consider “stale” science, not the cutting edge stuff happening right now.

janetdstemwedel: Part of that: as a philosopher, how do you tell which bits of the stuff happening right now are gonna turn out to be important to the long-term trajectory of the science

benjaminrcohen: aha — you bring in social, political, cultural studies of science! you recognize the always political dynamics of not just scientific organization and perceptions of scientific knowledge, but scientific knowledge production itself

janetdstemwedel: science DOES at least want good predictions, good explanations, good options for manipulating things to produce the outcomes you want. That might be truthy enough.

janetdstemwedel: But also, maybe scientists don’t NEED to know the answers to the philosophical questions to succeed at what they’re doing. And maybe philosophers don’t want their success to be judged based on whether SCIENTISTS see anything useful in what they’re doing

benjaminrcohen: well that’s a nice observation

janetdstemwedel: Disciplinary autonomy, baby! If chemistry can use it so as not to have to answer to the physicists, philosophers of science should be able to use it to tell the scientists to go back to the lab and leave them alone.

janetdstemwedel: Not that I have hostile relationships with the scientists of my acquaintance

benjaminrcohen: but the public might need to know some of this, or the scientist who are making public cases, or appeals to government, or against anti-climate changers, or sex abstinence arguments, maybe they need to not rest their case that global warming is real because of Popperian falsification

benjaminrcohen: this came up in my recent post about Dawkins, the symmetry issue

benjaminrcohen: at least in your framing, its a disciplinary symmetry issue, not a knowledge straight-up symmetry issue

janetdstemwedel: Yeah, the problem with the closed-shop/trade-secrets approach to the scientific method (and if nothing else, philosophy of science shows the scientific method and its implementation is WAY more complicated than the HS textbooks would make it seem), is that the non-scientists are in no position to evaluate the good scientific findings from the shoddy ones

janetdstemwedel: They’re left with “Trust us, we’re scientists!”

benjaminrcohen: and again with all those science studies works, on trust and authority and credibility

janetdstemwedel: And part of the charm of science was that it WASN’T supposed to be an appeal to authority (like the church), but to facts anyone could get hold of given well-functioning sense organs and their wits.

janetdstemwedel: Yeah, you have to build knowledge based on testimony (and that means appeals to authority) ’cause otherwise you have to do the whole job of figuring out the universe yourself.

benjaminrcohen: that’s why the history of science is so fun for me — you find those same battles going on always, just in different places and against different political backdrops and with differing degrees of already-attained or yet-to-be-attained authority

janetdstemwedel: But, but, but — science is still DIFFERENT, at least from what religions have traditionally been like

benjaminrcohen: that’s true — though aren’t religions different amongst themselves too?

janetdstemwedel: Yeah, same deal as with sciences — probably shouldn’t generalize from too few data points!

janetdstemwedel: As I understand it, religion is supposed to be about knowledge that comes from different sources than what is publicly available to experience with your sense organs. That changes the game significantly

janetdstemwedel: The idea is the science could change the whole story next week.

benjaminrcohen: gets us to a reflexive plea — that what we say of ourselves as scientists (or technical researchers, lets say) applies both ways. and does this mean we are straying from Popper or is this all getting back to him?

janetdstemwedel: I think it does get back to Popper. Popper’s not a fan of faith. All you can know is where you’ve been wrong

benjaminrcohen: religious faith, you mean

janetdstemwedel: The universe could pull the rug out from under you by changing how things work next week.

benjaminrcohen: “All you can know is where you’ve been wrong” you mean religion or Popper?

janetdstemwedel: Any faith — can’t have faith even in your best hypotheses. Popper doesn’t let you have faith that the hypothesis that survived 1000 tests will survive the next one. That kind of faith smacks of inductive inference

benjaminrcohen: damn him. and we were so close.

janetdstemwedel: Well, the scientists I know who are also religious are pretty clear that what they believe as scientists has to meet a certain standard of proof against very particular kinds of evidence (usually data), and that others have to be able to check it. What they “know” in their faith lives they know a very different way, and it can’t be interrogated (or supported) by others in the same manner.


  1. #1 Dave Carlson
    December 7, 2006

    Whoa. That was a trip.

  2. #2 GP1
    December 10, 2006

    Very stimulating dicussion.

    janetdstemwedel: “Would any scientist tolerate generalizing on the basis of one data point?”

    Ever since Newton Doctors of Philosophy have been in the habit of establishing so-called laws on scant data to save Newton’s authority. Coulomb’s law is based on only three measurements that could not be duplicated.

    But what is more troubling in the ways Doctors of Philosophy have been conducting their work is the assumed connection between measurements and the theory. I call this experimental by association. This is how cosmogonic speculations are marketed as observational science. Doctors loosely associate a popular cosmogonic scenario to white noise of radiation obtained from a NASA science payload, for instance.

    This is why I believe that academic physics is not science, it is a stealth religion impersonating science.

    We do not need Popper’s theories to classify physics to be outside science. This is more a legal issue because what physicists and cosmologists are doing is fraud.

    Thanks for this discussion. I’ll read it more carefully.

  3. #3 amy
    December 10, 2006

    hey guys, could you post the date of the conversation and make it more accurately citable? (Well, you did post it, you know.)

    Also, did you/Janet/Ben clean this up before posting? And — Janet/Ben — did you have the conversation all in one go? It’s interesting; it’s the least IMlike IM session I’ve ever seen.

    thanks –


  4. #4 Janet D. Stemwedel
    December 10, 2006

    Ben and I chatted on October 25, 2006, from 12 noon – 1 pm Eastern time.

    What was posted was the entire chat except for the hellos and goodbyes. We cleaned up spelling and punctuation, and made some logic-driven choices about transcript order in the instances where we were both typing furiously at the same time.

    There were no emoticons used in the chat at all. I take that as evidence that the both of us must be kind of old for IM.

  5. #5 BRC
    December 10, 2006

    although, i think we actually did cut out a two line back-and-forth where we made a joke about our inability to use emoticons, and, worse yet?, that we tried to laugh out loud, but not on the floor, and there was no rolling, none of that. ben

  6. #6 amy
    December 11, 2006

    Hey, thanks for the quick reply. I’d been looking around to see how chemists & biochemists play with ideas online post-usenet (and yes, I’m aware of the stabbing-own-foot nature of the search). Looked to me like humanities, soc-sci, and comp-sci people were pretty reliable bloggers, but working chemical-sciences people tended to avoid the scene. Not seeing much chem on ScienceBlogs either.

    Carl Gutwin has an article from a couple of years ago about how much academic conversation is still stuck in listservs because of the archivability; followed open-source projects to see what differences existed between live & remote team conversation & relationships, and found that project managers were shutting down IRC and other media to funnel conversation back to the lists. An HCI friend & I were talking about how unusually long the posts are on academic lists; people don’t seem to write email of that length anymore. Wondered if it was a function of older profs having a lock on conversation.

    But I hadn’t thought of IM. I guess maybe part of the problem with IM archiving is it depends on the participants & fragments the archive. Kind of like, you know, letters…hey, maybe it’s gmail-driven. You have to be some kind of nerd in the first place to have a gmail account, and then the sidebar makes nerd chat convenient.


    ps laughing upright, don’t do, higher odds of soda coming out your nose. ROF=adaptation.

  7. #7 amy
    December 11, 2006

    um. HCI=human-computer interaction, not acid.

  8. #8 wmock
    December 14, 2006

    It has been a while since I read both volumes of Popper’s “The Open Society and Its Enemies,” so I cannot cite specific examples, but I do recall feeling disappointed upon completion of the tomes. Essentially, Popper attempted to discredit Karl Marx’s brand of communism by discrediting Marx’s underlying historicism. Unfortunately, Popper approached the task from an historicist perspective, effectively discrediting his own arguments.

    Popper’s frames of reference should always be taken into account when evaluating his work. You have, of course, touched upon that in your references to physics as the ascendent form of inquiry during the time that he was writing about science. Have you also considered the effects of Popper’s social, economic, and political views upon his philosophy of science? If so, your opinions on those matters would provide outstanding blog posts.

    From time to time, the writings of Popper and, for that matter, any other science philosopher should be reexamined, challenged, and evaluated from a contemporary standpoint. Your exchange offers a valuable contribution to that process.

    Thank you.

  9. #9 Blader
    December 20, 2006

    In my experience, there are two practical difficulties that come up time and again in working as a scientist:

    1) Stating a testable hypothesis

    2) Devising an experiment that DIRECTLY tests the validity of that testable hypothesis.

    In other words, Popper is correct for the most part and an ideal worth working towards.

    I mean, if there should be any adjustment for the pragmatic reality of how science is actually practiced and its ideas disseminated, we should recognize that neither the hypothesis nor its test are always immediately obvious, but we must still do some work…collect data, probe around the edges, get some peaks at the systems we are studying, mull it over, what have you.

    To see the good, stark hypothesis and its test requires inspiration, and since we are only human, inspiration only comes in fits and starts.

    Someone out there no really wishes s/he would have hypothesized that Vioxx can promote atherosclerosis and enhanced arterial clotting. But the inspiration never really came until people realized bodies were falling.

  10. #10 Blader
    December 20, 2006

    I should add….

    I suggest there is a difference between lacking the inspiration to come up with some sort of good, testable hypothesis that if tested can really nail an issue, and stating an awful, convoluted, completely untestable ‘hypothesis’.

    By my estimation, at some 50%-75% of NIH grant applications fall into the latter category. Speaking in very general terms, the investigators who write proposals for those sorts of “hypotheses” are generally the ones out to “prove” some conception of how things must work. Teleologists, if you will.

  11. #11 CANDICE30Kent
    November 19, 2011

    It is good that we are able to get the loan and it opens new chances.

  12. #12 Wesley Kerfoot
    April 12, 2012

    If the author managed to read some of Conjectures and Refutations he would know that a theory being falsified does NOT mean it is all false according to Popper’s words. It just means the theory needs to be refined.