How can we justify science?: Sokal and Lynch debate epistemology

March 14, 2012 • 5:20 am

I’m not a philosopher, though I’ve read a fair amount of philosophy and took courses in it in college.  And I respect the discipline, at least insofar as it helps clarify our thinking—especially about ethical problems. But sometimes philosophical lucubrations seem pretty useless, and that’s the case in a recent exchange between Michael P. Lynch and Alan Sokal in The New York Times.

Lynch is a professor of philosophy at the University of Connecticut, while Sokal is a professor of mathematics at University College London and of physics at New York University. Sokal is also, as you know, the author of the most famous satire of postmodernism, a phony but convincing-sounding paper paper on “postmodern physics” published in Social Text in 1996.

Their debate, “Defending science: an exchange,” is based on an earlier essay by Lynch in the NYT, “Reasons for reason.”  In both pieces, Lynch bemoans the fact that we don’t seem to have any first principles that can justify the use of science to attain knowledge as opposed to other methods, especially religion. The discussion is motivated by creationists who reject science in favor of scripture (I’ll use Lynch’s quotes from both of the pieces):

. . . the public debate over evolution isn’t just about evolution. It is also about which sources or methods we should trust — science or scripture — when it comes to the history of life on this planet.

And the problem, says Lynch, is that we can’t justify using science to understand the world any more than we can justify using scripture:

Every one of our beliefs is produced by some method or source, be it humble (like memory) or complex (like technologically assisted science). But why think our methods, whatever they are, are trustworthy or reliable for getting at the truth? If I challenge one of your methods, you can’t just appeal to the same method to show that it is reliable. That would be circular. And appealing to another method won’t help either — for unless that method can be shown to be reliable, using it to determine the reliability of the first method answers nothing. So you end up either continuing on in the same vein — pointlessly citing reasons for methods and methods for reasons forever — or arguing in circles, or granting that your method is groundless. Any way you go, it seems you must admit you can give no reason for trusting your methods, and hence can give no reason to defend your most fundamental epistemic principles. . .

Debates over epistemic principles sound abstract, but they have enormous practical repercussions. For instance, in order to decide policy matters (like what to put in our textbooks and what to teach in science classrooms) we need to decide on the facts. But in order to decide on the facts, we need to decide on the best ways for knowing about those facts. And to do that, we need to agree on our epistemic principles. If we can’t, stalemate ensues. Each side looks at the other as if they inhabit a completely different world — and in a sense, they do.

This is an old debate, and one used by theologians to show that science is, at bottom, no better than faith. In fact, Lynch notes that both “methods” of attaining truth could be seen as based on faith:

According to many people, what the problem of justifying first principles really shows is that because reasons always run out or end up just going in circles, our starting point must always be something more like faith.

In contrast, Sokal argues that the conflict between faith and science doesn’t simply reflect a difference in epistemic principles, but the use of supplementary epistemic principles by the faithful.  After all, religious people live their everyday lives as if they trust reason and empiricism: they fly in planes, use computers, and take antibiotics when they’re sick:

The point is, simply, that fundamentalist Christians’ epistemic principles are not, at bottom, so different from ours.  They accept as evidence the same types of sense experience that the rest of us do; and in most circumstances they are attentive, just like the rest of us, to potential errors in the interpretation of sense experience.

The trouble is not that fundamentalist Christians reject our core epistemic principles; on the contrary, they accept them. The trouble is that they supplement the ordinary epistemic principles that we all adopt in everyday life — the ones that we would use, for instance, when serving on jury duty —  with additional principles like “This particular book always tells the infallible truth.”

But then we have a right to inquire about the compatibility of this special epistemic principle with the other, general, epistemic principles that they and we share.  Why this particular book? Especially, why this particular book in view of the overwhelming evidence collected by scholars (employing the general epistemic principles that we all share) that it was written many decades after the events it purports to describe, by people who not only were not eyewitnesses but who also lived in a different country and spoke a different language, who recorded stories that had been told and retold many times orally, and so on.  Indeed, how can one possibly consider this particular book to be infallible, given the many internal contradictions within it?

Sokal notes that our methods of finding stuff out are the result of evolution, and therefore are generally reliable (this is also Dan Dennett’s argument against Alvin Plantinga‘s claim that our senses can’t give us reliable information about the world).  Lynch’s response is that this is not a philosophical justification for science, but a practical one:

You point out that certain forms of reasoning are likely to promote survival. Og and his buddies had a greater chance of sticking around and making little Ogs if they relied on induction and observation to get by in the world. No quarrel there. But that is just my point: defending scientific principles of rationality by appeal to their survival value is to cite practical, not epistemic, reasons in their defense. Of course, survival value is hardly the only sort of “practical reason” we can cite on their behalf. We can endorse their usefulness in helping us build bridges and cure diseases. And we can — although this is a longer story — also defend them as having a more democratic character. What I’ve been arguing we can’t do is defend epistemic first principles with more epistemic principles.

Here Lynch is getting near my solution: we justify science rather than faith as a way of finding out stuff not on the basis of first principles, but on the basis of which method actually gives us reliable information about the universe.  And by “reliable,” I mean “methods that help us make verified predictions that advance our understanding of the world and produce practical consequences that aren’t possible with other methods”.  Take a disease like smallpox.  It was once regarded as manifestations of God’s will or displeasure. Indeed, inoculation was once opposed on religious grounds: that to immunize people was to thwart God’s will.  You can’t cure smallpox with such an attitude, or by praying for its disappearance. The disease was cured by scientific methods—the invention and testing of inoculations—and completely eradicated on this planet by the use of epidemiological methods. Science gets us to the Moon; religion can do no such thing.

Scientific understanding advances with time; religious “ways of knowing,” even by the admission of theologians, don’t bring us any closer to the “truth” about God. We know not one iota more about the nature or character of God than we did in 1300, nor are we any closer to proving that a god exists!  In what sense, then, has religious epistemology brought us any closer to truth?

And do we even need a philosophical justification for using the methods of science to understand the universe?  Why isn’t it enough to show that science produces understanding and religion doesn’t?  Philosophers like Lynch tear out their hair in frustration because they can’t justify, a priori, why to use science rather than religion.  Well, that’s how they earn their living, but I find those efforts a waste of time—at least for scientists’ own work, or for helping resolve the science vs. religion debate. You can’t do that by philosophically justifying why the methods of science are superior to those of faith (Lynch produces no such philosophical justification, by the way). Can you imagine converting creationists to evolution by presenting them with such a philosophical justification?

When Lynch asserts that “debates over epistemic principles sound abstract, but they have enormous practical repercussions,” he’s simply wrong, and merely defending his turf. These debates have no practical repercussions, because a) scientists ignore them, and rightly so, and b) the public won’t pay attention to them, either. They’re important only to philosophers.

This, while people like Lynch bemoan the lack of justification for the epistemology of science, scientists blithely ignore them and go on their merry way, curing diseases, making better crops, and understanding the evolution of both the universe and of life on earth.  In this sense, Richard Feynman was right: “Philosophy of science is about as useful to scientists as ornithology is to birds.”  (Note: I am not denigrating philosophy as a whole here, merely its obsession with this particular problem.)

As I said, Lynch argues that the resolution of the epistemological divide between science and faith must ultimately rest on the “democratic character” of science that can produce practical results.  But even here he gets it a bit wrong:

Yet this very fact — the fact that a civil democratic society requires a common currency of shared epistemic principles — should give us hope that we can answer the skeptical challenge. Even if, as the skeptic says, we can’t defend the truth of our principles without circularity, we might still be able to show that some are better than others. Observation and experiment, for example, aren’t just good because they are reliable means to the truth. They are valuable because almost everyone can appeal to them. They have roots in our natural instincts, as Hume might have said.

Well, everyone can appeal to religious dogma as well, or to revelation. Observation and experiment aren’t just good because “everyone can appeal to them,” for many people don’t. (In fact, 64% of Americans would accept their faith over science if a scientific fact were shown to contravene their faith.)  Perhaps Lynch means that “everyone who shares the scientific epistemology can appeal to the facts,” but that becomes circular, too.  You simply aren’t going to convince people to abandon their faith in the scripture by citing philosophy to them, any more than you can convince them by showing them the fossil record.

(By the way, Lynch shows a remarkable ignorance of paleontology, claiming that we can’t settle questions about the fossil record because “we can’t travel back in time and use observation [another commonly shared method] to settle who is right and who isn’t about the distant past.”  Of course we can! We can absolutely show that all modern groups were not created at once, and that fish evolved before mammals.)

The “democratic” nature of science is that scientists, who already accept scientific epistemology, can all appeal to the same experiments and observations (or repeat them) to determine what is true or false about the universe—noting, of course, that all scientific truth is provisional.  Our society does not democratically share epistemological principles, and Lynch can’t make that happen through philosophical rumination.

When someone like Lynch or Alvin Plantinga goes after science because we can’t justify its methods through a priori philosophical reasoning, thereby justifying religious epistemology (and, to be fair, Lynch rejects religious ways of knowing, though he doesn’t really explain why), I ask them to answer the following question:

You have the choice of living in one of two worlds: a world like ours in which science had come into being but religion never appeared, or a world in which religion had appeared but science never did. Which would you choose?

I doubt that many people except crackpots would choose the religious world, for in that one they’d die at age 25 of an absessed tooth while praying for recovery. And if you favor the science world, do we really need a philosophical justification?  Who benefits from such a justification besides philosophers?

In the end, Hawking is right: Science will win because it works.

170 thoughts on “How can we justify science?: Sokal and Lynch debate epistemology

  1. The epistemological problems with science apply to all our knowledge about the external world.

    You are justified in querying evolution, say, on such epistemological grounds only to the extent that you are justified in wondering whether the sun really *will* rise tomorrow.

    1. Exactly. If evangelicals are so keen on rejecting evolution, they should also reject every other scientific advance not covered in their infallible book.

      1. For that matter, they should reject everything they think is stated in their “infallible” book as well. How can they even know their book exists, or that it says what they think it says, or that it won’t turn into a fish the next time they look, without relying on the same sort of inference from repeatable observation that they disparage as the “faith” of scientists?

        Fundamentally, science is the minimalist assumption here; you can’t even get to existing, infallible, non-transfiguring-into-fish books whose words must be accepted on faith without it. The Bible is not an alternative to science, it’s an useless and crippling add-on, the adware of human inquiry.

        1. Your first paragraph was similar to my thought, supposing a postmodern operator is distributive over all systems:

          pomo ( Science & Scripture )
          = ( pomo Science ) & ( pomo Scripture )

          Problem, pomo Scripture?

    1. Eric beat me to it.

      I was going to say, “there’s an xkcd for that”.

      It’s at least as ubiquitous as “there’s an app for that.”


    2. My physics is rusty (graduated 30 years ago), but this looks like Planck’s solution to the problem of black body radiation.

      What we have here is more than saying science works: it corrects it’s own mistakes! This formula was the resolution of the “ultra-violet catastrophe”, the false prediction based on assuming classical electromagnetic oscillators.

      How long would it take religion to prove that Jesus isn’t really the son of God? lol.

  2. You are so right on here. We need to recognize that this statement, “In fact, 64% of Americans would accept their faith over science if a scientific fact were shown to contravene their faith,” should be rephrased as “64% of Americans SAY they would accept their faith over science…” because when push comes to shove they actually don’t accept faith over science. Very few refuse medical treatments that are based on evolutionary principles. I’ve never heard of anyone who refused to make use of any of the advances in materials, technologies (satellite TV, for one obvious example), or understandings that have resulted from the space program, which would never have been started if we had stuck to the biblical view of the universe (or have had such success if that view were true). If most of the people declaring their disdain for scientists and science actually lived like the Amish, I would have more respect for them, even while disagreeing with their viewpoint. As it is, they prefer to close their eyes to the history of the many aspects of modern life that directly benefit them.

  3. “…a popular quote attributed to physicist Richard Feynman goes, ‘Philosophy of science is about as useful to scientists as ornithology is to birds.’ In response, some philosophers (e.g. Jonathan Schaffer) have pointed out that it is likely that ornithological knowledge would be of great benefit to birds, were it possible for them to possess it.”
    Source-Philosophy of Science-Wikipedia

      1. Just curious, under what category name would you place, e.g., “critical thinking” skills, if not “Philosophy”? “Logic”? “Symbolic Logic” (“Mathematics”)?

      1. That Isn’t the point. The point Schaffer makes is that while ornithology Isn’t useful to birds because they can’t understand it, philosophy of science is useful to scientists because they can. All scientists have a personal philosophy of science, and most have a broader one informed by natural and social philosophy. Attacking philosophy itself, rather than its bad ideas (when they pop up), is to play the same knowledge-hating game played by religious and philosophical zealots.I say this as a scientist, and by the way, Feynman was wrong here – ornithology is extremely useful to birds, by informing conservation efforts.

    1. Well, I can actually imagine circumstances where ornithologists and their knowledge might be of some benefit to birds, not that the birds would be aware of it.

      Philosophers of science, on the other hand…

    2. Birds need to know how to find food, and build nests, etc. — which is not ornithology, which was Feynman’s point.

      What was Shaffer’s point (via Wikipedia)? What parts of ornithology would benefit birds, if they could know it?

      1. Can ornithology benefit birds? Well, obviously not directly (which was Feynman’s point). Indirectly, I imagine it has benefited bird populations many times in the setting up of reserves or protected nesting grounds etc.

        OTOH, I’m not sure that philosophy of science has ever helped a scientist make a discovery or apply for a grant. The downstream practical uses of scientific discoveries – or even just the aim of adding to the store of knowledge – are all the justification science has or needs in society.

  4. These are very interesting issues. I’ll confine myself to two related points and only a couple of replies per day at most, up to the thread depth limit.

    (1) There is a meta-debate here between realism and instrumentalism about science. Does science aim at discovering knowledge, or merely at making predictions that are confirmed within the system? Instrumentalists will have no problem with your view, but realists might. If science is supposed to give us knowledge, we’ll probably need a philosophical justification of science, since track record arguments are circular as well.

    (2) Without a justification of induction (which can only be philosophical), we aren’t even justified in appealing to track records. Yes, science has a cure for polio. To be justified in believing this cure will continue to work tomorrow, we need a justification of induction, and again, science cannot supply a non-circular one. Thus even the results of science are eternally in question without a justification of induction.

    In short, philosophy can explain (and has explained) why science is better than religion, but science cannot in a non-circular way.

    1. I don’t understand why these problems are important to anyone but philosophers. Scientists go on doing their thing regardless of whether philosophers justify scientific methodology or not. The scientific enterprise would be just as far along had this kind of philosophy never existed. Are you saying that if we can’t justify induction (and you imply that we can’t), then “we can’t appeal to track records”? That’s foolish, for we do appeal to track records, and using the methods that have given us past successes continues to give us successes.

      I bet fewer than 1% of scientists know anything about the philosophical justification of induction, but those are the people who are giving us knowledge and improving the world. In other words, science progresses completely independently of philosophy.

      I still don’t understand why science needs this kind of philosophical justification. Or, if it’s not intended for scientists, whom is it for?

      1. Thanks for your reply.

        I take it that the point of a track record argument is to say, ‘This method or approach has worked in the past; therefore, we should continue to use it.’ That argument shouldn’t persuade us unless we’re justified in accepting induction, right? If we’re not justified in accepting induction, then we have just as much reason to believe science will stop working today.

        I agree that there is an instrumentalist sense in which science continues to progress: it continues to generate predictions that it then verifies within itself. But if we think science is also intended to generate justified beliefs, then I think we need more than just internal consistency.

        When you say science would be “just as far along,” that of course presupposes that science has made any progress at all, so we need a definition of ‘progress.’ If progress is merely instrumentalistic, I don’t have a quarrel with that here. But if progress is supposed to be in generating justified beliefs, then I think these sorts of criticisms are potent.

        If you want to be able to argue that science is more trustworthy than religion, it would be helpful to have a non-circular justification of science, wouldn’t it? Religionists will be more than happy to offer you thoroughly circular “justifications” of their religion, citing “progress” that is “verified” within their own system.

        1. Except we don’t have just as much reason to believe science will stop working today.

          It’s a math problem. The sun has “risen” in the east for 4.7 billion years. Man has been observing the sun rise in the east for a few million (depending on how you define “man”). Modern humans have been observing the sun rise in the east for maybe 50,000 years.

          The probabilistic track record is then — at minimum — 18,270,000 to one that the sun will rise in the east tomorrow. The outside odds are 1.7 trillion to one (it’s actually more, since the Earth’s rotation was faster prior to now).

          If you claim that the odds of the sun not rising in the east are equivalent to them rising in the east, you are engaging in a logical fallacy of false equivalence. It’s the same error Christian apologists like William Lame Craig use when they claim the odds of a god existing are at worst 50-50.

          1. Kevin,

            Thanks for your reply.

            I understand that we can use induction to predict with very high certainty that the sun will rise in the east tomorrow. The Problem of Induction asks: Why should past observations be any guide to future cases? Certainly they have been in the past, but to appeal to that track record (the track record of induction itself) in order to justify induction would be a circular argument. One way you can see this is to imagine a counter-inductionist, someone who reasons according to the opposite of induction. She would reason, ‘The sun has risen in the east over and over again; therefore, it will rise in the west tomorrow.’ How do you argue that induction is better than counter-induction? If you appeal to induction, she’ll just appeal to counter-induction.

        2. I’m no philosopher, so I probably am missing something major. But here is why this seems silly to me.

          If we are sitting in a chair in the corner of a room with a closed door, and we want to leave, must we a priori wrestle with Xeno’s paradox and provide a non-circular justification for why we can walk out of the room? The scientist just stands and walks out, while the philosopher is still seated and wringing his hands over the a priori proof that it can be done.

          All metaphysics is based on a priori propositions, and as such they are completely contained within a closed system, the human mind. The scientist uses a posteriori propositions to break out of this closed loop. The matter of induction is handled by repeated observations which provide some kind of probabilistic asymptotic approach to reality that justifies substantial confidence in the otherwise unfounded assumption of induction.

          To totally discount this asymptotic approach, as philosophers seem to want to do, is like saying that a fantasy role playing game is equally likely to provide real knowledge as physics.

          Should we sit in a dark room and fret over the fact that we can’t deduce from first principles which direction the sun travels in the sky, or even that the sun exists, or should we just go outside and watch what happens?

          The obsession with a priori circularity bears a suspicious resemblance to idealistic solipsism. Why continue to struggle to pull yourself off the ground by your bootstraps when you can just climb a ladder or a tree?

          1. Jeff,

            Thanks for your response here.

            I agree that obsessing over philosophical problems might be generally impractical. This is related to the debate between realism and instrumentalism about scientific progress that I mentioned in my initial reply to Jerry.

            If scientists simply want to be able to generate predictions that are verified within their own systems, then they are free to assume anything they want, including that induction is justified. But if scientists want to generate justified beliefs about the world, they should probably in turn justify the assumptions they’re using. This is especially true if they want to be epistemologically superior to religionists or theologians, who might also like to import unjustified assumptions or employ circular arguments. They might mirror what you said: ‘Why all this fuss about justifying religion? While the atheists are obsessing about proof, we just keep on believing in God, flying planes into buildings, and so on.’

            I’m not yet satisfied with your justification of induction in your third paragraph, because it doesn’t yet explain why we are justified in expecting frequent observations to predict the future. On this point, please see my reply to Kevin.

      2. When you say that science progresses, you’re making a normative statement, the rational justification of which is philosophical rather than scientific. So, rather ironically, just when you imply that philosophy is useless because scientists do their thing without any use for philosophy, you give us reason to think that philosophy is needed, because philosophy rather than science rationally justifies the preference for the social effects of science and technology, compared to those of moralistic religious dogma.

      3. When Coyne says that science progresses, he’s making a normative statement, the rational justification of which is philosophical rather than scientific. So, rather ironically, just when he implies that philosophy is useless because scientists do their thing without any use for philosophy, he gives us reason to think that philosophy is needed, because philosophy rather than science rationally justifies the preference for the social effects of science and technology, compared to those of moralistic religious dogma.

        This is what I call the Positivist’s Pattern, which is the oblivious science worshipper’s inevitable vindication of philosophy.

        1. …philosophy rather than science rationally justifies the preference for the social effects of science and technology, compared to those of moralistic religious dogma.

          I assume the word “preference” refers to the preferences of human beings. Can philosophy rationally justify these preferences a priori, or can it only invent such justifications after observing what it is that real people actually prefer? If the latter, why then does science need this philosophy?

          1. I meant that philosophy (ethics) is the *attempt* to rationally justify normative preferences. Clearly, we haven’t figured out how to do this decisively, leaving no room for rational disagreement. And obviously, observing what people do prefer carries no normative weight and tempts us to commit the naturalistic fallacy. But whether a satisfying a priori rational justification can be made of Enlightenment humanism and a complementary condemnation of, say, Santorum-style conservatism is another matter.

            An oblivious science-worshipper is a stereotypical positivist who foreswears philosophy even while he presupposes philosophy’s value, by ironically engaging in philosophy even in his very belittling of that discipline. He’s oblivious in that he doesn’t know what philosophy is, and thus doesn’t recognize the mark of philosophy in his own meta preference for a science-centered society.

            1. I don’t see any worship going on here.

              Also, you seem to have things backward; you talk as if the meta-preference for a science-centered society requires philosophy, or is somehow founded upon or dependent upon philosophy.

              I would say that philosophy is a linguistic conceptual system for studying, describing, and understanding such phenomena, but you seem to be over-reaching by trying to insert philosophy as necessary to the phenomena being considered by philosophy. That would be like physicists claiming that their quantum theory is what causes elementary particles to behave as they do.

              Certainly scientists do things that philosophy can describe and illuminate, especially when they engage in mathematics and logic, or when they form concepts based on sensory observation or measurements performed with instruments.

              But scientists don’t need philosophy to get started. What you are saying sounds to me like saying that the play-write and actors cannot perform without the help of the critics. Certainly critics add something, as do philosophers, but it’s really an egotistical self-aggrandizing critic who claims that the performance would be impossible without his contribution.

              1. The meta-preference for a certain society doesn’t require philosophy. As I’ve said throughout this Comments forum, what requires philosophy is the rational justification of that normative preference.

              2. But I thought the whole issue here was whether or not an a priori rational justification was necessary to say whether science or religion was epistemelogically sound or preferable.

                Forgive me for being philosophically naive, but here is what I’ve been able to glean from the discussion, and perhaps you can correct this:

                It seems like philosophy wants to insist on an priori justification because a posteriori propositions run aground on induction; their weakness is in their particularity as opposed to the generality of a priori justification.

                But, per Kant, all metaphysics depends on synthetic a priori judgments. This seems like an endlessly circular closed-loop where you are forever trapped inside the closed system that is the human mind. In this system of a priori judgement every hypothesis, whether it’s flying spaghetti monsters, pink unicorns, or a monotheistic intelligent creator, is equally probable. You may as well try to determine what direction the sun moves by remaining forever in a dark room, and good luck.

                Scientists do something quite different: based on a posteriori synthetic judgments they establish what are the most probable hypotheses by falsifying or not falsifying possible candidates using observation. By repetition this amounts to an algorithm for asymptotically approaching true knowledge, and for establishing a high degree of probabilistic confidence in induction.

                So what is epistemelogically preferable: to have a reliably repeatable algorithm capable of producing successively better approximations of natural reality, or sitting in a dark room where everything you can imagine is equally probable, and continuing to fret over the fact that no a priori judgement can improve the situation or ever produce absolute knowledge (except for analytic judgments)?

                Are philosophers in this case doing something other than scoffing at extremely good approximations while futilely lamenting the inability to attain perfection?

      4. Jerry,

        You might not think it’s a serious problem, but think of replacing a single word in what you wrote: “Theologians go on doing their thing regardless of whether philosophers justify [their] methodology or not. … In other words, theology progresses completely independently of philosophy. ”

        Yes, they will go on doing what they do, and theology progresses even when the philosophers tell them they’re giving the wrong reasons, but should we prefer theology over science? It’s pretty obvious (at least to me and you) that scientific methodology is to be preferred over theological methodology, that scientific progress is different than theological progress, but many things we think are obvious that are wrong. We need a bit of rigor. That’s where philosophy comes in. You might not like it, but it’s possible that when we’re being rigorous, we might learn that some background assumptions we take for granted contradict one another.

        For instance, if you want to be rigorous and you think that preferences (for methods or theories) must be justified by giving positive reasons, then you’re caught in all sorts of problems — Hume’s problem of induction, Goodman’s riddle of induction, the Münchhausen Trilemma, the problem of underdetermination, the problem of the criterion, and so on. Here we have an apparent contradiction between philosophical rigor and positive reasons for belief.

        If this contradiction holds and we want to be rigorous, then we might have to change how we think why we preferred science over theology in the first place.

        Maybe it’s the function of a community’s attempt to solve problems? Lakatos might have a point about progressive and degenerative research programs, for instance. Maybe it’s ethical? See: virtue ethics and virtue epistemology. Maybe it doesn’t have to do with correspondence with the facts but with the willingness of scientists to revise their beliefs? Who knows? But we never get around to asking these questions if we keep giving bad arguments. Besides, if you want to crush theologians, then crush them with a good argument, not with a bad argument that feels good.

        1. We need a bit of rigor. That’s where philosophy comes in.

          Unfortunately… no, that’s where philosophy is supposed to come in. However, too many philosophers don’t worry about getting anywhere near the rigor of mathematics.

    2. “Without a justification of induction (which can only be philosophical), we aren’t even justified in appealing to track records.”

      Exactly. What I said. You should doubt science if you doubt the sun will rise tomorrow.

      1. (This will be my other reply for today; I’ll try to continue any threads tomorrow.)


        I agree that it’s inconsistent to doubt science without doubting that the sun will rise tomorrow.

        I think it’s obvious that the sun will rise tomorrow, because I think the intuitive (and thus non-scientific) credentials of induction are obvious. And I think the same about science: that the intuitive (and thus non-scientific) credentials of science are almost as obvious. Rationalist philosophers are perfectly happy to affirm the trustworthiness of induction and science; they just think that that trustworthiness requires a philosophical or a priori argument.

    3. The circularity isn’t vicious, however–there’s no reason that the use of a method to justify itself will automatically provide validation, especially when judged upon ability to yield successful predictions. A method can successfully find its own limitations. Alvin Goldman makes this same point in his _Epistemology and Cognition_, which argues for a form of naturalized epistemology. There’s still a residue of what “must be philosophical,” but it has to be informed itself by empirical science.

      1. Jim,

        I think epistemic circularity is a very interesting and complicated topic. And I think the ultimate justification for some theory of virtuousness of circles, or maybe coherentism in general, would be ultimately philosophical. I don’t think you disagree with that.

        As for this particular circle, I don’t yet see how we have a justification for trusting that induction will continue to work. Yes, it could have been the case that induction didn’t work sometime in the past, and luckily, it has kept working. Clearly there would be more reason to reject induction if it had failed in the past. But absence of evidence is not evidence of absence. Why think this reliability will continue?

        Suppose my theory says that induction works until today’s date, and your theory says that induction works up to today’s date but also after. Both seem as virtuously circular as each other, and so shouldn’t we be neutral between them? (You could respond that my theory is less likely to be correct for some other reason, but I suspect that reason will appeal to some sort of a priori evidence.)

        1. Tom,
          If you are willing to board an airplane, then I cannot take you at your word that you believe induction could fail today; you in fact believe the opposite. 7 billion hearts keep beating second after second; if we feared induction were false then we would live in fear of discontinuity, rampant chaos and death.

          The only reason we are capable of a priori logic (because we have brains) is because induction has not-failed for long enough to allow evolution to take place and to create and shape those brains. We have brains that can conceive of 2 + 2 = 4 because of the natural fact that two objects with continuous physical extent, when placed beside two other objects with continuous physical extent, form a pattern we recognize as 4 objects with continuous physical extent; that physical reality shaped our brains in such a way that it can represent in subjective logic that external reality our senses perceive; our logic is a projection of external natural reality into the structure and capabilities of our brain.

          It seems philosophers work from the hypothesis that the subjective brain precedes external objective reality, that somehow external physical reality depends on what our brains can do, and that the ability to doubt induction somehow takes precedence over the actual behavior of nature that created the brain that is capable of even conceiving of induction. Perhaps we can conceive of induction because it is true. It seems you philosophers abstract yourselves into a reversal of actual precedence: physical reality shaped the brain which works in ways according to the nature of the physical reality that shaped it.

          1. Jeff,

            I don’t believe induction will fail, because I trust a priori, philosophical evidence.

            It does seem likely that we can use logic because we evolved to have powerful brains.

            1. I’m a little confused by “I trust a priori, philosophical evidence.”

              I gathered that the lack of a priori justification for induction was the problem being discussed, so it isn’t clear to me what evidence you mean.

              But it does seem that abb3w has been making mathematical arguments that provide some rigor for what I’m talking about, a kind of probabilistic confidence. Unfortunately I don’t know enough about the deeper foundations of set theory, or it’s application to the completeness and consistency of mathematical axioms to fully appreciate abb3w’s points. Maybe with a bit of work.

              It seems like Lynch is reaching for some kind of unreachable logical perfection; it’s as if the subjective methods we have of representing the objective world, whether by visual metaphor or analogy, or by analysis, logic and other linguistic descriptions, is fated to forever be incomplete because of it’s fundamental lack of identity to its objective target. It seems there is always an inherently absent logical keystone that can never complete the arch to bridge the gap between the subjective and the objective.

              But this seems to be a very poor reason to continue, as Lynch seems to do, pretending there is some kind of equivalence between rational inquiry into the nature of our surroundings (via empirical measurement supplemented by logical analysis), and the seemingly purely imaginary subjective products of metaphysical longing.

              1. Jeff,

                In my original post, I wrote, “Without a justification of induction (which can only be philosophical), we aren’t even justified in appealing to track records …. philosophy can explain (and has explained) why science is better than religion, but science cannot in a non-circular way.”

                I think there actually is some a priori evidence for induction. In fact, I suspect it’s not that much different from what abb3w is hinting at. It requires the fundamentally philosophical supposition that the laws of mathematics are necessarily sound: they actually do apply to the world, rather than merely being a self-consistent formal system. (If you want to read about it, there’s a description here.)

                I agree that it would be a mistake to conclude that since science is epistemically unjustifiable (that is, unjustifiable in the sense required for knowledge) without philosophy, all philosophy (including, as you say, “imaginary subjective products of metaphysical longing”) is justified. But what I want to emphasize here is simply that the atheists’ typical critiques of religion have much less bite if those atheists import their own unargued assumptions about the trustworthiness of science.

              2. Jech’s “Set Theory” provides a solid (if hard) introduction to set theory’s foundations. Linz (ISBN 978-1449615529) provides some additional background on complexity that might be helpful.

                The philsophical supposition required is not that the laws of mathematics actually do apply to the world. Rather, it is the supposition that on the one hand the laws of mathematics do give rise to a self-consistent formal system (IE: a language), and on the other that some sub-system corresponds.

                Philosophically, it’s as axiomatically valid to take refutation on this. It just leaves you hopelessly out to (Ramsey) sea – which alternative does not help theology one bit.

                I suspect the reason that most atheists leave such foundations unargued is in part that most atheists have trouble with the math — but that most theists have even more so. This means such arguments will have no effect towards changing the minds of most people, which by practical criteria leaves the effort a waste of time.

                I’d also emphasize that science does not require ALL of philosophy for it’s justification; rather, saving only a single axiom to start talking about “evidence”, it only requires the philosophy of mathematics. (Engineering requires one further.) This is a far cry from needing to give a damn about Kant, Heidegger, or Nietzsche.

              3. Jech’s “Set Theory” provides a solid (if hard) introduction to set theory’s foundations.

                And bloody expensive! £108 for the forthcoming third edition on!!!


        2. The Vitanyi-Li MDLI theorem form of Occam’s razor provides means for determining which is more probable.

          The theorem does not require a priori evidence; rather, it requires an a priori axiomatic Assertion. (That experience has pattern recognizable with ordinal degree of Turing hypercomputation.) As it is done axiomatically, without justification from priors, you can instead take the Refutation… but are then stuck with consequences from doing so. (Alternately, you can avoid any axioms that would yield the theorem; similarly problematic.)

          In either case, the trick is for the philosopher to get anywhere at all thereafter.

          1. abb3w,

            I would hope for something more solid: a justification of induction like this one, which goes beyond mere axiomatic assertion, as long as we accept the philosophical argument that the laws of mathematics are necessarily sound.

            I agree that a philosopher who rejects induction will not live a very happy life. But, of course, many people claim that someone who rejects theism will lead a comparatively unhappy life. Whether this is true, we can at least recognize that ‘not-p makes me unhappy’ is not an argument that p is true or epistemically justified.

            If we’re instrumentalists about science, we might not care whether science is epistemically justified. But once again, as I and a few other commenters have tried to emphasize, this makes atheists’ criticisms of religion substantially weaker in turn, since the atheists tend not to be criticizing religion from a similar, instrumentalist position.

            1. The linked justification is actually weaker than the MDLI theorem on several levels. Foremost, “relative frequency” implicitly presupposes the ability to count — which means most of ZF or an analog has to be slipped in the back door. So, it’s making several axiomatic assertions; you’re just not paying close attention to what they are. The Vitanyi-Li paper instead rests explicitly on ZF, and an axiomatic assumption that reality produces evidence with RE-complexity (or, in extended form, of complexity recognizable by ordinal degree hypercomputation, with RE being degree zero). Williams and Stowe aren’t (or at least, the presentation of them isn’t) even that careful to be explicit in their requirements; and require the more specific assumption of a finite universe for sampling from, which is a sub-case.

              In short, you appear to to protest the assumptions only when they are introduced honestly, rather than skillfully slipped from the philosopher/magician’s sleeve.

              Also, “necessarily sound” may be the wrong way to put it. ZF gives rise to mathematics as a pure abstraction; there are several others, however, that give rise to analogous but distinct tools, much like Euclidean versus non-Euclidean geometry. For the interesting alternatives, translation back and forth is possible, such that there will be similar relational sub-structures in each.

              As to the “substantially weaker”… not really, since the religious position can be shown to be one of
              1) a sub-case of the Assertion, giving rise to science
              2) a sub-case of the Refutation, where a separate axiomatic act of faith is required each time one distinguishes a hawk from a handsaw
              3) internally inconsistent, and thus “False”.

    4. Without a justification of induction (which can only be philosophical)

      Somewhat, in so far as Mathematics is a branch of Philosophy.

      For a mathematical justification of induction, see “Minimum Description Length Induction, Bayesianism, and Kolmogorov Complexity” by Vitanyi and Li (doi:10.1109/18.825807). It depends on the existence of such an ordinal and on the standard axioms of ZF (and thus, the axiom of infinity), but not on the Axiom of Choice. The principle therein can be extended to higher (non-zero) ordinals of Turing hypercomputer model, rather than just Recursively Enumerable complexity (the 0 case). That experience is produced with some specific AH-complexity needs to be taken as an axiom.

      Refutation is equally valid, in which case the appearance of order is a local island in an adequately large Ramsey sea of chaos. However, there appears to be no prospect for justifying going to get a pastrami sandwich. Thus, any philosopher who gets up for a sandwich rather than sits in place indefinitely has rejected such Refutation.

      Science can be expressed as a pseudo-algorithm resulting from the theorem.

      1. abb3w,

        I think there’s a sense in which mathematics is a branch of philosophy, or at least requires philosophy. Mathematics seems to be a language of logic, and there’s always the metalogic question of whether the language of logic is sound: whether the theorems of that logic actually “apply to” reality. Yes, our logic says that 2+2=4, but how do we know that will always be the case in real life?

        I think the rationalist has a relatively easy answer here, that it’s inconceivable that 2+2 not equal four, or the intuition that necessarily, 2+2=4, and so on. These are ultimately a priori and thus philosophical reasons.

        The empiricist might try to appeal to induction, but there’s good reason to think that won’t justify non-contingent truths such as that 2+2=4, and in any case, would just widen the circle here: now we’re appealing to induction to justify mathematics, which in turn we appeal to to justify induction.

        1. Sigh. “WordPress login required” ate the long response. Retyping short highlights….

          0) You’re either confusing or equivocating between two senses of “induction” at the end; while only incidental, it’s grating.
          1) 2+2=4 isn’t an a priori; it’s a theorem, derived from axioms about abstract relations of abstract entities. Under a perverse enough axiom schemata, it won’t hold.
          2) “Whether it will always be the case” is effectively the wrong question; it negelects the mapping between abstraction to instantiation. While we can map the abstractions to the instantiations, an instance “not the case” means the map was incorrect, not that the theorem is wrong.
          3) Axioms aren’t “justified”. They’re just taken. “These are my principles; If you don’t like them, I have others.”
          4) Math isn’t so much a language, as a class of languages.
          5) To reject mathematics, you have to avoid any tools inherited, and any tools that might re-introduce it. The correspondence of Chomsky Type-0 grammars to Church-Turing Automata recognized languages makes that difficult. (Most philosophical journals limit publication to finite strings of symbols, for one thing….)

          1. abb3w,

            (0) By ‘induction’ in that paragraph I mean ‘reasoning that unobserved cases are probably like observed cases.’ I don’t understand how I can only use that sense in at most two of the three occurrences of ‘induction.’

            (1, 2, and 3) By ‘2+2=4’ I mean the corresponding soundness-of-logic claim: that (crudely) “in real life,” 2+2 will always equal 4. My worry might be recast as indeed that the map might be incorrect. As for justification, again, I meant the soundness claim, not the justification of axioms of a language of logic.

            More importantly and generally:

            How could a hardcore empiricist ever justify the claim that arithmetic is sound, let alone that our arithmetic is necessarily sound?

            (Here of course the designator “our arithmetic” is functioning rigidly.)

            (4) If that’s the terminology you use, I can use that too.

            (5) I don’t know what those things are. But I certainly don’t think we should reject mathematics, because I accept a priori, philosophical evidence.

            1. The usual mathematical sense of induction refers to the process of showing for a base case, and then showing case k implies case k+1. The formal justification of this (under ZF) boils down to the a property resulting from construction of the natural numbers via the Axiom of Infinity. Such induction, therefore, results from a proposition explicitly taken as Affirmation; it also involves unary probability in the conclusion. The scientific sense of induction is in part derived from this, as that theorem relies on such mathematical induction for some precursor lemmas; however, in so far as it involves inference about unobserved evidence from observed evidence, the justification also involves (and principally results from) taking an additional non-ZF axiom to assign mathematical reference to “evidence”.

              Addressing your recast worry: yes, it’s possible the map might be wrong. What the scientific sense of induction allows showing is that it’s probably not. (Though that’s a slight oversimplification to translate the math into English.)

              To address your main question:

              How could a hardcore empiricist ever justify the claim that arithmetic is sound, let alone that our arithmetic is necessarily sound?

              Depends what you mean by “hard core”. I will presume, however, that I suffice as an example. (Dr. Coyne almost certainly has a different position.) In terms of the Münchhausen Trilemma, my approach is primarily foundationalist (though it can be topologically recast into infinitism with trivial effort, or nearly coherentist form with a bit more). In my case, the justification is roughly that I explicitly take the soundness of jointwise logical affirmation of the ZF axioms as an abstractly sound starting point — which can be characterized as the “faith”. However, that is essentially the only “faith” needed for mathematics (as the rest follows as consequent inferences and attaching of names). Someone who wants to dispute it, can point out which Axiom they wish to take refutation of… and proceed from there. At which point the problem then becomes finding something which is does not hide some ordinary form of mathematics wearing Groucho Marx glasses as a disguise, and which allows showing that the disputant is not actually a cabbage.

              Of course, you may have a different sense of “sound” — that is, not whether it is valid in abstraction, but whether it has a correspondence to “reality”. That involves the non-ZF axiom mentioned earlier; again, taking Refutation is an option, but leaves one hopelessly unable to distinguish a hawk from a handsaw. There are too many patterns mathematics can describe to readily avoid a correspondence.

              Wikipedia has a bit on the Chomsky grammar types. Essentially: anything that you might think of as a language is actually a cleverly disguised form of math that talks about patterns. Church-Turing Automata are more commonly called “Turing Machines”; the idiosyncratic use is to avoid confusion with the notional program from Turing’s AI thought experiment. Linz (ISBN ) gives a good technical introduction.

  5. please, Mr. Lynch, do cease being such a hypocritical ass and actually live like your supposed philosophy requires. How can we know anything at all? Indeed, you should have no more consternation in using a pigeon to cure leprosy than you would using modern medicine.

  6. If there is, in fact, only one method for determining what is true, then we plainly would not, in principle, be able to justify it without self-referencing.

    It’s not entirely clear to me that this would be wrong, either. Surely another word for this approach is *consistency*. Appealing to, for example, a supernatural justification could be (would be, in the case of science) *inconsistent* with the mechanisms outlined by ‘the method’, to establish what’s true.

  7. Indeed, its not at all clear why anyone should care about this. Similarly:

    I can never be sure that I am not a brain in a tank.

    A child can always respond to any explanation of anything with “But why?”.

    So what?

    1. Primitive religious tribalists can also “do things” without abstruse philosophical justifications: they can create a moralistic communal or hierarchical society that empowers religious officials who maintain their social order with dogmas and superstitions. When you want to rationally justify your normative preference for the social effects of a modern, humanistic, pragmatic, technoscientific society (including its threat to all life on Earth), you’ll be doing philosophy, not science. Just so you know…

      1. I would be interested to know what you would consider to be epistemic justification for heliocentrism.

      2. science does not need justification

        overshoot is the result of our evolutionary succces: we transform DNA from non-human into human form because we are most resent product of evolution

        Science is manifestation of our deliberative capability

        our social organization is very primitive and is completely pre-science

        for us (mankind as organism-whole) to evolve further and handle our evolutionary success that has become the source of our problems means taking the science and bringing it into government

        that is to make science shepherd of human condition as opposed to institutionalized ignorance as it currently is

        but don’t worry: two or three generations of die-off will make us all scientists and non-believers

        it is just a matter of time and the fianl sustainable population: the less time we take to rebalance the more of human DNA we will have at sterady state; the more time we waste on philosophy aand religion (or beliefs in general) the less complex DNA (human and non-human) we will have at an attractor state of human condition

        that’s all

        1. One confusion I had in discussion with Mr. Hue is that there are two separate issues here: one is the question of epistemelogical basis for religion and science as ways of knowing, which is the main topic of Jerry’s post.

          The other, which Mr. Hue seems to have latched onto, is related to Jerry’s final question of which society we would prefer to live in. All of Mr. Hue’s posts are related to that last question, and discussing the philosophical basis for a rational choice. This is a matter of ethics, not epistemology.

  8. When people give arguments like this to show that science is no different than religion I want to throw up my arms in despair. Anyone who uses this explanation would be impossible to appeal to. Science must use the natural would as proof for a supposition whereas religion will use nothing more than religion itself and their deep feelings of transcend truth as proof. These people are totally delusional.

  9. Science and religion…….

    By their fruits Ye shall know them.

    I think science’s fruit, no matter how imperfect to some philosophers, is still superior to anything religion has produced in 3000 years.

    1. Woah! Wouldn’t think I would run into someone else that think justificationism just doesn’t work. It’s why I’m a bit worried whenever Jerry or another New Atheist claims that all rational beliefs must rest on sufficient reasons. It’s just not good philosophy. Hans Albert knew this; Bartley knew this. Blah.

      1. » d:
        Hans Albert knew this; Bartley knew this.

        Um, yeah, because they learnt it straight from Popper… 🙂

        But you’re right. The radical epistemological inversion—much like Darwin’s “strange inversion of reasoning”—hasn’t exactly sunk in yet.

        1. And Popper is indebted to Kant and the Presocratics. So it goes. What matters is that, as far as we can tell, the nonjustificationist position is to be preferred.

          The ‘strange inversion of reasoning’ has started to make a dent in epistemology and philosophy of science, especially in virtue epistemology and some externalist versions of justificationism, but there’s a long way to go.

  10. The problem with philosophy is twofold.

    1 – You cannot deduce things about the universe by just sitting in a chair and thinking really hard about it.

    2 – Philosophers seem unable to accept anything other than deductions made from sitting in a chair and thinking.

    If your discipline cannot even find a way to acknowledge that the universe exists, and that observations and experiments can tell us more about it, I don’t really, really don’t know what you expect philosophy to tell us at all. It has always seemed to me that some philosophers just reached the limits of what can be done in their comfy chair and declared no one could do anything ever.

    1. Ooh, harsh but fair.

      To some extent this is true by definition, in the sense that once people get out of the comfy chair they don’t tend to get called philosophers any more.

    2. To divorce careful observations from epistemology constitutes a grave malady. Thereafter the patient becomes overwhelmed by a marching army of illusions.

      1. Which illusions? The very simple fact that I can place an object somewhere, then I can communicate to another person what the object is and where it is, and they can independently, using my verbal message, go find in that location an object with properties that match my description is already a huge step toward establishing that there is an objective reality, which allows us to form very adequate approximate concepts of what it is and how it behaves.

        It seems like in 30 seconds a pair of children can do most of what epistemology might like to do, and it only remains for epistemology to catch up by dressing our common experiences with rigorous conceptual foundations. But the mere fact that we have evolved and exist with such a lengthy continuous history of life over geologic timescales means that, at least in the locality of earth and this universe’s time scale we need not worry about the philosopher’s worst epistemelogical nightmares of strange discontinuities or inconsistencies between our conceptual models and physical reality.

    3. 2- is misleading. An individual philosopher can retreat to the world inside their head. Academics in the philosophy department of a university cannot.

  11. Science advances, not by proving ideas to be true, but by weeding out those ideas that are incorrect.
    It might be correct to state that neither religion or science have a definite way to show something is true but only one of them has a means of determining whether a particular idea is false.

    1. » Sigmund:
      Science advances, not by proving ideas to be true, but by weeding out those ideas that are incorrect.

      As I say, an idea that we owe to Popper. We should remember his ideas more often, I think.

    2. That’s the Sherlock Holmes school: If you eliminate the impossible, whatever remains, however improbable, must be the truth. But I disagree, Sigmund: sometimes we test ideas that we think are true, and when they pass the test, we gain more confidence in those ideas. When evolution makes predictions about what we find in the fossil record, and then we see that (e.g. birdlike dinosaurs about 140 mya), we are confirming something at the same time we’re dispelling creationism. To advance scientifically, we need not only to weed out bad hypotheses, but find repeated support for the alternative good hypotheses.

      1. » Jerry:
        But I disagree, Sigmund: sometimes we test ideas that we think are true, and when they pass the test, we gain more confidence in those ideas.

        But that’s not to disagree, Jerry, that is, in effect, to make Sigmund’s point: Only if an idea passes a test do we gain confidence. And logically, such a test must entail the possibility of failure, so that every real test becomes a de facto attempt at refutation. I keep dropping Popper’s name, but that’s because his ideas are really indispensable in this regard.

        1. …though I’d argue Popper’s emphasis on Falsification neglects how it is a sub-case of Simplicity. (If your explanation does not give the correct answer, it is not the simplest explanation that gives the correct answer.) Ultimately, he underplays the philosophical import of Simplicity, confusing some anthropologically practical reasons it’s used with foundational philosophical justification. (It’s used because it’s easy; but that’s not why it works.

      2. But testing good hypotheses involves making sure that the ‘positive’ hypothesis is not incorrect and the best available explanation. The standard process of reviewing a scientific paper often involves a reviewer thinking of an alternative explanation for the results shown (perhaps enzyme A doesn’t really degrade protein B, it actually degrades protein C that stimulates the expression of protein B) and then asking for control experiments that would rule out the alternative hypothesis. Even the positive experiment can be viewed as disproving an alternative incorrect explanation – as in your point about disproving creationism.
        I guess this sounds a little like a semantic argument but I don’t think it is that simple. Many ‘positive’ results can actually be explained by alternative mechanisms and progress is made by additional experiments that show one of these mechanisms is incorrect.

      3. To advance scientifically, we need not only to weed out bad hypotheses, but find repeated support for the alternative good hypotheses.

        David Desutsch, one of the early founders of the theory of Quantum Computing, wrote a book, The Fabric of Reality expounding on a similar opinion: that the main job of science is not falsification of bad ideas, but to come with an explanatory framework for true ones. In the book he uses Evolution, Computability Theory and Multiverses as examples (though I personally am not so sanguine about the last one).

        1. *David Deutsch. Also, if I remember correctly from a discussion I had earlier on this weblog, one of the regular readers (probably Ant Allan?) is a big fan of Deutsch’s ideas, so maybe (s)he can correct me if I inadvertently misrepresented any of them in my post above.

          1. It is I, but I don’t remember well enough to correct or confirm. (I can check later.)


            PS. He. Ant is short for Anthony.

        2. That is indeed somewhat misrepresenting Deutsch’s thesis, tbh. Deutsch is firmly on board the falsificationist boat, the only thing he adds is an insistence on good explanations as a first step in the whole process, which possible goes a little further than Popper’s ideas did. I’d have to check that, though.

          1. Yes, Deutsch’s does emphasise the explanatory power of a “theory” (which, in TFoR, he seems to be using in the idiomatic sense; I’d say, “hypothesis”): “theories that are capable of giving more specific explanations are automatically preferred” (p. 66 in the 1998 Penguin pb edition).

            There’s probably more discussion of this later in the book, but I need to get back to work!


          2. (Sorry, I have to use a different pseudonym because of the new sign-on process)

            If I remember correctly, he puts “Popperian” epistemology of falsification as one of the pillars of his vision of a Grand Explanatory Theory, the other pillars being Computability, Evolution and the Multiverse interpretation. In that sense, I do think he gives higher weight to explanatory power than to falsification.

    3. Continual validation gives an ever higher confidence level in scientific truth.

      Note that the title of this bl— website is not Why Evolution Is Not False! 😉


      1. » Ant:
        Note that the title of this bl— website is not Why Evolution Is Not False!

        In a sense, though, that is what science does: to find explanations that are not-obviously-false. We are still pretty certain that our theories only approximate reality, and only after having passed rigorous tests do we confer upon a theory the tentative status of being true. But “true” here means just that: a good explanation that has passed the most rigorous tests we could think of. Ultimately, every theory that has been true will be superseded by another that meets even the new challenges the old one(s) failed to address.

        1. » Peter: Ultimately, every theory that has been true will be superseded by another that meets even the new challenges the old one(s) failed to address.

          Well, no. The theory of evolution by natural selection has been superseded by the theory of evolution by natural selection, genetic drift and [a couple of other things I can’t quite remember off the top of my head]. Does that mean the theory of evolution by natural selection is false? No (well, profoundly unlikely). Just incomplete.

          In a very real sense, it may be wrong to say it’s been superseded; rather, it’s been extended and refined.

          See Asimov.


          1. Ant, we are actually saying pretty much the same thing. ‘True’ means ‘tested with respect to a particular problem-situation’, and with respect to the problem-situations prevalent in the 18th and 19th centuries, Newton’s laws are still true in that sense. Except, of course, for the fact that the explanations imbedded in Newton’s theories have in fact turned out to be false, and this part of what it means to do science (to provide good explanations) shouldn’t be forgotten.

            1. ‘True’ means ‘tested with respect to a particular problem-situation’

              That’s a rather limited notion of truth. And surely it is far too limited for science. “The earth is flat” is “true” by that definition under some reasonable “problem situations” (e.g., firing artillery shells a few hundred yards).

              1. I take it you can offer a more satisfactory alternative for ‘truth’, or at least state why my definition is too limited?

                And “problem situation” is a technical term, meaning a conflict between ideas that is deemed for some reason to warrant resolution. And in any case, firing artillery shells a few hundred yards is not just a mere technical problem (and not an explanatory one) but more importantly doesn’t depend on any theory as to the non-local shape of the planet (as opposed to, say, ICBMs).

    1. That’s a nice expression of philosophical pragmatism. Are you unaware that faith works in the sense that it efficiently creates an alternative sort of society, one that’s highly moralistic, dogmatic, and communal or tribal? When you favour the social effects of science and technology over those of spiritualism, mysticism, and religious dogma, do you think that choice of yours is self-justifying or that philosophy isn’t the sole means of rationally justifying it?

      1. If we could go back over time and observe human cultural evolution, I suspect we would find that morality, dogmatism, community, and tribalism all preceded faith. Faith was probably a kind of surrender to the fact that humanity was unable to fully explain and understand all the mysteries they faced in their environment.

        So I would say it’s totally false to say that faith created morality, dogmatism, and community. Faith was probably created by a moral communal people’s inability to do good science.

        1. And let me clarify that these people did decent science when they created stone axes, harnessed fire, and made slings and arrows and spears and developed hunting technology and the means to use animal skins for shelter.

          It’s just that their developing brains eventually enabled them to ask questions that they could not answer, so they punted by inventing faith. We are now learning to discard that temporary place holder.

          1. Jeff Johnson,

            Now you’re talking about something called “faith,” not religion. You’re getting hung up a definitional issue here. You seem to be identifying religion with organized religion. Again, I’m no historian, but as I understand the anthropology of religion, primitive religion, in the sense at least of a ritualistic burial of the dead that entails a theology of the afterlife, is just about the oldest human cultural expression for which we have any archeological evidence. Shamanic religions are less hierarchical and codified than organized, institutional ones, but they’re religions nonetheless.

            Either way, my point stands that whatever you want to call religion works in its own right by contributing to the creation of a different sort of society, requiring a normative choice between that sort and, say, the humanistic, pragmatic, science-centered one. And a rational justification of that choice in turn requires philosophy.

            1. “whatever you want to call religion works in its own right…”

              Sure, but that’s not what’s under discussion here. Religion might “work” in its own right, but it does not work in the sense of “finding things out about the world”. It’s not hard to redefine the word “works” such that the statement “X does Y, therefore X works” is a tautology, but that’s not really relevant to the topic at hand.

      2. John:
        Scientific methodology has yielded reliable, reproducible, and predictably accurate results. That’s why I say that science works. Faith has done none of these. As you say, faith creates an alternative sort of society (how efficiently is up for discussion), but since there is no one agreed upon faith methodology, then the results of this creation are inconsistent one from another. It’s a garbage in, garbage out logic stream with no initial basis in evidence.

        1. I understand why you say that science works, because it works in a pragmatic, instrumental sense, empowering us to control the forces of nature, raising our standard of living in a materialistic and otherwise quantifiable way (our lifespan), etc. The question is whether you understand that when Coyne says this, he contradicts himself when he also says that philosophy is useless on this matter of justifying first principles. What Coyne misses is that his preference for a pragmatic, science-centered society presupposes humanistic, Enlightenment-style values, assuming he’s interested in rationally justifying that preference. This is because the notion that one sort of society is better than another is normative and value-laden, and is thus hardly provable by science. Philosophy is the last rational resort for adjudicating normative issues.

          1. No, science “works” in the sense that it does what it says on the tin — it gives us ways of making reliable predictions about what the universe is going to do under any given set of circumstances. Religion claims to do all kinds of stuff like making people more moral, helping them go to heaven, describing the origin of the universe, blah blah blah, and it simply doesn’t do any of that stuff, as far as anyone can tell. It may have certain effects on the way people interact with each other, but the things it claims to do are only loosely connected to the things it actually does. So it’s not value-laden at all to say that science works and religion doesn’t, it’s just taking the two approaches at their word and comparing their results to their claims.

            Not only that, but religion depends fundamentally on the same methods that science depends on in order to operate. In order to believe that this book right here is the word of God and that we have to do what the book says, you have to be able to have confidence that there *is* a book here, that it says the same thing every time you open it, and that it won’t turn into a fish the next time you look at it. Only then can you begin to take its commands as a basis for faith. In other words, you have to be willing to make an inductive inference based on your repeatable past observation of how books, and this book in particular, work. Furthermore, in order to put the book’s commandments into practice, you have to be willing to make inductive inferences based on your repeated past observations of how you are able to interact with the world around you, and to choose your actions on this basis.

            As far as anyone has ever been able to discern, one simply *must* accept the fundamental principles of science in order to be able to do anything at all. No-one has ever proposed a reasonable alternative to this — a retreat into the solipsism of refusing to draw any inferences whatsoever from your past experiences means that you literally have no way of deciding what to do to accomplish your goals and must remain frozen in indecision or act entirely at random (and even those are decisions). This holds true even if you’re actually a brain in a jar, since the only way you can ever possibly find this out is to examine your experiences extremely carefully and try to find inconsistencies which would prove that they’re not real. If those inconsistencies are non-existent or in principle too subtle for you to detect, you might as well treat the solipsistic world as if it’s reality, because you will literally never be able to inhabit anything else.

            Religion is *not* an alternative to science, it’s an add-on, so it’s absurd to argue that the two are equally good. Religious people don’t just base everything on faith, they base everything on the same ideas on which science is based, and in some cases add an additional assumption that their understanding of their religious beliefs is also accurate. So it’s absurd to argue that the two are simply alternative and equally good ways to look at the world. Religion fails Occam’s Razor by multiplying assumptions without need, and it fails the “it works” test by not doing what it claims to do (testable by the same type of inference from repeatable observation that everybody has to follow just to live — and if you don’t think everybody has to do this you better be ready to propose an alternative). It’s just not on an equal intellectual footing with science, and it’s a ridiculous bit of sophistry to try to pretend that it is.

            1. Actually… I want to make a stronger claim than that we have to follow scientific principles just to live. I want to claim that we literally don’t know any other way of thinking other than to draw inferences from our past experience. Even “I think therefore I am” is an inference from our recollections of our past experience of thinking and being — we have never experienced thinking without being or being without thinking (at some level) and so we infer that the two are identical. Logical axioms, such as the principle of non-contradiction, are similar — we have no experience of a thing and its negation existing simultaneously, and no experience of even being able to conceptualize such a state, and so we rule it out. Even when we formulate a purely logical argument we must have a certain trust in the ability of our own minds to formulate the argument correctly, and that trust is an inference from our past experience of having done so (as far as we can tell).

              Without inference from past experience, I don’t see how we would even be able to think, and, at it’s root, science is just a rigorous formalization of that process of inference from past experience. We think, therefore we do science. Consequently, we need no more justify that process than my laptop needs to justify doing its operations in binary. That’s just how the machine between our ears works, and nobody has ever proposed a good way of getting away from that. Religion is the interloper requiring justification, in the same way that you’re gonna have to talk pretty fast to convince me that I need to be running Windows on my machine.

  12. There are many religions, but no way of telling which (if any) is the most true.

    There are many philosophies, but no way of telling which (if any) is the most true.

    There is only one ‘imperfect’ science, which appears to be an increasingly accurate way of describing the natural world.

    Can Michael P. Lynch prove that he exists through religious or philosophical thought?

  13. Yessir, accumulated practical track-records (accumulated results of empirically inter-subjective demonstrations) rightly trump philosophical ponderings as a (fallible, yet nonetheless stunningly reliable) way of knowing.

    And yes, all empirically-derived objective knowledge is forever fallible (and thus forever subject to revision and even abandonment as future discoveries may warrant), but prevailing point-in-time empirical demonstrations (despite lacking rigorous philosophical justification as certainly and unconditionally true) ARE the point-in-time “way to bet.”

    Sure, occasionally such an empirically-based bet will subsequently be (empirically!) shown to be a losing bet (which can THEN be modified or abandoned with warrant), but that is as good as it gets or even CAN get; thus far no non-empirical epistemology has produced a track-record that even begins to approach the spectacularly successful historical track-record of (admittedly fallible) empirical science as a means of producing reliable knowledge of the physical world.

    [For starters, thoughtfully read Karl Popper’s OBJECTIVE KNOWLEDGE; or not (most do not).]

  14. Coyne asks, “if you favor the science world, do we really need a philosophical justification?” This is embarrassingly clueless. Deciding between a world in which science and technology empower us over natural forces and a world in which religious institutions are empowered as most people are enslaved by religious dogma is a normative, ethical, and thus ultimately philosophical choice, not a scientific one. Coyne’s question is philosophical, and reasons given to prefer reliability as a criterion of truth and science’s practical advantages over religion are likewise nonscientific and—-when expressed as rigorously as possible–philosophical.

    Positivists love to stomp their feet and shout, “Just look at science: it works and religion doesn’t. End of story!” But both science and religion “work” in that each contributes to a real process. Science gives us technology which elevates many people’s standard of living, but also endangers all life on Earth, threatens the ecosystem, kills off most other species, and so on. Religion creates a moralistic and rationally primitive society that’s communal and peaceful or hierarchical, tribal, and aggressive. Again, the choice between these two worlds is determined by our values, character, and other normative factors.

    Coyne merely begs the question in favour of philosophical pragmatism when he says that between science and religion, only science works because science is far more reliable in letting us predict the future and thereby gain practical advantage. Machiavellian survival at all costs, with no thought to morality and no spiritual perspective is merely one normative choice among many, and ethics is the philosophical field of rationally weighing such practical options. When you denigrate philosophy at the same as you presuppose an ethical stance (instrumentalism, pragmatism, Enlightenment humanism, or the empowerment of humans over nature), you evince the classic, stereotypical positivism, that is, the absurdity of philosophical antiphilosophy, which positivists love to say doesn’t exist.

    Sure, Coyne is quick to qualify his remarks by saying he respects philosophy as a whole and especially ethics, but he’s actually protesting too much, defending himself before he’s accused in a most revealing way. If he respects ethics, why does he pretend that philosophy and specifically ethics isn’t needed to decide between the social consequences of humanistic pragmatism and of moralistic religion? Surely he doesn’t think science justifies any such choice or somehow proves that we should be pragmatists rather than conservative religious dogmatists. The preference for pragmatism over spiritualism as your first principle becomes philosophical as soon as you become aware that there are alternative norms, which prompts you to think of reasons to justify the pragmatic lifestyle.

    1. The debate issue in question is one of epistemology, of scientific/empirical-based epistemology vs. nonscientific/nonempirical based epistemology. The debate issue in question is NOT whether philosophy has much, or some, or little, or no relevance and input into human wisdom, life choices or the consequences of life choices.

      If the relative demonstrable efficacies of competing epistemologies are not of useful value in ranking those competing epistemologies, then by what means can we sort physical or philosophical wheat from chaff? On what grounds would you (or ANYone) argue that a purely philosophical conclusion is superior to or should trump (or is in any sense at least equally meritorious with) an opposing but empirically intersubjectively demonstrable conclusion?

      I think Coyne “protests” claimed merits of philosophical conclusions that counter or contradict (forever tentative in some non-zero degree or another) scientific/empirical conclusions with precisely warranted balance and focus.

      1. Sorry, Frank Lovell, I didn’t see any sign that helpfully tells us all what exactly are the issues at stake. Nor did I hear any moderator declare that the sentence I quoted from Coyne’s post, which summarizes much of the rest of that post, is somehow irrelevant or immune from criticism. I assumed it’s up to us as grown-ups to interpret what we read and to respond to it as best we can, bringing our own assumptions to bear. If you can’t see the relevance of my criticism of Coyne’s post, are you sure the fault lies with my criticism?

        Regarding the rest of what you say, I’m afraid philosophy wins out over science by default on this specific issue, since between philosophy and science, philosophy is the *only* rational way of coming to a decision on *any* normative, value-laden, prescriptive issue. The deeper issue, raised by my criticism of Coyne’s post, is whether the normative preference for a humanistic, pragmatic, science-centered society over a dogmatic, superstitious, moralistic, religious one can be rationally justified without philosophy. You’re free to think that there’s no rational justification at all one way or the other with respect to that normative matter. But I don’t see how you can maintain that science alone tells us what counts as social progress, without committing the naturalistic fallacy.

        1. Yes, I’m pretty sure it’s your fault. I’m not sure why you keep bringing up philosophy’s usefulness in determining whether or not science makes the world a better place. No one else did. It’s a completely separate and, dare I say, unrelated, topic.

        2. Sorry, John Hue, but we seem NOT to be in the same conversation! I did not even express (let alone try to “maintain”) that “science alone tells us what counts as social progress” (who the heck thinks THAT???). But I can tell from what you last said and the way you said it that in your infallible mind our failure to communicate me falls solely on me, so I acquiesce (without yielding) and apologize for having wasted some of your time (I regret wasting some of mine as well).

    2. Religion creates a moralistic and rationally primitive society that’s communal and peaceful or hierarchical, tribal, and aggressive. Again, the choice between these two worlds is determined by our values, character, and other normative factors.

      You really think religion does that ?

      1. Sounds like putting the cart before the horse, doesn’t it? I think everything he says religion creates was there before religion, and then that society finally created religion as a convenient system for encoding certain known values. And this was done at a time of relative ignorance.

        1. When you talk about religion as a system of codifying traditions, you’re referring to what I called the hierarchical form of religion. This is merely a definitional matter, but as I understand the anthropology of religion, there’s a distinction between that more sophisticated form and the earlier, more shamanic and often communal form (the Essenic Jews, Native Americans, etc).

          1. But notice your earlier phrase “Religion creates”. I don’t see how you can justify that as having any truth or meaning. I think perhaps “religion is an expression of” might be more accurate. The existence of religion is contingent upon the evolution of a complex set of human capabilities. And there is no reason why that contingent system can’t have been one of temporary utility prior to the descendants of those humans developing other newer capabilities.

            1. As far as I can tell, you’re raising another quibble here, Jeff Johnson. I also said in this Comments forum that religion “contributes” to the creation of a different sort of society than does science. Note also that by saying this, I’m just following up on the end of Coyne’s post, where he poses his question to philosophers about which sort of society they’d prefer. That is, I’m arguing by way of reductio ad absurdum: I’m deriving a contradiction from Coyne’s own statements. If you don’t think religion has any independent impact on society, take it up with Coyne’s thought experiment.

              Of course, science’s own impact on society likewise is “contingent upon the evolution of a complex set of human capabilities,” but this is neither here nor there. The point of Coyne’s thought experiment is surely the modest one that science is socially better than religion. And my criticism is that, to the extent there’s any rational justification of such a normative judgment, about what counts as social progress, that justification is philosophical, not scientific. Hence, Coyne’s belittling of philosophy ironically reveals philosophy’s usefulness. Do you see my point now?

    3. I don’t appreciate “embarrassingly clueless” and I don’t appreciate “Coyne.” This is my website, and you could be a little more polite about it, especially when you are trying to dominate a thread.

      1. My apologies. Do you prefer for your readers to call you by your full name? Or Mr Coyne? Honestly, I meant no disrespect by calling you by your last name, since we’re obviously not on a first name basis.

        As for “embarrassingly clueless,” I’m afraid I stand by that judgment of the statement of yours which I quoted. I don’t see what dominating a thread has to do with anything. Is it too much to ask you to stop presupposing philosophical (ethical, normative) assumptions about social progress in your very act of belittling philosophy (“turf defense,” etc)?

        1. Sorry, this website is like my house, and I expect a certain level of decorum from my guests. Rude language towards me and other commenters violates that decorum. Like a puppy with persistent incontinence, you have wetted my carpet, and therefore will have to sit outside for a while.

    4. Deciding between a world in which science and technology empower us over natural forces and a world in which religious institutions are empowered as most people are enslaved by religious dogma is a normative, ethical, and thus ultimately philosophical choice, not a scientific one.

      Specifically, an engineering choice, rather than purely a scientific one.

  15. This just in, a results driven system produces results!

    There may not be such a thing as intellectually honest complete certainty, but we need not reject everything that does not have complete certainty. In fact we cannot, unless we are to reject everything.

  16. Quoth whyevolutionistrue:

    Philosophers like Lynch tear out their hair in frustration because they can’t justify, a priori, why to use science rather than religion.

    This is because too many Philosophers these days (IE: since Gödel) seemingly are no longer able to follow the necessary math. I’ve ranted on this before….

    Science as an algorithm can be derived from mathematics. (Though yes, science as a body of knowledge can’t; what body of knowledge results depends on what experiential data the universe you have produces. A universe with an inverse-cube law of gravitation would give rather different data than one with an inverse-square law.) While there are alternative mathematical schema to choose between (ZF versus vNGB versus…), the standard way to introduce one is to show that it can model and be modeled by ZF (the current gold standard), which does not obviously map to a trivial semantic lattice where FALSE and TRUE are just two names for the only one element. Having done that, it’s simply a “pick one” to start with, like deciding whether to discuss philosophy in Chinese or English.

    If Lynch wants to present a significant challenge to the underlying methods of mathematics, he has to present an alternative that is NOT contained in the scope of mathematics… which (courtesy Gödel, Turing, and Chomsky) pretty much rules out anything expressible in language. Similarly, additional principles (such as his example of “This particular book always tells the infallible truth.”) people want to introduce principle P to the starting epistemic set A have to be willing to let P be weighed against ¬P if set A allows deriving a means of weighing.

    Of course, a more practical difficulty is not that Philosopher’s can’t follow the math, but the average Joe can’t either. (A further lesser practical difficulty is that even many scientists would have trouble — and almost all would have better things to do with their time. For foundations of an epistemology that are solid, worrying about that solidity is like worrying about whether the planet you are standing on is about to spontaneously explode when it isn’t.)

    The odd part is that it would appear from his credentials that Sokal probably CAN follow the math. I suspect this means he just hasn’t encountered it, or hasn’t seen the point to doing so.

  17. To summarise what (IMHO) a lot of us are saying:

    If philosophy can’t find a justification for science then that is a problem for philosophy, not a problem for science.

    This applies to both the usual usages of “problem”.

  18. I fear that Jerry is wrong if he thinks that our scientific civilization can persevere and prosper with a muddled public understanding of scientific epistemology. Carl Sagan tried (I once heard him say it needed more attention, but added that “my plate is full”). Jacob Bronowski did the best job, I believe, in clarifying it for the public in The Ascent of Man, especially in “Knowledge or Certainty.” (If you haven’t seen it, you should; it is on YouTube).

    Science is approximate knowledge, with justification that is not circular, but is a network of cross-references. We bring all of our senses and instruments to bear on a question, and all we can hope to find is a sketch, an abstract of reality that captures some of its truth, but never all of it. And that is the strength of science, not a weakness. I don’t see that either Lynch or Sokal understand that.

    1. Well, our scientific civilization IS persevering and prospering, and in fact is healthier than ever, despite your assertions that that can’t be the case if the public misunderstands scientific epistemology.

      1. I disagree that it is healthier now than ever. Maybe it’s just my own personal experience, but American attitudes toward science were, in my opinion, much healthier back in the 1950’s and 1960’s, when my own attitudes as a teenager were influenced by the efforts of both American industry (General Motors Caravan of Progress, the Bell System Science TV series) and government (the Biological Sciences Curriculum Study) in promoting science education. Now, I have been following Republican politics as it disowns science as well as education, and I work for a government science agency that is being fiscally squeezed out of existence by Obama’s latest budget, and colleagues talk about getting out while the getting is good.

  19. In the end, Hawking is right: Science will win because it works.

    Not quite. Science will win as long as it works. When we find a methodology that works better, we’ll use that one instead. But since we’ll probably call that method ‘science’ too, the whole issue may be moot. 🙂

  20. Hmmm…I can’t seem to “reply” so my apologies for starting a new thread, as it were.

    Tom @7:

    To be justified in believing this cure will continue to work tomorrow, we need a justification of induction, and again, science cannot supply a non-circular one.

    Meh. So the method isn’t justified to our satisfaction. Until someone comes along with a better justification for some other method, I will live with philosophical disappointment.

    That is, after all, better than the alternative. Which would be to sit, paralyzed into inaction, until philosophy comes up with a fully justified method.

    So I guess my question to you, Tom, is – what do you expect people to do while we seek this justification? Sit on their hands? Spin a dial every morning to decide what method they will use to evaluate the world?

    In some ways this complaint about science is similar to creationist complaints that evolution has holes. As long as the alternatives are even holier (heh), it doesn’t provide any reason for me to change my behavior.

    Thus even the results of science are eternally in question without a justification of induction.

    Of course they are eternally in question! That’s why we call our conclusions tentative and say they are subject to revision.

    It seems to me that all current methods of knowledge-production share this trait. Science is superior to revelatory methodologies because it at least acknowledges this limitation. Look at the trouble the RCC has put itself in – both past and present – by pretending otherwise.

  21. I think it’s pretty ridiculous to say that epistemological justification of science is ridiculous. Of course from a pragmatic perspective it doesn’t matter: scientists will just keep doing what they do. But as people have pointed out, it also doesn’t matter from a pragmatic perspective to religious people, who also keep doing what they do. But what the hell is the whole point of this website and similar places that argue about the truth of evolution, or the big bang, or whatever, and the falsity of religion? The point is not to say something pragmatic about having a job that pays the bills. The point is to actually make a truth statement about the world, isn’t it? And if you’re attempting to make a truth statement, then you have to deal with epistemology.

    And seriously, saying that “science works because it works” is as irritating as a religious person saying “Jesus existed because the bible says so”.

    1. ‘saying that “science works because it works” is as irritating as a religious person saying “Jesus existed because the bible says so”.’

      That’s a straw man. Science works, period. As can be demonstrated in innumerable ways. Every component in this here laptop wot I am typing this on, for example, works because it’s based on technology which comes directly from scientific discoveries. Religion, demonstrably not – show me a Christian who has managed to post their message by appealing to Jesus.

      Only a philosopher could contrive to ignore the difference.

      1. In what sense is it a straw man? You are literally saying “I am justifying the methods of science (induction, repeatability, etc.) by using the methods of science (induction, repeatability, etc.)”. Is there any other way to translate what you said?

        Contrast that with a Christian appealing to Jesus’ existence: “I am justifying the existence of Jesus (in the bible) by using the existence of Jesus (in the bible).” It’s almost the exact same sentence.

        The problem is that both of these arguments are inherently circular! Please point out where the difference is if you don’t think that the scientific argument is circular. By saying “I posted this message because science works”, you are doing exactly what I said: using science to justify science. Is that so hard to see?

        This is not to say that science isn’t a better way of knowing than religion (I think that it is) but rather that you have to appeal to NON-SCIENTIFIC reasoning to justify scientific knowledge; otherwise, you are stuck with a pragmatic view of science (“it works, who cares if it’s true?”)

        So my final question to you is: do you care if scientific knowledge is true (or at least gets closer and closer to the truth)? If you do, I don’t see how people can so tacitly dismiss epistemological analysis of scientific methods but then so gleefully apply epistemic condemnation to religious people.

        1. Well, that phrasing of the argument is circular.

          Contrariwise, there’s an alternative foundationalist argument involving an axiomatic supposition, which allows deriving the validity mathematically. Having the such derivation allows explicitly dismissing the epistemological analysis as “trivial”.

        2. No I did NOT literally say what you quoted there. (Check a dictionary for ‘literally’). I am justifying the METHODS of science by using the RESULTS of science. Which work, regardless of whether some philosopher has given them permission to. And that is an entirely pragmatic view, I think.

          I’d also say that the fact it does work strongly suggests it’s ‘true’, for all practical purposes. (For example I could suggest Newton’s laws which are near enough ‘true’ for most technology and all everyday applications; they are of course a simplification of relativistic laws when we get to astronomical speeds and phemonena.)
          And, contra your quote (“it works, who cares if it’s true?” I do care whether facts are accurate (for example I hate maps that have errors in them).

          On the other hand, very little in religion can be shown to ‘work’ in a physical sense, their beliefs contradict each other, and many of their holy books are self-contradictory. ‘The Bible is infallible because it says so’ is the epitome of a circular argument.

    2. » Josh:
      And seriously, saying that “science works because it works” is as irritating as a religious person saying “Jesus existed because the bible says so”.

      Which, of course, is not what Jerry is saying. He is saying that he doesn’t care about an a priori justification when there is a track record of science’s success at doing what it is trying to do, which Jerry might, for all I know, call ‘finding out how nature really works’.

      1. Last time I checked, part of the methodology of science is that if an observation is repeated multiple, independent times, we get more confidence in it. e.g. you might not trust just one study that reports that smoking cigarettes causes cancer, but when everyone who’s ever looked finds the same thing, you get more confidence that smoking cigarettes causes cancer. Is that a misrepresentation?

        Given that “having a track record” is part of the methodology that science uses to attempt to get at the truth, it is therefore circular and tantamount to saying “science works because it works” to say that “science works because it has a track record”.

        Please don’t misunderstand me: I definitely think science works and is the best way to get at the truth. I’m notorious among my friends (in particular, the other grad students in my biology department) for fervently defending the view that science is the only way we know of that will actually get us closer to the truth. But the reason I have such confidence in that position is NOT because I make a post-hoc circular justification of science. It’s because I actually took the time to read and appreciate the kinds of stuff that goes on in philosophy of science.

        1. » Josh:
          Given that “having a track record” is part of the methodology that science uses

          I think you’re confused here. Reproducibility is indeed part of the methodology, but “having a track record” is not: it is an outcome of applying the methodology.

  22. Let’s see – science works, religion doesn’t, but we can’t say one is better than the other? Michael Lynch is a moron of the lowest degree.

    1. No, he’s just a typical philosopher. I find them more annoying than creationists. Creationists at least make statements that can be disproved. Philosophers will suffocate you in hairsplitting, logic-twisting obfuscations until any point there was to be made is lost in a cloud of verbiage. And then they pretend that science needs them in order to validate its discoveries.

  23. For anyone who doubts the trustworthiness or reliability of science due to there not being a philisophical justification for it, as Lynch does, I simply ask that they drive a car towards a concrete wall at a rate equivalent to the level of their doubt. My bet is that the laws of physics are reliable and trustworthy and will still work.

  24. It all comes down to searching for Truth. If you care about truth, about whether evolution is true for example, then you should be concerned upon learning the our methodology for finding out the truth cannot be justified in a non-circular manner. That doesn’t mean it’s false, and it doesn’t mean you should abandon your intuitions that it’s a good methodology and that religion isn’t – but if you care about holding true, well-justified, positions, you should be somewhat concerned.

    I think philosophy helps clears thought. In this case, it clarifies that our scientific methodology works IF there is sufficient order. [This is a key assumption in the mathematical proofs alluded to above, BTW.] One cannot justify it, but one can determine the conditions under which it will work.

    We have no means to test whether this order actually exists, since we can only do so scientifically, and that is circular… but we can say that IF it exists than our method is justified, and that it is a very WEAK kind of order – for example, if the sun will stop rising tomorrow that truth will STILL be discovered empirically; but if our memories and records will get jumbled tomorrow to indicate that it didn’t rise today then we could not discover that scientifically. Indeed, for the most part the requirements must be correct for rational thought itself, rather than empiricism, to be possible.

    And we can furthermore say that in contrast, the religious methodology works only given very specific models of reality that do not at all conform to how reality otherwise works (even within these same models, and certainly in scientific ones).

    So there are definite a priori advantages in terms of simplicity and rationality to accepting science over religion as a “way of knowing”, even though there is no and there cannot be any actual justification for believing its foundational epistemological principles.


  25. Seems like they’re just trying to bring science down to the level of religion without noting that science is based on observations and works when religion has neither of those.

    It’s like these philosophers are sitting in Plato’s cave arguing over shadows on the wall that are made from us scientists and engineers outside flying robots to Mars in spaceships.

  26. I’m not very well-versed in philosophy or epistemiology, so this might be a dumb question. Why do we (or maybe just the philisophers) need a ‘first principle’?

    By the looks of it, given that all our knowledge etc is based on prior assumptions and sensory input that we cannot know 100% to be true, what could ever be the basis of such a first principle? It sounds to me like a device invented by philosophers to keep themselves occupied.

    Science works like evolution: you don’t know a priori which of 1 million gnus on the African plains is the fittest. Neither do you know which genes make this so. None of the gnus care either. You just set them off, and after several generations, it will become clear in and of itself: the genes that were the best survivors are the genes that survived best. Same goes for scientific theories. The best predictive theories are those that do the best predicting.

    Circular? Hell yeah! But does it work? Of course it does!

    1. We don’t necessarily need a first-principle, but we do need to understand where we stand – what it is that we’re assuming and why. I think anyone that cares about truth and clear thinking should care about that.

      It all comes down to David Hume’s “Principle of Uniformity” – science works because the universe is (sufficiently) uniform, in key aspects. For example, no predictive theory will win-out, as in your scenario, if the universe was so complex that prediction would have been practically impossible.

      There is nothing irrational about making simple assumptions such as the principle of uniformity. The only thing is that you need to be clear about what it is that you assume, for your own sake – so that you’ll know where you stand.

  27. Philosophers like Lynch tear out their hair in frustration because they can’t justify, a priori, why to use science rather than religion.

    But the a posteriori “justifications” of science are ultimately inductive, and induction can’t be established as truth in one certain sense – the analytic, absolute sense.

    I think it’s okay to move on to tentative knowledge and say “these things I know, even though they might conceivably be wrong”, but Coyne doesn’t seem to think the difference between “might conceivably be wrong” and “could not logically be otherwise” is germane to whether something can be called true or not.

    1. Induction (in the scientific sense) can’t be established as analytic, absolute truth. However, it can analytically and absolutely be established via mathematics as a basis for maximally probabilistic truth.

  28. Being unable to adequately justify SCIENCE doesn’t undermine the significant trust it’s built up in the eye’s of other university departments (philosophy, ethics, history, etc).

    Whereas, being unable to justify BIBLICAL MIRACLES profoundly undermines any plea for Faith in church teachings and religious moralising.

Leave a Reply