Mistake of the Week: Explain Everything (Or Don’t Bother Explaining Anything)

In today’s post I was going to announce the winners of my Unparadox Contest. But then I noticed the Lake Wobegon unparadox: if the total winnings are zero, then no one’s winnings are below average and in that sense, everyone’s a winner!

So instead of that, I thought I’d contribute to the general shnoodification of humankind, by discussing the same thing every other science blogger’s discussing: Paul Davies’s New York Times op-ed.

Over the years I have often asked my physicist colleagues why the laws of physics are what they are. The answers vary from “that’s not a scientific question” to “nobody knows.” The favorite reply is, “There is no reason they are what they are — they just are.” The idea that the laws exist reasonlessly is deeply anti-rational. After all, the very essence of a scientific explanation of some phenomenon is that the world is ordered logically and that there are reasons things are as they are. If one traces these reasons all the way down to the bedrock of reality — the laws of physics — only to find that reason then deserts us, it makes a mockery of science.

Now, I know Paul Davies: he took me out to a nice dinner in Iceland, and even quoted me next to Ludwig Wittgenstein in the epigraph of one of his papers. And I know for a fact that his views are much more nuanced than you’d think, if the above passage was all you were going on. I can assure you that, if his claim that physics without metaphysics is “a mockery of science” reminds you of those hooded monks from Monty Python and the Holy Grail, pounding their heads with wooden boards in between mystic incantations, then you’ve read his piece too superficially and have failed to grasp its subtler message.

But even so, reading his op-ed made me wonder: when did we, as a civilization, have a similar conversation before? Then I remembered: the early 1600’s!

Galileo: Hey, I’ve discovered that Jupiter has moons! And that objects in free fall follow parabolic trajectories! And that…

Jesuit schoolmen: Ah, foolish one, but you have told us nothing about the underlying causes of motion, or what it is that imbues the lunar bodies with their lunarity. Of what use are your so-called “explanations” if they rest on a foundation that is itself unexplained? One can hardly build a pyramid on sand!

One imagines the schoolmen feeling sorry for the naïve Galileo, with his rampant scientism and countless unexamined presuppositions. In their minds, if Galileo hadn’t explained everything then he hadn’t really explained anything — and hence they themselves (who had explained nothing) were the wiser by far.

Four hundred years after the scientific revolution, most people still think like the Jesuit schoolmen did:

How does a toaster work?

By converting electrical energy into heat.

But what is electricity?

The movement of electrons through a wire.

But what are electrons?

Fundamental particles with spin 1/2, negative charge, mass of 10-27 grams…

But why do particles exist? Why does anything exist?

Well, those are excellent and profound questions, and you see…

Aha! Aha! So science doesn’t have all the answers! Ultimately, then, science is just another form of faith!

The schoolman glances at the intermediate steps — how a toaster works, what electricity is, what electrons are — and is not only profoundly unimpressed, but baffled and annoyed that anyone thinks he should be impressed. What are these so-called “answers” but irrelevant distractions from the Answer? What are they but the build-up to the punchline, stepping-stones on the road to the metaphysical abyss?

Science, in the schoolman’s mind, is just a massive con game: an attempt to distract people from the ultimate questions of essence by petty conjuring tricks like curing diseases or discovering the constituents of matter. Even pure math is part of the con: all Wiles did was reduce Fermat’s Last Theorem to some supposedly “self-evident” axioms. But why bother with such a reduction, if you can’t justify the axioms or the laws of logic themselves?

I frequently encounter the schoolmen even in my little corner of the world. People will ask: isn’t computational complexity theory a colossal failure, since all you ever do is prove “this problem is as hard as that other one,” or “this problem is hard relative to an oracle,” and never really prove anything is hard?

Let’s leave aside the factual misunderstandings — we can prove certain problems are hard, etc. etc. — and concentrate on the subtext, which is:

Don’t waste my time with the accumulated insights of the last half-century. If you haven’t solved the P versus NP problem — and you haven’t, right? — then aren’t you, ultimately, just as ignorant about computation as I am?

Of course, “does P=NP?” differs from “where do the laws of physics come from?” in that we know, at least philosophically, what an answer to the former question would look like. And yet, if complexity theorists ever do prove P≠NP, I’m guessing the schoolmen will switch immediately to saying that that was merely a technical result, and that it doesn’t even touch the real question, which is something else entirely.

The schoolmen’s philosophy leads directly to a fatalist methodology. What causes polio? If you say a virus, then you also have to explain what viruses are, and why they exist, and why the universe is such that viruses exist, and even why the universe itself exists. And if you can’t answer all of these questions, then your so-called “knowledge” rests on a foundation of arbitrariness and caprice, and you’re no better off than when you started. So you might as well say that polio is caused by demons.

Yet so long as the schoolmen are careful — and define the “ultimate explanation for X” in such a way that no actual discovery about X will ever count — their position is at least logically consistent. I’ll even confess to a certain sympathy with it. I’ll even speculate that most scientists have a smidgen of schoolman inside.

All I really object to, then, is the notion that tracing every question down to what Davies calls “the bedrock of reality” represents a new, exciting approach to gathering knowledge — one at the cutting edge of physics and cosmology. Say whatever else you want about the schoolman’s way, it’s neither new nor untried. For most of human history, it’s the only approach that was tried.

71 Responses to “Mistake of the Week: Explain Everything (Or Don’t Bother Explaining Anything)”

  1. Greg Kuperberg Says:

    On the same theme: All healthy children pass through the phase, which can linger on for a long time, of asking “Why?” in response to every answer.

  2. Tim Says:

    I don’t think it’s a coincidence that most scientists have a smidgen of the schoolmen in them; science is a direct descendant of the schoolmen’s approach. We’ve made so much progress in spite of falsifying so many “final” theories that we’ve come to recognize that final theories are not really necessary, so it doesn’t actually matter that we don’t (and possibly, probably, can’t) have them. But as you say, there are still a lot of people who haven’t quite gotten the memo. It certainly took me a while.

  3. Koray Says:

    For a moment the title made me think that this post was about the D-Wave Sudoku demo…

  4. Job (UG) Says:

    I think the difference is that, if we think of a tree of knowledge, then whereas in science the root of that tree is in human experience, in religion it’s somewhere else. So whereas in religion i need to have faith in that somebody got the tree rooted correctly, in science i only have to believe in that fact as much as i believe in my own existence.

  5. Jonathan Vos Post Says:

    Job (UG) : “in science the root of that tree is in human experience, in religion it’s somewhere else.”

    Galileo, thinking of how Giordano Bruno had been executed, was concerned that the tree of science must be refreshed from time to time with the blood of philosophers and heretics. The schoolmen had more than mere rational suasion in their armamentarium.

    I’ve paraphrased:

    “The tree of liberty must be refreshed from time to time with the blood of patriots and tyrants.” Thomas Jefferson, in letter to William S. Smith, 1787.

    I think that many of us have a metaphysical stance that differs from what Davies calls “the bedrock of reality.” That is, except for those (such as Einstein) who accept that the bedrock may be Spinoza’s “mind of God.”

    The default metaphysics of Science is to assume that Science has observation, experiment, and the whole process of the Scientific Method as bedrock, rather than either an axiomatic or theological bedrock.

  6. Job (UG) Says:

    Paul Davies says that science requires faith in that nature is “rational and intelligible”, but doesn’t that fact follow from observation through physical means, or am i getting caught up in some circular logic?

  7. cody Says:

    i had a friend that did this to me for a while, (the incessant ‘why’ and ‘then you must not understand’). i believe what he finally accepted was a quote i found somewhere on the internet a long time ago, supposedly by Bertrand Russell:

    …“Electricity”, Bertrand Russell says, “is not a thing, like St. Paul’s Cathedral; it is a way in which things behave. When we have told how things behave when they are electrified, and under what circumstances they are electrified, we have told all there is to tell.”

    looking for it now, i only found it here:

    i also subscribe to Feynman’s response to the question, ‘are you looking for the ultimate theory?’ in ‘the pleasure of finding things out’. he seemed to think that maybe the universe is ultimately explainable and maybe it is not; he also seems to be quite aware of the futility of rushing the matter, and maybe the utility of just investigating what you can, for your own pleasure.

  8. Tyler DiPietro Says:

    Setting aside the nuance and subtlety Paul Davies may exhibit elsewhere, his argument is spectacularly underwhelming. Substance-wise it differs little from the Cartesian dream game: “Can’t prove that your subjective experience isn’t a big dream? Aha! You really know nothing!” You can’t defeat the argument there is nothing that could disconfirm even in principle. Religious apologists who want their “faith” reinforced, no matter how badly, will find solace in it. I find it annoying and useless.

  9. Tyler DiPietro Says:

    “You can’t defeat the argument [because] there is nothing that could disconfirm [it] even in principle.”


  10. Jack in Danville Says:


    You start out by saying how easy it would be to superficially read Davies’ op-ed, then go on to deliver a criticism of the superficial reading! I would like to pick up a tangent to Davies’ thesis, in that he makes the distinction that asking “why?” today is not the same as asking “why?” in Galileo’s time. Why? All the progress in science up until today, and the outlook for further progress leaves the future of science, arguably, in a different light than ever before in history. The standard model, general relativity, and the great limiting theories of the 20th Century not only explain just about everything that any layman or engineer could care about, but also put some definable limits on what can be known.

    Perhaps in your lifetime mankind can reasonably expect to answer P=NP? and unite general relativity and the standard model, but will those achievements advance technology or only dot i’s and cross t’s? We can’t know until it happens, but I suspect the latter. Heat engines were already transforming human existence before Carnot and Kelvin; and Edison and Tesla figured out lots of neat things to do with electricity before quantum theory. Those scientific advances opened the door for even greater technological advance, but is there any significant technology today that doesn’t have a decent scientific explanation? We kind of know what the scientific playing field looks like now.

    Scientist themselves, not the schoolmen, are increasingly asking “why?” because that may be the future road of science, but it may be a road unlike the one traveled up until now: no falsifiable theories, and you have to be in an elite club of those with the aptitude and time to digest a lot of advanced math to have any understanding at all. …hmmm sort of like the schoolmen?

    Thanks for the reply on logs, btw.

  11. Scott Says:

    You start out by saying how easy it would be to superficially read Davies’ op-ed, then go on to deliver a criticism of the superficial reading!

    That’s correct! I do so because I’m pretty sure the superficial reading is the reading >98% of readers would’ve read (unless I’m badly misreading).

  12. Tyler DiPietro Says:

    “Perhaps in your lifetime mankind can reasonably expect to answer P=NP? and unite general relativity and the standard model, but will those achievements advance technology or only dot i’s and cross t’s?”

    I think it takes an astonishing myopia to conclude the latter. “P versus NP” may be a bit abstract, but a closely related problem, the existence of one-way functions, is strongly related to public-key cryptography.

    You can go back further than Edison and Tesla to see the sort of thing you’re talking about. We were able to develop archery and various forms of projectile weaponry before the development of classical mechanics, and we were able to figure out that certain forms of armor were better than others before we had a deep understanding of materials and chemistry. I don’t think it says anything particularly profound about science that crude engineering applications can preempt it. They are only becomming increasingly rare because the phenomena we are studying are becomming increasingly difficult to exploit. I think it’s premature to conclude from that, however, that significant technological accomplishments won’t be made possible once we understand them.

  13. Chris W. Says:

    Neither Galileo nor the Jesuit schoolmen had much inkling of the scope and power of the requirement of point-of-view invariance, when properly formulated.

    I think Davies is suggesting that, given the developments of the last hundred years, we are in a position to sensibly ask what “explains” the laws of physics, and the question might actually be fruitful, rather than a sterile scholastic exercise. The largely negative reactions to his essay are arguably due to neglect of this background for his musings.

  14. Jack in Danville Says:

    I’m pretty sure the superficial reading is the reading >98% of readers would’ve read

    But I would have thought readers of your blog would be disproportionately represented by the 2%

  15. David Moles Says:

    Scott, when you’re done with that, do you want to have a go at Krauss and Dent‘s suggestion that measuring dark energy could “reset the quantum mechanical clock” of the observable universe and make us all more likely to die in a false vacuum collapse?

  16. Coin Says:

    David Moles: Krauss actually claims that everyone’s misreading/misquoting him (he appears in the comment section at that link) and that wasn’t what he was trying to say at all. Apparently Krauss and Dent’s comments were trying to express something more along the lines that the things we measured about dark energy imply false vacuum decay occurs sooner, rather than trying to claim that the act of measuring causes false decay to occur sooner. The “reset the clock” comment Krauss gave to the New Scientist reporter appears to have been intended as metaphorical.

  17. John Sidles Says:

    “Progress in science depends on new techniques, new discoveries and new ideas, probably in that order of importance.” — Sydney Brenner

    It is characteristic of human cognition—of all primate cognition—that Brenner’s quote is memorable mainly because it asserted a rank-ordering.

    If Brenner had said “Progress in science depends mutually on new techniques, new discoveries and new ideas” his statement would have been more accurate, but less memorable.

  18. David Orban Says:

    It is an axiom of Science that the Universe can be explained, that it is rational. As any formal system, Science can and should periodically re-evaluate its foundations. Rationality does not have to be a dogmatic stance consequently, when we explore the laws of nature. We might find, as for me Chaitin’s results are indicating, that most of the Universe is actually not rational. That to me means that what we are left with is a greater degree of freedom, where we actually choose reason, instead of taking it for granted. It will be science’s gain, and the dogmatic position’s (including theists’) loss. The proof of this will be relatively simple, in the generative, and explanatory power of the two worldviews. My guess is that the rational stance is going to be the winning one 🙂

    Job UG

    “…doesn’t that fact follow from observation…”.

    It doesn’t actually, in my opinion. If not all of the Universe is rational, then, since we use reason to explore it, and live in its “rational subset”, what we observe, and progressively explain, is its rational part, leaving the rest untouched.

    Tyler DiPietro:

    “…They are only becomming increasingly rare..”.

    If we would plot a series of technological advances of the past, like the ones you quote, as archery, one after the other, and compare their sequence to those of today, what we’d see is a closer progression of more frequent engineering innovations. We are too accustomed to see this acceleration actually. That is why we are so outraged that physicists have not made fundamental steps ahead in the last 30 years in their quest for unification.

  19. Charlie C Says:

    It seems to me that we are currently wedged about three why’s up from the bottom and three meta’s down from the top. But the tools continue to improve and maybe we’ll achieve four before too long. The fun is in the journey! Enjoy it! It’s a wondrous world that we live in.

  20. harrison Says:

    I would have thought the answer Davies got would be more along the lines of “because if they were different, your repetitive line of questioning wouldn’t be so damn annoying!”

    OK, I’ll admit that, after some point, asking “Why?” over and over becomes silly and meaningless. On the other hand, that point isn’t fixed but fluid. Imagine the following conversation:

    Copernicus: Guess what! The Earth moves around the sun instead of the other way around!
    You: Cool! Why?
    Copernicus: *looks annoyed*
    Newton: Guess what! The Earth moves around the sun because of gravity!
    You: Cool! Why does gravity act on them?
    Newton: Because they both have mass!
    You: Cool! Why do they have mass?
    Newton: You’re an idiot.
    Dalton: The Earth has mass because subatomic particles have mass!
    You: Cool! Why?
    Dalton: *looks annoyed*
    Physicists possibly next year: Guess what! Subatomic particles have mass because of the Higgs boson!
    You: Cool! Why?

    My point is that if people didn’t ask “why” incessantly, science wouldn’t advance, or (more likely) would advance but less quickly. It’s annoying, yes, but I’ll contend that it’s necessary.

    Apologies for the long comment; just chalk it up to the laws of physics.

  21. Not even right Says:

    I heard the following questions from some students:

    Why does total internal reflection occur?

    If gravitational waves pass through an object, the object will change its shape and dimension. Why is it so?

    Why is the law of gravitation an inverse square law but not other laws like

    gravitational force ~ r^(-1) or r^(-3)?

    I don’t have good answers myself. But I’d think if questions are kept being asked down to the foundation of every thing, only the creator of this universe could know the answers.

    What do you think?

  22. Scott Says:

    Why is the law of gravitation an inverse square law but not other laws like

    gravitational force ~ r^(-1) or r^(-3)?

    That particular question actually has excellent answers! (Where by “answers,” I mean explanations in terms of other things that are arguably more basic.)

    The first answer is, because space is 3-dimensional! So if you imagine the gravitons spreading out in concentric spheres, then they’re necessarily going to dissipate like 1/(4πr2). (This idea becomes more formal in GR, where you can derive 1/r2 from the Einstein field equations, which themselves talk only about the bending of local patches of spacetime.)

    The second answer is more “anthropic”: one can show that, if the law were 1/rp for any p≠2, then planets wouldn’t have stable orbits. They’d either spiral into the sun or fly off into space. Draw your own conclusions. 🙂

  23. Scott Says:

    Incidentally, in case it wasn’t clear from the post: I’m a huge fan of asking “why” about everything, always have been. What I object to is the surprisingly-common idea that, if you’ve explained X in terms of Y but can’t yet explain Y in terms of anything deeper, then your explanation of X was somehow a sham. It’s not a sham: it explains X in terms of Y.

  24. David Moles Says:

    Coin @16: I see V2 of the paper is much more sensible than V1, but boy howdy, the way V1’s written, Krauss should have seen that coming.

  25. mir Says:

    There is a circle hidden in deep inside the number pi. At least in the novel Contact 🙂

  26. Scott Says:

    David Moles: I’ve long suspected that magazines like New Scientist trawl the arXiv for the stupidest thing any respectable physicist can be caught saying in a paper, which they can then distort further, strip from all surrounding context, and print in a banner headline. That suspicion has now been spectacularly confirmed.

    Krauss and Dent, of course, bear a large part of the responsibility for this fiasco: in the last paragraph of v1, they were begging to be misunderstood.

    But the fact remains that there are thousands of physicists in the world, and sometime, somewhere, one of them will say something that makes no sense. And when it happens, New Scientist and The Telegraph will be there to record it.

  27. RubeRad Says:

    Note how your fictional dialogue with Galileo shifted from “how does a toaster work?” to “why does anything exist?” Perhaps you intend “why” in the sense of “how do they come to exist?” (which science might be able to answer) rather than “what is their purpose for existence?” (which science can never pretend to answer), but I think it is important to understand the distinction that science is able only to address “how” questions, not “why” questions.

    I don’t see how Davies’ goal that “the laws should have an explanation from within the universe and not involve appealing to an external agency” could in principle be possible. How would such a beastie be immune from Davies’ own criticism of the explanatory weakness of multiverse theory?

    This process will require its own laws, or meta-laws. Where do they come from? The problem has simply been shifted up a level from the laws of the universe to the meta-laws of the multiverse

  28. Nagesh Adluru Says:

    You truly are a scientist who cares about limits in the most likely conceivable and profitable way:) I can hardly ever get tired of being your fan:)

  29. Scott Says:

    Nagesh, good to know you’re on my side.

  30. Nagesh Adluru Says:

    Yes I have always been from the moment I knew about your work.

  31. Cynthia Says:

    Scott, not to sound like I’m belittling you, but I do think you’re missing something big here: the big thing being that Davies’ op-ed piece was written on average for a simple-minded audience, not for a complex-minded one. Let me explain…

    Even if bio-bloggers can’t get beyond the surface of Davies’ message, I think it’s wrong for you to dismiss them outright as a bunch of simpletons. Be mindful, biologists (unlike physicists, especially complexity theorists) are still actively battling IDers who are still aggressively trying to gain a foothold in life-science curriculums across the US. And believe me, just because the ID movement lost the Dover Case doesn’t mean that they have thrown in the towel, declaring themselves the losers. In fact, from the looks of this pro-ID film, the IDers are just getting warmed up. Needless to say, “Expelled” reeks heavily of ID propaganda laced with a lot of cold hard cash:


    Maybe it’s true that biologists don’t have a clue about the complexities surrounding the origins of the Universe, but they sure as hell know about the complexities surrounding the politics of science–at least at the grassroots, that is… So when a mainstream scientist writes an op-ed piece conflating faith and science, most biologists (unlike most physicists) are acutely aware that this sort of rhetoric only adds an air of legitimacy to everything ID!

  32. Scott Says:

    Cynthia, that was exactly my point!! That’s why, after quickly explaining that the “surface message” that 98% of readers will take away can’t possibly be what Davies meant, I then spent the entire rest of the post responding to the surface message.

    I do think that Davies, who has plenty of experience in the public arena, should’ve known better than to write something that could be so easily misinterpreted by creationists and others hungry to misinterpret.

  33. Robert Wolfersteig Says:

    My interpretation of Dr. Davies’ editorial is that science has to understand that reality may be more and/or different than the one that science has access to. Science is also a cultural enterprise and that scientific bias cuts many ways.

  34. Cynthia Says:

    I apologize for misreading your message: misreading it the point of reading it backwards, in fact…:~(

    Apologies aside: I still think that if one’s message means one thing on the surface but something entirely different deeper down, then one should always make it clear somewhere (preferably in the beginning) that the message contains two opposing meanings–unless the message is either pure fiction or pure Onion!

    And maybe it’s alright to convey messages with opposing meanings in specialized settings, such as Shteti-Optimized, but it’s not alright to be doing such things in general settings, such as The New York Times.

  35. Chris W. Says:


    If you going to trot out that tired assertion—“science has to understand that reality may be more and/or different than the one that science has access to”—then you have to come to grips with fact that science has arguably done a better job of enlarging the scope of its access to, and grasp of, reality than any other alleged path to knowledge. It does this by challenging its own body of assertions, not just with questions, but with empirical tests—tests that can fail even in violation of the expectations of the people conducting them.

    Perhaps the saddest irony of the pervasive role of the products and professionalization of scientific investigation in modern society is that science is seen by many people mainly as a source of authority and a path to success, rather than a self-questioning enterprise. Einstein said about the recognition of his own work: “To punish me for my contempt for authority, fate made me an authority myself.”

  36. Vladimir Levin Says:

    In Reply to Cody’s comment about Feynman “i also subscribe to Feynman’s response to the question, ‘are you looking for the ultimate theory?’ in ‘the pleasure of finding things out’. he seemed to think that maybe the universe is ultimately explainable and maybe it is not; he also seems to be quite aware of the futility of rushing the matter, and maybe the utility of just investigating what you can, for your own pleasure.”

    I don’t know where in The Pleasure Of Finding Things Out Feynman says something like this, but there is a spot in Feynman’s Rainbow by Leonard Mlodinow where he says something similar. His comment is promted by a question about string theory’s explanatory power, and Feynman says something to the effect that as far as he is concerned, there may not be one single theory to explain everything. Nothing in nature tells him there *must* exist such a theory. I’m sorry, I can’t find my copy of the book, so I can’t quote a particular page.

  37. Robert Wolfersteig Says:


    I believe that you misinterpreted my comment, but that’s fine.

    I do agree with your statement about science as a “self-questioning enterprise.” However, many in the scientific community do believe they are the only source of authority and truth.

    Anyway, this isn’t the appropriate forum for a scrum.

  38. Coin Says:

    Coin @16: I see V2 of the paper is much more sensible than V1, but boy howdy, the way V1’s written, Krauss should have seen that coming.

    What does kind of boggle my mind is that the New Scientist actually called up Krauss for quotes and managed to complete that entire interview without either party realizing that they were each simultaneously misunderstanding every word the other one said. It must have been like one of those old European farce comedies.

  39. Tyler DiPietro Says:

    This discussion has veered off without me, but I’d just like to clarify something for David above: I wasn’t talking about engineering applications in general, only one’s that predate a thorough scientific understanding.

  40. John Sidles Says:

    Tyler DiPietro says: … “I don’t think it says anything particularly profound about science that crude engineering applications can preempt it.”

    … “Crude?” Ouch. 🙂

  41. Bad Says:

    I don’t think Davies has been misinterpreted: I think he was indeed, just very sloppy. What was misrepresented was what science is and does, and how.

  42. Cynthia Says:

    Scott, don’t know if you already know this, but Brad DeLong did a post on you! BTW, Brad is one of favorite econ-bloggers…:~)


  43. cody Says:

    Vladimir Levin: there is a very good chance i am wrong, as i do not know specifically which bits of Feynman i picked up where. though i do know there was some variation in the videos of him. i dont recall exactly, but i think there were two versions of it, and one version contained more material than the other. i think the shorter one might have been split into two episodes as well, though i dont recall. either way, in the instance referred to i think it is Richard himself who says “people ask me, ‘are you looking for the ultimate theory?’ and i say, ‘no, i am not…'”, i cannot quote exactly, but he then suggests the universe may be like an onion, with layer after layer and you never get the the bottom. ill try to find it and let you know.

  44. cody Says:

    okay, Vladimir Levin, its at about 45 min 20 sec in the 50 minute version of “the Pleasure of Finding Things Out” that i have on my computer.

  45. Job Says:

    Science is peer-to-peer, religion is server-based. 😀

  46. Vladimir Levin Says:

    Thanks Cody, I’ve made a note of it! Just for the record, I wasn’t suggesting you were wrong in your original comment. I just happened to recall something similar in the Feynman’s Rainbow book. I figured I’d mention it in case anyone wanted a reference that might be a bit easier to track down. In general everything I’ve read about/by Feynman suggests he was very comfortable with the idea that we are all very ignorant and there are lots of things we, as in humanity, simply don’t know. As Scott points out, just because there are lots of things we don’t know, that doesn’t mean we don’t know anything at all. As for faith, we all assume the things we remember about our lives are more or less accurate, that they really did happen. In the same way science and mathematics relies on certain basic assumptions. We assume that if a theory is reliably verified by experiment no matter how hard we try to prove it wrong, that at least for the time being we can consider it to be “right.” Maybe later on we’ll find a new theory that’s even more general, that subsumes the earlier one as in the case of relativity vs. newton’s theory of gravity, or we may even find out that the theory is fundamentally flawed as in the case of the ptolemaic earth-centric system of planets vs. the copernican sun-centered one.

    What the formal methods of science and math try to do is to clearly define what those assumptions are, so they’re out in the open and can at times be challenged – like Euclid’sfifth axiom. That’s all really. Of course we all have to have some kind of faith that we really do exist in a real world and that it’s not all a figment of our imagination – and at that, perhaps it all is: We can all recognize that possibility, but it usefulness is pretty much nil.

    I must say I am surprised by how gentle Scott was with this Davie’s fellow, all things considered. Ok, end of rant. 🙂

  47. Sam Says:

    You have to choose your battles, and I think focusing on this one is a big mistake.

    1. Philosophically, the author is on strong ground. Maybe it isn’t phrased very well, but I certainly can’t disagree with his broad philosophy.
    2. He isn’t so much attacking science as he is trying to relate the scientific method to the faith that most people are more familiar with. This is positive outreach.
    3. If you insist on seeing the article as on attack on science, then the only part of science it could be seen as attacking is arguments for physical laws based on the anthropic principle. And such arguments fully deserve to be attacked, as they are not scientific. They are not testable and they do not even give a more concise set of equations for what we already know.

  48. Erick Chastain Says:

    I agree that conflating the issues surrounding faith and science with his real point was a mistake.

    His real point though, if I understand it correctly, is a profound problem in naive descriptive models of phenomena. A great deal of physics is susceptible to this, as Davies points out.
    This issue actually comes up a lot in biophysics and neuroscience as well. One can either say that the Hodgkin-Huxley Equations exist as they are because they fit the data well enough, or one can search for a better explanation. The better explanation need not be magical or religious or philosophical. In fact, it could be that as Adrienne Fairhall believes, the hodgkin-huxley equations have dynamics that implement a constrained range of functions adaptively tuned to natural statistics in our environment. Now which explanation would you prefer, as a scientist?

    That was Mr. Davies’ point. I would add that my personal opinion is that “why” behind descriptive models that work well, at least in biological systems, seems to be in terms of principles of optimality. Normative theories.

  49. Vladimir Levin Says:

    I can’t resist one more comment. The most absurd part of Davies article is here:

    “Clearly, then, both religion and science are founded on faith — namely, on belief in the existence of something outside the universe, like an unexplained God or an unexplained set of physical laws, maybe even a huge ensemble of unseen universes, too…”

    If you find that putting money in a bank account means you can reliably retrieve it later on – it doesn’t disappear, no one steals it – then you develop an *expectation* that it’s ok to put your money in the bank. This expectation may be further reinforced by some knowledge of how the banking system works as well as by recognizing that while it may not be foolproof, it’s better than putting the money under your mattress or in a big pile on the street in front of your house. That’s not “faith.” Leaving the money hanging on a clothesline in your backyard, because you believe that people are “good” and wouldn’t steal it, that’s “faith.”

    We may be able some day to explain the underpinnings of physical law to our satisfaction, and we may not. No one really expected the elements to fall into a neat and tidy periodic table that explains so much about them. For a long time, scientists believed that elements were what they were and we’d never have a single explanatory model. Ultimately such a model was found, by Mendeleev and others, but it wasn’t based on faith, it was based on careful observation over decades followed by careful confirmation that also took years, even decades.

  50. Job Says:

    It may be unescapable that science requires some faith, if i put money in a bank account then i’m not placing as much faith in people as i would if i were storing it on the clothesline, but i’m placing some.

    Science does rely on a peer-review system which isn’t guaranteed to work, and though it’s possible in theory to verify the claims, i’m not realisticaly able to confirm every observation and argument that an accepted theory relies on. On the other hand, the probability of error or corruption in the system decreases with the number of peers. So even if we can’t evade the faith requirement completely, we’re able to approximate faithlessness arbitrarily through the addition of more peers.

  51. Tapan Parikh Says:

    Interesting thread. I have my own thoughts, but I will defer to Steven Weinberg – from the conclusion to “The First Three Minutes”:

    However all these problems may be resolved, and whichever
    cosmological model proves correct, there is not much of comfort in any of this. It is almost irresistible for humans to believe that we have some special relation to the universe, that human life is not just a more-or-less farcical outcome of a chain of accidents reaching back to the first three minutes, but that we were somehow built in from the beginning… It is very hard to realize that this all is just a tiny part of an overwhelmingly hostile universe. It is even harder to realize that this present universe has evolved from an unspeakably unfamiliar early condition, and faces a future extinction of endless cold or intolerable heat. The more the universe seems comprehensible, the more it also seems pointless.

    But if there is no solace in the fruits of our research, there is
    at least some consolation in the research itself. Men and women are not content to comfort themselves with tales of gods and giants, or to confine their thoughts to the daily affairs of life; they also build telescopes and satellites and accelerators, and sit at their desks for endless hours working out the meaning of the data they gather. The effort to understand the universe is one of the very few things that lifts human life above the level of farce, and gives it some of the grace of tragedy.

    A humble scientist with a sense of humor. Now thats something really special. Now, if we could only develop a sense of purpose…

  52. Job Says:

    Incidentally, our only hope for a “faithless science” might be computers, and only if P = NP at that, or no? If P != NP then are we able to efficiently automate the scientific process with a 100% success rate?

  53. Tapan Parikh Says:

    OK, since I alluded to my own thoughts, I might as well give them, un-scientific as they are.

    I wish faith and science weren’t always juxtaposed like this. Both science and religion contain elements of belief and logic, mixed in different ways. Drawing fundamental distinctions between the two seems (to me) to be contrived.

    Both seek fundamental understanding of the universe and humanity’s role in it – albeit using different methods. The experimental method is a vast improvement on the intuition and imposition method that preceded it – but that doesn’t mean that science doesn’t contain elements of belief also.

    However, what religion provides to the masses that science hasn’t so far (with some notable exceptions) is a real sense of purpose, that can speak to humanity’s deepest motivations – the desire to live, love and procreate.

    Thats why I’m an engineer at heart.

  54. Job Says:

    You know how Apple moved into the Intel architecture and will probably become more popular than ever because of the move?

    Religion should do the same. Instead of religion trying to appeal to the hardcore PowerPC population it should recognize that its ideals are hardware independent. It should just say “hey, the bible was written so that a very unscientific population could understand, and it should not be taken literally”. This way it will no longer conflict with the more popular Intel architecture for no good reason, and it can run on both scientists and republicans.

  55. roland Says:

    speaking of provably hard problems –

    One might call a specific question hard, if no circuit of reasonable size solves it. To proof something like that, one would need a non asymptotic bound. If MA/1 includes problems that need circuits of high degree polnomial size, this surely includes hard problems, but one never knows for what size the asymptotic behaviour really sets in.

    Or is that wrong?

  56. Len Ornstein Says:

    On Peter Woit’s blog, he got upset when the discussion of Paul Davies letter to the Times got into generalities about science and faith.

    However, I believe that it’s probably only by looking at the roots of the ‘problem’, that it can be ‘completely’ understood:

    Language (and its special variations; logic and mathematics) can only work when their users commit to the discipline of ‘trust’ or ‘belief’ in a set of ‘axioms’, rules and definitions. Such commitment is necessary to make it possible to model our mental ‘images’ in a form that can be communicated. And that commitment is basically indistinguishable from an act of faith. In this sense, all inductive reasoning can be viewed as faith-based. Until Gödel, the truth associated with careful deduction was believed to be at least tautologically absolute…but now we know better.

    Hume taught that the belief in extrapolation (or interpolation), from observation of the parts of a class or process, to the whole, could not be logically justified; that inductive reasoning also necessarily depends upon ‘faith’ of a sort.

    So the issue of faith, belief or truth…as distinguishing science from religion, is really a straw man, despite the literature of the Positivists. Paul Davies was was making a pointless point.

    Rather it’s the degree of commitment to models, that provides the distinction. Differences among disciplines, with respect to the degree of commitment, can be enormous. Most theists demand absolute commitment…as do some political ideologies. Science, on the other hand, calls for skepticism, often quantitatively related to ‘confidence-interval-like’ computations (e.g., through processing measured variations in repeated observations, and using models like the Central Limits Theorem) .

    For science, there are no absolute truths; only degrees of tentative ‘truth’.

    Even the most carefully constructed models remain indistinguishable from good science fiction…until they’re confirmed or refuted by experiment…or other pertinent observation. String theories, multiverses, the anthropic landscape and even the Higgs boson fall in this category. And it’s not unreasonable (even though, in rare cases, it may ultimately turn out to be wrong) for confidence in a model to diminish, the longer it takes to establish confirmation…or even confirmability, (again, a sort of dependence on Central-Limits-like theories).

    Unfortunately some scientist consider themselves priestly prophets, whose models are absolutely true. It’s often the responsibility of other scientist to inform the public about the uncertainties…or sometimes, apparent unconfirmability…of such models.

  57. Bad Says:

    How is Davies on strong ground philosophically? He basically mangles the philosophy of science, to begin with, and his philosophy of faith isn’t much more useful.

    His picture of science is just far more grandiose than it actual is or should strive to be.

    And the claim that scientists are just giving up and not looking because “that’s the way it is” is posturing nonsense. No one is giving up or stopping asking why. What they are telling Davies, and he is apparently not getting, is that we should not necessarily expect that a) we cannot expect that answers will be practically available or b) that philosophically there must be a reason for everything.

    What Davies is doing, and what they are responding to, is trying to DRAW CONCLUSIONS about the fact that things like the constants are the way they are. What the people who are responding to him are saying is that you can’t draw those conclusions, and you certainly can’t call it science when you do. Science has methodological limits. Davies seems to think these limits are ideological. It’s he who is misrepresenting people who he characterizes.

  58. Scott Says:

    Roland: Given pretty much any asymptotic circuit lower bound, one can (with a lot of grunt work!) extract a lower bound on the circuit size needed to solve some concrete problem on a fixed input size n.

    A famous example is due to Meyer and Stockmeyer. They showed that, to decide the truth or falsehood of a sentence of length 610 in a formal system called “WS1S”, you need a circuit with at least 10125 gates.

  59. Hanamaru Says:

    >For most of human history, it’s the only approach that was tried.

    Faith. Then faith and logic (in this order). Then only logic. Now it seems that logic and faith (in this order) is the (fashionable) way to go.

    I can certainly see some advantages on using faith whenever in need. What about this line of reasoning: Davies conclusions are wrong AND “Thus you will know them by their fruits” (Matt 7:15-23) AND Davies article is the offspring of string theory => string theory is wrong?

    Do you believe in oracles? Then P=NP!!! Oh, wait…

  60. roland Says:

    thank you, scott.
    I’m still 30 years behind in complexity theory.

  61. Vladimir Levin Says:

    Scott, this one’s for you! 🙂


  62. Peter Sheldrick Says:

    People will ask: isn’t computational complexity theory a colossal failure, since all you ever do is prove “this problem is as hard as that other one,” or “this problem is hard relative to an oracle,” and never really prove anything is hard?

    This sounds like something a four-year old would say along the lines of:

    I want hot coco now!!!

    Sure, when we prove that a problem is NP-complete, in a philosophical sense we never actually prove that it is hard. But it in practical sense by proving NP-completeness we do prove that the problem is hard, where the word “hard” is much more precisely defined. You could almost say: It is hard by observation.

    Compulsory quote by Francis Bacon (1561–1626):

    Men have sought to make a world from their own conception and to draw from their own minds all the material which they employed, but if, instead of doing so, they had consulted experience and observation, they would have the facts and not opinions to reason about, and might have ultimately arrived at the knowledge of the laws which govern the material world.

    (quote according to Wikipedia)

  63. John Sidles Says:

    David Orban says: “…we are outraged that physicists have not made fundamental steps ahead in the last 30 years in their quest for unification.”

    That is too simple for me! First, I think physicists *have* made fundamental steps ahead … and it’s not the fault of physicists that the world is inherently complex.

    Supposing, as seems likely, that we live in a world in which the fundamental laws are somehow contingent or evolved. If this doesn’t bother the biologists, why should it bother the physicists?

    As for the second great deliverable of physics—the first deliverable being fundamental insights, the second deliverable being practical technologies—IMHO the mathematicians, physicists, and chemists have collectively been delivering exponential improvements in quantum simulation capability for more than 40 years, and there seems to be no fundamental reason that this “Moore’s Law of quantum simulation” is going to end anytime soon.

    In short, we’re living in a Golden Era of scientific discovery and enterprise.

    Geez … why are so many folks grumpy about it?

  64. Abel Says:

    “The job of the scientist, we were told, is to discover the laws and apply them, not inquire into their provenance. The laws were treated as “given” — imprinted on the universe like a maker’s mark at the moment of cosmic birth — and fixed forevermore.”

    I’m surprised that Davies has not read Kuhn.

  65. Peter Shor Says:

    Scott says:

    People will ask: isn’t computational complexity theory a colossal failure, since all you ever do is prove “this problem is as hard as that other one,” or “this problem is hard relative to an oracle,” and never really prove anything is hard?

    Interestingly enough, I have been told by a Russian that some Soviet bureaucrats’ reactions to completeness results was similarly negative. Paraphrasing:
    “You showed that we aren’t able to do something. This isn’t progress; progress can only be made by showing that we are able to do something.”

    Unfortunately, I don’t remember whether this was in connection with Levin’s NP-completeness results, or Mnev’s universality results, or both.

  66. Koray Says:

    Peter Shor:

    I read somewhere about a research where they played “what number between 1 and 100 am I thinking of” with children. When the kids asked if the number was greater than 50, they would cheer if the answer was yes and moan if the answer was no. It appears to me that people just don’t like hearing negatives, even if the utility value is equal.

  67. Prempeh Says:

    I’m an average human being. All around me there is a profusion of ideas and prescriptions for life: various philosophies, folk wisdom, scientific theories … I have a seemingly simple problem: I need to relate better with people; this appears more important to my enjoyment of life than any other factor. I rely on bits of insight from friends, motivational speakers, spirituality (not the same as the hypocrisy that is most religion), and other sources for help. I’m really no happier because of knowing that a phenomenon called quantum entanglement exist. Now, you say, this phenomenon has the potential to enable super-powerful computing, teleportation, … I say, until science helps me with a comprehensive, provable, repeatable methodology for using it’s results to make me (and everyone who wants to be) happy, I really do not see it as significantly more helpful than faith.

    NB: Any chance that a unification theory could help the poor stave off devastating climate change caused in part by the profligacy of the west? End the brutality of war? Stop child sexual exploitation? Remove corruption, greed, racism, …

    This is not a rhetorical question

  68. Serdar Says:

    “Any chance that a unification theory could help the poor stave off devastating climate change caused in part by the profligacy of the west? End the brutality of war? Stop child sexual exploitation? Remove corruption, greed, racism, …”

    No, because that’s not what it’s for. If I went to my local car service station and groused at them for not offering billiards, I would probably not get a lot of sympathy either.

  69. Herb Ivorous Says:

    The basic characteristics of life do not change, regardless of the development of science, mathematics, and technology. Read Buddhism and Schopenhauer to learn that life is basically suffering. Super-powerful computing and teleportation will not change the fundamental characteristics of life. The suffering can be temporarily alleviated, however, by the distraction of mental or physical occupation. Hence the popularity of the Internet, cinema, novels, television shows, mathematical puzzles, and sports.

  70. David Feldman Says:

    A skeptic may well complain that computer scientists
    never really prove anything is hard without any
    trace of a schoolman’s foundational infinite regress.
    Mathematicians and theoretical computer scientists
    ruthlessly discard superseded results. Tourists
    usually want to see permanent monuments, so for
    most, marking something “under construction” or
    “slated for demolition” means check back on your next
    visit. An altogether different value system holds for
    players, those who might hope to achieve the next
    monument. When an obviously provisional result
    (e.g. Vinogradov’s three primes theorem) resists
    significant improvement for a long time, the achievement gradually acquires the air of monumentality. But in
    the culture of mathematics, Vinogradov’s stock will
    crash the day someone proves Goldbach, and thus
    his result has less savor for the layman than a result
    which tells us what the experts had previously
    construed as “what we really want to know.”

  71. Fred Says:

    This seems a lot like a variant of Zeno’s paradox. For each answer to “why” set another “why” between what is known and knowing everything.