A Fight for the Soul of Science

https://www.quantamagazine.org/20151216-physicists-and-philosophers-debate-the-boundaries-of-science/

Physicists typically think they “need philosophers and historians of science like birds need ornithologists,” the Nobel laureate David Gross told a roomful of philosophers, historians and physicists last week in Munich, Germany, paraphrasing Richard Feynman.

The crisis, as Ellis and Silk tell it, is the wildly speculative nature of modern physics theories, which they say reflects a dangerous departure from the scientific method. Many of today’s theorists — chief among them the proponents of string theory and the multiverse hypothesis — appear convinced of their ideas on the grounds that they are beautiful or logically compelling, despite the impossibility of testing them. Ellis and Silk accused these theorists of “moving the goalposts” of science and blurring the line between physics and pseudoscience. “The imprimatur of science should be awarded only to a theory that is testable,” Ellis and Silk wrote, thereby disqualifying most of the leading theories of the past 40 years. “Only then can we defend science from attack.”

30 thoughts on “A Fight for the Soul of Science

  1. Excellent choice of topic, Petrushka. I’ve thought about this quite a bit, and regret that I can’t join in. I’ll remark on the fact that speculative “physics” — speculative philosophy done by physicists — is exactly the precedent that the ID movement needs, to argue that its crypto-creationism is science. Yet the ID movement slams speculative “physics” at every turn, for pretty much the same reasons that we slam crypto-creationism. It’s not only irony that abounds, but also hypocrisy.

  2. Good thread.
    Once again it shows that philosophy and history of science is essential for what is science. That feynman guy was plain wrong and shows he only knew his limited field of subject.
    These speculative pgysic stuff in a subject very testable demonstrates why the error of evolution as being a scientific theory has continued or existed.
    Evolutionism is just in the equation of these questioned physic speculations presented as good science.
    AHA says the creationist.
    If science is anything then it must be more then a good hunch. Even if true.
    Science must be about backing up a hypothesis/hunch before its claimed to be a scientific conclusion.
    Evolutionism never was a scientific conclusion, not just because its wrong, but because it was not based on biology. It wasn’t testable on biology but instead geology, comparative anatomy and genetics, and biogeography etc were its evidences. Even if true evolution would still not be a bio sci theory and so not a sci theory.
    There is something wrong about what people think science is.
    EVEN in hard cold physics it comes up. how much more in historic sciences?
    String theory might be a creationist best friend. A line of reasoning.

  3. This topic keeps coming up, so I guess I’ll keep recycling this comment:

    KN,

    String theory and multiverses make me grumpy. If the constraint isn’t testability, then what’s the difference between science and mythology?

    We’ve had this conversation before. A June comment:

    KN, to Neil:

    I mostly agree with those assessments. Superstring theory looks like a lot of fun mathematics with no experimental support, so I just shrug my shoulders at it. Likewise, I have no idea what could confirm or disconfirm the multiverse; it makes for good science fiction but I don’t see it as a scientific hypothesis.

    I think everyone agrees that falsifiability is desirable, but let me sketch out a scenario in which string theory, with its concomitant multiverse, could come to be regarded as our best theory even if the multiverse remains unfalsifiable.

    String theory is largely motivated by the desire to unite quantum mechanics and general relativity by explaining gravity in quantum terms. Let me stress that IANAP, but my understanding is that string theory is currently the most promising approach toward that unification. String theory happens to imply a multiverse, but this is an implication, not an assumption (which puts to rest the common ID claim that the multiverse is a transparent attempt to obviate the Designer).
    If at some point string theory succeeds in uniting QM and GR, matches our experimental results to date, and we can find no other theory that does so without implying the existence of a multiverse, then I would say that we are justified in accepting the existence of the multiverse despite being unable to test it directly. As with all scientific conclusions, our acceptance would be provisional.

    Of course the hope is that we will come up with a way to test the multiverse directly. Here’s one proposal, although I don’t know if the multiverse being tested for here is the same kind as the one predicted by string theory:

    How Do You Test The Multiverse? With Bubbles

  4. My own take is that physical theories are mathematical models, and the most parsimonious models are the best.

    Of course we still use Newton when we are in a hurry and the velocities do not require corrections for relativity.

    Not unlike the way video game coders might use lookup tables, because they are faster than calculations.

  5. petrushka,

    My own take is that physical theories are mathematical models, and the most parsimonious models are the best.

    All else being equal, yes, parsimonious models are best. If all else is equal, then the predictions of the competing models are the same, and the more complicated model gains you nothing.

    That only holds when the predictions are the same, however. A complicated model that unites quantum mechanics and general relativity is superior to a simpler but incomplete model.

    It’s also important to distinguish the complexity of the model from the complexity of its predictions. People often complain that the multiverse is extravagant, but that’s irrelevant if the multiverse is a prediction of your model rather than an assumption.

  6. It really is true that complaints about what science is helps creationism.
    We are accused of not doing science and we, yec,like me, accyse evolutionusm of not being scientific. i make a more careful case that its not science because its evidences are not based on biology but other subjects.
    Science to be something other then regular ordinary thinking before conclusion mUST be about methodology. Everyone is tempted to say science is about accurate conclusions. I trespassed with this also. I SAY if evolution is not true it couldn’t possibly have evidence to back it up and herefore is not a scientific theory. I think this was wrong. I retreat thats its not science because its not based on biology as i said.
    Wrong conclusions is irrelevant to whether something is scientific. Accurate conclusions is irrelevant to what is scientific if methodology is not behind it.
    A space alien video of earths biology origins HOWEVER PERFECT accurate of the truth would NOT be science.
    iD is science. YEC in criticisms is science depending on the subject questioned. YEC assertions are greatly based on a witness(bible) and so not scientific. Evolution is not scientific as stated in its big points. Physics in this string stuff etc is probably scientific.
    Experimentation, falsifying, etc are only tools in doing science. The scientific methodology can be applied without them.
    Its simply a high standard of investigation before conclusions are drawn.
    a agreed standard.
    The questioning in such a rock solid science as physics shows the creationists are right, or an option, that science is not being understood and misued as a concept for centuries now.

  7. “For every complex problem there is an answer that is clear, simple, and wrong.”
    H. L. Mencken

    String theory is not exactly clear and simple, but unless it can be tested, one has to assume that it’s wrong. Replace “simple” with “parsimonious”, and “clear” with “logically compelling” or “beautiful”, and it all falls into place.

    We’re still in that situation where there are a nearly infinite number of ways to be wrong, and very few ways to be right. There’s no substitute for testability.

    Anyway, as I understand it, when somebody talks about string theory, isn’t the proper response to ask “which one”?

  8. llanitedave,

    We’re still in that situation where there are a nearly infinite number of ways to be wrong, and very few ways to be right.

    I would argue that there are infinitely many ways to be right. I think that for any candidate theory, there are an infinite number of alternative theories that make exactly the same predictions.

  9. keiths,

    While that may be the case for String Theories, I can’t think of anything else that would support your statement. Certainly there are not many theories that make exactly the same predictions as General Relativity, are there?

  10. keiths:

    I would argue that there are infinitely many ways to be right. I think that for any candidate theory, there are an infinite number of alternative theories that make exactly the same predictions.

    llanitedave:

    Certainly there are not many theories that make exactly the same predictions as General Relativity, are there?

    I think there are.

    General Relativity holds that the presence of matter warps spacetime. Here are a couple of ontologically distinct theories that make the same empirical predictions:

    1. Matter doesn’t directly warp spacetime. Instead, the presence of matter induces vortices in the (inaccessible to us) 5th, 6th, and 7th dimensions. These vortices in turn cause the warping of spacetime, giving the appearance that it follows directly from the presence of matter.

    2. God observes the universe and carefully and systematically warps spacetime wherever he finds matter.

    We reject these two theories not because they clash with observation, but because they are needlessly complex. We prefer parsimonious explanations for practical and esthetic reasons, but that doesn’t mean that the more baroque explanations are wrong.

  11. keiths:
    We reject these two theories not because they clash with observation, but because they are needlessly complex. We prefer parsimonious explanations for practical and esthetic reasons, but that doesn’t mean that the more baroque explanations are wrong.

    I would rather argue that, unless the “needlessly complex” theories can be tested in isolation, i.e., their additional statements can be tested independently, then they are indeed more likely to be incorrect than a more parsimonius theory that makes the same predictions. If the additional statements make no additional predictions, then it’s hard to even consider them as separate theories, but merely as meaningless embellishments of the more fundamental theory. Putting lipstick on a pig, as the ever-popular metaphor suggests, does not create a new creature.

    Unless the embellishments are carefully constructed post hoc, it’s highly likely that they will indeed introduce new predictions, and require new tests.

  12. llanitedave,

    I would rather argue that, unless the “needlessly complex” theories can be tested in isolation, i.e., their additional statements can be tested independently, then they are indeed more likely to be incorrect than a more parsimonius theory that makes the same predictions.

    That depends on your assumptions regarding the probability distributions of possible worlds. But what justifies those assumptions?

    If the additional statements make no additional predictions, then it’s hard to even consider them as separate theories, but merely as meaningless embellishments of the more fundamental theory.

    You’re a positivist? I reject positivism because I see no reason that reality should be constrained by what is or isn’t observable to us.

    Unless the embellishments are carefully constructed post hoc, it’s highly likely that they will indeed introduce new predictions, and require new tests.

    Agreed. My point is simply that our preference for parsimony is a heuristic. By itself, parsimony cannot tell us which of two theories is correct.

    Summing up, I would say that:

    1. Our preference for parsimonious theories is a heuristic, and it’s been quite successful. However, there is no reason, given two theories that make identical predictions, why the more complex theory cannot be true.

    2. Simpler theories are easier to work with and esthetically more appealing. A more complicated theory with identical entailments might actually be true, but we have no way of determining that, and therefore no reason to choose it over its simpler counterpart.

    3. Even for heuristic purposes, it is the parsimony of the theory, not of its entailments, that matters.

  13. keiths:
    You’re a positivist?I reject positivism because I see no reason that reality should be constrained by what is or isn’t observable to us.

    That may be so, but if that’s your approach, it seems to me you have no reason to criticize the statements of WJM, Phoodoo, Mung, or even Byers. Every superstition relies on the existence of an untestable “reality”, and if you invoke one of your own, you’re in the same boat as they are.

    Anything that’s not observable to us is unknowable by us, and is a poor choice to use as the basis of any existential statement. We’re not usually arguing about what reality is, but rather about what reality can be asserted to be. The unknowable cannot be assertable.

  14. llanitedave: Anything that’s not observable to us is unknowable by us

    When exactly did you observe that?

    It’s like you all don’t even read what you write 😉

    peace

  15. llanitedave,

    Anything that’s not observable to us is unknowable by us, and is a poor choice to use as the basis of any existential statement.

    Yes. Ironically, I’m taking that principle more seriously than you are.

    If two theories make identical predictions, then we can’t determine through observation which, if either, is true. Our reasons for favoring the simpler theory may be practical, esthetic, or philosophical, but they are not empirical.

    Every superstition relies on the existence of an untestable “reality”, and if you invoke one of your own, you’re in the same boat as they [WJM, phoodoo, Mung, Byers] are.

    I’m not invoking an untestable reality. I’m simply pointing out that there is no empirical basis for ruling them out, provided that the theories invoking them do not clash with observation. Occam’s Razor is a heuristic, not a law of nature.

    Let me stress that the non-empirical reasons for favoring parsimony, all else being equal, may be quite compelling. I’m not saying they aren’t compelling — I’m simply saying that they aren’t empirical.

  16. fifthmonarchyman: When exactly did you observe that?

    Today, yesterday, the day before that. In fact, every day of my life has provided discreet support for this observation.

    It’s like you all don’t even read what you write

    peace

    Actually, it appears you don’t understand the insight of that observation and why there is no way to provide a counter example.

  17. keiths:
    llanitedave,

    Yes. Ironically, I’m taking that principle more seriously than you are.

    If two theories make identical predictions, then we can’t determine through observation which, if either, is true.Our reasons for favoring the simpler theory may be practical, esthetic, or philosophical, but they are not empirical.

    I’m not invoking an untestable reality.I’m simply pointing out that there is no empirical basis for ruling them out, provided that the theories invoking them do not clash with observation. Occam’s Razor is a heuristic, not a law of nature.

    Let me stress that the non-empirical reasons for favoring parsimony, all else being equal, may be quite compelling.I’m not saying they aren’t compelling — I’m simply saying that they aren’t empirical.

    I’m going to disagree vigorously on this one, while at the same time hoping you had a very good holiday so far. I know I have!

    I’ll reiterate my “lipstick on a pig” example. It still stands. Putting lipstick on a pig does not give you a new creature. Adding non-functional, untestable, and gratuitous embellishments to a theory does not give you a new theory. That’s not a matter of aesthetics or philosophy, it’s a simple identity.

    You cannot claim that there’s a difference between two theories unless they give you different predictions. In this case, the simpler one will always win, because you CAN empirically show that if the factors that make the simpler theory are required, then the embellishments are not required. And if they are not required they are not part of the theory.

    E = mc^2 is a fundamental theoretical statement

    E = (ma/a)*c^2 is not a different theory. That’s not a statement of philosophy.

    Now, if you’re postulating a unique sequence of events, you may be able to argue that any one of several pathways could have led to the same current circumstances. But those aren’t really different theories. The theory would consist of the underlying process that enabled the pathways. Those unique pathways are a one-off. The underlying theory is a general statement about the world, not a one-off. And that theory would have to be general enough to allow either of the pathways you think are viable.

  18. llanitedave,

    I’m going to disagree vigorously on this one, while at the same time hoping you had a very good holiday so far. I know I have!

    Likewise! And I welcome the disagreement — it’s one of the reasons I come to TSZ.

    I’ll reiterate my “lipstick on a pig” example. It still stands. Putting lipstick on a pig does not give you a new creature. Adding non-functional, untestable, and gratuitous embellishments to a theory does not give you a new theory. That’s not a matter of aesthetics or philosophy, it’s a simple identity.

    To show why I disagree, let me employ an example I’ve used before to argue against ID:

    Bob is walking through the desert with his friend, a geologist. They come across what appears to be a dry streambed. After some thought, Bob states that every rock, pebble, grain of sand and silt particle was deliberately placed in its exact position by a Streambed Designer. His friend says “That’s ridiculous. This streambed has exactly the features we would expect to see if it was created by flowing water. Why invoke a Streambed Designer?”

    Who has the better theory, Bob or his friend?

    To my mind, the geologist’s theory is far better than Bob’s, but by your criterion, the two theories are identical, since both predict a classic desert streambed. That seems absurd to me. How can a theory involving the painstaking placement of particles by an intelligent agent be identical to one in which this is accomplished by the physics of flowing water?

  19. keiths: Aqueducts don’t look like dry desert streambeds.

    How do you know that? And who cares anyways? If it’s an aqueduct, Bob is right. If you want to stack the deck against Bob you can’t blame Bob. Bob didn’t know he was playing against a stacked deck.

  20. Bob is walking through the desert with his friend, a geologist. They come across what appears to be a dry aqueduct. After some thought, Bob states that every rock, pebble, grain of sand and silt particle was deliberately placed in its exact position by an Aqueduct Designer. His friend says “That’s ridiculous. This is just a streambed and it has exactly the features we would expect to see if it was created by flowing water.. Why invoke an Aqueduct Designer?”

    Who has the better theory, Bob or his friend?

  21. keiths: Aqueducts don’t look like dry desert streambeds.

    So? Are you claiming that aqueducts don’t exist in the desert? How do you know that Bob and his geologist friend didn’t stumble across an aqueduct?

  22. keiths:
    llanitedave,

    Likewise!And I welcome the disagreement — it’s one of the reasons I come to TSZ.

    To show why I disagree, let me employ an example I’ve used before to argue against ID:
    Bob is walking through the desert with his friend, a geologist. They come across what appears to be a dry streambed. After some thought, Bob states that every rock, pebble, grain of sand and silt particle was deliberately placed in its exact position by a Streambed Designer. His friend says “That’s ridiculous. This streambed has exactly the features we would expect to see if it was created by flowing water. Why invoke a Streambed Designer?”

    Who has the better theory, Bob or his friend?

    To my mind, the geologist’s theory is far better than Bob’s, but by your criterion, the two theories are identical, since both predict a classic desert streambed. That seems absurd to me.How can a theory involving the painstaking placement of particles by an intelligent agent be identical to one in which this is accomplished by the physics of flowing water?

    Well, here we’re not seeing two theories with identical predictions. We’re not even seeing two theories! What Bob is offering is not a theory, but a bare assertion. It makes no actual predictions, it involves no testability, and supplies no general processes.

    An actual fluvial theory would be able to explain the size and shapes of the cobbles, any imbrication of the grains with respect to stream flow, the layering of sediment in the banks, the composition of the grains with respect to source areas, the stream gradient, catchment volume, and local precipitation regime, and the effect of the types of materials that it flows across along the channel. It would be able to explain not only this particular desert wash, but other stream channel features as well, both in deserts and wetter environments. It would predict specific similarities and differences based on the different starting landforms and climates. Bob, on the other hand, would merely continue to mumble “Goddidit”.

    No, we are definitely not looking at identical theories or predictions here. As I said, a theory is not a one-off. It is a general, process explanation that covers a class of phenomena, not just an individual event.

  23. llanitedave,

    As I said, a theory is not a one-off. It is a general, process explanation that covers a class of phenomena, not just an individual event.

    Not so. For example, the Big Bang theory was introduced to explain a one-off event — the birth of our universe.

    Also, I think you’re focusing on the definition of ‘theory’, which is just semantics, when the real issue is what the evidence tells us, and more importantly what it can’t tell us, about reality.

    To use my favorite example, the Rain Fairy hypothesis — in which a superhuman Rain Fairy flits about creating the weather — can be invoked to explain every weather observation ever made on earth. Yet all of us, including ID proponents, prefer modern meteorological theory to the Rain Fairy hypothesis.

    Why? It’s not because modern meteorological theory fits the evidence better. In fact, the Rain Fairy hypothesis fits the evidence perfectly, and always will. Whatever weather happens, happens because the Rain Fairy makes it so.

    We prefer modern meteorological theory not because of a superior fit to the evidence, but because it makes far fewer ad hoc assumptions. We reject the Rain Fairy hypothesis on philosophical, not empirical, grounds. It isn’t falsified outright, but we favor modern meteorological theory for other compelling reasons.

    (Likewise with Intelligent Design as an explanation of the pattern of life on earth.)

  24. To summarize:

    1. If two theories/hypotheses/explanations/thingamajigs have identical entailments, no observation can ever distinguish the two.

    2. One or the other can still be favored on philosophical or esthetic grounds.

    3. As a corollary to #1, no observation by itself can support the claimed existence of untestable elements in either of the theories/hypotheses/explanations/thingamajigs.

    4. Likewise, no observation by itself can support the denial of the existence of untestable elements in either of the T/H/E/Ts.

  25. keiths:
    llanitedave,

    Not so.For example, the Big Bang theory was introduced to explain a one-off event — the birth of our universe.

    Also, I think you’re focusing on the definition of ‘theory’, which is just semantics, when the real issue is what the evidence tells us, and more importantly what it can’t tell us, about reality.

    To use my favorite example, the Rain Fairy hypothesis — in which a superhuman Rain Fairy flits about creating the weather — can be invoked to explain every weather observation ever made on earth. Yet all of us, including ID proponents, prefer modern meteorological theory to the Rain Fairy hypothesis.

    Why?It’s not because modern meteorological theory fits the evidence better. In fact, the Rain Fairy hypothesis fits the evidence perfectly, and always will.Whatever weather happens, happens because the Rain Fairy makes it so.

    We prefer modern meteorological theory not because of a superior fit to the evidence, but because it makes far fewer ad hoc assumptions.We reject the Rain Fairy hypothesis on philosophical, not empirical, grounds.It isn’t falsified outright, but we favor modern meteorological theory for other compelling reasons.

    (Likewise with Intelligent Design as an explanation of the pattern of life on earth.)

    That’s a common misconception, but the Big Bang Theory really does not explain the beginning of the universe. The beginning of the Universe was a singularity, which remains unexplained. There is no theory as to why this singularity appeared or began expansion. The Big Bang theory covers the physical events that followed the beginning of that expansion, only after energy existed in a recognizable form.

    We are aware of only this universe, but the Big Bang Theory does not explain the singularity, therefore it predicts that if other singularities did appear, and did began to expand in such a way that a super-condensed energy field was created, the same Big Bang Theory would apply to these new phenomena as well.

    And yes, I’m concerned with semantics here. Semantics is “meaning” in the most fundamental sense of defining your terms and creating coherent and understandable explanations. You could not have science without it, and as long as we are “Fighting for the Soul of Science”, then I’m going to insist that semantics are important, and a “theory” requires a specific and consistent definition.

  26. llanitedave,

    The beginning of the Universe was a singularity, which remains unexplained.

    True, but my point remains: the Big Bang theory is a theory despite being concerned with a one-off event.

    You could not have science without it, and as long as we are “Fighting for the Soul of Science”, then I’m going to insist that semantics are important, and a “theory” requires a specific and consistent definition.

    You’re missing the larger point, which is that the precise boundaries of the term ‘theory’ are not at issue here. Whether you consider something to be a theory, a hypothesis, an explanation, a model, or even merely a story doesn’t matter. If it has observable entailments, then it can be falsified if those entailments do not obtain.

    If two theories/hypotheses/explanations/models/stories have identical entailments, then no observation by itself can rule out either one.

    A T/H/E/M/S in which a Streambed Designer painstakingly situates every rock, pebble, grain of sand and silt particle is different from one in which the work is done by flowing water. The explanations are (quite!) distinct, but the entailments are the same.

    We reject the Streambed Designer, the Rain Fairy, and the Intelligent Designer, and our reasons for doing so are solid — who would argue for the Rain Fairy, after all? — but it’s not because they have been falsified, nor is it because the simpler of two explanations is necessarily the truer.

Leave a Reply