10 ms·

Why falsificationism is false

54 points by viburnum 7y ago

whatshisface 7y ago
If there's no conceivable experiment that could contradict your claim then it's meaningless. Just because there might be other ways for a claim to be vacuous doesn't mean the necessary condition above doesn't hold.
- smallnamespace 7y ago
  There may be some difficulties with the criterion that you just laid out, one practical and another logical: 1. 'Conceivable experiment' is a changing definition, since the realm of what humans can conceive is always expanding with the growth of technology and knowledge. If we had been around 2000+ years ago, would you have considered Democritus's atomic theory to be meaningless, given the lack of any conceivable path to test it at the time? Either you have to take quite an expansive view of what is 'conceivable', or admit that 'meaningless' here refers more to practical difficulties that may be temporary, rather than anything inherent in a particular claim. 2. If we turn your statement back on itself, statements about meaning are also statements about the world (after all, what is 'meaning' if it doesn't involve what humans think and believe?). Is your statement that 'only falsifiable claims are meaningful' itself falsifiable? If so, what sort of experiment would you set up to test it?
  - whatshisface 7y ago
    >would you have considered Democritus's atomic theory to be meaningless, given the lack of any conceivable path to test it at the time? Strictly speaking Democritus's atomic theory is still untestable today because it doesn't include the actual scale of the atoms. For all we know the apparently continuous quantum fields are themselves made of "atoms" (in his sense, not the modern meaning of the word atom). If he had said "things will get clearly atom-y around a nanometer" then it would have been at least conceivable that someone could eventually look that closely. >you have to take quite an expansive view of what is 'conceivable' That's right, I was laying down a very liberal criteria (and saying that it was a necessary but insufficient condition). >If so, what sort of experiment would you set up to test it? That's kind of like asking what sort of experiment could disprove the law of the excluded middle: not an unreasonable question, but the answer is going to sound pretty silly. To test my claim, you could iterate though every claim about the universe, checking to see if it simultaneously could not contradict with any observation, and yet predicted what you would observe. Good luck actually doing it; but it does technically meet my criterion.
    - smallnamespace 7y ago
      > That's right, I was laying down a very liberal criteria > but the answer is going to sound pretty silly. Yes, I was trying to illustrate that the line between 'science' and 'metaphysics' is necessarily a blurry one, and the choice of (and justification for) where to draw it itself lies in the realm of metaphysics. The broader the universe of claims you admit as scientific, the more 'testability' and 'falsifiability' lie closer to the realm of theory. However, even theories (and paradigms) that can't be directly placed on the experimenter's bench today can point towards fruitful lines of investigation. On the flip side, we can't but to make metaphysical (and therefore non-scientific) claims and hold them as true, despite our best efforts ... the strong version of 'scientism' (that only scientific claims can have meaning) is a self-defeating one.
  - klyrs 7y ago
    > Is your statement that 'only falsifiable claims are meaningful' itself falsifiable? If so, what sort of experiment would you set up to test it? As a mathematician, I've always looked on this issue with some bemusement. Physicists are particularly egregious in their abuse of mathematical notation, yet get very fussy about what constitutes "proof." If we consider a result measured to "five nines" of uncertainty -- scratch that -- if we consider a million results measured to "five nines" of uncertainty, than we're almost certain to have "proved" a falsehood. And that's assuming that everybody is using complete models with no missing variables (protip: they're not). And then they turn around and heap scorn on "soft sciences" which deign to apply scientific methodology to systems too small, and far too complicated for such simplistic modeling, where experimentalists jump through extraordinary hoops to avoid ethical disasters. The need for falsifiability is a strongly-held opinion. And don't get me wrong, it's a great rule of thumb. From a mathematical basis, I can take a stab at your question. If you have a million data points, you can find a million-degree polynomial to fit it. If you're allowed to adjust the polynomial every time we make a new observation, it's unfalsifiable. However, that approach is rightly seen as problematic because the predictive power (that is, meaningfulness) is zero. But of course, this is only an existence proof (aka, an anecdote), that some unfalsifiable theories are meaningless. It doesn't prove that all unfalsifiable theories are meaningless.
  - mannykannot 7y ago
    Democritus might have said that his theory was falsifiable in principle, as there were different theories being proposed by others. All of those theories, however (Democritus' included) lacked any empirical evidence for or against them, so they were all metaphysical speculation, not science (a claim that has been made about some aspects of fundamental physics today.) I agree, however, that it was not meaningless - that should be reserved for ideas that are unfalsifiable in principle, such as p-zombies With regard to point 2, if 'only falsifiable claims are meaningful' is not falsifiable, then it might indicate that it is not a scientific theory -- but no surprise there, as it is just a personal opinion, as is whether or not you agree with Popper. As it happens, I think it is falsifiable, in general - mathematical identities are not falsifiable, if you include the relevant axioms, and they are not meaningless. It's pretty clear, however, that whatshisface's comment was intended in the context of science.
    - whatshisface 7y ago
      >mathematical identities are not falsifiable Math papers say, "If you follow these induction rules from these definitions and axioms you will arrive at these results." That's falsifiable, in fact it is found to be false every now and then.
      - mannykannot 7y ago
        But is that a falsifiable claim about a proposition, or the identity itself (in those cases where the claim is not false)? The Platonist in me suspects that these are different things, though I could probably be induced to change my mind by a persuasive counter-argument.
        whatshisface 7y ago
        The Platonist idea that mathematical statements correspond to something "out there" in metaphysical space is a great example of an idea that's not falsifiable. However that's not an idea that I would consider part of mathematics.
- logicchop 7y ago
  The response (from Duhem, Quine and others) is this: Suppose you think A contradicts B. You would surely admit that, there is likely some other C that, if C were true, A wouldn't contradict B. But if that's the "logic" of science then it isn't straightforward deduction. If you want an example, consider an ordinary empirical observation. Someone claims the ball is red; you look and you see green. Classic contradiction, right? Except if lighting conditions are wrong, or there is a pane of colored glass between you, or.. The point is that there are additional claims that, if they were true, would remove the "contradicts" relation between the two original claims. But ordinary logical relationships shouldn't work like that.
  - whatshisface 7y ago
    That's just the common mistake of incorrectly encoding one's beliefs in logic and then complaining logic doesn't work. If you think there might be a glass pane between you and a colored ball, your true belief is (No glass pane implies green ball), not (green ball). If you don't mind me getting on my soap box, it kind of seems like the recipe for coming up with a renowned counterargument to a technical tool goes like: - Find a technical device that you don't understand. - Misuse it because you don't understand it. - Notice that it isn't working. - Conclude that logic doesn't work, math doesn't work, science doesn't work, the bug is in the compiler, take your pick.
    - logicchop 7y ago
      It's not clear what you mean or are complaining about. Feel free to provide your own example. Just come up with two empirical claims (A and B) such that you think A contradicts B. It should then be relatively easy to come up with a C that, if C were true, A would no longer contradict B. It's obvious once you start laying things out carefully.
      - whatshisface 7y ago
        Here's a specific example. Claim A: "If nothing is wrong with your experiment or analysis, you will observe that the speed of light in a vacuum is the same in all inertial reference frames up to experimental uncertainty." Claim B: "If nothing is wrong with your experiment or analysis, you will observe that the speed of light in a vacuum is different in every inertial reference frame up to experimental uncertainty." Every experimentalist realizes that their real result is "I didn't make a mistake -> the answer is X," not "the answer is X." That's the correct way to encode an experiment in logic, leaving aside the high sophistication of how "I didn't make a mistake" gets treated in practice.
- yters 7y ago
  There could be experiments that confirm the claim even if none contradict it. For example, the immaterial man cannot be detected in any empirical way, and there is no experiment that can falsify his existence. However, he can choose to materialize in front of everyone, confirming his existence.
  - pessimizer 7y ago
    > However, he can choose to materialize in front of everyone, confirming his existence. Not confirming his existence as an immaterial man, unless you're willing to take his word for the fact that he was recently immaterial. If I appear in your country and tell you that I'm from the country of "Balanava," I'm not a confirmation of the claim that the country of "Balanava" exists. At best I'm a support.
    - yters 7y ago
      His immateriality is immaterial :) The point is he exists and his existence cannot be falsified (while immaterial) but his existence can still be confirmed. Let's call him person X. So, the statement "person X exists" is empirically meaningful even though it cannot be falsified.
lisper 7y ago
This article attacks a straw man. Being falsifiable is necessary but it is not sufficient to be considered a scientific hypothesis. Such a hypothesis also, and more importantly, has to provide a better explanation of some phenomenon than the current best theory. Experimental evidence is only brought to bear to decide among plausible alternative theories after the vast majority of candidate theories have been eliminated for not providing good explanations. It's easy to see that this must be true because we can only ever have a finite amount of data, and that will always be consistent with an infinite number of falsifiable theories (c.f. Russell's teapot). So data cannot possibly help us choose from among those.
- logicchop 7y ago
  Popper heavily criticizes the value of explanation. He gives examples of really great "explanations" that have almost zero empirical value (things like psychoanalysis). Hard to see how it could be a "straw man."
  - akvadrako 7y ago
    I don’t think that’s true. One of the most prominent figures to continue developing popper’s philosophy is David Deutsch, and he emphasizes good explanations. Being a good explanation means, among other things, it has a wide range of applicability and is hard to adjust if predictions don’t work out. Psychoanalysis wasn’t a good answer.
    - logicchop 7y ago
      It's easy to go read Popper. Look for his discussions of Adler.
  - vixen99 7y ago
    But as we know, these examples are not 'great', they're lousy explanations because they are not testable. Good explanations are another matter altogether as Deutsch points out in his criticism of the instrumentalist view of quantum theory.
  - 7y ago
nerdponx 7y ago
What is the point of this article? The content is interesting: falsification is more complicated than it seems at first, and scientists don't practice philosophically-pure falsificationism. But it's wrapped up in some kind of finger-pointing "the emperor has no clothes!" type of exposé format that doesn't make sense and doesn't lead to a satisfying conclusion.
- dang 7y ago
  The article explains clearly and at length why the author wanted to write it. But the best thing you say here is "The content is interesting". That's why it belongs on HN! https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html
  - nerdponx 7y ago
    Sorry I was unclear. My comments were meant for the author, not the poster.
    - dang 7y ago
      Yes, and he goes into that at length in the article. The HN thing was just my addition; shop talk basically.
viburnum 7y ago
Well that is definitely not the title I submitted.
- dang 7y ago
  The site guidelines call for using the original title unless it is misleading or linkbait (https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html). Judging by the first couple of comments that appeared here, the title is baity, because it was generating shallow reactions—to the top-level dramatic claim, rather than to the rich content in the article, such as its final paragraph. When the comments in a thread are generated from the most-obvious, already-familiar information in an article—which often shows up in the title—than we get a discussion of people reciting things they already know or believe (not interesting!), instead of a learning process (much more interesting). I changed the title to "There are strong arguments against falsificationism", which is a phrase from the article body. (When we change titles, we always look for representative language in the article itself.) A couple minutes later, I wasn't sure, so reverted it. This is a bit of an edge case. To some extent, what to do with the title depends on which direction the thread takes: shallow "I know better" dismissals? or thoughtful responses to the less obvious?
outlace 7y ago
This article attacks falfsificationism in part on the basis that scientists don’t seem to actually behave according to “naive falsficationism” in which you supposedly should immediately throw out a theory as soon as you get some conflicting empirical data. The article notes that collecting empirical data to falsify a theory depends on auxiliary theories (eg testing the idea that smoke causes cancer in rats depends on our ability to precisely define and identify when cancer is present) and so most scientists will question these auxiliaries rather than the main theory if conflicting data comes out. This is presented as evidence that scientists don’t actually practice falsficationism. This whole discussion completely ignores uncertainty and the fact that theories describe not only relationships between variables but also involve parameters that need to be estimated. If I fill a rat cage with smoke and the rats don’t get cancer, I don’t immediately assume all previous studies were wrong. This isn’t ignoring my duty to falsify, it’s realizing that the theory that smoking causes cancer describes a causal relationship between variables (smoking -> cancer) it also implicitly or explicitly depends on parameters (in this case, how much and how long a rat must be exposed to smoke to cause cancer at some incidence rate). Given that many other studies found the causal relationship, it is more likely that my experiment messed up the parameters of the causal model (eg I didn’t administer smoke long enough) rather than messed up the causal relationship. And parameter estimation is subject to uncertainty. Now if I’m absolutely sure I’m getting all my model parameters correct, and I can only be confident if other people get the same results. Then I’m moving toward falsifying the theory. The point is that science involves positing theories, and theories are descriptions of causal relationships and generally involve parameters. Much of science is just estimating model parameters given that we assume some model to be tentatively true. You assume your model to be tentatively true, eg smoking causes cancer, now you need to estimate the effect size which is a parameter problem.
noego 7y ago
The author makes some good points, but does not propose his own criteria for what distinguishes a good theory from a bad theory, which seems like a cop-out to be frank. I imagine a big appeal of falsifiability in its early days, stemmed from its opposition to religious dogma. Imagine dealing with someone whose answer to every question comes back to God. "Why did the patient die?" "Because God decreed it." Talking about germ-theory must have been a massive improvement over theology, hence why ideas like falsifiability must have arisen to provoke people to think in more critical ways. My own standard for evaluating theories comes down to the following: The simplest explanation that best fits existing data, and makes useful predictions for the future. "Simplest" because we need to prevent overfitting. "Fits existing data" because the whole point of a theory is to explain what we see around us. And "useful predictions" because it will actually provide practical benefit to us. As opposed to a theory like "God decreed it", which may be extremely simple and error-free, but provides us with no practical benefit at all. As I see it, falsifiability is a side-effect of the second and third requirements. If a theory is making useful predictions, we might as well check to see if it's accurate. And if it's not, there's no point in holding on to a theory that's making faulty predictions. Falsifiability, in this perspective, isn't a requirement in and of itself - it's a natural consequence of the requirement that theories be useful and fit any data at hand.
- thethirdone 7y ago
  > The simplest explanation that best fits existing data, and makes useful predictions for the future. My main issue with that definition is what makes it a useful prediction. Does a theory (for particle physics) that predicts the Higgs Boson more useful than one that didn't? If not, then before we collected the data to reject other theories they all would have been equal. I would rather suggest that the properties of a good theory are best fitting existing data and having specific predictions for future measurements. > Falsifiability, in this perspective, isn't a requirement in and of itself - it's a natural consequence of the requirement that theories be useful. If you are of the opinion that any good theory is falsifiable (in the sense described in the article), then presumably you disagree with the rejection of falsifiability in the article. Where would you say the error in logic in the article is?
thethirdone 7y ago
> This means that a theory is never falsifiable simpliciter, but only relative to a set of background assumptions. Therefore, if we say that a theory is only scientific if it’s falsifiable, then it follows that no theory, not even a theory as successful as Newton’s law of universal gravitation, is scientific. Of course, this is absurd, so falsificationism is false. You can consider the union of the theory and background assumptions to be a single theory. The premises do not imply that no theory is scientific. An analogy to mathematics is pretty apt. The Riemann hypothesis (RH) is falsifiable in the crude sense described, but cannot be proven or disproven without an axiomatic basis. If you were able to derive a contradiction from RH and some other accepted theorems, you would think that falsifies RH. However, if you wanted a more concrete picture of what was proved, you might reduce the other theorems to Zermelo–Fraenkel set theory (ZF) so you have the result that RH + ZF is inconsistent/false. Maybe in some other axiomatization RH is true, but the general community accepts ZF so that means RH + ZF is false effectively means RH is false. I do agree with some of the sentiment of the article. There should be thought about how to actually test what we say we are testing (not having a reliance on background assumptions) and to accept negative results rather than just find something to blame.
RyanAF7 7y ago
This article is a garbled mush of ego entrism. Appeals to authority out the ass with no real argument.
- dang 7y ago
  Maybe so, but please don't post unsubstantive comments here. https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html
  - RyanAF7 7y ago
    I would argue that stating the flaws of an article is substative. Seems you just don't like how I said it.
    - dang 7y ago
      Ok, I see what you mean. Let me try explaining this a different way. The site guidelines ask you not to call names in comments. Your comment was full of name-calling, in the sense that they use the term. Could you please review them and take the spirit of this site more to heart? https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html A comment like that one encourages the worst in others, and we're trying to avoid sinking into a downward spiral here. Downward spirals are the default outcome on the internet, so avoiding them takes energy. We need everyone to contribute. Here's another way to look at this. Name-calling adds noise to a comment (and encourages others to add still more noise), which lowers the signal/noise ratio. As the signal/noise ratio gets lower, a community like HN deteriorates. This process happens faster than linearly: thoughtful users leave because of the noise, causing the posts to get even worse, driving even more thoughtful users away. That would be a scorched-earth outcome, i.e. an extinction event for HN. Our job is to avoid that. A much better way to post would be to raise the signal/noise ratio. "Signal" here means adding new information, information we can learn from. In the case of your comment, that would mean pointing out more clearly, in more detail and more neutrally, the flaws you see in the article. If you can't or don't want to do that, or you just think the article is so bad it doesn't deserve the effort, then please don't post to the thread at all—instead, flag the article as one that you think doesn't belong on HN, and move on. There's a karma threshold of 30 before you get to flag, but it should only take a few good submissions or comments to get there.

54 points by viburnum 7y ago

If there's no conceivable experiment that could contradict your claim then it's meaningless. Just because there might be other ways for a claim to be vacuous doesn't mean the necessary condition above doesn't hold.

smallnamespace 7y ago
There may be some difficulties with the criterion that you just laid out, one practical and another logical: 1. 'Conceivable experiment' is a changing definition, since the realm of what humans can conceive is always expanding with the growth of technology and knowledge. If we had been around 2000+ years ago, would you have considered Democritus's atomic theory to be meaningless, given the lack of any conceivable path to test it at the time? Either you have to take quite an expansive view of what is 'conceivable', or admit that 'meaningless' here refers more to practical difficulties that may be temporary, rather than anything inherent in a particular claim. 2. If we turn your statement back on itself, statements about meaning are also statements about the world (after all, what is 'meaning' if it doesn't involve what humans think and believe?). Is your statement that 'only falsifiable claims are meaningful' itself falsifiable? If so, what sort of experiment would you set up to test it?
- whatshisface 7y ago
  >would you have considered Democritus's atomic theory to be meaningless, given the lack of any conceivable path to test it at the time? Strictly speaking Democritus's atomic theory is still untestable today because it doesn't include the actual scale of the atoms. For all we know the apparently continuous quantum fields are themselves made of "atoms" (in his sense, not the modern meaning of the word atom). If he had said "things will get clearly atom-y around a nanometer" then it would have been at least conceivable that someone could eventually look that closely. >you have to take quite an expansive view of what is 'conceivable' That's right, I was laying down a very liberal criteria (and saying that it was a necessary but insufficient condition). >If so, what sort of experiment would you set up to test it? That's kind of like asking what sort of experiment could disprove the law of the excluded middle: not an unreasonable question, but the answer is going to sound pretty silly. To test my claim, you could iterate though every claim about the universe, checking to see if it simultaneously could not contradict with any observation, and yet predicted what you would observe. Good luck actually doing it; but it does technically meet my criterion.
  - smallnamespace 7y ago
    > That's right, I was laying down a very liberal criteria > but the answer is going to sound pretty silly. Yes, I was trying to illustrate that the line between 'science' and 'metaphysics' is necessarily a blurry one, and the choice of (and justification for) where to draw it itself lies in the realm of metaphysics. The broader the universe of claims you admit as scientific, the more 'testability' and 'falsifiability' lie closer to the realm of theory. However, even theories (and paradigms) that can't be directly placed on the experimenter's bench today can point towards fruitful lines of investigation. On the flip side, we can't but to make metaphysical (and therefore non-scientific) claims and hold them as true, despite our best efforts ... the strong version of 'scientism' (that only scientific claims can have meaning) is a self-defeating one.
- klyrs 7y ago
  > Is your statement that 'only falsifiable claims are meaningful' itself falsifiable? If so, what sort of experiment would you set up to test it? As a mathematician, I've always looked on this issue with some bemusement. Physicists are particularly egregious in their abuse of mathematical notation, yet get very fussy about what constitutes "proof." If we consider a result measured to "five nines" of uncertainty -- scratch that -- if we consider a million results measured to "five nines" of uncertainty, than we're almost certain to have "proved" a falsehood. And that's assuming that everybody is using complete models with no missing variables (protip: they're not). And then they turn around and heap scorn on "soft sciences" which deign to apply scientific methodology to systems too small, and far too complicated for such simplistic modeling, where experimentalists jump through extraordinary hoops to avoid ethical disasters. The need for falsifiability is a strongly-held opinion. And don't get me wrong, it's a great rule of thumb. From a mathematical basis, I can take a stab at your question. If you have a million data points, you can find a million-degree polynomial to fit it. If you're allowed to adjust the polynomial every time we make a new observation, it's unfalsifiable. However, that approach is rightly seen as problematic because the predictive power (that is, meaningfulness) is zero. But of course, this is only an existence proof (aka, an anecdote), that some unfalsifiable theories are meaningless. It doesn't prove that all unfalsifiable theories are meaningless.
- mannykannot 7y ago
  Democritus might have said that his theory was falsifiable in principle, as there were different theories being proposed by others. All of those theories, however (Democritus' included) lacked any empirical evidence for or against them, so they were all metaphysical speculation, not science (a claim that has been made about some aspects of fundamental physics today.) I agree, however, that it was not meaningless - that should be reserved for ideas that are unfalsifiable in principle, such as p-zombies With regard to point 2, if 'only falsifiable claims are meaningful' is not falsifiable, then it might indicate that it is not a scientific theory -- but no surprise there, as it is just a personal opinion, as is whether or not you agree with Popper. As it happens, I think it is falsifiable, in general - mathematical identities are not falsifiable, if you include the relevant axioms, and they are not meaningless. It's pretty clear, however, that whatshisface's comment was intended in the context of science.
  - whatshisface 7y ago
    >mathematical identities are not falsifiable Math papers say, "If you follow these induction rules from these definitions and axioms you will arrive at these results." That's falsifiable, in fact it is found to be false every now and then.
    - mannykannot 7y ago
      But is that a falsifiable claim about a proposition, or the identity itself (in those cases where the claim is not false)? The Platonist in me suspects that these are different things, though I could probably be induced to change my mind by a persuasive counter-argument.
      - whatshisface 7y ago
        The Platonist idea that mathematical statements correspond to something "out there" in metaphysical space is a great example of an idea that's not falsifiable. However that's not an idea that I would consider part of mathematics.
logicchop 7y ago
The response (from Duhem, Quine and others) is this: Suppose you think A contradicts B. You would surely admit that, there is likely some other C that, if C were true, A wouldn't contradict B. But if that's the "logic" of science then it isn't straightforward deduction. If you want an example, consider an ordinary empirical observation. Someone claims the ball is red; you look and you see green. Classic contradiction, right? Except if lighting conditions are wrong, or there is a pane of colored glass between you, or.. The point is that there are additional claims that, if they were true, would remove the "contradicts" relation between the two original claims. But ordinary logical relationships shouldn't work like that.
- whatshisface 7y ago
  That's just the common mistake of incorrectly encoding one's beliefs in logic and then complaining logic doesn't work. If you think there might be a glass pane between you and a colored ball, your true belief is (No glass pane implies green ball), not (green ball). If you don't mind me getting on my soap box, it kind of seems like the recipe for coming up with a renowned counterargument to a technical tool goes like: - Find a technical device that you don't understand. - Misuse it because you don't understand it. - Notice that it isn't working. - Conclude that logic doesn't work, math doesn't work, science doesn't work, the bug is in the compiler, take your pick.
  - logicchop 7y ago
    It's not clear what you mean or are complaining about. Feel free to provide your own example. Just come up with two empirical claims (A and B) such that you think A contradicts B. It should then be relatively easy to come up with a C that, if C were true, A would no longer contradict B. It's obvious once you start laying things out carefully.
    - whatshisface 7y ago
      Here's a specific example. Claim A: "If nothing is wrong with your experiment or analysis, you will observe that the speed of light in a vacuum is the same in all inertial reference frames up to experimental uncertainty." Claim B: "If nothing is wrong with your experiment or analysis, you will observe that the speed of light in a vacuum is different in every inertial reference frame up to experimental uncertainty." Every experimentalist realizes that their real result is "I didn't make a mistake -> the answer is X," not "the answer is X." That's the correct way to encode an experiment in logic, leaving aside the high sophistication of how "I didn't make a mistake" gets treated in practice.
yters 7y ago
There could be experiments that confirm the claim even if none contradict it. For example, the immaterial man cannot be detected in any empirical way, and there is no experiment that can falsify his existence. However, he can choose to materialize in front of everyone, confirming his existence.
- pessimizer 7y ago
  > However, he can choose to materialize in front of everyone, confirming his existence. Not confirming his existence as an immaterial man, unless you're willing to take his word for the fact that he was recently immaterial. If I appear in your country and tell you that I'm from the country of "Balanava," I'm not a confirmation of the claim that the country of "Balanava" exists. At best I'm a support.
  - yters 7y ago
    His immateriality is immaterial :) The point is he exists and his existence cannot be falsified (while immaterial) but his existence can still be confirmed. Let's call him person X. So, the statement "person X exists" is empirically meaningful even though it cannot be falsified.

lisper 7y ago

This article attacks a straw man. Being falsifiable is necessary but it is not sufficient to be considered a scientific hypothesis. Such a hypothesis also, and more importantly, has to provide a better explanation of some phenomenon than the current best theory. Experimental evidence is only brought to bear to decide among plausible alternative theories after the vast majority of candidate theories have been eliminated for not providing good explanations. It's easy to see that this must be true because we can only ever have a finite amount of data, and that will always be consistent with an infinite number of falsifiable theories (c.f. Russell's teapot). So data cannot possibly help us choose from among those.

logicchop 7y ago
Popper heavily criticizes the value of explanation. He gives examples of really great "explanations" that have almost zero empirical value (things like psychoanalysis). Hard to see how it could be a "straw man."
- akvadrako 7y ago
  I don’t think that’s true. One of the most prominent figures to continue developing popper’s philosophy is David Deutsch, and he emphasizes good explanations. Being a good explanation means, among other things, it has a wide range of applicability and is hard to adjust if predictions don’t work out. Psychoanalysis wasn’t a good answer.
  - logicchop 7y ago
    It's easy to go read Popper. Look for his discussions of Adler.
- vixen99 7y ago
  But as we know, these examples are not 'great', they're lousy explanations because they are not testable. Good explanations are another matter altogether as Deutsch points out in his criticism of the instrumentalist view of quantum theory.
- 7y ago

nerdponx 7y ago

What is the point of this article? The content is interesting: falsification is more complicated than it seems at first, and scientists don't practice philosophically-pure falsificationism. But it's wrapped up in some kind of finger-pointing "the emperor has no clothes!" type of exposé format that doesn't make sense and doesn't lead to a satisfying conclusion.

dang 7y ago
The article explains clearly and at length why the author wanted to write it. But the best thing you say here is "The content is interesting". That's why it belongs on HN! https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html
- nerdponx 7y ago
  Sorry I was unclear. My comments were meant for the author, not the poster.
  - dang 7y ago
    Yes, and he goes into that at length in the article. The HN thing was just my addition; shop talk basically.

viburnum 7y ago

Well that is definitely not the title I submitted.

dang 7y ago
The site guidelines call for using the original title unless it is misleading or linkbait (https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html). Judging by the first couple of comments that appeared here, the title is baity, because it was generating shallow reactions—to the top-level dramatic claim, rather than to the rich content in the article, such as its final paragraph. When the comments in a thread are generated from the most-obvious, already-familiar information in an article—which often shows up in the title—than we get a discussion of people reciting things they already know or believe (not interesting!), instead of a learning process (much more interesting). I changed the title to "There are strong arguments against falsificationism", which is a phrase from the article body. (When we change titles, we always look for representative language in the article itself.) A couple minutes later, I wasn't sure, so reverted it. This is a bit of an edge case. To some extent, what to do with the title depends on which direction the thread takes: shallow "I know better" dismissals? or thoughtful responses to the less obvious?

outlace 7y ago

This article attacks falfsificationism in part on the basis that scientists don’t seem to actually behave according to “naive falsficationism” in which you supposedly should immediately throw out a theory as soon as you get some conflicting empirical data. The article notes that collecting empirical data to falsify a theory depends on auxiliary theories (eg testing the idea that smoke causes cancer in rats depends on our ability to precisely define and identify when cancer is present) and so most scientists will question these auxiliaries rather than the main theory if conflicting data comes out. This is presented as evidence that scientists don’t actually practice falsficationism. This whole discussion completely ignores uncertainty and the fact that theories describe not only relationships between variables but also involve parameters that need to be estimated. If I fill a rat cage with smoke and the rats don’t get cancer, I don’t immediately assume all previous studies were wrong. This isn’t ignoring my duty to falsify, it’s realizing that the theory that smoking causes cancer describes a causal relationship between variables (smoking -> cancer) it also implicitly or explicitly depends on parameters (in this case, how much and how long a rat must be exposed to smoke to cause cancer at some incidence rate). Given that many other studies found the causal relationship, it is more likely that my experiment messed up the parameters of the causal model (eg I didn’t administer smoke long enough) rather than messed up the causal relationship. And parameter estimation is subject to uncertainty. Now if I’m absolutely sure I’m getting all my model parameters correct, and I can only be confident if other people get the same results. Then I’m moving toward falsifying the theory. The point is that science involves positing theories, and theories are descriptions of causal relationships and generally involve parameters. Much of science is just estimating model parameters given that we assume some model to be tentatively true. You assume your model to be tentatively true, eg smoking causes cancer, now you need to estimate the effect size which is a parameter problem.

noego 7y ago

The author makes some good points, but does not propose his own criteria for what distinguishes a good theory from a bad theory, which seems like a cop-out to be frank. I imagine a big appeal of falsifiability in its early days, stemmed from its opposition to religious dogma. Imagine dealing with someone whose answer to every question comes back to God. "Why did the patient die?" "Because God decreed it." Talking about germ-theory must have been a massive improvement over theology, hence why ideas like falsifiability must have arisen to provoke people to think in more critical ways. My own standard for evaluating theories comes down to the following: The simplest explanation that best fits existing data, and makes useful predictions for the future. "Simplest" because we need to prevent overfitting. "Fits existing data" because the whole point of a theory is to explain what we see around us. And "useful predictions" because it will actually provide practical benefit to us. As opposed to a theory like "God decreed it", which may be extremely simple and error-free, but provides us with no practical benefit at all. As I see it, falsifiability is a side-effect of the second and third requirements. If a theory is making useful predictions, we might as well check to see if it's accurate. And if it's not, there's no point in holding on to a theory that's making faulty predictions. Falsifiability, in this perspective, isn't a requirement in and of itself - it's a natural consequence of the requirement that theories be useful and fit any data at hand.

thethirdone 7y ago
> The simplest explanation that best fits existing data, and makes useful predictions for the future. My main issue with that definition is what makes it a useful prediction. Does a theory (for particle physics) that predicts the Higgs Boson more useful than one that didn't? If not, then before we collected the data to reject other theories they all would have been equal. I would rather suggest that the properties of a good theory are best fitting existing data and having specific predictions for future measurements. > Falsifiability, in this perspective, isn't a requirement in and of itself - it's a natural consequence of the requirement that theories be useful. If you are of the opinion that any good theory is falsifiable (in the sense described in the article), then presumably you disagree with the rejection of falsifiability in the article. Where would you say the error in logic in the article is?

thethirdone 7y ago

> This means that a theory is never falsifiable simpliciter, but only relative to a set of background assumptions. Therefore, if we say that a theory is only scientific if it’s falsifiable, then it follows that no theory, not even a theory as successful as Newton’s law of universal gravitation, is scientific. Of course, this is absurd, so falsificationism is false. You can consider the union of the theory and background assumptions to be a single theory. The premises do not imply that no theory is scientific. An analogy to mathematics is pretty apt. The Riemann hypothesis (RH) is falsifiable in the crude sense described, but cannot be proven or disproven without an axiomatic basis. If you were able to derive a contradiction from RH and some other accepted theorems, you would think that falsifies RH. However, if you wanted a more concrete picture of what was proved, you might reduce the other theorems to Zermelo–Fraenkel set theory (ZF) so you have the result that RH + ZF is inconsistent/false. Maybe in some other axiomatization RH is true, but the general community accepts ZF so that means RH + ZF is false effectively means RH is false. I do agree with some of the sentiment of the article. There should be thought about how to actually test what we say we are testing (not having a reliance on background assumptions) and to accept negative results rather than just find something to blame.

RyanAF7 7y ago

This article is a garbled mush of ego entrism. Appeals to authority out the ass with no real argument.

dang 7y ago
Maybe so, but please don't post unsubstantive comments here. https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html
- RyanAF7 7y ago
  I would argue that stating the flaws of an article is substative. Seems you just don't like how I said it.
  - dang 7y ago
    Ok, I see what you mean. Let me try explaining this a different way. The site guidelines ask you not to call names in comments. Your comment was full of name-calling, in the sense that they use the term. Could you please review them and take the spirit of this site more to heart? https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html A comment like that one encourages the worst in others, and we're trying to avoid sinking into a downward spiral here. Downward spirals are the default outcome on the internet, so avoiding them takes energy. We need everyone to contribute. Here's another way to look at this. Name-calling adds noise to a comment (and encourages others to add still more noise), which lowers the signal/noise ratio. As the signal/noise ratio gets lower, a community like HN deteriorates. This process happens faster than linearly: thoughtful users leave because of the noise, causing the posts to get even worse, driving even more thoughtful users away. That would be a scorched-earth outcome, i.e. an extinction event for HN. Our job is to avoid that. A much better way to post would be to raise the signal/noise ratio. "Signal" here means adding new information, information we can learn from. In the case of your comment, that would mean pointing out more clearly, in more detail and more neutrally, the flaws you see in the article. If you can't or don't want to do that, or you just think the article is so bad it doesn't deserve the effort, then please don't post to the thread at all—instead, flag the article as one that you think doesn't belong on HN, and move on. There's a karma threshold of 30 before you get to flag, but it should only take a few good submissions or comments to get there.