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.
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?
>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.
> 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.
> 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.
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.
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.
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.
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.
That's a good point, though I don't think you have to be a Platonist to think that a claim about a proposition is not the same as that proposition.
It occurred to me this morning that one idea, that is looked on with interest by scientifically- and technically-minded people, is that we're living in a simulation. While there are ways that the truth of this might become apparent, I am not sure it is falsifiable, and that might be good reason to ignore it, but it is not simply a meaningless idea.
When you think of a simulation, you might imagine a programmer, some server racks, and maybe a broader purpose. However you also have to include the possibility that the simulation popped into existence out of nothing, that the simulation is mathematically inevitable, that God created the simulation, that the simulation was created by beings that live in simulation... In so many words, the idea that you live in a simulation leaves as many options open as the idea that you live in the universe. You don't get any claims about how the simulation got there, how it behaves, or even what it is
Insofar as your objections are meaningful (in particular, I am not so sure about the last), it would seem they could equally be raised about non-simulation physical theories of the universe. Would that make those meaningless?
I would think that, on the contrary, the fact that you can raise these objections indicates that the simulation hypothesis is not without meaning, as does Scott Aaronson's argument that it is unfalsifiable[1] (note: he is not saying that that it is meaningless), and as does Bostrom's argument that if it is possible, we are likely in a simulation[2].
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.
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.
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.
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.
There's always something between you and the ball. The era of completely defined and easily observable experiments is long gone. Today we are always composing inherited theories.
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.
> 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.
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.
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."
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.
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.
Which is rather ironic, considering how long he continued to oppose the big-bang theory even after its successful prediction of background radiation (and the explanation of isotope ratios is as good as a prediction, given that the processes are no longer happening.) I'm wondering if he made that statement specifically to avoid these issues?
I heard Fred Hoyle when he came to lecture at my school in the 80s (a prestigious, rich boys school). It was one of the few places willing to give him the time as his views were vaguely in line with Christian creationism though he was an atheist, ironically. It was doubly ironic in that the school was Catholic and Catholic Christianity is not biblically literalist. In fact the inventor of the Big Bang theory eventually became a Catholic priest. hahaha!
> Popper heavily criticizes the value of explanation
You have misunderstood Popper. Psychoanalysis is an example he cites of a bad explanation (because it's not falsifiable). He doesn't criticize explanation in general. That would be throwing out the baby with the bathwater.
Furthermore, if I have understood Popper's criticism of psychoanalysis, he thought it was treating these bad or unsupported 'explanations' / just-so stories as if they provided, or could substitute for, real evidence.
"[Popper] discovered the psychoanalytic theories of Freud and Adler (he served briefly as a voluntary social worker with deprived children in one of the latter’s clinics in the 1920s), and listened entranced to a lecture which Einstein gave in Vienna on relativity theory. The dominance of the critical spirit in Einstein, and its total absence in Marx, Freud and Adler, struck Popper as being of fundamental importance: the pioneers of psychoanalysis, he came to think, couched their theories in terms which made them amenable only to confirmation, while Einstein’s theory, crucially, had testable implications which, if false, would have falsified the theory itself."
>he dominance of the critical spirit in Einstein, and its total absence in Marx, Freud and Adler
What's ironic about this is not only that Marx and Freud inspired the entire school of critical theory today, but that Marx, Freud and Nietzche are the "big three" when it comes to the culture of suspicion and criticism. I'm not sure how Popper could claim the lack of the critical spirit in Marx when Marx's major thesis was a critique of political economy and capitalism which had not been attempted an nearly such a deep and rigorous level until Marx. Successive Marxists after Marx have been very critical of Marxism too. It makes me wonder whether Popper ever read Marx at all when he thinks like this.
And for all the talk of the lack of critical spirit, it's odd that scientists dogmatically cling to the theory of falsifictaionism, even in Popper's sufficient and necessary criterion.
> Marx and Freud inspired the entire school of critical theory today
You have focused too much on the word "critical". It's not enough to be critical. You have to be critical in the right way, and in particular, you have to be critical of (and only of) inadequate explanations. The word you should be focused on is "explanation", not "criticism" (though both are important).
> dogmatically cling to the theory of falsifictaionism
That's not a dogma, that's the definition of science. Following that "dogma" is what makes science effective.
But as to what this "right way" is has been at the center of the defence of critical theory, see for instance the positivism disputes and the accusations of reductionism from both sides. To say that Marx wasn't critical of inadequate explanations is false, since again his whole project was to criticize the ad-hoc solutions of the political economists (such as their use of the labour theory of value in which some goods were conveniently fixed in value to explain how wages for the same work can vary). You might think that Marx's reflections on his own explanations are more lackluster, but at least he claimed to be accepting criticism[0] and to say his own explanations were wrong is different from saying that his mode of explanation is wrong. Marx claimed that his results had more explanatory power than those previous theories. To say that Freud wasn't critical is similarly inaccurate, since before Freud the explanations offered were poor (especially when one considers that Freud practically founded the discipline of the study of the mind). It's a little known fact that the same standard Popper attached to Freud's theories, of a "metaphysical research programme" he also attached to Darwin's. Even if it isn't science, the claim that it is unfalsifiable has met considerable resistance[2].
>That's not a dogma, that's the definition of science.
As the article points out there have been many definitions of science and ways of going about it; consider, for example, the German philosophers who contended that science was rigorous and systematic study. To say that a particular (Popperian) vision of science ought to claim authority over science is ironically a claim lacking in much explanation itself. In any respect, the dogmatism is when scientists use a method which has faced sustained criticism from philosophers of science without paying heed to the criticisms. The issue actually seems to be political[1].
[0] "Every opinion based on scientific criticism I welcome. As to prejudices of so-called public opinion, to which I have never made concessions, now as aforetime the maxim of the great Florentine is mine: Segui il tuo corso, e lascia dir le genti."
[2] "Popper’s charge about the unfalsifiability of key Freudian claims is anyhow highly questionable. He himself focused on Freud’s claim that all dreams are wish-fulfillments, arguing that one can always find grounds for maintaining such a claim but one cannot specify what evidence would falsify it. This is incorrect, as I have argued at greater length elsewhere (Michael 2015). There are numerous logically possible observation statements that would falsify Freud’s wish-fulfillment claim (at least, if one accepts certain background theories, as Popper’s theory requires). For example, if it were found that the limbic system and associated neural structures essentially involved in motivation were (even relatively) inactive in the dreaming brain, that would falsify Freud’s wish-fulfillment claim in the sense required by Popper’s theory. Similarly, if it were found that no or few young children (ages 2–4) had straightforwardly wish-fulfilling dreams, that would falsify Freud’s claim in Popper’s sense." https://link.springer.com/article/10.1007/s11406-018-0020-8#...
> To say that Marx wasn't critical of inadequate explanations is false
You'll have to take that up with Popper, not me.
> science was rigorous and systematic study
That's ridiculous on its face: theological study can be rigorous and systematic, but it's clearly not science.
> To say that a particular (Popperian) vision of science ought to claim authority over science is ironically a claim lacking in much explanation itself.
No, that's not true. One of the reasons we can have confidence that Popper's version is correct is that we can apply it to itself. It is actually possible to explain why explanations are so powerful in terms of concepts from computability and complexity theory. The details are too long for an HN comment, but if you want to find out more read David Deutsch's "The Fabric of Reality", particularly chapter 6.
Fair enough, it seems Popper's disillusionment while working with Adler was the start of his insight into falsifiability. To be clear, I agree that Popper does not seem to be anti-explanation in principle (and if he was, we are under no obligation to agree.)
Yes, that's a very important point. What matters is the idea, not who originated it (except insofar as it's good social policy to properly credit the originator of an idea because it helps motivate people to come up with original ideas).
"It’s the crude version that most people, especially scientists, have in mind when they talk about falsificationism. I will be satisfied if all I have achieved with this post is to convince you that, at least in this crude version, falsificationism is false."
It isn't a straw man if he's right that the crude version is what most people believe.
Even the author calls it "naive falsificationism". That's the strawman. The author is right that it's a crude version of falsificationism, but I doubt that most people believe it.
I suspect most people believe the version that the author calls reasonable: first confirm the experiment and its conclusions before questioning a proven theory.
In that case you and the author have an empirical disagreement about how many people believe which version. Given how often "naive falsificationism" shows up in internet comments, I wouldn't bet against the author. But isn't it moot anyhow? He states up front that not everyone believes this, but those who do are the ones he's writing for. That's not a straw man argument.
"Naive falsificationism" shows up in Internet comments from the kind of people who think vaccines cause autism or global warming is a hoax, and cling to a single isolated result as "proof". Even on the Internet they aren't a majority.
If the author merely wants to argue against that kind of thinking, that's fine, but given the extensive discussion of Karl Popper and physicists, I don't think that's the author's intent.
You should come to my workplace and tell this to actual scientists, who seem to have no trouble hypothesizing things they are unable to falsify, who pick up methods because "other people are doing this", not because of some underlying epistemological criterion, and whose epistemologically broken hypotheses have no trouble getting published in top-flight journals.
I also object to you saying "better explanation" without defining what that is. What makes something "better"? It's more intellectually satisfying? It feels "right"? It has a lower p-value? The ultimate subjectivity of this criterion, which is really at the root of all scientific endeavor, is why philosophy of science remains so difficult to pin down.
In the end what makes up scientific knowledge is that a bunch of people we call "scientists" approve of it, and pass it on in their writing and speech. Philosophers of science are welcome to come up with theories about how science operates, but I suggest they might benefit from some anthropological approaches, because scientists are sure not reading and applying their philosophy books.
This. And that’s always what these criticisms miss. Science’s strength is that any one individual or even groups of individuals can be bias. That bias might persist (theories of static universe, vitalism, etc etc) for a long time within the scientific community.
But science is the only method that eventually will rut out the biases. It’s self correcting, but you need to look at it from the long view.
That's generally right. However, a non-trivial number of well-educated, scientifically minded persons do not take a long enough view. A possible explanation: it is difficult for humans to care about research timelines that may or may not be longer than the average human life. Maybe it's some sociological theory X, and you think, "well, yeah, X has some kinks, but they should be worked out within the next decade or so." In reality, you should essentially have no attachment to said theory. And in reality, far too many people have attachments to one or more theories that have yet to be proved out fully, and could be proved out fully in a decade/century/etc. or worse... never proved out at all.
In the case of currently 'proved out' (whatever we take that to mean) theories, a non-trivial number of persons have excessive attachment to one or more of them, since, taking a long view, said theories could be overturned or reinterpreted, etc.
>But science is the only method that eventually will rut out the biases. It’s self correcting, but you need to look at it from the long view.
This is just dogmatic nonsense. Absent a mechanism, how does this vaunted self-correction operate? If the individuals that make up the scientific community can't be trusted to be free of bias, and that bias can persist for generations, what, in point of fact, is the mechanism for rejecting it?
The fact is that insight does not emerge from any particular method, and science really has no foolproof way to advance inexorably towards the truth. This is why scientific advancement is frequently stalled - because it isn't fucking foolproof. You can claim otherwise, but you're at odds with both any kind of theoretical basis and with the observed history of science.
> This is just dogmatic nonsense. Absent a mechanism, how does this vaunted self-correction operate?
There is indeed a mechanism. All it takes is a single individual that lacks the bias to develop a competing hypotheses that can be tested through experiment. A single person can take down years of group bias. The reason this is possible is because the scientific method requires evidence and the ability to falsify competing hypothesis. If for instance you went against the current “bias” that now takes Darwinian evolution for a given, and proved your hypothesis with data, then the scientific community would be forced (and indeed would honestly be delighted) to embrace the new understanding.
> This is why scientific advancement is frequently stalled - because it isn't fucking foolproof. You can claim otherwise, but you're at odds with both any kind of theoretical basis and with the observed history of science.
History shows an inexorable march forward in understanding the universe brought about via the scientific method. At no point in the history of the scientific method has this been reversed. It may be true that advancement stalls out in certain areas, but that’s to be expected. No other method has ever shown anything close to the scientific methods ability to understand how the universe actually functions.
Any given hypothesis, any given experiment, any given consensus is not “foolproof”. But the scientific method over the long run is indeed “foolproof”, because ultimately data and it’s ability to falsify hypothesis will always overturn mistakes.
“I have steadily endeavoured to keep my mind free so as to give up any hypothesis, however much beloved, as soon as facts are shown to be opposed to it.” — Charles Darwin
>History shows an inexorable march forward in understanding the universe brought about via the scientific method. At no point in the history of the scientific method has this been reversed.
Again, this is just dogma. Science is frequently wrong and sometimes remains wrong for decades. You can say "well eventually someone will correct it, and we'll be open to that possibility", but that is JUST DOGMA. It is not incontrovertible or inexorable.
As for "at no point in history has this been reversed", let's take a few simple examples. One good one is the recent serotonin theory of depression, which enjoyed a good run and racked up a few billion in drug sales before evidence began to accumulate that it is wrong. That is a reversal; people introduced a faulty theory and proceeded with it for decades. Even if we accept that Kirsch and others are overturning this, it was STILL WRONG for a long time.
Or let's take another prominent recent example: fat. For decades we were told that eating fat is bad for you and causes obesity, but sugar was okay. This is all garbage science that persisted because (1) people were incentivized to look the other way and (2) the epistemology was weak.
Or let's take decades of bad science on female physiology that persisted merely because science was controlled by men who were uninterested in women's health. This didn't change because of some self-correcting mechanism in science, it changed because feminism resulted in women taking up the role.
>No other method has ever shown anything close to the scientific methods ability to understand how the universe actually functions.
This is not the question under issue, and obviously if I really felt otherwise I wouldn't be doing science every day. But I don't believe there are foolproof 'scientific methods' that give us an ability to understand anything, there is just our commitment to observation and examining evidence. That is it. We do NOT have a foolproof epistemology.
Where this becomes especially important is where our epistemology is weakest. Nutrition and psychology are good candidates because (1) what we are trying to examine is poorly understood in the first place and (2) we lack good methodologies for dissecting the underlying questions.
In general biology is not as tractable as physics, which is governed by fundamental physical laws that can be modeled using a few parameters. Biological complexity is much greater and can't be attacked using the same methods - you can't do thought-experiments to figure out how ribosomes operate. The mathematical tools required are mostly statistical and are often inadequate to model the complexity we observe. This means our epistemology just fails most of the time - without the necessary tools we can't generate the insights required, nor can we set up experiments to test the propositions we wish to make.
> ultimately data and it’s ability to falsify hypothesis will always overturn mistakes.
Data has no abilities, and falsification is only possible if you're able to create hypotheses in the first place. Creating a hypothesis requires insight; insight requires human understanding. If humans are unable to grasp the basic concepts at work (e.g., how cytokine cascades modulate immune responses) because they are too complex and are not easily amenable to controlled experiment, we can't do "science" to move ahead on these problems.
Your examples are headline understandings of science, and not how science actually progresses. Science has an issue with public perception of how it functions, hence when knew data demonstrates that old hypothesis were incomplete or indeed wrong, people like you say “See, the scientific method doesn’t work” when in fact these are beautiful examples of how science progresses.
New understandings of the “serotonin theory of depression” for instance, which is more accurately termed “monoamine hypothesis of depression” does not mean that science doesn’t work.
Science makes the best of what data is available. Monamines like serotonin substantial impact the course of clinical depression. Current understanding via the scientific method reveals that early hypothesis don’t explain everything. This is a refinement. It does not discount the effects of SSRIs it other antidepressants which currently help millions of people. Could we do better once we have a more complete picture of depression. Yes. And the scientific method is how we will get there.
Einstein’s theory of general relativity updated Newton’s laws of physics. Was Newton and the scientific method wrong? No, he simply didn’t have all the data, and the scientific method allowed us to refine those hypothesis.
I urge anyone who has trouble with the scientific method to read Carl Sagan’s fabulous book, Demon Haunted World.
>Your examples are headline understandings of science, and not how science actually progresses.
I do science every day. I have done science for all of my adult life, at the best research institutions in America. I have a very intimate understanding of how science works.
My guess from your writing is that you are NOT a scientist, you're an engineer with a deep technical education. You have learned a lot of science but you don't spend much time actually doing it. In point of fact science is a culture; its practitioners are subject to all sorts of weaknesses, moral and rational, that end up in its method.
Let me give you an example from my field - my current PI invented CTLA4 blockade, a cancer immunotherapy, in 1996. His PhD advisor, Jim Allison, just won the Nobel Prize for this work. Since about 2014 people have become interested in immunotherapy following several successful new therapies, but it was largely ignored since 1996 - why? Because the cancer establishment wasn't interested. People who wished to study the immune component of cancer were shunned or ridiculed, and what is now an entire field was reduced to a rump study of the "tumor microenvironment" that could never receive funding.
Why is it receiving funding now? Because drug companies are interested now that there are successful therapies on the market. The actual scientific evidence, meanwhile, was produced a generation ago.
What this means is that whatever epistemology is at work in "science" is not merely evidence-driven, it's subject to all sorts of human foibles, being done by humans.
Your story is a shining example of the scientific method. It sucks it took so long. But that’s the whole point of science. Those human foibles are unavoidable. But in the end the science won out.
I guess I’m not sure what you are arguing? Science has to contend with human bias. Any method has to. That’s not going away. Isn’t it still the best method we have, short of altering humanity?
I’m always tickled pink at people who like to debate the validity of the Scientific Method online. You do realize that the scientific method is what made your computer and the the internet possible, right? The insights in physics, chemistry, etc. Everything that involves the behavior of electrons in things like semiconductors relies on quantum mechanics.
Yes? Jesus i hate this argument. You realize people were inventing devices long before they had science, right? You realize that making transistors says nothing about your ability to make valid claims about psychology, right? Having understood one principle does not mean you can understand anything.
This argument doesn’t jive with the facts that the scientific method was required for us to have computers. It’s a simple fact, which you can read about. The complexity of solid state physics meant that early electronics had issues that couldn’t be resolved until a complete understanding of quantum mechanics was made.
Also making transistors doesn’t have to say anything about psychology. The field stands on its own. So does the advance of medicine and psychology via the scientific method.
“We can try nearly futile psychoanalytic talk therapy on the schizophrenic patient, or we can give him 300 to 500 milligrams a day of clozapine. The scientific treatments are hundreds or thousands of times more effective than the alternatives.” — Carl Sagan
>“We can try nearly futile psychoanalytic talk therapy on the schizophrenic patient, or we can give him 300 to 500 milligrams a day of clozapine. The scientific treatments are hundreds or thousands of times more effective than the alternatives.” — Carl Sagan
This is a particularly bad example; clozapine is not much of a treatment, it is just a tranquilizer. It doesn't cure schizophrenics, it chemically lobotomizes them. When the tranquilizers stop, psychosis resumes. This work also didn't really come out of scientific studies, it came out of the system of mental institutions, and it isn't really about treating disease, it's about controlling and successfully institutionalizing mental patients.
As for science, schizophrenia has remained stubbornly intractable to it, because science keeps looking for things like brain abnormalities (finding none in decades of searching - other than the stunting that results from long-term use of neuroleptics), since this is what it is able to operate on epistemologically. Meanwhile, outcomes for schizophrenics are actually WORSE in terms of recovery in the developed world than in the developing world, probably because of the use of antipsychotics, discussed here: https://www.psychologytoday.com/us/blog/mad-in-america/20100....
In other words, this is a domain where it is pretty clear that the scientific establishment has bit the big one, and alternative treatments based on things more like talking therapy (e.g. Soteria houses) and understanding schizophrenics as requiring stability, empathy, etc., rather than chemical tranquilizers, are more likely to be promising.
What's especially damaging here is the attitude of superiority, that "science knows best", that it has already understood something better, that whatever science is doing is implicitly better. This sort of arrogance is purely dogmatic, and leads to error. I seriously suggest you abandon it.
First, I appreciate the debate here, and don't take anything personally. Hopefully you feel the same.
So here you cite an article that uses the scientific method to determine if the current medical guidelines for care of schizophrenia could be better. The takeaway from the article is not to abandon science in the care of those with schizophrenia, but to better apply the data that current treatment has provided.
New hypothesis: blanket use of clozapine has worse outcomes
Data: comparing countries with lower use versus higher
Result: skepticism and the scientific method opens new areas of research into the treatment of schizophrenia
So I still don't understand your argument.
It really seems like you are arguing that people make mistakes, they are bias, and they often can take the results of science as some holy word and ignore new data that conflicts with their beliefs
That is all TRUE, and I wouldn't argue against that. But that is a result of poor education of science, even within the scientific community. Being a scientist does not insulate one from anything, even a lack of understanding of how the scientific method should function.
The answer isn't to abandon the scientific method. It's to educate people, scientist or layperson, about what science really provides.
It doesn't provide ultimate truths that are unquestionable. It doesn't result in anything close to perfect understanding that is mistake free. There is no superiority, that "science knows best". Instead science is humble and says the results of the scientific method are ALWAYS open to revision in the face of conflicting data.
As Carl Sagan says, the scientific method only says the following:
"First: There are no sacred truths; all assumptions must be critically examined; arguments from authority are worthless. Second: whatever is inconsistent with the facts must be discarded or revised."
Science, and the results of the scientific method are not sacred truths. If that is your argument, then we agree. Arguments from scientists based on them being scientists (from authority) are worthless. If these biases are encountered within the scientific community (which honestly they must be, since we are all human) the solution is more rigorous application of the scientific method, and better education on how the scientific method functions. Not abandonment of science.
"We accepted the products of science; we rejected its methods." --Carl Sagan
My argument is that basically all of what Sagan says is wrong.
The argument from authority is paramount in science; in fact it is the entire basis of the scientific enterprise. This is why it is important for you to ridicule other ways of knowing, as you did above ("You'll have to use the scientific method to research whether your alternative methods result in better outcomes for patients. Otherwise its just faith or some other nonsense."), implying that there are no other valid or useful epistemologies besides science. Of course this is not true, because the vast majority of people are not scientists, and yet are able to gain useful knowledge about the world without making use of "scientific methods".
You have continually asserted that all it takes to reject incorrect scientific knowledge is for one person to show up with new observations controverting a theory. But this is not true; it is not sufficient for you to do an experiment showing X to be false or Y to be true - because science produces knowledge by consensus. The basic requirement of science, above all else, is that your result must be peer-reviewed. That is it; there are no requirements of method, no statistical thresholds that must be crossed, no rules about the kinds of evidence that are admissible. The entry of your experiment into the domain of science requires you to cross this threshold. Even after that your experiment faces further review and scrutiny from members of your community.
This immediately means there is an important social dimension to doing science, based on a principle of authority: who gets to do peer review? The answer is: those who have been accorded respect by the community of scientists.
What this means is that powerful scientists are gatekeepers on what is considered to be "true" according to science. It is their judgement that determines what is scientifically valid and invalid, above and beyond the evidence. In general we expect that these people are correctly motivated and believe ideologically in the project of science, are good moral actors, scrupulous, etc. - but all of these things must be true in order for science to succeed.
Without this foundation of authority, science has no mechanism for validating evidence, which, unfortunately, is not self-evident. It is up to these people, and their judgment, to make the determination that some hypothetical is consistent or inconsistent with the facts - that is, whether something is true. It has been observed many times that this can hold up the progress of science.
And, to underscore, this is the extent of the scientific method. There are no rules for what the reviewer must do; they are merely handed a paper, told to critique it, and their determinations are handed back for further argumentation. The bounds of the disputation are up to the disputants. What this means is that the actual epistemology (how we decide "what is true") is continually shifting. One might summarize this by saying that the "scientific method" at any given moment consists of the set of instruments (physical or conceptual) that the community of scientists brings to bear on a problem. Right now there is a vogue of meta-analysis papers; meanwhile, multiple-test correction didn't even exist before 1960.
This is just a taste, I'm not really prepared to develop what should be a book-length argument in a HN comment, but it indicates basically what I am saying: there is a lot, lot more to the "scientific method" than falsificationism can present to us, and the sociological dimension of scientific knowledge is equally important to understand, but underappreciated by people like Popper.
Most of my thinking here has been shaped by two books, both seminal works: one is "Leviathan and the Air-Pump" by Steven Shapin and Simon Schaffer; the other is "Against Method" by Paul Feyerabend. I recommend you read them both.
> In other words, this is a domain where it is pretty clear that the scientific establishment has bit the big one, and alternative treatments based on things more like talking therapy (e.g. Soteria houses) and understanding schizophrenics as requiring stability, empathy, etc., rather than chemical tranquilizers, are more likely to be promising.
How will you evaluate if these alternate therapies are effective? You'll have to use the scientific method to research whether your alternative methods result in better outcomes for patients. Otherwise its just faith or some other nonsense.
Also, you're conflating a current treatment (clozapine) with the scientific method. The scientific method does not dictate that treatments need be pharmaceutical. Just that your treatment efficacy, whether its talking, empathy, Chinese herbs, etc etc, is evaluated via facts that can falsify your treatment hypothesis.
Again, science is not a straight line up. It has hiccups, it makes mistakes. But rigorously apply the scientific method (analyse the data of schizophrenia treatment in detail, accept the facts, revise understanding, challenge the current dogma), and eventually you'll come to a better understanding.
Using words like "inexorably towards the truth" makes it sound like you're conflating "scientific truth" (which is a progressively less distorted view of reality facilitated by actual measurements and vigorous competition among competing explanations) with "religious truth" or "philosophical truth" which are much more difficult to pin down.
> You should come to my workplace and tell this to actual scientists
I would be happy to do that if your workplace is not too far away and you want to invite me to speak.
> I also object to you saying "better explanation" without defining what that is. What makes something "better"?
The short version of the answer to that is: a superior explanation is one that explains more phenomena with fewer ad hoc assumptions and free parameters. (There are other characteristics of good explanations. For example, they are hard to change without losing their explanatory power. The classic example of a bad explanation that runs afoul of this criterion is Ptolemaic epicycles: why circles? Why not some other shape? It turns out that any shape can actually be used to construct Ptolemaic epicycles because what Ptolemy actually discovered was Fourier analysis, though he didn't realize it at the time.)
The long version of the answer is: read Popper. Or David Deutsch.
True - it's a rather undergraduate article for a PhD candidate! Nothing new or interesting, just rehashing old complaints about Popper.
I find his other stuff concerning ... he shows a clear history of setting up straw men and picking them apart ... generally implying that "liberals" of one flavour or another are poor mistaken fools.
Just because Lemoine can pick apart a straw man, or a single statement, doesn't mean that there is no racism, nor that we should ignore climate change, nor that blatant dishonesty by politicians is a good or acceptable thing. Perhaps he's aiming for a career as a Republican stooge?
I've seen similar critiques of philosophy of science, and it's usually clear that the critique is just a subterfuge for validating the person's own alternative opinions.
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.
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!
The site guidelines call for using the original title unless it is misleading or linkbait (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?
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.
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.
> 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?
You can't possibly know what will match experimental results in the future so that doesn't help determine good theories in the present.
And I would say that Newtonian gravity was a useful theory despite not matching experimental currently because when it was the best it mostly explained observations.
> Newtonian gravity was a useful theory despite not matching experiment
"Useful" in this context means useful for a future theory or experimentation, not utility. Newtonian theory certainly predicted many results, and all of them were useful in the above sense. A prediction doesn't have to exactly match observations to be useful, but that the theory produces a prediction which _could_ be tested against.
Even flat-earth theory makes useful predictions - albeit already proven false as time and time again, it predicts the wrong results.
What isn't useful is an unfalsifiable theory, which means the predictions it makes is not able to be tested, or the results of such a test could be construed to match if you squint a bit. A theory like creationism, for example.
I am a bit confused by your comment. I'm not sure if you agree with gliop or not. I am simply saying that Newtonian gravity is useful (in pretty much any sense) and that it did not match all experiments. Particularly in the orbit of Mercury could not be explained.
> Even flat-earth theory makes useful predictions - albeit already proven false as time and time again, it predicts the wrong results.
I definitely agree usefulness is independent of fitting experimental data. gliop would seem to say flat earth theories are not useful.
I don't think the value of a theory can be evaluated in isolation. It can only be evaluated relative to its competition. During its time, Newtonian gravity, even if it didn't perfectly fit all data, was still superior to its competition on an aggregate of the 3 criteria mentioned. Only once Einstein's theories were put forward, were we able to replace Newtonian gravity with a superior alternative.
> Suppose that, contrary to what actually happened, Cavendish’s experiment had yielded a result that was qualitatively incompatible with the law of universal gravitation. The reaction of physicists would certainly not have been to reject Newton’s theory of gravitation and start looking for another one.
Looking for another theory is precisely what physicists do, once experimental error, miscalculation, and misunderstanding the theory are ruled out. Hence relativity and quantum physics, and the current search for a theory that explains dark matter and dark energy.
And another:
> From a purely logical point of view, what scientists do when they blame the failure of a prediction on the background assumptions instead of theory they take to be well-established is no different from what charlatans peddling pseudoscience do...
This is arguing that "falsificationism" requires discarding a theory after a single failed experiment, but forgets that even experiments can be falsified (by finding an error in the experiment or the associated conclusions).
The difference between scientists and "charlatans peddling pseudoscience" is that when a prediction proves to be false, scientists look for an explanation.
The author forgets that the explanation might be found in the experiment or the prediction, and scientists look there first before questioning a theory that's proven itself in the past.
> 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.
That's not the purpose of the article, though. The purpose is to show that the most popular purported solution to the problem (naive falsificationism) is so severely limited as to be untenable, and that's really useful in itself. The title is "Why falsificationism is false", not "What should replace falsificationism".
>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.
Only better answers than "Because God decreed it." were known and handled, by philosophers and early scientists, even by thelogicians, since antiquity.
And there were tons of nuanced epistemologies (that included practical falsifiability) by the time Popper came up with his idea already.
> 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.
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?
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.
Sounds like you might have a degree in communications?
Well written.
The comment I made, poorly spelled at that, had no name calling in it.
Saying something is filled with ego centrism* or is a garbled mush* is no different than pointing out the overuse of the word "I" in a text, or the poor argument structure of a faulty conclusion.
It lends weight and contributes to the discussion by describing the perspective and feel of the piece.
Again, it seems as if you don't like how I said something because it lacked respect. That, I can agree with and in the future reconsider how something is phrased.
Please, consider the brevity of my comment and it's efficient conveyance of my opinion on the piece versus something like this thread; overly detailed and descriptive for the single purpose of reconciling differing opinions on how one should comment on posts.
Appreciate the input, and the earnestness with which you address the want for civility.
90 comments
[ 3.1 ms ] story [ 125 ms ] thread1. '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?
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.
> 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.
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.
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.
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.
It occurred to me this morning that one idea, that is looked on with interest by scientifically- and technically-minded people, is that we're living in a simulation. While there are ways that the truth of this might become apparent, I am not sure it is falsifiable, and that might be good reason to ignore it, but it is not simply a meaningless idea.
When you think of a simulation, you might imagine a programmer, some server racks, and maybe a broader purpose. However you also have to include the possibility that the simulation popped into existence out of nothing, that the simulation is mathematically inevitable, that God created the simulation, that the simulation was created by beings that live in simulation... In so many words, the idea that you live in a simulation leaves as many options open as the idea that you live in the universe. You don't get any claims about how the simulation got there, how it behaves, or even what it is
Maybe it is a meaningless idea.
I would think that, on the contrary, the fact that you can raise these objections indicates that the simulation hypothesis is not without meaning, as does Scott Aaronson's argument that it is unfalsifiable[1] (note: he is not saying that that it is meaningless), and as does Bostrom's argument that if it is possible, we are likely in a simulation[2].
[1] https://www.scottaaronson.com/blog/?p=3482
[2] https://www.simulation-argument.com/simulation.html
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.
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.
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.
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.
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.
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.
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.
You have misunderstood Popper. Psychoanalysis is an example he cites of a bad explanation (because it's not falsifiable). He doesn't criticize explanation in general. That would be throwing out the baby with the bathwater.
"[Popper] discovered the psychoanalytic theories of Freud and Adler (he served briefly as a voluntary social worker with deprived children in one of the latter’s clinics in the 1920s), and listened entranced to a lecture which Einstein gave in Vienna on relativity theory. The dominance of the critical spirit in Einstein, and its total absence in Marx, Freud and Adler, struck Popper as being of fundamental importance: the pioneers of psychoanalysis, he came to think, couched their theories in terms which made them amenable only to confirmation, while Einstein’s theory, crucially, had testable implications which, if false, would have falsified the theory itself."
What's ironic about this is not only that Marx and Freud inspired the entire school of critical theory today, but that Marx, Freud and Nietzche are the "big three" when it comes to the culture of suspicion and criticism. I'm not sure how Popper could claim the lack of the critical spirit in Marx when Marx's major thesis was a critique of political economy and capitalism which had not been attempted an nearly such a deep and rigorous level until Marx. Successive Marxists after Marx have been very critical of Marxism too. It makes me wonder whether Popper ever read Marx at all when he thinks like this.
And for all the talk of the lack of critical spirit, it's odd that scientists dogmatically cling to the theory of falsifictaionism, even in Popper's sufficient and necessary criterion.
You have focused too much on the word "critical". It's not enough to be critical. You have to be critical in the right way, and in particular, you have to be critical of (and only of) inadequate explanations. The word you should be focused on is "explanation", not "criticism" (though both are important).
> dogmatically cling to the theory of falsifictaionism
That's not a dogma, that's the definition of science. Following that "dogma" is what makes science effective.
But as to what this "right way" is has been at the center of the defence of critical theory, see for instance the positivism disputes and the accusations of reductionism from both sides. To say that Marx wasn't critical of inadequate explanations is false, since again his whole project was to criticize the ad-hoc solutions of the political economists (such as their use of the labour theory of value in which some goods were conveniently fixed in value to explain how wages for the same work can vary). You might think that Marx's reflections on his own explanations are more lackluster, but at least he claimed to be accepting criticism[0] and to say his own explanations were wrong is different from saying that his mode of explanation is wrong. Marx claimed that his results had more explanatory power than those previous theories. To say that Freud wasn't critical is similarly inaccurate, since before Freud the explanations offered were poor (especially when one considers that Freud practically founded the discipline of the study of the mind). It's a little known fact that the same standard Popper attached to Freud's theories, of a "metaphysical research programme" he also attached to Darwin's. Even if it isn't science, the claim that it is unfalsifiable has met considerable resistance[2].
>That's not a dogma, that's the definition of science.
As the article points out there have been many definitions of science and ways of going about it; consider, for example, the German philosophers who contended that science was rigorous and systematic study. To say that a particular (Popperian) vision of science ought to claim authority over science is ironically a claim lacking in much explanation itself. In any respect, the dogmatism is when scientists use a method which has faced sustained criticism from philosophers of science without paying heed to the criticisms. The issue actually seems to be political[1].
[0] "Every opinion based on scientific criticism I welcome. As to prejudices of so-called public opinion, to which I have never made concessions, now as aforetime the maxim of the great Florentine is mine: Segui il tuo corso, e lascia dir le genti."
[1] https://www.jstor.org/stable/4224749?seq=1#page_scan_tab_con...
[2] "Popper’s charge about the unfalsifiability of key Freudian claims is anyhow highly questionable. He himself focused on Freud’s claim that all dreams are wish-fulfillments, arguing that one can always find grounds for maintaining such a claim but one cannot specify what evidence would falsify it. This is incorrect, as I have argued at greater length elsewhere (Michael 2015). There are numerous logically possible observation statements that would falsify Freud’s wish-fulfillment claim (at least, if one accepts certain background theories, as Popper’s theory requires). For example, if it were found that the limbic system and associated neural structures essentially involved in motivation were (even relatively) inactive in the dreaming brain, that would falsify Freud’s wish-fulfillment claim in the sense required by Popper’s theory. Similarly, if it were found that no or few young children (ages 2–4) had straightforwardly wish-fulfilling dreams, that would falsify Freud’s claim in Popper’s sense." https://link.springer.com/article/10.1007/s11406-018-0020-8#...
You'll have to take that up with Popper, not me.
> science was rigorous and systematic study
That's ridiculous on its face: theological study can be rigorous and systematic, but it's clearly not science.
> To say that a particular (Popperian) vision of science ought to claim authority over science is ironically a claim lacking in much explanation itself.
No, that's not true. One of the reasons we can have confidence that Popper's version is correct is that we can apply it to itself. It is actually possible to explain why explanations are so powerful in terms of concepts from computability and complexity theory. The details are too long for an HN comment, but if you want to find out more read David Deutsch's "The Fabric of Reality", particularly chapter 6.
Yes, that's a very important point. What matters is the idea, not who originated it (except insofar as it's good social policy to properly credit the originator of an idea because it helps motivate people to come up with original ideas).
"It’s the crude version that most people, especially scientists, have in mind when they talk about falsificationism. I will be satisfied if all I have achieved with this post is to convince you that, at least in this crude version, falsificationism is false."
It isn't a straw man if he's right that the crude version is what most people believe.
No, he doesn't. The word "explanation" doesn't even appear in the article.
I suspect most people believe the version that the author calls reasonable: first confirm the experiment and its conclusions before questioning a proven theory.
If the author merely wants to argue against that kind of thinking, that's fine, but given the extensive discussion of Karl Popper and physicists, I don't think that's the author's intent.
I also object to you saying "better explanation" without defining what that is. What makes something "better"? It's more intellectually satisfying? It feels "right"? It has a lower p-value? The ultimate subjectivity of this criterion, which is really at the root of all scientific endeavor, is why philosophy of science remains so difficult to pin down.
In the end what makes up scientific knowledge is that a bunch of people we call "scientists" approve of it, and pass it on in their writing and speech. Philosophers of science are welcome to come up with theories about how science operates, but I suggest they might benefit from some anthropological approaches, because scientists are sure not reading and applying their philosophy books.
Some might not be doing science at all, even though it says so on the building and their business cards.
But science is the only method that eventually will rut out the biases. It’s self correcting, but you need to look at it from the long view.
In the case of currently 'proved out' (whatever we take that to mean) theories, a non-trivial number of persons have excessive attachment to one or more of them, since, taking a long view, said theories could be overturned or reinterpreted, etc.
This is just dogmatic nonsense. Absent a mechanism, how does this vaunted self-correction operate? If the individuals that make up the scientific community can't be trusted to be free of bias, and that bias can persist for generations, what, in point of fact, is the mechanism for rejecting it?
The fact is that insight does not emerge from any particular method, and science really has no foolproof way to advance inexorably towards the truth. This is why scientific advancement is frequently stalled - because it isn't fucking foolproof. You can claim otherwise, but you're at odds with both any kind of theoretical basis and with the observed history of science.
There is indeed a mechanism. All it takes is a single individual that lacks the bias to develop a competing hypotheses that can be tested through experiment. A single person can take down years of group bias. The reason this is possible is because the scientific method requires evidence and the ability to falsify competing hypothesis. If for instance you went against the current “bias” that now takes Darwinian evolution for a given, and proved your hypothesis with data, then the scientific community would be forced (and indeed would honestly be delighted) to embrace the new understanding.
> This is why scientific advancement is frequently stalled - because it isn't fucking foolproof. You can claim otherwise, but you're at odds with both any kind of theoretical basis and with the observed history of science.
History shows an inexorable march forward in understanding the universe brought about via the scientific method. At no point in the history of the scientific method has this been reversed. It may be true that advancement stalls out in certain areas, but that’s to be expected. No other method has ever shown anything close to the scientific methods ability to understand how the universe actually functions.
Any given hypothesis, any given experiment, any given consensus is not “foolproof”. But the scientific method over the long run is indeed “foolproof”, because ultimately data and it’s ability to falsify hypothesis will always overturn mistakes.
“I have steadily endeavoured to keep my mind free so as to give up any hypothesis, however much beloved, as soon as facts are shown to be opposed to it.” — Charles Darwin
Again, this is just dogma. Science is frequently wrong and sometimes remains wrong for decades. You can say "well eventually someone will correct it, and we'll be open to that possibility", but that is JUST DOGMA. It is not incontrovertible or inexorable.
As for "at no point in history has this been reversed", let's take a few simple examples. One good one is the recent serotonin theory of depression, which enjoyed a good run and racked up a few billion in drug sales before evidence began to accumulate that it is wrong. That is a reversal; people introduced a faulty theory and proceeded with it for decades. Even if we accept that Kirsch and others are overturning this, it was STILL WRONG for a long time.
Or let's take another prominent recent example: fat. For decades we were told that eating fat is bad for you and causes obesity, but sugar was okay. This is all garbage science that persisted because (1) people were incentivized to look the other way and (2) the epistemology was weak.
Or let's take decades of bad science on female physiology that persisted merely because science was controlled by men who were uninterested in women's health. This didn't change because of some self-correcting mechanism in science, it changed because feminism resulted in women taking up the role.
>No other method has ever shown anything close to the scientific methods ability to understand how the universe actually functions.
This is not the question under issue, and obviously if I really felt otherwise I wouldn't be doing science every day. But I don't believe there are foolproof 'scientific methods' that give us an ability to understand anything, there is just our commitment to observation and examining evidence. That is it. We do NOT have a foolproof epistemology.
Where this becomes especially important is where our epistemology is weakest. Nutrition and psychology are good candidates because (1) what we are trying to examine is poorly understood in the first place and (2) we lack good methodologies for dissecting the underlying questions.
In general biology is not as tractable as physics, which is governed by fundamental physical laws that can be modeled using a few parameters. Biological complexity is much greater and can't be attacked using the same methods - you can't do thought-experiments to figure out how ribosomes operate. The mathematical tools required are mostly statistical and are often inadequate to model the complexity we observe. This means our epistemology just fails most of the time - without the necessary tools we can't generate the insights required, nor can we set up experiments to test the propositions we wish to make.
> ultimately data and it’s ability to falsify hypothesis will always overturn mistakes.
Data has no abilities, and falsification is only possible if you're able to create hypotheses in the first place. Creating a hypothesis requires insight; insight requires human understanding. If humans are unable to grasp the basic concepts at work (e.g., how cytokine cascades modulate immune responses) because they are too complex and are not easily amenable to controlled experiment, we can't do "science" to move ahead on these problems.
New understandings of the “serotonin theory of depression” for instance, which is more accurately termed “monoamine hypothesis of depression” does not mean that science doesn’t work.
Science makes the best of what data is available. Monamines like serotonin substantial impact the course of clinical depression. Current understanding via the scientific method reveals that early hypothesis don’t explain everything. This is a refinement. It does not discount the effects of SSRIs it other antidepressants which currently help millions of people. Could we do better once we have a more complete picture of depression. Yes. And the scientific method is how we will get there.
Einstein’s theory of general relativity updated Newton’s laws of physics. Was Newton and the scientific method wrong? No, he simply didn’t have all the data, and the scientific method allowed us to refine those hypothesis.
I urge anyone who has trouble with the scientific method to read Carl Sagan’s fabulous book, Demon Haunted World.
https://en.wikipedia.org/wiki/The_Demon-Haunted_World
I do science every day. I have done science for all of my adult life, at the best research institutions in America. I have a very intimate understanding of how science works.
My guess from your writing is that you are NOT a scientist, you're an engineer with a deep technical education. You have learned a lot of science but you don't spend much time actually doing it. In point of fact science is a culture; its practitioners are subject to all sorts of weaknesses, moral and rational, that end up in its method.
Let me give you an example from my field - my current PI invented CTLA4 blockade, a cancer immunotherapy, in 1996. His PhD advisor, Jim Allison, just won the Nobel Prize for this work. Since about 2014 people have become interested in immunotherapy following several successful new therapies, but it was largely ignored since 1996 - why? Because the cancer establishment wasn't interested. People who wished to study the immune component of cancer were shunned or ridiculed, and what is now an entire field was reduced to a rump study of the "tumor microenvironment" that could never receive funding.
Why is it receiving funding now? Because drug companies are interested now that there are successful therapies on the market. The actual scientific evidence, meanwhile, was produced a generation ago.
What this means is that whatever epistemology is at work in "science" is not merely evidence-driven, it's subject to all sorts of human foibles, being done by humans.
I guess I’m not sure what you are arguing? Science has to contend with human bias. Any method has to. That’s not going away. Isn’t it still the best method we have, short of altering humanity?
I’m always tickled pink at people who like to debate the validity of the Scientific Method online. You do realize that the scientific method is what made your computer and the the internet possible, right? The insights in physics, chemistry, etc. Everything that involves the behavior of electrons in things like semiconductors relies on quantum mechanics.
https://en.wikipedia.org/w/index.php?title=Semiconductor#Ear...
Also making transistors doesn’t have to say anything about psychology. The field stands on its own. So does the advance of medicine and psychology via the scientific method.
“We can try nearly futile psychoanalytic talk therapy on the schizophrenic patient, or we can give him 300 to 500 milligrams a day of clozapine. The scientific treatments are hundreds or thousands of times more effective than the alternatives.” — Carl Sagan
This is a particularly bad example; clozapine is not much of a treatment, it is just a tranquilizer. It doesn't cure schizophrenics, it chemically lobotomizes them. When the tranquilizers stop, psychosis resumes. This work also didn't really come out of scientific studies, it came out of the system of mental institutions, and it isn't really about treating disease, it's about controlling and successfully institutionalizing mental patients.
As for science, schizophrenia has remained stubbornly intractable to it, because science keeps looking for things like brain abnormalities (finding none in decades of searching - other than the stunting that results from long-term use of neuroleptics), since this is what it is able to operate on epistemologically. Meanwhile, outcomes for schizophrenics are actually WORSE in terms of recovery in the developed world than in the developing world, probably because of the use of antipsychotics, discussed here: https://www.psychologytoday.com/us/blog/mad-in-america/20100....
In other words, this is a domain where it is pretty clear that the scientific establishment has bit the big one, and alternative treatments based on things more like talking therapy (e.g. Soteria houses) and understanding schizophrenics as requiring stability, empathy, etc., rather than chemical tranquilizers, are more likely to be promising.
What's especially damaging here is the attitude of superiority, that "science knows best", that it has already understood something better, that whatever science is doing is implicitly better. This sort of arrogance is purely dogmatic, and leads to error. I seriously suggest you abandon it.
So here you cite an article that uses the scientific method to determine if the current medical guidelines for care of schizophrenia could be better. The takeaway from the article is not to abandon science in the care of those with schizophrenia, but to better apply the data that current treatment has provided.
New hypothesis: blanket use of clozapine has worse outcomes Data: comparing countries with lower use versus higher Result: skepticism and the scientific method opens new areas of research into the treatment of schizophrenia
So I still don't understand your argument.
It really seems like you are arguing that people make mistakes, they are bias, and they often can take the results of science as some holy word and ignore new data that conflicts with their beliefs
That is all TRUE, and I wouldn't argue against that. But that is a result of poor education of science, even within the scientific community. Being a scientist does not insulate one from anything, even a lack of understanding of how the scientific method should function.
The answer isn't to abandon the scientific method. It's to educate people, scientist or layperson, about what science really provides.
It doesn't provide ultimate truths that are unquestionable. It doesn't result in anything close to perfect understanding that is mistake free. There is no superiority, that "science knows best". Instead science is humble and says the results of the scientific method are ALWAYS open to revision in the face of conflicting data.
As Carl Sagan says, the scientific method only says the following:
"First: There are no sacred truths; all assumptions must be critically examined; arguments from authority are worthless. Second: whatever is inconsistent with the facts must be discarded or revised."
Science, and the results of the scientific method are not sacred truths. If that is your argument, then we agree. Arguments from scientists based on them being scientists (from authority) are worthless. If these biases are encountered within the scientific community (which honestly they must be, since we are all human) the solution is more rigorous application of the scientific method, and better education on how the scientific method functions. Not abandonment of science.
"We accepted the products of science; we rejected its methods." --Carl Sagan
The argument from authority is paramount in science; in fact it is the entire basis of the scientific enterprise. This is why it is important for you to ridicule other ways of knowing, as you did above ("You'll have to use the scientific method to research whether your alternative methods result in better outcomes for patients. Otherwise its just faith or some other nonsense."), implying that there are no other valid or useful epistemologies besides science. Of course this is not true, because the vast majority of people are not scientists, and yet are able to gain useful knowledge about the world without making use of "scientific methods".
You have continually asserted that all it takes to reject incorrect scientific knowledge is for one person to show up with new observations controverting a theory. But this is not true; it is not sufficient for you to do an experiment showing X to be false or Y to be true - because science produces knowledge by consensus. The basic requirement of science, above all else, is that your result must be peer-reviewed. That is it; there are no requirements of method, no statistical thresholds that must be crossed, no rules about the kinds of evidence that are admissible. The entry of your experiment into the domain of science requires you to cross this threshold. Even after that your experiment faces further review and scrutiny from members of your community.
This immediately means there is an important social dimension to doing science, based on a principle of authority: who gets to do peer review? The answer is: those who have been accorded respect by the community of scientists.
What this means is that powerful scientists are gatekeepers on what is considered to be "true" according to science. It is their judgement that determines what is scientifically valid and invalid, above and beyond the evidence. In general we expect that these people are correctly motivated and believe ideologically in the project of science, are good moral actors, scrupulous, etc. - but all of these things must be true in order for science to succeed.
Without this foundation of authority, science has no mechanism for validating evidence, which, unfortunately, is not self-evident. It is up to these people, and their judgment, to make the determination that some hypothetical is consistent or inconsistent with the facts - that is, whether something is true. It has been observed many times that this can hold up the progress of science.
And, to underscore, this is the extent of the scientific method. There are no rules for what the reviewer must do; they are merely handed a paper, told to critique it, and their determinations are handed back for further argumentation. The bounds of the disputation are up to the disputants. What this means is that the actual epistemology (how we decide "what is true") is continually shifting. One might summarize this by saying that the "scientific method" at any given moment consists of the set of instruments (physical or conceptual) that the community of scientists brings to bear on a problem. Right now there is a vogue of meta-analysis papers; meanwhile, multiple-test correction didn't even exist before 1960.
This is just a taste, I'm not really prepared to develop what should be a book-length argument in a HN comment, but it indicates basically what I am saying: there is a lot, lot more to the "scientific method" than falsificationism can present to us, and the sociological dimension of scientific knowledge is equally important to understand, but underappreciated by people like Popper.
Most of my thinking here has been shaped by two books, both seminal works: one is "Leviathan and the Air-Pump" by Steven Shapin and Simon Schaffer; the other is "Against Method" by Paul Feyerabend. I recommend you read them both.
How will you evaluate if these alternate therapies are effective? You'll have to use the scientific method to research whether your alternative methods result in better outcomes for patients. Otherwise its just faith or some other nonsense.
Also, you're conflating a current treatment (clozapine) with the scientific method. The scientific method does not dictate that treatments need be pharmaceutical. Just that your treatment efficacy, whether its talking, empathy, Chinese herbs, etc etc, is evaluated via facts that can falsify your treatment hypothesis.
Again, science is not a straight line up. It has hiccups, it makes mistakes. But rigorously apply the scientific method (analyse the data of schizophrenia treatment in detail, accept the facts, revise understanding, challenge the current dogma), and eventually you'll come to a better understanding.
I would be happy to do that if your workplace is not too far away and you want to invite me to speak.
> I also object to you saying "better explanation" without defining what that is. What makes something "better"?
The short version of the answer to that is: a superior explanation is one that explains more phenomena with fewer ad hoc assumptions and free parameters. (There are other characteristics of good explanations. For example, they are hard to change without losing their explanatory power. The classic example of a bad explanation that runs afoul of this criterion is Ptolemaic epicycles: why circles? Why not some other shape? It turns out that any shape can actually be used to construct Ptolemaic epicycles because what Ptolemy actually discovered was Fourier analysis, though he didn't realize it at the time.)
The long version of the answer is: read Popper. Or David Deutsch.
I find his other stuff concerning ... he shows a clear history of setting up straw men and picking them apart ... generally implying that "liberals" of one flavour or another are poor mistaken fools.
Just because Lemoine can pick apart a straw man, or a single statement, doesn't mean that there is no racism, nor that we should ignore climate change, nor that blatant dishonesty by politicians is a good or acceptable thing. Perhaps he's aiming for a career as a Republican stooge?
https://news.ycombinator.com/newsguidelines.html
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?
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.
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.
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?
And I would say that Newtonian gravity was a useful theory despite not matching experimental currently because when it was the best it mostly explained observations.
"Useful" in this context means useful for a future theory or experimentation, not utility. Newtonian theory certainly predicted many results, and all of them were useful in the above sense. A prediction doesn't have to exactly match observations to be useful, but that the theory produces a prediction which _could_ be tested against.
Even flat-earth theory makes useful predictions - albeit already proven false as time and time again, it predicts the wrong results.
What isn't useful is an unfalsifiable theory, which means the predictions it makes is not able to be tested, or the results of such a test could be construed to match if you squint a bit. A theory like creationism, for example.
> Even flat-earth theory makes useful predictions - albeit already proven false as time and time again, it predicts the wrong results.
I definitely agree usefulness is independent of fitting experimental data. gliop would seem to say flat earth theories are not useful.
> Suppose that, contrary to what actually happened, Cavendish’s experiment had yielded a result that was qualitatively incompatible with the law of universal gravitation. The reaction of physicists would certainly not have been to reject Newton’s theory of gravitation and start looking for another one.
Looking for another theory is precisely what physicists do, once experimental error, miscalculation, and misunderstanding the theory are ruled out. Hence relativity and quantum physics, and the current search for a theory that explains dark matter and dark energy.
And another:
> From a purely logical point of view, what scientists do when they blame the failure of a prediction on the background assumptions instead of theory they take to be well-established is no different from what charlatans peddling pseudoscience do...
This is arguing that "falsificationism" requires discarding a theory after a single failed experiment, but forgets that even experiments can be falsified (by finding an error in the experiment or the associated conclusions).
The difference between scientists and "charlatans peddling pseudoscience" is that when a prediction proves to be false, scientists look for an explanation.
The author forgets that the explanation might be found in the experiment or the prediction, and scientists look there first before questioning a theory that's proven itself in the past.
That's not the purpose of the article, though. The purpose is to show that the most popular purported solution to the problem (naive falsificationism) is so severely limited as to be untenable, and that's really useful in itself. The title is "Why falsificationism is false", not "What should replace falsificationism".
Only better answers than "Because God decreed it." were known and handled, by philosophers and early scientists, even by thelogicians, since antiquity.
And there were tons of nuanced epistemologies (that included practical falsifiability) by the time Popper came up with his idea already.
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.
Appeals to authority out the ass with no real argument.
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.
Well written.
The comment I made, poorly spelled at that, had no name calling in it.
Saying something is filled with ego centrism* or is a garbled mush* is no different than pointing out the overuse of the word "I" in a text, or the poor argument structure of a faulty conclusion.
It lends weight and contributes to the discussion by describing the perspective and feel of the piece.
Again, it seems as if you don't like how I said something because it lacked respect. That, I can agree with and in the future reconsider how something is phrased.
Please, consider the brevity of my comment and it's efficient conveyance of my opinion on the piece versus something like this thread; overly detailed and descriptive for the single purpose of reconciling differing opinions on how one should comment on posts.
Appreciate the input, and the earnestness with which you address the want for civility.