7 ms·

> Science is anti-fragile. The harder you attack it, the sturdier it becomes. This is true for the subset of "scientific professionals, devoting significant am

by ethbr0 5y ago

> Science is anti-fragile. The harder you attack it, the sturdier it becomes. This is true for the subset of "scientific professionals, devoting significant amounts of time to their field". That's not all of the public. That's not even all of most scientific disciplines. The scientific method is perfect. The scientific method as practiced by fallible and irrational humans is subject to starts and stops, groupthink and cargo culting, and politics ad nauseam. The even more common case of "laypeople, with limited technical background knowledge, devoting limited amounts of time to understanding issues, who want to appear informed and intelligent about issues" is significantly more fraught with danger. And the source of most ineffective outcomes.

LocalH 5y ago
>The scientific method is perfect. Nonsense. Nothing is perfect, especially nothing ever created by a human. The scientific method is the best we have developed, in my opinion, but it's not perfect. The scientific method itself was developed by "fallible and irrational humans". Even with the best of intentions and the utmost in effort, all humans are susceptible to fallibility and irrationality. You are, I am, everyone is.
- rcoveson 5y ago
  The word/concept "perfect" was "developed" by "fallible and irrational humans" as well, so we do have the authority to apply it. The 3/4/5 right triangle is perfect. The perfect fifth in music is (tautologically) perfect. "p or not p" is perfect. Perhaps none of these things can be reproduced in the physical world in a perfect manner, but the concepts really are perfect. Anything in idea-space can in fact be perfect, and the scientific method lives in idea-space. In the realm of perfection, there isn't really any difference between "developing" something and "discovering" it. Did we "develop" the scientific method or "discover" it? I think that's a meaningless distinction. It's the loop that describes how inductive reasoning is applied by limited observers to a structured universe. People were doing it before it was "developed".
  - angelzen 5y ago
    This is confusing models with reality. It is the application of any specific model to reality that is not perfect.
    - rcoveson 5y ago
      Yes, one might say: > Perhaps none of these things can be reproduced in the physical world in a perfect manner, but the concepts really are perfect.
marginalia_nu 5y ago
The scientific method cannot be separated from the humans that practice it. It doesn't have some separate existence as a platonic ideal. Ideas like these must be judged based on their actual real world performance, and not on the lofty promises of the ideologues that promote the ideas. Luckily science doesn't require perfectly rational practitioners. It still works even if practiced human beings.
marivilla 5y ago
>The scientific method is perfect. >The scientific method as practiced by fallible and irrational humans This is giving me flashbacks to Catholic school :) 'Christ is perfect, his earthly church is not. Because the church consists of fallible human clergy and laity it is vulnerable to corruption, ambition, lust, and politics, oh the politics' It's interesting to know that secular students get the same arguments just in defense of scientism instead of Christianity. All apologists sound alike. Not a criticism, just interesting
- beebmam 5y ago
  Metaphors/analogies aren't an argument, and if you find them persuasive that's your problem
  - rajin444 5y ago
    This isn't really a metaphor or an analogy. Modern scientific leaders and religious leaders both act as liaisons between the public and knowledge. Scientists tend to produce a lot more useful knowledge, but the issues faced by both are the same. It's obvious they both end up developing similar patterns for resolving communication issues. The parent noticed this similarity because both are in the same business.
  - marivilla 5y ago
    I said 'Not a criticism, just interesting' in the hopes of emphasizing that I was making an observation, not an argument. Sorry I failed to communicate that more clearly! That said, there is something to be said for https://en.wikipedia.org/wiki/Argument_from_analogy https://en.wikipedia.org/wiki/Argument_from_analogy
    - ohwellhere 5y ago
      I view your observation as interesting and worth making, and I also believe there's an implicit argument there against that line of thinking. The word "perfect" is problematic to me. It's ascribed to God to contrast with man, as most of us would agree that nothing we do is perfect. Using it implies the same religious status of being above man's foibles. I like the word "useful" here, or "anti-fragile" in the original comment. Further, I like defining context. Useful in what class of problems? For I don't think the scientific method is regarded as particularly useful in all domains. Consciousness and the subjective are valid realms of human experience worth dedicated inquiry, for which the scientific method seems woefully lacking.
iammisc 5y ago
The scientific method is a human invention. It is not 'perfect'. This is ridiculous religious dogmatism.
drbojingle 5y ago
If the scientific method is so perfect why does it fail when used by imperfect operators
- thephyber 5y ago
  Does it? You are making an assertion without any facts. The method itself is simple. It is agnostic to bad faith implementations, to bad measurements, to completely fallacious conclusions, etc. It is intended to be iterative; individual errors wash out in time with enough new experiments. What non-scientists learn from secondary or tertiary sources is irrelevant — those misunderstandings do not mean the Method “failed”.
  - cogman10 5y ago
    Meh, the scientific method is akin to "agile" in software development. It's good principles but really is less a method and more guiding principles. That is the good and the bad of the scientific method. Implementing any one of the steps poorly and you've got crap studies which technically followed the scientific method but are ultimately "unscientific". You get into the same sorts of arguments as a result. "Agile is pefect, but scrum sucks" "The scientific method is perfect, but nutrition survey studies suck". The part that the scientific method lacks is determining what a good implementation looks like.
    - thephyber 5y ago
      > The part that the scientific method lacks is determining what a good implementation looks like. That seems like a feature, not a bug, to me. Descriptions of good implementations are likely subjective and vary based on what is being tested. It may even change over time as human knowledge improves.
  - drbojingle 5y ago
    Context man, I'm referring the the parents comment that the scientific method is perfect and operated by infallible humans leads to bad results and I would contend that if process is truly perfect it can't be corrupted. The scientific method is useful not perfect. Nothing is perfect.
devilduck 5y ago
Calling it perfect is absurd and makes me skeptical of your understanding :)
notahacker 5y ago
Yep, and you can add in a worse case still which is "people with enough background knowledge to sound like they know something, who have a particular motivation for wanting a particular scientific theory to be incorrect, who will never, ever admit to error" taking up the mantle of scepticism about current scientific theory. In the "marketplace of ideas" practising the scientific method isn't antifragile in the face of such reckless certainty, it's downright fragile, because it's requires being the only side of the argument which will admit that its hypotheses aren't proven and some of its previous theories were wrong. Worse still, when the motivated sceptic has their "stopped clock" moment, practising the scientific method requires conceding they are right. The more science you practice, the more the opponent gets to point to the stuff you've conceded wasn't quite as expected. Science might become sturdier in terms of better theory in response to attacks (where said attacks are accompanied by experimentation) but it's more vulnerable to losing trust
native_samples 5y ago
The scientific method really isn't perfect, unfortunately. It's a collection of heuristics, traditions and methods that's been assembled over time and which is still being assembled. That's why there's such tremendous uncertainty about what the scientific method precisely consists of and requires, and why nobody anywhere in the scientific system is responsible for ensuring scientists actually use the scientific method. Here are just a few questions that expose the imperfection of the scientific method: 1. Does it require you to do experiments? 2. What P value threshold does the scientific method require? 3. What can it be applied to? Just the natural world? People? Economies? Artificial machines? Is AI research using the scientific method, or is it just engineering? 4. How should uncertainty be treated? Does the scientific method require the use of subjective or objective Bayesian theory, or does it not matter at all? 5. If you build a statistical model, are you allowed to use assumptions? If yes, how many and of what type? If there are no restrictions on it, what stops astrology being scientific? 6. Does the scientific method require pre-registration of studies? If it does, then does it mean every study before this became a common approach were unscientific? If not, then why are scientists encouraged to do it? Etc.

by ethbr0 5y ago

LocalH 5y ago

>The scientific method is perfect. Nonsense. Nothing is perfect, especially nothing ever created by a human. The scientific method is the best we have developed, in my opinion, but it's not perfect. The scientific method itself was developed by "fallible and irrational humans". Even with the best of intentions and the utmost in effort, all humans are susceptible to fallibility and irrationality. You are, I am, everyone is.

rcoveson 5y ago
The word/concept "perfect" was "developed" by "fallible and irrational humans" as well, so we do have the authority to apply it. The 3/4/5 right triangle is perfect. The perfect fifth in music is (tautologically) perfect. "p or not p" is perfect. Perhaps none of these things can be reproduced in the physical world in a perfect manner, but the concepts really are perfect. Anything in idea-space can in fact be perfect, and the scientific method lives in idea-space. In the realm of perfection, there isn't really any difference between "developing" something and "discovering" it. Did we "develop" the scientific method or "discover" it? I think that's a meaningless distinction. It's the loop that describes how inductive reasoning is applied by limited observers to a structured universe. People were doing it before it was "developed".
- angelzen 5y ago
  This is confusing models with reality. It is the application of any specific model to reality that is not perfect.
  - rcoveson 5y ago
    Yes, one might say: > Perhaps none of these things can be reproduced in the physical world in a perfect manner, but the concepts really are perfect.

marginalia_nu 5y ago

The scientific method cannot be separated from the humans that practice it. It doesn't have some separate existence as a platonic ideal. Ideas like these must be judged based on their actual real world performance, and not on the lofty promises of the ideologues that promote the ideas. Luckily science doesn't require perfectly rational practitioners. It still works even if practiced human beings.

marivilla 5y ago

>The scientific method is perfect. >The scientific method as practiced by fallible and irrational humans This is giving me flashbacks to Catholic school :) 'Christ is perfect, his earthly church is not. Because the church consists of fallible human clergy and laity it is vulnerable to corruption, ambition, lust, and politics, oh the politics' It's interesting to know that secular students get the same arguments just in defense of scientism instead of Christianity. All apologists sound alike. Not a criticism, just interesting

beebmam 5y ago
Metaphors/analogies aren't an argument, and if you find them persuasive that's your problem
- rajin444 5y ago
  This isn't really a metaphor or an analogy. Modern scientific leaders and religious leaders both act as liaisons between the public and knowledge. Scientists tend to produce a lot more useful knowledge, but the issues faced by both are the same. It's obvious they both end up developing similar patterns for resolving communication issues. The parent noticed this similarity because both are in the same business.
- marivilla 5y ago
  I said 'Not a criticism, just interesting' in the hopes of emphasizing that I was making an observation, not an argument. Sorry I failed to communicate that more clearly! That said, there is something to be said for https://en.wikipedia.org/wiki/Argument_from_analogy https://en.wikipedia.org/wiki/Argument_from_analogy
  - ohwellhere 5y ago
    I view your observation as interesting and worth making, and I also believe there's an implicit argument there against that line of thinking. The word "perfect" is problematic to me. It's ascribed to God to contrast with man, as most of us would agree that nothing we do is perfect. Using it implies the same religious status of being above man's foibles. I like the word "useful" here, or "anti-fragile" in the original comment. Further, I like defining context. Useful in what class of problems? For I don't think the scientific method is regarded as particularly useful in all domains. Consciousness and the subjective are valid realms of human experience worth dedicated inquiry, for which the scientific method seems woefully lacking.

iammisc 5y ago

The scientific method is a human invention. It is not 'perfect'. This is ridiculous religious dogmatism.

drbojingle 5y ago

If the scientific method is so perfect why does it fail when used by imperfect operators

thephyber 5y ago
Does it? You are making an assertion without any facts. The method itself is simple. It is agnostic to bad faith implementations, to bad measurements, to completely fallacious conclusions, etc. It is intended to be iterative; individual errors wash out in time with enough new experiments. What non-scientists learn from secondary or tertiary sources is irrelevant — those misunderstandings do not mean the Method “failed”.
- cogman10 5y ago
  Meh, the scientific method is akin to "agile" in software development. It's good principles but really is less a method and more guiding principles. That is the good and the bad of the scientific method. Implementing any one of the steps poorly and you've got crap studies which technically followed the scientific method but are ultimately "unscientific". You get into the same sorts of arguments as a result. "Agile is pefect, but scrum sucks" "The scientific method is perfect, but nutrition survey studies suck". The part that the scientific method lacks is determining what a good implementation looks like.
  - thephyber 5y ago
    > The part that the scientific method lacks is determining what a good implementation looks like. That seems like a feature, not a bug, to me. Descriptions of good implementations are likely subjective and vary based on what is being tested. It may even change over time as human knowledge improves.
- drbojingle 5y ago
  Context man, I'm referring the the parents comment that the scientific method is perfect and operated by infallible humans leads to bad results and I would contend that if process is truly perfect it can't be corrupted. The scientific method is useful not perfect. Nothing is perfect.

devilduck 5y ago

Calling it perfect is absurd and makes me skeptical of your understanding :)

notahacker 5y ago

Yep, and you can add in a worse case still which is "people with enough background knowledge to sound like they know something, who have a particular motivation for wanting a particular scientific theory to be incorrect, who will never, ever admit to error" taking up the mantle of scepticism about current scientific theory. In the "marketplace of ideas" practising the scientific method isn't antifragile in the face of such reckless certainty, it's downright fragile, because it's requires being the only side of the argument which will admit that its hypotheses aren't proven and some of its previous theories were wrong. Worse still, when the motivated sceptic has their "stopped clock" moment, practising the scientific method requires conceding they are right. The more science you practice, the more the opponent gets to point to the stuff you've conceded wasn't quite as expected. Science might become sturdier in terms of better theory in response to attacks (where said attacks are accompanied by experimentation) but it's more vulnerable to losing trust

native_samples 5y ago

The scientific method really isn't perfect, unfortunately. It's a collection of heuristics, traditions and methods that's been assembled over time and which is still being assembled. That's why there's such tremendous uncertainty about what the scientific method precisely consists of and requires, and why nobody anywhere in the scientific system is responsible for ensuring scientists actually use the scientific method. Here are just a few questions that expose the imperfection of the scientific method: 1. Does it require you to do experiments? 2. What P value threshold does the scientific method require? 3. What can it be applied to? Just the natural world? People? Economies? Artificial machines? Is AI research using the scientific method, or is it just engineering? 4. How should uncertainty be treated? Does the scientific method require the use of subjective or objective Bayesian theory, or does it not matter at all? 5. If you build a statistical model, are you allowed to use assumptions? If yes, how many and of what type? If there are no restrictions on it, what stops astrology being scientific? 6. Does the scientific method require pre-registration of studies? If it does, then does it mean every study before this became a common approach were unscientific? If not, then why are scientists encouraged to do it? Etc.