How does science work?

How does science work? And what's all this about quantum mechanics?

Moderators: AMod, iMod

uwot
Posts: 6093
Joined: Mon Jul 23, 2012 7:21 am

Re: How does science work?

Post by uwot »

RCSaunders wrote: Sat Jul 20, 2019 7:05 pmThe hypothesis is, "the bridge is safe enough to drive over." The test for the falseness of this hypothesis is to drive over the bridge. When the test is performed, the collapse of the bridge and car plunging into the ravine proves the hypothesis is false. If the bridge did not collapse and the car safely drove to the other side, (the test for falseness failed), the hypothesis is proved true.
So could the next person to drive over the bridge conclude that it's safe? Sooner or later, the bridge will collapse.
User avatar
RCSaunders
Posts: 4704
Joined: Tue Jul 17, 2018 9:42 pm
Contact:

Re: How does science work?

Post by RCSaunders »

uwot wrote: Sat Jul 20, 2019 8:01 pm
RCSaunders wrote: Sat Jul 20, 2019 7:05 pmThe hypothesis is, "the bridge is safe enough to drive over." The test for the falseness of this hypothesis is to drive over the bridge. When the test is performed, the collapse of the bridge and car plunging into the ravine proves the hypothesis is false. If the bridge did not collapse and the car safely drove to the other side, (the test for falseness failed), the hypothesis is proved true.
So could the next person to drive over the bridge conclude that it's safe? Sooner or later, the bridge will collapse.
I wouldn't conclude that it was safe without additional tests. The example is only to illustrate what I meant, but we could start with a different hypothesis, "The bridge will still be safe to drive over for (a year, ten years, a century)." Today's engineering technology could surely devise tests that would prove the hypothesis false, if it is false. Such testing is actually done (though not enough) for both bridges and tunnels, just not from that negative viewpoint. They are the same tests, however.
User avatar
RCSaunders
Posts: 4704
Joined: Tue Jul 17, 2018 9:42 pm
Contact:

Re: How does science work?

Post by RCSaunders »

uwot wrote: Sat Jul 20, 2019 8:01 pm
RCSaunders wrote: Sat Jul 20, 2019 7:05 pmThe hypothesis is, "the bridge is safe enough to drive over." The test for the falseness of this hypothesis is to drive over the bridge. When the test is performed, the collapse of the bridge and car plunging into the ravine proves the hypothesis is false. If the bridge did not collapse and the car safely drove to the other side, (the test for falseness failed), the hypothesis is proved true.
So could the next person to drive over the bridge conclude that it's safe? Sooner or later, the bridge will collapse.
I wouldn't conclude that it was safe without additional tests. The example is only to illustrate what I meant, but we could start with a different hypothesis, "The bridge will still be safe to drive over for (a year, ten years, a century)." Today's engineering technology could surely devise tests that would prove the hypothesis false, if it is false. Such testing is actually done (though not enough) for both bridges and tunnels, just not from that negative viewpoint. They are the same tests, however.

If you want to point out that such tests are not absolutely reliable, I agree, but the point was made on the basis that the test had to be a reliable one. The actual driving over the bridge is such a reliable test, and would be for the other cases as well, a year, ten years, a century. It's just unlikely anyone will want to be the one who makes the test.

As you said, "Sooner or later, the bridge will collapse." Being the one who disproves the hypothesis will probably not be that satisfying.
uwot
Posts: 6093
Joined: Mon Jul 23, 2012 7:21 am

Re: How does science work?

Post by uwot »

RCSaunders wrote: Sat Jul 20, 2019 8:23 pm
uwot wrote: Sat Jul 20, 2019 8:01 pm
RCSaunders wrote: Sat Jul 20, 2019 7:05 pmThe hypothesis is, "the bridge is safe enough to drive over." The test for the falseness of this hypothesis is to drive over the bridge. When the test is performed, the collapse of the bridge and car plunging into the ravine proves the hypothesis is false. If the bridge did not collapse and the car safely drove to the other side, (the test for falseness failed), the hypothesis is proved true.
So could the next person to drive over the bridge conclude that it's safe? Sooner or later, the bridge will collapse.
I wouldn't conclude that it was safe without additional tests.
What tests would you conduct other than driving over the bridge?
RCSaunders wrote: Sat Jul 20, 2019 8:23 pmThe example is only to illustrate what I meant, but we could start with a different hypothesis, "The bridge will still be safe to drive over for (a year, ten years, a century)."
Well yeah, you can make up any hypothesis you like, but one day the the bridge will fail, and every scientist who predicted a different day is wrong.
RCSaunders wrote: Sat Jul 20, 2019 8:23 pmToday's engineering technology could surely devise tests that would prove the hypothesis false, if it is false. Such testing is actually done (though not enough) for both bridges and tunnels, just not from that negative viewpoint. They are the same tests, however.
So should we hold scientists and engineers responsible when they get it wrong?
User avatar
RCSaunders
Posts: 4704
Joined: Tue Jul 17, 2018 9:42 pm
Contact:

Re: How does science work?

Post by RCSaunders »

You make good points.
uwot wrote: Sat Jul 20, 2019 8:56 pm So should we hold scientists and engineers responsible when they get it wrong?
If the engineers worked for me they would definitely be held responsible. I don't know who "we" is.
User avatar
-1-
Posts: 2888
Joined: Thu Sep 07, 2017 1:08 am

Re: How does science work?

Post by -1- »

I don't know the author who originally wrote:...This demanded that anything that could not be supported by empirical evidence or strict logic was metaphysics and had no place in science (or indeed, anywhere else). One major problem – which in fairness the logical positivists were well aware of – is that no amount of empirical evidence (or logic) can prove a scientific claim. The classic example is that a million white swans do not prove that every swan is white. Popper’s innovation was to point out that it only takes one black swan to prove that the proposition ‘all swans are white’ is false. So the evidence could show you either what was only likely to be true, or what was definitely false. Therefore, as an endeavour seeking certainty, science should commit itself to trying to prove its own theories wrong. This is Popper’s principle of falsification.
Yes, there is the element of lacking certainty in findings of scientific truths. You should not throw them away, and most people trust them, because they are useful.

A not so original example of showing this is the scientific truth, or finidng, that a proper amount of arsenic will kill you. So how many people do you hear say, "Humbug, it's not true, it's only evidence-based, therefore not proven. So tonight I and my entire family will have arsenic stew with cyanide powder to make it go down easier."

Evidence is stronger than simply a high likelihood. It's precisely half way between very likely and absolute truth. (This was a joke.)
Skepdick
Posts: 14366
Joined: Fri Jun 14, 2019 11:16 am

Re: How does science work?

Post by Skepdick »

RCSaunders wrote: Sun Jul 21, 2019 1:28 am If the engineers worked for me they would definitely be held responsible.
And what good is that? If people are working on ground-breaking research (e.g nothing anybody has ever done before) and they make an error, what is the point of "holding people responsible"?

Do you think it has any effect on preventing errors?
Skepdick
Posts: 14366
Joined: Fri Jun 14, 2019 11:16 am

Re: How does science work?

Post by Skepdick »

RCSaunders wrote: Sat Jul 20, 2019 7:05 pm The hypothesis is, "the bridge is safe enough to drive over." The test for the falseness of this hypothesis is to drive over the bridge. When the test is performed, the collapse of the bridge and car plunging into the ravine proves the hypothesis is false. If the bridge did not collapse and the car safely drove to the other side, (the test for falseness failed), the hypothesis is proved true.
That simply indicates that you have very low standards.

What you seem to be asking with the question "is the bridge safe enough?" is tantamount to the question "Will the bridge collapse on the next crossing?".

Is an airplane that survives one flight "safe"?
Is a parachute that works one time "safe"?

Surely safety is more than single-use?
uwot
Posts: 6093
Joined: Mon Jul 23, 2012 7:21 am

Re: How does science work?

Post by uwot »

-1- wrote: Sun Jul 21, 2019 2:51 am
I don't know the author who originally wrote:...This demanded that anything that could not be supported by empirical evidence or strict logic was metaphysics and had no place in science (or indeed, anywhere else). One major problem – which in fairness the logical positivists were well aware of – is that no amount of empirical evidence (or logic) can prove a scientific claim. The classic example is that a million white swans do not prove that every swan is white. Popper’s innovation was to point out that it only takes one black swan to prove that the proposition ‘all swans are white’ is false. So the evidence could show you either what was only likely to be true, or what was definitely false. Therefore, as an endeavour seeking certainty, science should commit itself to trying to prove its own theories wrong. This is Popper’s principle of falsification.
That was me.
-1- wrote: Sun Jul 21, 2019 2:51 amYes, there is the element of lacking certainty in findings of scientific truths. You should not throw them away, and most people trust them, because they are useful.
Well yeah. For practical purposes, all that matters is that a model works. It doesn't really matter whether the theory is true.
-1- wrote: Sun Jul 21, 2019 2:51 amA not so original example of showing this is the scientific truth, or finidng, that a proper amount of arsenic will kill you. So how many people do you hear say, "Humbug, it's not true, it's only evidence-based, therefore not proven. So tonight I and my entire family will have arsenic stew with cyanide powder to make it go down easier."
Bon appetit!
-1- wrote: Sun Jul 21, 2019 2:51 amEvidence is stronger than simply a high likelihood. It's precisely half way between very likely and absolute truth. (This was a joke.)
What did it die of?
User avatar
RCSaunders
Posts: 4704
Joined: Tue Jul 17, 2018 9:42 pm
Contact:

Re: How does science work?

Post by RCSaunders »

Skepdick wrote: Sun Jul 21, 2019 8:27 am
RCSaunders wrote: Sun Jul 21, 2019 1:28 am If the engineers worked for me they would definitely be held responsible.
And what good is that? If people are working on ground-breaking research (e.g nothing anybody has ever done before) and they make an error, what is the point of "holding people responsible"?

Do you think it has any effect on preventing errors?
We weren't talking about engineers working on, "ground-breaking research:" we were talking about engineers whose failure allowed a bridge to collapse. The good is, that after they are fired, I won't have to continue paying incompetent engineers, and they won't cause my company any more trouble. That's the point of holding people responsible.

By the way, you can forgive errors if you choose, but reality never does.
Last edited by RCSaunders on Sun Jul 21, 2019 2:39 pm, edited 1 time in total.
User avatar
RCSaunders
Posts: 4704
Joined: Tue Jul 17, 2018 9:42 pm
Contact:

Re: How does science work?

Post by RCSaunders »

Skepdick wrote: Sun Jul 21, 2019 8:30 am Surely safety is more than single-use?
I answered this for uwot here.
User avatar
-1-
Posts: 2888
Joined: Thu Sep 07, 2017 1:08 am

Re: How does science work?

Post by -1- »

uwot wrote: Sun Jul 21, 2019 11:30 am
-1- wrote: Sun Jul 21, 2019 2:51 amEvidence is stronger than simply a high likelihood. It's precisely half way between very likely and absolute truth. (This was a joke.)
What did it die of?
HA, HA, HA! This gave me a belly-laugh. Thanks. ...Aaah...
Skepdick
Posts: 14366
Joined: Fri Jun 14, 2019 11:16 am

Re: How does science work?

Post by Skepdick »

RCSaunders wrote: Sun Jul 21, 2019 2:29 pm We weren't talking about engineers working on, "ground-breaking research:" we were talking about engineers whose failure allowed a bridge to collapse. The good is, that after they are fired, I won't have to continue paying incompetent engineers, and they won't cause my company any more trouble. That's the point of holding people responsible.
You aren't even going to bother trying to understand why the bridge collapsed before assigning blame? You are happy to jump to a conclusion without allowing for an alternative hypothesis: something went wrong that was outside of anybody's control. Was it even the engineers' fault? Was it process failure ? Material defect? Poor manufacturing? Budget cuts? Expedient delivery deadlines at the expense of quality? Something else entirely?

And the most important question of all: Is changing engineers going to prevent the next bridge from collapsing in a similar fashion?
If the answer is "no" then you are just scapegoating your engineers.

Any company with you at the helm is already in trouble. If your engineers are incompetent - who is to be held responsible for hiring incompetent engineers? Who is be held responsible for delegating bridge-building to incompetent engineers?

Given your attitude, it sounds like a cultural, not an engineering failure. If you owned an engineering business - I would be sure to keep my business away from you.
User avatar
RCSaunders
Posts: 4704
Joined: Tue Jul 17, 2018 9:42 pm
Contact:

Re: How does science work?

Post by RCSaunders »

Skepdick wrote: Mon Jul 22, 2019 8:14 am You aren't even going to bother trying to understand why the bridge collapsed before assigning blame?
No, after all the analysis is done, of course. Why would you assume the worst.
Skepdick wrote: Mon Jul 22, 2019 8:14 am Given your attitude, it sounds like a cultural, not an engineering failure. If you owned an engineering business - I would be sure to keep my business away from you.
I suspect my company would not be losing much.
Skepdick
Posts: 14366
Joined: Fri Jun 14, 2019 11:16 am

Re: How does science work?

Post by Skepdick »

RCSaunders wrote: Mon Jul 22, 2019 2:02 pm No, after all the analysis is done, of course. Why would you assume the worst.
I am not assuming the best or worst. I am merely observing that your narrative contains a single conclusion of dozens possible.

Notice that you ignored my key point. Even if analysis showed human error on behalf of your engineers, who ought to be held responsible for hiring incompetent engineers? Who ought to be held responsible for having insufficient checks&balances in their end-to-end process?
RCSaunders wrote: Mon Jul 22, 2019 2:02 pm I suspect my company would not be losing much.
Lets just say that if your company was trading publicly, I'd short your stock.
Post Reply