1. Methods: testing an idea

I'm a little concerned about events where folks seem to take discussion about models personally. I think that partly this is because we are conditioned by popular culture to make black white distinctions and not consider shades of gray. So folks are conditioned to assume that when a discussion of the limitations of a model or solution method happens that it implies that the model or solution are completely wrong. Then folks get testy or provocative. Maybe it would help if I shared how I think about proposed solutions? This is a process I use to criticize my own ideas. Hopefully sharing it will decrease sensitivities.

Once an idea is proposed:

1) The first thing to do is subject it to a "smell test". Does the idea or solution permit energy, momentum and angular momentum to be conserved? If so OK, if not it can be discarded as wrong. Also check and make sure it allows entropy to remain constant or increase for closed systems. Betting against the 2nd Law of Thermodynamics is IMO the worst bet one can make in physics. It's safer to bet against Newtonian dynamics or General Relativity.

2) If the idea passes the smell test, criticize it so that the limits of its validity become clear. What implicit approximations does it make? When does it break down? The fact that it does break down doesn't make the approximation completely useless or wrong. It just means that you understand the conditions under which you shouldn't try to use the idea. By trying to understand a problem so that you know what a "better" solution would look like, one can be comfortable about how valid the simpler approach is, what it's limits are. Always take a solution method or idea one step farther than necessary so you can appreciate it's limitations. That way you understand how your idea might be "good enough" but not make the mistake of accepting "good enough" for "absolutely right".

3) "Absolutely right" never happens. Not even in math. There are cases when 1 + 1 does not = 2.

2. The Following 6 Users Say Thank You to not_Fritz_Argelander For This Useful Post:

astroval (01-18-2014),chas53 (01-16-2014),j.gardavsky (01-16-2014),KathyNS (01-16-2014),kencrowder (01-16-2014),Seeker725 (01-16-2014)

3. Re: Methods: testing an idea

While I agree with you, it is a consequence of years of my being taught to think that way. There are very few disciplines that demand such skepticism. More often the one who is most adept at debate will carry the scepter. In physics we are trained to test any speculation against theory that is considered probably reliable as a consequence of decades of testing. And "probably reliable" is the highest accolade we can award any theory.

The eagerness of journalism to report new "findings" and the need to get public support for funding often masks the science. I'm often amused when reading a paper that a "commercial" appears in the conclusions just to appease the funding source.

I suspect there are very few beginners that can appreciate the statement, "no theory can be proven true but one test may prove it false."

4. The Following 4 Users Say Thank You to kencrowder For This Useful Post:

astroval (01-18-2014),chas53 (01-16-2014),not_Fritz_Argelander (01-16-2014),Seeker725 (01-16-2014)

5. Re: Methods: testing an idea

Agreed. There may be a scientific reason for the following:

Originally Posted by kencrowder
I suspect there are very few beginners that can appreciate the statement, "no theory can be proven true but one test may prove it false."
We have a hard time tolerating uncertainty. We'd rather have a simple answer than a nuanced one. To some extent scientific thinking as a whole (not just in astrophysics) runs counter to how we evolved as a species.

This blog The certainty of uncertainty | quantum shifting has an interesting speculation:

As Thayer et al write, there is “an evolutionary advantage associated with the assumption of threat” and that our “‘default’ response to uncertainty, novelty, and threat is the sympathoexcitatory preparation for action commonly known as the fight or flight response”. Essentially, because we have inherited a certain vigilance to our environment, when faced with uncertainty, we unconsciously prepare for the worst. While useful for survival if we are about to be attacked by a lion, it’s hardly the most progressive state to be in if we want to thrive. This goes for businesses living in uncertain times as well as individuals.
Please note I do not subscribe to the political views in the blog above. I am quoting it only for the physiology of uncertainty bit.

The Thayer et. al. reference is here: http://wagerlab.colorado.edu/files/p...pfc.NBR.12.pdf

Our brains are hard wired to crave the illusion of certainty. That is what must be overcome to be able to think about science in general or astrophysics in particular.

6. The Following 4 Users Say Thank You to not_Fritz_Argelander For This Useful Post:

astroval (01-18-2014),chas53 (01-17-2014),j.gardavsky (01-16-2014),kencrowder (01-16-2014)

7. Re: Methods: testing an idea

Originally Posted by not_Fritz_Argelander
Agreed. There may be a scientific reason for the following:

We have a hard time tolerating uncertainty. We'd rather have a simple answer than a nuanced one. To some extent scientific thinking as a whole (not just in astrophysics) runs counter to how we evolved as a species.

Our brains are hard wired to crave the illusion of certainty. That is what must be overcome to be able to think about science in general or astrophysics in particular.
IMO, uncertainty is wired to be associated with fear and that triggers "flight or fight." That could be the cause of the response you are concerned about. We are wired to overcome fear due to uncertainty by defending our beliefs. I applaud beginners willing to make the effort to make an exception for the sake of physics.

8. The Following 3 Users Say Thank You to kencrowder For This Useful Post:

astroval (01-18-2014),chas53 (01-17-2014),not_Fritz_Argelander (01-16-2014)

9. Re: Methods: testing an idea

Maybe it is useful to remember that frequently with uncertainty there is opportunity?

10. The Following 3 Users Say Thank You to OleCuss For This Useful Post:

astroval (01-18-2014),chas53 (01-18-2014),not_Fritz_Argelander (01-17-2014)

11. Re: Methods: testing an idea

Originally Posted by OleCuss
Maybe it is useful to remember that frequently with uncertainty there is opportunity?
If only folks could feel that in their guts, scientific illiteracy would diminish greatly. Sound bites would lose their charm.

12. The Following User Says Thank You to not_Fritz_Argelander For This Useful Post:

chas53 (01-18-2014)

13. Re: Methods: testing an idea

There are two interesting blog articles of Sean Carroll about "Which scientific ideas should be retired?"

Falsifiability needs to be retired
and What Scientific Ideas Are Ready for Retirement? It will be interesting any comments about it with point of view of testing an idea.

14. The Following 2 Users Say Thank You to astroval For This Useful Post:

chas53 (01-19-2014),not_Fritz_Argelander (01-18-2014)

15. Re: Methods: testing an idea

Originally Posted by astroval
There are two interesting blog articles of Sean Carroll about "Which scientific ideas should be retired?"

Falsifiability needs to be retired
and What Scientific Ideas Are Ready for Retirement? It will be interesting any comments about it with point of view of testing an idea.
It is an interesting and provocative idea, that falsifiability should be discarded from science.

It is also utterly mistaken.

Falsifiability, the possibility that an idea or theory can be tested and contradicted in an objective and reproducible way is precisely what distinguished science from mythologies or artistic products.

The article claims that some ideas are more central to science: empirical and definite prediction. But I would say that music is also empirical. We hear it. It also can be quite definite. Pieces can be notated and performed. Interpretations can differ, but only in nuance.

Falsifiability is precisely what distinguishes science from other endeavors.

Mythologies can't be falsified. All of the religions make empirical sense of the world in terms of the agencies of natural or supernatural forces. But mythologies and religions are not falsifiable.

I find that the dismissal of "falsifiability" as being something of importance only to "armchair philosophizing" is quite a wrongheaded piece of armchair philosophizing itself.

One might somewhat cynically speculate that the "anti falsifiability" faction is promoting that agenda because doing the experiments that can provide falsification of ideas is becoming more and more expensive. There is a heavy burden on theorists these days to come up testable theories. It is becoming increasingly hard. I'm sure that some would rather pretend that spinning a mathematical mythology that isn't testable is still doing science.

I don't believe it is. I think it's an indolent cop out. Falsifiability requires you risk being wrong. Newton, Einstein, Schroedinger, Heisenberg all risked being wrong. If you can't risk being wrong, you're not a scientist. IMO.

16. The Following 4 Users Say Thank You to not_Fritz_Argelander For This Useful Post:

astroval (01-18-2014),chas53 (01-19-2014),jerryTheC (01-18-2014),kencrowder (01-18-2014)

17. Interesting - doesn't the ability to make definite predictions imply falsifiability? If the prediction gets the wrong result, then the theory is wrong (or incomplete). If it consistently gets the right result then it's useful (but may still be wrong). Or am I missing something?

18. The Following 4 Users Say Thank You to jerryTheC For This Useful Post:

astroval (01-18-2014),chas53 (01-19-2014),kencrowder (01-18-2014),not_Fritz_Argelander (01-18-2014)

19. Re: Methods: testing an idea

Originally Posted by jerryTheC
Interesting - doesn't the ability to make definite predictions imply falsifiability? If the prediction gets the wrong result, then the theory is wrong (or incomplete). If it consistently gets the right result then it's useful (but may still be wrong). Or am I missing something?
I don't think you are missing anything, of course. I think that the reasons for retreating from 'falsifiability' are emotional and unfortunate.

Think of the current many competing string theories for unifying all of physics. It is a beautiful arena for mathematical play. So far it has been practically impossible to come up with tests for them. Some folks have suggested that because there is no practical way to falsify that they aren't science, rather they are a branch of philosophy or theology. If you are a string theorist you may prefer to play in your mathematical garden and ignore the need to test the theory. (Personally I hope that some clever person will come up with low energy lower cost observational tests. as long as there is hope of falsification I'd hold out for string theory being science.)

There are also folks who want to "rig the books" so that physics is always right. If you forget about falsifiability then the tests of General Relativity don't falsify Newtonian theory.

No. Science finds improved versions of "what is the truth" by daring to be wrong and being comfortable with that.

20. The Following 4 Users Say Thank You to not_Fritz_Argelander For This Useful Post:

astroval (01-18-2014),chas53 (01-19-2014),jerryTheC (01-19-2014),kencrowder (01-18-2014)

Page 1 of 2 12 Last

Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•