A theory is only as successful as...

I believe myself to be open-minded.  My background is science, but I've also studied a dose of philosophy. 

Core to my way of thinking is the scientific method.  A recommended book on the subject is "The Structure of Scientific Revolutions" by Thomas S. Kuhn.

On this page I'd like to describe what I understand is meant by the scientific method.  

It's not straightforward to define.  This is evidenced by the fact that the current Wikipedia article on it is not brilliant (at time of writing) and the fact that there is (was?) an open request for people to re-write it.

This brilliant video of Richard Feynman describes how scientists come up with a new physical law.

.. but it's not quite the whole story to me.

I am not attempting to correct Feynman, but I have a few bits of context to add to that, so here's my attempt - not so much a description of the scientific method, but more a description of how I see that it works:

1. In the very beginning there are observations.  Not necessarily scientific observations, simply people observing things.
2. People come up with explanations for the observations, the explanations take the form of describing how the world works.  This is called a theory.  This is Feynman's "Guess it"
3. The theory may be good, that is to say that it explains things and it makes predictions about other things that may be seen in the world.  (If the theory doesn't explain things or predict anything then it's not really very good so we should dismiss it).
4. Later on, the theory starts to get a bit shakey at the edges as things are observed that don't seem to fit the theory.  New instruments may be able to make more precise measurements, or a theory may be tested in new places or under different conditions (for example at different temperatures).
5. People often try to rescue the old theory when they find cases where it doesn't work.  Sometimes people blame the researchers who are finding the bad results, sometimes people propose additions to the original theory to try to save it.
6. People keep testing the old theory harder than ever before (sometimes this is with a motivation to try to prove it is right after all).
7. Eventually so much evidence mounts up that people accept that the old theory is wrong (or let's say instead say that the theory has clearly identified areas where it is wrong).  At this point the community who are interested in the theory ends up in crisis because the theory has been shown to have errors - and to say a theory has errors is to say that there is not a working theory.
8. A new theory is hopefully put forward that seems to explain the previous results and also explains the things that the old theory couldn't explain. (This would be Feynman making another more educated guess, this time based on further information).
9. (Note that new theories are frequently not widely welcomed till everyone agrees the old theory really can't be rescued.  People often tend to cling to things that they used to find valuable).
10. If the new theory becomes welcome then people will flesh out its details to find that it makes some predictions of totally new events, never before observed, that, of course, are open to being tested.
11. The predictions of these entirely new events turn out to be correct.  Quite often in practice this is much to everyone's surprise and we learn something genuinely new about the world (or universe).
12. ..... or the predictions may fail in which case the theory is wrong and clearly a different new theory is needed.
13. If the predictions worked then the world gets behind the successful new theory because it explains the results of experiments and has told us something genuinely new about the world.

[Deep breath!  Sorry that was so long.]


OK, so Feynman talks about theories being wrong. "If it disagrees with experiment, it's wrong!"

He's correct, but there can still be ways of understanding the world that work most of the time.

There were bridges and buildings before Einstein proved Newton's laws to be "wrong".

How physical objects exerted forces on each other and how they moved, like how billiard balls bounced off each other and how the planets orbited the sun was described by Isaac Newton in his Principia Mathematica published in 1687 (not perfectly for one of the planets, but extremely close).

Newton's laws of motion are really pretty good.  They work incredibly accurately most of the time.  Of course Feynman would say they're wrong and strictly they are wrong and have been proven to be wrong, but they are actually genuinely pretty good until you push the limits.  (The problems with Newton's laws mostly arise when things are moving at speeds close to the speed of light).

Because of all this, I like to distinguish between theories that are "good", theories that are "successful" and theories that are "useful".

A "good" theory
A "good" theory is one that makes predictions about the world that can be tested.  It does this by providing predictions of things, ie by providing some direction for how to find out something genuinely new about the world - so that we can check if this prediction turns out to be true.
Fundamentally for a theory to be good it needs to make predictions that can be tested.  It needs to be possible to prove it wrong.  If there's no way to prove it wrong then it simply isn't a theory.
Importantly a "good" theory doesn't have to be "right".

A "useful" theory 

A useful theory is a theory that is "good" and that makes predictions that are found to be correct.  It therefore gives direct instructions of how to prove it wrong and it turns out to make predictions that are correct.

A useful theory may turn out to be wrong when pushed past its limits.  It doesn't stop it being useful.  Newton's laws are most definitely still useful despite having been proven wrong.

A "successful" theory 

A successful theory is one that is good (ie it makes predictions), is useful (ie the predictions turn out to be correct), and where it has not yet been proven to fail.

In other words a successful theory is one that still stands after people have made repeated attempts to prove it wrong.

To me a theory's success should be measured by the number and quality of the attempts to prove that what it predicts is wrong.

It's important to note that a "successful" theory is still not necessarily "right", it just may not have been possible to test everything it predicts yet.  (Case in point: The Standard Model of Particle Physics.  The Large Hadron Collider, one of the most ambitious and expensive experiments ever conceived, has been built partially to test whether the Standard Model still holds true at its limits).

To add one further point: A good theorist is one who attempts to come up with theories that describe the world but who does not mind being proven wrong.

I count myself as following the scientific method.  Anything I say could be wrong and I welcome feedback.  I'm hoping to write some posts that are "good", aspiring for some to be "useful".  I doubt very much that any will be "right".

I'd very much welcome any constructive feedback or even just being shown that anything on here is wrong, because ultimately:

         A theory is only as successful as the failed attempts to falsify it.

Unfalsified Theories

This blog is called "Unfalsified Theories."

This is "just a blog" of no special significance.  It will likely carry discussions of some scientific subjects, though other areas may also be covered.

Theories or ideas will not be posted here if they are known to be incorrect, though the title is intended to highlight that the pointing out of factual errors is welcome.  It is assumed that there will be errors.

The web address "disceptic" was chosen very late one night as a derivation from the Latin word:
   disceptatio  [-onis  f.]  -  debate , discussion, controversy

Any reader who identifies incorrect or missing facts or who has other genuine contributions to the discussion is wholeheartedly encouraged to make their point by adding a comment.

A theory is only as good as the failed attempts to falsify it.


I hope you enjoy reading and contributing.