How can you prove or disprove a theory in physics

[5ymmetrica1]

If you have a theory on a certain principle in physics, how can you determine it's validity?

is a thought experiment and some equations showing proof enough to solidly prove a theory in physics?
or is it more through experiments, physical evidence and data?

and if you have a theory how do you know if it's nonsense or not?

 

[Simon Bridge]

If you have a theory on a certain principle in physics, how can you determine it's validity?

look up "empiricism".

is a thought experiment and some equations showing proof enough to solidly prove a theory in physics?
or is it more through experiments, physical evidence and data?

It is more that you have to demonstrate that the theory is worth spending money on the experiment.

and if you have a theory how do you know if it's nonsense or not?

That is what empiricism is all about.

Generally, you can't.
However, you can eliminate known nonsense from your theory.
If the result manages to include some advance on known theory, and has results which are falsifiable, then it may be worth doing the experiment.

You are asking at HS level?
Did you have a particular theory in mind?

 

[5ymmetrica1]

thanks for the response Simon
I have been working on a theory for quite a while and know that at the present time, I'm willing to accept that the nonsense in it probably far outweighs the facts since my level of experience in physics is not yet at the level it should be to be theorizing such advanced concepts as I do.

So I'm interested in knowing how I can "eliminate known nonsense" as you said and take the thought experiments, equations, ect that I have so far and begin to justify and prove or disprove it as a theory.

 

[Simon Bridge]

I'll tell you what - pm me from my profile and we can discuss it in private.
Personal theories are frowned upon in PF... here we are supposed to be helping people understand existing physics.

In general, try to express your idea in mathematical form - in terms of things you can measure.
Figure out what kind of experiment would disprove the theory - and conduct the experiment.

If the theory contradicts well-established "Laws" of physics then the chance it has a terminal amount of nonsense is quite high.

 

[5ymmetrica1]

thanks Simon, I wasn't aware personal theories were frowned upon here, but thanks for the heads up, and I can see the reason with so many personal theories it would be easy to lose the real physics amongst all the theorists!

 

[Simon Bridge]

More that we get a lot of pseudoscience promotions, perpetual motion and so on.
Such people tend to get upset when their ideas are soundly challenged.

 

[chill_factor]

There's no proof in physics. Proof is for mathematics. Physics has evidence for or against.

The difference is that proof follows from logic, evidence follows from experiments, observations and designed systems.

 

[5ymmetrica1]

So does a theory in physics absolutely need solid measurable evidence to have validity, or is good mathematics enough to qualify a theory as valid (but inconclusive) until evidence rises either for or against?

 

[Drakkith]

So does a theory in physics absolutely need solid measurable evidence to have validity, or is good mathematics enough to qualify a theory as valid (but inconclusive) until evidence rises either for or against?

It's hit and miss on this. To be "accepted as fact" you need both math and evidence. To be "generally accepted by most people" you can make do with math if your underlying theory is sound. For example, we have never directly observed the event horizon of a black hole, yet we strongly suspect that it exists based on the math of General Relativity since GR has been shown to be extremely accurate and has passed every test we've thrown at it so far.

 

[5ymmetrica1]

So is there a logical way of determining if your math is "good enough" or is it largely guesswork that's dependent on either you or someone down the track finding conclusive evidence?

What I mean to ask is, is it possible to determine if there will be any real evidence for your math? Or is it just that some get lucky and have both their equations work and end up finding solid evidence? (such as the case in general relativity)

 

[Drakkith]

So is there a logical way of determining if your math is "good enough" or is it largely guesswork that's dependent on either you or someone down the track finding conclusive evidence?

Math itself is formulated using mathematical logic, which is different than "normal" logic if you will. But doing math just to do math is pretty pointless unless you are working with physical phenomena or are doing advanced theory.

What I mean to ask is, is it possible to determine if there will be any real evidence for your math? Or is it just that some get lucky and have both their equations work and end up finding solid evidence? (such as the case in general relativity)

You should be working on a specific problem that can't be explained by current theory, and should have tangible proof that will tell you your math is correct. Otherwise...what are you doing? Doing math just to do math typically makes you a mathematician, not a scientist.

 

[Simon Bridge]

It sometimes happens that someone does some math for, basically, the fun of it ... and someone else finds a physical problem that the math is needed for.

Usually a physicist is expected to have some physical motivation for the math ... the math is expected to, at some stage, describe something in nature. An example, I guess, would be the Cosmological Constant.
http://cosmology.carnegiescience.edu/timeline/1917
http://en.wikipedia.org/wiki/Fudge_factor

There are ways to tell if a particular bit of math is junk science, but there is no systematic way to be certain that any mathematical statement is clear of all junk. The more complex the statement the harder it is. It is not, however, a matter of chance either - it is a systematic and sustained effort. You'll find lots of examples if you frequent the skeptical forums.

Scientific theories stand, not so much on the preponderance of evidence in support, but on the cunningness and quantity of the attempts to disprove them.

 

[mrspeedybob]

This is a question Dr. Feynman addressed some decades ago. His explination is still correct.

 

[Dali]

I am personally fascinated by this video where physics legend Richard Feynman gives a lecture on the very basics of the scientific method: http://www.youtube.com/watch?v=EYPapE-3FRw. This would be so good viewing for a lot of people today... :-)

Free transcription from the introduction:
"In general we look for a new law [in physics] by the following process:
- first we guess it (don't laugh, it's really true!)
- then we compute the consequences to see what it would imply
- and then we compare those computation results to Nature or experiments
We compare it directly with observations to see if it works. If it disagrees with experiment, it is wrong! In that simple statement is the key to science."

[Oops, doublepost of the exact same video as in mrspeedybob's post above. The definite video to link here apparently!]

 

[5ymmetrica1]

I just got finished reading Richard Feynman's "The character of physical law" which goes into some detail on the subject of how physicists and mathematicians create laws and theorems.
It's the first Feynman book that I've had the pleasure of reading and I must say that he is a master of creating simple analogies to explain extremely complicated subjects.

Thanks for recommending the video dali and mrspeedybob, appreciate the link!

 

[harrylin]

If you have a theory on a certain principle in physics, how can you determine it's validity?

is a thought experiment and some equations showing proof enough to solidly prove a theory in physics?
or is it more through experiments, physical evidence and data?

and if you have a theory how do you know if it's nonsense or not?

In addition to what has been mentioned already: you cannot "solidly prove a theory in physics". One can reasonably prove a theory to be good for many applications by showing its usefulness. For a theory to be solidly proved, there must be consensus that never any exceptions can be found. From experience (think of Newton's mechanics) we know however that one can never be sure.

You can solidly disprove a theory by means of either a solid mathematical proof of inconsistency, or by means of a number of repeated independent experiments which demonstrate that the theory makes wrong predictions. This is how new theories are tested.

 

[plexel]

You have to start from certainties, here are some:-

A) There is experience
B) There is thinking
C) There is imagination
D) There is smell
etc

Everybody knows these and no one doubts them. They don't require proof, being stronger than proof. Proofs can now be derived from them perhaps but maybe not the type you immediately want.

The place the experience comes from may not be in the physical world, i.e. you couldn't locate it by cutting up the brain etc. Senses come from the physical world but sensations may not.

You couldn't even really say A) ==> there is a source of experience since you will need B) to define the word source. To get physical proofs you will need to use B) but I'm not sure how.

 

[Simon Bridge]

You have to start from certainties, here are some:-

A) There is experience
B) There is thinking
C) There is imagination
D) There is smell
etc

Everybody knows these and no one doubts them.

Speak for yourself.

 

[plexel]

Speak for yourself.

Well if you don't know you experience things then you are dead.

You experienced typing or something you thought was typing in order to write your reply.

How do you thing those certainties could be doubted then?

 

[Drakkith]

Well if you don't know you experience things then you are dead.

You experienced typing or something you thought was typing in order to write your reply.

How do you thing those certainties could be doubted then?

The issue is that you are proposing extremely generalized things and saying "everyone knows these things" and that they are "certainties". That simply isn't true. Everything in your list is a result of the mind and consciousness, of which we know far less about than we wish we did. And even then, to think that these things are "certainties" is simply wrong. I know there are people out there without a sense of smell for example. Even everyday people with smell have wildly varying sensitivities. Some people can smell a skunk from 5 miles away while some couldn't smell it unless it was inside their car. So you can't make any "certain" claims using smell, it is subjective.

 

[Simon Bridge]

Also this is not a philosophy forum and the list is very philosphyesque.
IRL: we seldom start from "certainties" about when we conduct empirical investigations - usually strong suspicions are good enough to go on with. Everything we may use to probe physics is fraught with uncertainty - which is why we have statistics.

 

[Evo]

The topic has been answered sufficiently.