[yor_on (OP)]

To me it is a term that contradict itself,

A falsifiable TOE is one where you by simple logic prove it to be wrong if I get the idea correctly. If it in any sense can be proven to be wrong, how can a TOE ever exist?

And if one demand of any physical theory is that it must be proven to be falsifiable to be true, then?

[Bogie_smiles]

To me it is a term that contradict itself,

A falsifiable TOE is one where you by simple logic prove it to be wrong if I get the idea correctly. If it in any sense can be proven to be wrong, how can a TOE ever exist?

And if one demand of any physical theory is that it must be proven to be falsifiable to be true, then?

You have the main point about the requirement that a theory has to be falsifiable, and that point is that in order for it to qualify as a theory, instead of speculation, it must be testable. If a theory is testable, then the theory can be falsified by conducting the test. If there is no proposed way to test the hypothesis, then it does not qualify as a theory in its own right.

Using that reasoning in regard to someone’s Theory of Everything just means that if there is no known way to falsify it, because there is no proposed test to prove or disprove it, means it remains speculation.

[chiralSPO]

Falsifiable does not mean logically false. Instead, it refers to a prediction or precondition (assumption) of the theory that can be experimentally tested, and at least one of the possible experimental outcomes can cast significant doubt on whether the assumptions are valid, or predictions are upheld.

For instance:

The theory of special relativity predicts that nothing with mass can be accelerated to c. If at any point a well-performed and reproducible experiment demonstrates the acceleration of a massive object to c or greater, then the theory has been falsified, and must be thrown out or refined.

Or, there was a point when it appeared that simple iron salts were observed to serve as phenomenal catalysts for a class of cross-coupling reactions, and several theories were developed to explain this unforseen phenomenon. Eventually studies came forward indicating that the activity was related to ppb to ppm levels of copper contamination. The assumption that iron was the active component was invalid, and therefore all of the theories developed to explain the activity of iron were also invalid.

[evan_au]

One candidate for a TOE is String Theory. However, investigators into String Theory are still trying to fit the multiple undefined parameters of String Theory into the universe we see, so it can't even describe phenomena that we do understand (like electrons and protons), let alone things we don't understand (eg Dark Matter). So as a predictive theory, it currently fails. But one day, it may be useful; we just aren't there, yet.
See: https://en.wikipedia.org/wiki/Theory_of_everything

Something which claims to be a Theory of Everything (TOE) must be able to make predictions about many things (indeed, about "everything"). So there should be many ways to test it, ie it is falsifiable.

So in principle, a TOE should be falsifiable.
See: https://en.wikipedia.org/wiki/Falsifiability

[yor_on (OP)]

It has to be deeper than that folks.


"A statement, hypothesis, or theory has falsifiability (or is falsifiable) if it can logically be proven false by contradicting it with a basic statement.[1][2] For example, the claim "all swans are white" is falsifiable since the basic statement, "In 1697, during the Dutch explorer Willem de Vlamingh expedition, there were black swans on the shore of the Swan River in Australia"[3] contradicts it. The concept is also known by the terms refutable and refutability. "

Notice that this statement actually is found false.
https://en.m.wikipedia.org/wiki/Falsifiability
=

Reading about it strongly reminds me of that theorem in where common mathematics contain its own negation, meaning that every true calculation also contain its own disproval. Don't remember the guys name now but it's a accepted theorem in mathematics, and truly weird, isn't it :)

so?
What does 'falsifiability' stands for?

[Kryptid]

I have to echo what everyone else has said. It seems pretty straightforward to me. Falsifiability is simply the condition where it is possible to determine if a theory is false via experimentation and observation. That doesn't necessarily mean that it will be false.

[yor_on (OP)]

You didn't read the citation if that is what you think Kryptid. Which doesn't mean that I disagree with you :)

"A statement, hypothesis, or theory has falsifiability (or is falsifiable) if it can logically be proven false by contradicting it with a basic statement."
=

either we have statements, theorems etc that can be proven to be a truth, or we don't?
If everything contain its own negation, what would then be true?

anyway, I think its a accepted demand of a theory, that you should be able to falsify it. It might be semantics though, I'm not entirely sure of that one.

[Kryptid]

You didn't read the citation if that is what you think Kryptid. Which doesn't mean that I disagree with you

Are you talking about the swan quote? "All swans are white" is a falsifiable statement, but that doesn't necessarily mean that it is false. A world where all swans are indeed white would be a possible world (just not this particular world). In that case, the statement "all swans are white" would be a true statement but, from the non-omniscient theorist's point of view, it would still be possible in principle to detect the presence of non-white swans if they had existed in that world.

either we have statements, theorems etc that can be proven to be a truth, or we don't?

In science, I don't think we have proofs in the way that we do in mathematics. It's more a matter of gathering evidence that is consistent with a given truth while being inconsistent with competing models.

If everything contain its own negation, what would then be true?

I don't see how they do. I know you said something about mathematics doing that, but I'm unfamiliar with it.

anyway, I think its a accepted demand of a theory, that you should be able to falsify it. It might be semantics though, I'm not entirely sure of that one.

If a theory is wrong, it should be possible to discover that it is wrong. Otherwise, it isn't scientific. If it is true, then you just won't ever discover anything wrong with it. It will keep passing attempts to falsify it.

[yor_on (OP)]

Ok, got it. the wikipedia express it illogically, using semantics. Well, to my mind then. To falsify something simply means that for any theory to be valid it should be able to withstand probing of it. And the way you probe a theory is by it having ways to be probed. So you set up a question that is inherent to the theory, then test it. If the answer to that question prove the theory to be wrong then you've falsified the theory. Actually, the logic here also states that you 'falsified' it even when finding your probes to validate the theory. Which is kind of confusing semantically (as well as logically), isn't it :)
=

And no Kryptid, was referring to first sentence there.
"A statement, hypothesis, or theory has falsifiability (or is falsifiable) if it can logically be proven false by contradicting it with a basic statement." .. that's what had me confused. that as there is a demand for theory's to be able to be 'falsified'.

[yor_on (OP)]

It's this I was thinking of Kryptid
https://plato.stanford.edu/entries/goedel-incompleteness/

And as I seem to remember it actually prove that mathematics contain its own negation/opposite for lack of better words. Long time ago though :) need to revisit it.
=

Here is something very readable about Turing and Gödel
http://chadpearce.com/home/BOOKS/221305209-Thinking-about-Godel.pdf

[yor_on (OP)]

Possibly one can simplify it even more?

As in stating that any idea, theory, hypothesis that you can't probe, by its very nature must belong to beliefs. That's what falsifying means to me,  'testability'.  But calling that falsifying gives me the wrong image.

[alancalverd]

Falsifiable hypotheses in physics seem to boil down to two types: A > B or A → B. One measurement exceeds another, or one event precedes another. Nothing to do with semantics, beliefs, or anything other than physics itself.

[jeffreyH]

Physical theories always involve numbers. These numbers are approximations. You can get exact solutions to a problem but you would need to contrive the data carefully. You can derive mathematical proofs, but that is an abstraction. Experiments deal in approximations. This is why proof is never a goal. We cannot get total accuracy. It is far easier to invalidate a proposition. The experimentation only has to show a significant deviation from the expected result. This is where statistical significance comes in. https://en.wikipedia.org/wiki/Statistical_significance

[alancalverd]

Hmm. Theories involve quantities and interactions rather than numbers. A > B is independent of any chosen basis of measurement and can always be falsified by experiment. Likewise  A → B. Neither statement involves an approximation: a single contrary observation will disprove it.

In the words of Rutherford "if you need statistics, you should have done a better experiment".

[jeffreyH]

How do you count your quantities?

[jeffreyH]

They should have done a better experiment.
https://www.hindawi.com/journals/physri/2010/808424/

[chiralSPO]

In the words of Rutherford "if you need statistics, you should have done a better experiment".

While I appreciate the sentiment on some level, I wouldn't use a single snarky comment of a notorious jackass (no matter how famous) to frame the discussion of epistemology in the physical sciences.

 Statistics is an integral part of the scientific method for any quantitative pursuit. As a chemist, I am blessed with having uniquely large datasets--who else can perform 10²⁰ identical experiments simultaneously? (They're not independent, so while precision is instrument-limited, accuracy depends on my not having done something stupid)

Effect sizes can vary widely from one experiment to the next, and careful statistical analysis can be very useful in determining whether or not one can reject a null hypothesis when the data are somewhat noisy (and it can be prohibitively expensive and time consuming to gather enough data for the results to speak for themselves as signal-to-noise often tracks with (datapoints)^–2, so one would have to perform an experiment 4 times as big to get twice the clarity.

There are also those who don't have the luxury of being able to experiment at all (purely observational research), like astronomers and astrophysicists, who (as far as I know) cannot manipulate stars and planets to set up good control experiments and instead must rely on finding examples of different scenarios to observe (which, depending on the rarity of the scenario, or how easily observed they are, can sometimes end up with only one or two observations of a phenomenon... eeek!)

Of course one has to beware the abuse of statistics. Methods like p-hacking, goal-oriented data filtering/selection, and fundamentally biased studies can generate beautiful graphics and clear conclusions--but cannot (meaningfully) push the field forward.

[alancalverd]

How do you count your quantities?

That's experiment, not hypothesis.

Same applies to Chiral's point. We can predict that molecule X has a lower potential energy than Y  or Z so should dominate the reaction product: the physics says A > B and the hypothesis is falsifiable by experiment. The reason I gave up organic chemistry was because the elegant and obvious physics was always falsified when I did the actual chemistry!

[chiralSPO]

The reason I gave up organic chemistry was because the elegant and obvious physics was always falsified when I did the actual chemistry!

Yup! Ain't it always the case? Of course, this is precisely why I chose to study organic chemistry, and left the physicists to their in silico calculations of perfectly rigid, spherical cows in vacuo!

[Bogie_smiles]

The reason I gave up organic chemistry was because the elegant and obvious physics was always falsified when I did the actual chemistry!

Yup! Ain't it always the case? Of course, this is precisely why I chose to study organic chemistry, and left the physicists to their in silico calculations of perfectly rigid, spherical cows in vacuo!

Nothing wrong with that, as long as its turtles all the way down.