How do know identify crackpot from real physics

[aeroboyo]

How do you keep it 'real' in physics?

Seems like some of the directions of modern physics are going into higher and higher dimensions like string theory for example. There's so many crazy sounding theories out there. How do you know identify a down to earth, well grounded theory from something that's 'crackpot' or 'cranky'?

I'm thinking about studying theoretical physics, so this is an important issue to me. I guess I'm asking about what are the axioms of a well grounded theory. From my very limmited understanding, a few such axioms might be:

- a theory shouldn't lead to contradictions with itself
- a theory should begin by stating the assumptions it makes

Thats about all i know so far about keeping it real. I'm tempted to add that, a theory shouldn't be drastically at odds with current scientific 'consensus' i.e, it shouldn't say that GR is wrong, or that other well proven theories are wrong. BUT I'm hesitant to write that because, that would rule out all revolutions in physics as being cranky until they're proven.

 

[neutrino]

For starters, you can use John Baez's Index

http://math.ucr.edu/home/baez/crackpot.html

 

[aeroboyo]

I've read his index. It makes some good points but it isn't written as something which is that serious. I'm hoping there is a more scientific, rigerous, text-book kind of guidline for what makes a theory a viable theory. The 'a theory can't lead to self contradictions' is an obvious one.

 

[neutrino]

It should come up with a good explanation(better/more "elegant" than the existing one) for things that have already been observed before making predictions.

 

[aeroboyo]

Perhaps another is that it should lead to testable predictions. How do you define better or more elegant. Those are very subjective terms... do you mean it should be mathematically simpler? Just today for example i was reading a thread here about DeSitter spacetime as a viable alternative to Minkowski spacetime. I don't know enought about DeSitter space to say that it's not better or more elegant, but it certainly doesn't seem simpler. Although if i understand it correctly, it predicts that the universe should be expanding which isn't something Minkowski space describes.

 

[Andre]

For starters, you can use John Baez's Index

http://math.ucr.edu/home/baez/crackpot.html

Supplemented by Carl Sagan's baloney detection kit

 

[neutrino]

perhaps another is that it should lead to testable predictions.

That may be debatable. Though this point is often stated, I'd like to know what exactly is a "testable" prediction. Testable today? Tomorrow? 10 centuries into the future? String theorists say stuff that may take quite a while to be verified. I'd like someone to enlighten me.

 

[aeroboyo]

Well i would think that a theory can't be considered anything but 'cranky' until it leads to a testable prediction. Once you start going of into the world of maths without any testable grounding in reality (like string theory) how is that any different from a dog chasing its own tail.

 

[Andre]

That may be debatable. Though this point is often stated, I'd like to know what exactly is a "testable" prediction. Testable today? Tomorrow? 10 centuries into the future? String theorists say stuff that may take quite a while to be verified. I'd like someone to enlighten me.

Here is a major problem. An idea lead to logical predictions but whenever those prediction proof to be correct, it does not proof that the hypothesis is correct. You cannot proof ideas, hypotheses and theories, you can only falsify them, check out the leading philosophy on that of http://www.eeng.dcu.ie/~tkpw/. If after arduous attempts, refutation stays out, and after many successful predictions, the idea /theory may obtain the status of physical law until it is refuted.

That's the tragedy of science, the slaying of a beautiful hypothesis by an ugly fact... Thomas Huxley

 

[aeroboyo]

So General Relativty for example, it isn't proven per se, it's just accepted as physical law because it hasn't been falsified yet AND has made many correct predictions.

That would explain the neccessity of peer review... giving many people a chance to falsify your theory. That makes me wonder what the goal of science SHOULD be. If it's true that you can never prove a theory, then science shouldn't strive for an ultimate theory of everything... that would be pointless. What would be usefull is to keep refining and improving our idea of reality so that we can get something useful out of it.

Maybe i should read 'The Logic of Scientific Discovery' by Popper?

 

[tehno]

How do you keep it 'real' in physics?

Seems like some of the directions of modern physics are going into higher and higher dimensions like string theory for example. There's so many crazy sounding theories out there. How do you know identify a down to earth, well grounded theory from something that's 'crackpot' or 'cranky'?

One has to have a good doseage of scepticism.
I ,for instance,don't believe anything without experimental confirmation.
Theory which makes sense but is still without experimental confirmation I call a (mathematical) model.Just that.

 

[aeroboyo]

Ok so a theory which seems to make sense but isn't experimentally tested is a mathematical model... but even once its tested experimentally it then gains the status of theory, and might even gain the status of physical law... but it will always be a theory at best.

Makes you wonder if... we could somehow postulate all possible theories and disprove all but one, then only then could you say that that theory isn't a theory but reality.

 

[neutrino]

Here is a major problem. An idea lead to logical predictions but whenever those prediction proof to be correct, it does not proof that the hypothesis is correct. You cannot proof ideas, hypotheses and theories, you can only falsify them, check out the leading philosophy on that of http://www.eeng.dcu.ie/~tkpw/. If after arduous attempts, refutation stays out, and after many successful predictions, the idea /theory may obtain the status of physical law until it is refuted.

Thanks for the link. I've heard about Popper but never actually got to read (about) any of his works.

 

[aeroboyo]

I think Popper might be healthy for a physicist to read, if it gives us a way to 'pinch ourselves' at times to make sure we keep it as real as possible.

 

[Danger]

The way I see it, science is like Nature in evolution. It's avowed purpose is to destroy itself, and that which survives the attempts is worthy of existence.

 

[robphy]

http://insti.physics.sunysb.edu/~siegel/warning.html (advance by clicking the lower-right corner of the webpage)
http://insti.physics.sunysb.edu/~siegel/quack.html

 

[arildno]

It is worth noting that several scientists express a certain dissatisfation with, say, string theory, because it does not produce enough testable predictions (as yet).

While the mathematics behind it is acknowledged, those scientists are not convinced that the theory is more than a mathematical fantasy, i.e, that it also can be said to describe something "real", in want of a better word.


However, one of the surest signs of crackpottery is the lack of anything resembling a logical framework like maths.
This is a clear distinction between what we could call scientific speculation and crackpottery.

Just another thing:
It would be wrong to regard scientific speculation as a border-line case between crackpottery and science proper; rather, scientific speculation is an essential feature of the research front, something that cannot be dispensed with when we are charting new, unkown territory.

 

[Andre]

scientific speculation or crackpottery?

Perhaps a good excercise in judgment

 

[vanesch]

It is worth noting that several scientists express a certain dissatisfation with, say, string theory, because it does not produce enough testable predictions (as yet).

The problem with things like string "theory" is that they are simply half-understood speculative ideas which are not yet entirely wrapped up in a "theory" (something with clear hypotheses, a mathematical framework, predictions and all that). Every theory, in embryonic stage, goes through that phase: there's a rough idea, some things don't work out (yet?) etc...
The question is: how long, and how much effort can be spend in this stage before it gets the smell of crankish ? Difficult to say.

Just another thing:
It would be wrong to regard scientific speculation as a border-line case between crackpottery and science proper; rather, scientific speculation is an essential feature of the research front, something that cannot be dispensed with when we are charting new, unkown territory.

Exactly. The question is: when should one draw a line ?

 

[arildno]

The problem with things like string "theory" is that they are simply half-understood speculative ideas which are not yet entirely wrapped up in a "theory" (something with clear hypotheses, a mathematical framework, predictions and all that). Every theory, in embryonic stage, goes through that phase: there's a rough idea, some things don't work out (yet?) etc...
The question is: how long, and how much effort can be spend in this stage before it gets the smell of crankish ? Difficult to say.

In this case, believing you are correct in your description (I haven't read any S.T, so I wouldn't know), then the element of "hand-waviness" and lack of mathematical rigour reduces the gap between the "theory" and crack-pottery.

 

[Dr Transport]

My definition of crackpot is:

If you can show good logical evidence that they are wrong/misguided and they still do not believe you, you have a crackpot on your hands and the best thing to do is walk away.