what are the requirements of a scientific theory?

[univeral theory]

i have been reading about scientific theory possibly here https://en.wikipedia.org/wiki/Scientific_theory and i have failed to conceptualise well the requirements of a scientific theory. can some one out there help to throw more explanation on this for me?

[mathematic]

One fundamental rquirement is that predictions of experimental outcomes can be tested.

Edited July 4, 2013 by mathematic

[EdEarl]

[Delbert]

I think Richard Feynman summed it up perfectly

 

I think he said something like:-

1/ You make a guess

2/ you compute the consequences of said guess

3/ you compare the consequences of said guess with experiment or experience

4/ if it doesn't compare with experiment or experience then the guess or theory is wrong. It doesn't matter how beautiful it is, how smart you are or the name of the person who said it, it is wrong.

Edited July 30, 2013 by Delbert

[Griffon]

Some nice explanations so far.

 

I would add that a scientific theory should be falsifiable. It should be possible in principle to find an observation that shows the theory to be untrue. That doesn't mean the theory is false.

 

Some explanations are not falsifiable in any known way, like supernatural explanations for example, so are not scientific.

[Strange]

This might be relevant: http://blogs.scienceforums.net/onablog/2012/10/15/an-open-letter-about-your-new-theory/

[Griffon]

Am I alone in thinking that many physicists are sloppy in their use if the word theory? As we've seen a scientific theory has quite a tight definition. Yet the word theory is often used to describe explanations weakly supported by evidence, such as some recent theories in cosmology, to the pillars of modern physics like quantum theory and relativity. In part I think this is why 'theory' has such a bad reputation in non-scientific circles. Hypothesis might be a better word, or maybe we need something between a hypothesis and a theory?

[swansont]

Yes, we can be sloppy, because of the mixing of the lay use and scientific use of the word in discussions. However, I would argue that a physicist is in a much better position to discern between the two definitions when s/he hears it from another physicist, than from a layperson who doesn't know the difference, or that there is a difference. The problem is in communicating to a lay audience and being sloppy, and the main problem there lies in a lack of training or practice in communicating with a lay audience.

[ajb]

As we've seen a scientific theory has quite a tight definition. Yet the word theory is often used to describe explanations weakly supported by evidence, such as some recent theories in cosmology, to the pillars of modern physics like quantum theory and relativity.

By theory we mean a mathematical model of nature. Even if there is little in the way of evidence that a theory is good, that is describes nature well, it is still a theory.

 

The problem is that the common use of the word is closer to speculation. Most "theories" proposed by laypeople on this forum are not theoreis in the scientific sense. Usually they lack any kind of mathematical framework and have no testiable features.

Hypothesis might be a better word, or maybe we need something between a hypothesis and a theory?

A hypothesis or a conjecture should be based on theory and/or observation. It maybe a better choice of nomenclature for some things, but I would expect just as much confusion as theory.

The problem is in communicating to a lay audience and being sloppy, and the main problem there lies in a lack of training or practice in communicating with a lay audience.

I agree with the sentiment here. I am not sure that all scientific work can be well explained to a general audience. However, more effort should be put into this for sure.

[Phi for All]

I agree with the sentiment here. I am not sure that all scientific work can be well explained to a general audience. However, more effort should be put into this for sure.

 

Trying to explain a scientific concept to a general audience just involves more and more questions, since you're often required to have knowledge of other concepts in order to understand the new one.

 

I'm also amazed at the number of "enthusiasts" who dabble in science and then feel qualified to declare a theory flawed and advance their own in its place. My favorite was the guy who claimed something like, "My area of expertise is purely theoretical physics, and I have a breakthrough concept. Mind you, this is only a theory...." How can someone claim expertise in something they so clearly don't understand?

[Griffon]

Yes. Oh how many times have I found myself arguing with some creationist or whatever who comes out with the famous 'but it's just a theory'? It doesn't seem to matter how patiently you try and explain the difference between a scientific theory and every day use of the word .... they often don't want to understand the difference.

[Ophiolite]

By theory we mean a mathematical model of nature.

Not exclusively. The mathematical aspects of plate tectonic theory, for example, are important - Euler and poles of rotation - but they are not central to the theory. You've spent too much time with cyclotrons.

[ajb]

Not exclusively. The mathematical aspects of plate tectonic theory, for example, are important - Euler and poles of rotation - but they are not central to the theory. You've spent too much time with cyclotrons.

It depends on what exactly you mean by a theory. I would not think of the overall idea of plate techtonics as a theory, in the sense use in physics. The mathematical models used to describe plate tectonics are the theory and these can empirically be tested.

 

But your point is taken and one has to take care depending on the branch of science in question.

[ydoaPs]

i have been reading about scientific theory possibly here https://en.wikipedia.org/wiki/Scientific_theory and i have failed to conceptualise well the requirements of a scientific theory. can some one out there help to throw more explanation on this for me?

Many of the replies have been about what the scientific methods are, rather than what the requirements are for a theory to be scientific. This is a topic in the philosophy of religion largely popularized by Karl Popper, but, despite his popularity outside of the philosophers, he was wrong about just about everything. He was right that falsification is a criterion, but he was extremely wrong when he claimed that it was the only one.

 

There is more or less consensus that Michael Ruse hit it on the head with his criteria in this paper.

[Griffon]

We can't read that article ...

 

Interesting that you say Karl Popper has fallen out of favour. I'd heard that of late. I'm not sure why, or why Ruse is critical of him. Popper was generally highly thought of when I started out in science 35 yeas ago.

[ydoaPs]

We can't read that article ...

 

Interesting that you say Karl Popper has fallen out of favour. I'd heard that of late. I'm not sure why, or why Ruse is critical of him. Popper was generally highly thought of when I started out in science 35 yeas ago.

Largely because of his main two things (falsification as the only demarcation criterion and his anti-inductivist beliefs) being wrong. Yeah, you'll need a subscription (which you need to buy or just log on through a university computer) to see. Try Googling "Creation Science is not Science" and see if some university has a pdf on their website freely available.

[Griffon]

Yes, we can be sloppy, because of the mixing of the lay use and scientific use of the word in discussions. However, I would argue that a physicist is in a much better position to discern between the two definitions when s/he hears it from another physicist, than from a layperson who doesn't know the difference, or that there is a difference. The problem is in communicating to a lay audience and being sloppy, and the main problem there lies in a lack of training or practice in communicating with a lay audience.

Yes. I think that's because working physicists are pragmatists and don't worry too much about what the word theory actually means. They use the term theory to cover everything from a speculative explanation, with maybe a smattering of supporting evidence, to relativity, which has been tested till the pips squeak. I don't think this is good enough. Many so called new "theories" that get reported to fill column inches of New Scientist or Scientific American are actually not theories with anything like the stature of Darwin's theory of the evolution of species, say, or the atomic theory of matter. We might term them somthing like proto-theories.

[swansont]

Yes. I think that's because working physicists are pragmatists and don't worry too much about what the word theory actually means. They use the term theory to cover everything from a speculative explanation, with maybe a smattering of supporting evidence, to relativity, which has been tested till the pips squeak. I don't think this is good enough. Many so called new "theories" that get reported to fill column inches of New Scientist or Scientific American are actually not theories with anything like the stature of Darwin's theory of the evolution of species, say, or the atomic theory of matter. We might term them somthing like proto-theories.

 

I think you have to include journalists in the blame for that, possibly more than the scientists.

[cladking]

Ultimately all theory and all science are based on observation and logic and if they can't make accurate predictions then they are worthless. If they make inaccurate predictions then they are wrong.

[Griffon]

I think you have to include journalists in the blame for that, possibly more than the scientists.

True. So perhaps scientists should be more careful.

 

So, should string theory be considered a scientific theory? Lee Smollin has his doubts. Perhaps it's a branch of mathematics?

 

PS How odd. I keep trying to add this comment on ST as separate post, but it keeps adding it to this post. Doesn't this forum allow you to post twice in a row?

Edited July 31, 2013 by Griffon

[ajb]

True. So perhaps scientists should be more careful.

So, should string theory be considered a scientific theory? Lee Smollin has his doubts. Perhaps it's a branch of mathematics?

I would say that it is a theory, though one in the process of being formulated. String theory does make some general predictions about our world, 10 dimensions, supersymmetry, small corrections to general relativity, a noncommutative nature of space-time, moduli fields etc. The trouble is that right now we do not know what soulution describes our Universe or indeed that sting theory actually contains such a solution. Thus, we cannot be sure about the details and make very specific predictions.

 

So I would say that because of the possibility that the right solution exists, string theory is a theory in the physical sense, just one that is not properly understood.

 

But it may turn out that string theory is not a good theory and does not describe nature well...

Edited July 31, 2013 by ajb

[Griffon]

I see what you're saying. But doesn't string theory make predictions that we can't test, and won't be able to test if ever because of the inaccessible energies required. Mind you, I suppose the same might be said of theoretical considerations of the conditions at the earliest time within the Big Bang. I suppose all we can do when experiments are not possible is to see how well the theory predicts what we see now.

[swansont]

True. So perhaps scientists should be more careful.

 

So, should string theory be considered a scientific theory? Lee Smollin has his doubts. Perhaps it's a branch of mathematics?

String theory will have to pony up sooner or later or be dropped. It's a rigorous investigation and is theory, but is it "a scientific theory" as we consider e.g. relativity to be? No, not IMO. It needs experimental confirmation first.

 

PS How odd. I keep trying to add this comment on ST as separate post, but it keeps adding it to this post. Doesn't this forum allow you to post twice in a row?

 

It automatically merges them. Fewer posts to keep track of in the database. And though not a big issue, it fight post-count trolls.

[ajb]

I see what you're saying. But doesn't string theory make predictions that we can't test, and won't be able to test if ever because of the inaccessible energies required.

Sure, but there are aspects that can be tested at lower energies. Even so, you might have to make some distinction between beeing able to test it now and being able to in principle test it. If there is just no way even in principle it can be tested then it cannot really be a theory.

 

Right now the best hope of "seeing strings" is in cosmology and people are woring in this area. See below also.

 

Mind you, I suppose the same might be said of theoretical considerations of the conditions at the earliest time within the Big Bang.

Right, but the imprint of this physics should be visiable today in observational cosmology.

Edited July 31, 2013 by ajb

[Delbert]

 

I would add that a scientific theory should be falsifiable. It should be possible in principle to find an observation that shows the theory to be untrue. That doesn't mean the theory is false.

A theory is false with just one experiment the contrary.

 

It would be a contradiction in terms to say it should be possible in principle to find observations showing a theory to be untrue, and at the same time saying a theory is true. Indeed, if such were the case, the world would be chaotic.

 

Some explanations are not falsifiable in any known way, like supernatural explanations for example, so are not scientific.

A so called supernatural explanation is a falsification.