- #1
potato123
- 13
- 0
I heard that in order for something to become a theory it must be tested.
A theory does not need to be tested in order to be theory.potato123 said:I heard that in order for something to become a theory it must be tested.
Demystifier said:A well-defined theory needs to have a well-defined set of general principles and assumptions, from which everything else should be derivable, at least in principle. String theory lacks this property, so people sometimes say that string theory is not really a theory, but only a theoretical framework.
That is not relevant. To see this, one can ask: according to standard QED, would there ever be a way to "directly" observe an electron? The answer is clearly no, and yet QED is extremely successful as a quantum theory. The point is that whether a particle is "small" or not (this needs of course to be defined, given that we are discussing quantum physics) is not directly related to whether the theory can be tested or not.serp777 said:Seems more like the string religion. No predictions, not that much evidence, and no apparent way to observe something 50 orders of magnitude that small. I mean considering it seems almost impossible to observe something on the 10^-32 meter magnitude, it seems even more impossible to observe something as small as a string. .
nrqed said:That is not relevant. To see this, one can ask: according to standard QED, would there ever be a way to "directly" observe an electron? The answer is clearly no, and yet QED is extremely successful as a quantum theory. The point is that whether a particle is "small" or not (this needs of course to be defined, given that we are discussing quantum physics) is not directly related to whether the theory can be tested or not.
That's a fallacy, also known as the false analogy. Math has concrete results in a variety of fields and it is used in physics to make predictions. The usefulness, productivity, and efficacy of math makes it completely different from string religion. Furthermore, are you seriously putting string religion and mathematics on the same level? A language is simply a way of communicating information and coming to conclusions based. String theory is a description of reality whereas mathematics is used to make theories which are descriptions of everything. It would be like asking if English was a religion because it was used to construct the bible. No sorry, that makes no sense.haushofer said:Is mathematics also a religion? Your labeling depends heavily on the fact that you classify string theory as a theory in physics. I think the border between physics and math is vading.
serp777 said:That's a fallacy, also known as the false analogy. Math has concrete results in a variety of fields and it is used in physics to make predictions. The usefulness, productivity, and efficacy of math makes it completely different from string religion.
Not necessarily. To be a valid theory, it must be tested and the results of the test shown to agree with the theory.potato123 said:I heard that in order for something to become a theory it must be tested.
It is not the theory of everything. It is the theory of anything.atyy said:String theory is not "a" theory. It is the theory. :P
serp777 said:Seems more like the string religion. No predictions, not that much evidence, and no apparent way to observe something 50 orders of magnitude that small. I mean considering it seems almost impossible to observe something on the 10^-32 meter magnitude, it seems even more impossible to observe something as small as a string. There's no good reason to believe that string theory is true at all. Its ambiguity and unfalsifiability make it equivalent to Yahweh or whatever. It also seems like there are way to many different versions of string theory, which makes it seem like the different sects of Christianity. I guess it does have one prediction though--a multiverse.
stevendaryl said:Are Newton's laws of motion falsifiable? I would say they are not. They become falsifiable when you add a specific hypothesis as to the nature of the forces involved in a particular problem. But in the absence of the knowledge of what forces might be relevant, to say that force is proportional to acceleration is consistent with absolutely any motion. Newton's laws are a framework that can be used to form a falsifiable theory, but it isn't a falsifiable theory in itself.
Haelfix said:Sometimes a subset of a given theories parameter space outputs a model with testable consequences, but other parts of that space are not verifiable thus nonverification of the former simply entails exclusion limits.
Don't apologize. These are important questions for physicist. Otherwise we are just sophisticated bookkeepers :PHaelfix said:Apologies this is veering off into metaphysics, but there is another sociological danger with too much reliance on 'falsifiability' and that is what is known as lamp post physics.
The analogy goes as follows:
Suppose you were looking for the proverbial needle in the haystack and you knew it was located somewhere in a dark street. Now at some point in time a street light would turn on, illuminating a small portion of the street. I could make a theory (or a set of theories) describing in detail how the needle (or set of needles) happened to be right where the light would illuminate. A perfectly sound, falsifiable idea.
But in the absence of any other reason to be there, my theory has no new information content even though I have now given myself some percentage of chance for having correctly described the situation (and winning an award). Meanwhile the idea that the needle might be closer to the seamstress's house seems to be a better idea, although of course it is completely unfalsifiable (b/c it stays in the dark).
A good example is general relativity. There, energy-momentum tensor of matter does not obey global conservation. But then one invents an ugly non-covariant object called energy-momentum pseudo-tensor, which does obey global conservation.Ben Niehoff said:Same goes for "energy". It's happened over and over again: Study some system in detail and find that energy seems to be lost? Just postulate a new kind of energy! Now energy is just changing form.
"Energy is conserved" is more or less a tautology; it only requires that you define energy appropriately.
Or to paraphrase Rutherford, otherwise physicists risk to become stamp collectors.haushofer said:Don't apologize. These are important questions for physicist. Otherwise we are just sophisticated bookkeepers :P
Haelfix said:All of these are examples of specific models which are constructed such that their predictions would be testable, and are subsets of much more general frameworks. They are all basically examples of lamppost physics.
This was the initial situation. Later it has become a theory. The papers of Heisenberg and Schrödinger, together with the Born rule which clarified that the wave function defines the probability have transformed this vague old "quantum theory" into a real physical theory.Smattering said:Hm ... was quantum theory derived from a well-defined set of general principles and assumptions in the first place? When I look at the history it rather seems that it was extended ad-hoc in some step-by-step manner.
Ilja said:This was the initial situation. Later it has become a theory. The papers of Heisenberg and Schrödinger, together with the Born rule which clarified that the wave function defines the probability have transformed this vague old "quantum theory" into a real physical theory.
No. Results about planets have been derived a long time using Newtonian mechanics. Then GR came, and after this all what we think about planets has been rederived based on GR. This will happen again if some better theory of gravity will be found.Smattering said:That's true. But you cannot change the original derivation of something at a later point in time. If a result was originally derived in a certain way, that will will stay the way it was originally derived until the end of time. It might turn out that there are additional ways to derive the same result, but that does not change the original derivation.
Ilja said:No. Results about planets have been derived a long time using Newtonian mechanics. Then GR came, and after this all what we think about planets has been rederived based on GR. This will happen again if some better theory of gravity will be found.
The same holds, of course, for quantum results. What has been derived in old quantum "theory" has been rederived later based on quantum theory as we use it today.
And, given that many quantum results are only approximate, they are recomputed all the time if a better approximation method or simply a more powerful computer or computation program appears. And in all these cases, the former derivation are no longer interesting except for history.