Why should we work with gauge theories

[the_pulp]

1) In one thread I saw that a Lagrangian that comes from a gauge theory principle is capable to generate interactions, and that would be why we should work with gauge theories. Nevertheless, any lagrangian which have multiplications of diferent fields generates interactions (or am I wrong?)
2) In some books I read that gauge theories are renormalizable, but some non gauge theories can be renormalizable too (so that would no be the reason either)
3) Finally, Symmetrys generates conservation of observables, so every observable that conservates after an interaction should have an undelying symmetry behind. And as we (well, to be honest "you") make experiments where we ("you") scatter particles and see what conservates, this should be the reason.
Is the reason 1), 2) , 3) or another?

Thanks!

 

[mike372]

I think the real reason is that they offer a very concise theoretical description of nature that AGREES WITH EXPERIMENT. After all this is THE key requirement of a successful theory of nature.

Since the advent of the standard model, it has been realized that the em, weak and strong force all can be described by gauge theories and this led to predictions that have since been verified, such as the running of the strong coupling, the ratio of W to Z mass etc. If you did not assume a gauge theory you would have no reason to understand why such quantities behave as they do. So there is now just a huge weight of evidence that this is the correct theory of nature.

In the early days, the fact that they are renormalisble was also seen as strong evidence but in the days of effective field theories this is no longer seen as such an important requirement for a low energy theory.

 

[the_pulp]

Ok, so, to sum up, the answer is that gauge theories are pretty (I really think they are). But something sounds to me not enough. In fact, Gauge theories only worked twice (electrodinamics and cromodinamics). For electroweak, we (in fact, "you") are not so sure that theory is working. But I don´t know, I thought that there was something more conclusive (isnt it the 3rd point I mentioned in the other msg).

Thanks!

 

[tom.stoer]

Gauge theories work rather well for the electroweak theory, too. In addition general relativity can be reformulated as a gauge theory (the dynamics looks different, but many structures are identical). So b/c there is nothing else in nature we have to describe I think gauge theories made a very good job.