Theoretical Physics - How to? Although I doubt somewhat whether this question is really appropriate for this site, I hope it gets answered anyways. I guess, what I'm wondering is:


*

*How does one get to work as a theoretical physicist and - probably more importantly - what do theoretical physicist actually do all day long?

*How are theoretical physicists distinguishable from mathematicians? Does a physicists day look very different from that of a mathematician?

*I have a great interest in physics, but I'm not really much interested in doing experiments: Would it be advisable to do my bachelor in mathematics and try to get into theoretical physics later on?

*Is there a real chance of getting into research afterwards? (not that any kind of answer to this question would ever stop me from trying...)
Well, I hope this question is acceptable. 
I think 1) might for example be answered by giving a link to a blog of a working theoretical physicist, who gives some insight into his or her everyday life, or some kind of an essay on the topic. Of course any other kind of answer is greatly appreciated.
Thanks in advance!
Kind regards,
Sam
Edit after several answers: Thanks a lot for all the responses! I found it very interesting and helpful to to get some input from you guys.
Although the opinions seem to differ a bit, one can definitely see many overlaps, too.
I do still have some time do decide, and will definitely look at some books suggested here, visit some lectures and try to get a feel for what it would be like to do either physics or maths.
Thanks again for your effort! :-)
 A: This was originally a comment under ptomato's answer but got too long and became point 3.
As for your questions:


*

*I am still just a student but I guess what you do is you mostly read papers and study books and try to work on some problems (both your thesis and things from physics.SE ;-) )...

*...which is probably the same thing mathematicians do, only the area of research and methods used differ. But I don't think there's a clear distinction. You have people like Witten who won Field's medal but is clearly a theoretical physicist. Now what's that about? :-)

*I don't think being good experimenter is important at all. Sure, it helps. So does good teaching skills, programming skills, communication skills, algebraic geometry skills, ... But none of that is really necessary for theoretical physics. All that is needed is that person likes science and math and has inquisitive mind that is curious about the ways nature works.

*Certainly. I know both of physics majors who got into math research and math majors who got into physics research. I think the required skills greatly overlap. The main difference is level of rigor and amount of physical intuition. You can adapt both to the environment you are working in. Of course, it might be excruciating for physicist to be forced to write formal proof where he just sees the answer, or for mathematician to watch hand-wavy derivations in physics.
Personally, I am more on the math side and I hated experiments. So we'll see if I can succeed in being a good theorist :-)
A: If you want to work as a theoretical physicist, it would be advisable to get a little bit of grounding in experimental physics anyway. So my answer to #3 is, if you want to get into theoretical physics, get a Bachelor's in physics, not mathematics, and take at least one or two experimental courses.
I work at a university where students often do shoddy work in the undergraduate experimental courses, because "I came here to study theoretical physics and I'm not interested in experiments." What they don't realize is that most of our professors in theoretical physics are of the opinion that to be an excellent theoretician, you primarily need to be a well-rounded physicist, with both theoretical and experimental skills. These students run into trouble when they're looking for internships or final projects, because no professor will accept them.
(I edited this answer, because I didn't mean to imply that you had to be an excellent experimenter to be a good theoretician. If you don't enjoy experimenting, then you don't enjoy experimenting. Just don't dismiss it or ignore it altogether.)
From Saturday Morning Breakfast Cereal:

A: Good question. I did my diploma in Theoretical Particle Physics and therefore worked one year as a theoretical pysicist.
To your questions:
1) The short answer: Find the error in your calculations. One guy at our institute once said: "Theoretical physicists spend half of their time in finding algebraic sign errors." This is to a degree true. 
Long answer: It depends on the field you are working at. Actually Theoretical Physics starts at the point the calculations become too complicated for experimental physics. So for the particle physics this is "Phenomenology". (hep-ph on arxiv.org -- don't be surprised if you don't understand anything there) These are the guys that calculate the - theoretical - cross sections for experiments that are tested at the LHC and other colliders. So this is in the end squaring Feynman diagrams, integrating them and analyzing the results. For some problems you can do that completely with pen and paper for other problems it makes sense to do 95% of that work on the computer. Depends on the research group and what they are working on.
On the other hand there is this completely theoretical physics area. Of course they also need to make sure at some point that their work corresponds with the experiments but they mostly concentrate on "model building". (hep-th on arxiv.org) For instance: We are seeing that Standard Model Higgs mechanism is probably not the way nature implemented the Electro-Weak-Force. Well, why don't we check out what happens if we have 2 Higgs fields instead of one? Or what happens if we change the symmetry group? Sounds interesting but you also pay a price: Most stuff you do is really abstract and will probably not be the way nature has implemented things. :D
Of course there is also Mathematical Physics (math-ph on arxiv), but that is usually done in a Mathematics department. Theoretical physcists usually don't have the time to check every mathematical detail. In particular nobody hasn't been able yet to formalize Quantum Field Theory (= Theoretical Particle Physics) like it has been done with non-relativistic Quantum Mechanics or even Newtonian Mechanics.
2) Physicists don't do mathematics in such a formal way. If they we're that detailed as mathematicians are, there wouldn't be able to do actual physics.
The physicists approach is: Let's do this calculation, it looks formally ok and makes sense. If our cross-checks are not consistent, let's check again and pay more attention to the details.
You'll probably want to ask a mathematician how his work day looks like. (stackexchange...)
3) Me not either. And no, please don't do the same mistake as I did by concentrating to much on the mathematics. Except you have enough time to learn all the important stuff about physics you want to work later on.
Make yourself clear: Physics and Mathematics are completely different subjects. Physics is a Natural science. Are you interested in that? Really? Study physics. Of course you need to do some experiments, they are physicists' source of information. But in the end it matters that you are informed about the important experiments that matter to your work.
Are you interested in abstract concepts and the physics doesn't matter to you? Study mathematics.
If you are in between, consider studying Mathematical physics. But make sure your university has a mathematical physics department. I recommend you to talk to students or professors to get some first hand information. Mathematical physics is also really interesting. Take dynamical systems for instance or google for Wightman axioms.
4) Yes of course. Is there a chance to become a professor? Of course, but you need to be focussed and make smart decisions. The process differs from country to country. About Germany I can say it is easy to get into research. But it's not easy to become a professor. Make yourself clear (in one or two years) whether you are a research guy or whether you are just interested about the subject.
Again I recommend you to visit a university and ask as many questions as possible to the mathematicians and physicians there. Most universities have regular events for students interested in studying at their university. At least in Germany there are a lot of student's organizations you can talk to and a lot of mathematics and physics departments have professors that are specialized in helping you with your questions.
EDIT: I want to be more precise about the rigor. Theoretical physicists need to be as rigor as mathematicians. But: Their rigor is focussed on the physics not on the mathematics.
A: First you must ask yourself what excites you.  Is it watching events happen?  Comparing what happens to what is predicted to happen?  Making the prediction?  Understanding the prediction?  Making a model for hypotehtical predictions?  Understanding what constitutes a model?  None of the above?  Unless you're just in the business for the money, your research interests should drive your education as much as possible.  Otherwise, why bother?
In my case, what excites me is the connection between the concrete and the abstract, which is why I always straddled the math-physics fence.
That said, the general pattern of any researcher is the same.  Make observations.  Ask questions.  Compare with previous questions and snwers.  Figure out how to find answers, when none exists.  The methods are very different, of course.  It is true that a theoretician and a mathematician have similar daily patterns, and even similar methods (talking, reading, writing e-mail, working things out on a chalkboard/tablet/whatever).  The nature of the research is different, but there's lots of overlap.  When people pose the "what are you?" question to me, I try to punt.
[Note added:  Back now to finish.  I cut my answer short earlier.]
The main difference between the fields is not so much one of rigor, though that is an obvious one.  The more relevant difference is the questions that drive research.  In mathematics, one tries to understand the structure behind constructions of objects.  In theoretical physics, one tries to understand the structures behind physical models and physical phenomena -- and these can often employ or reveal mathematical structures.  Sometimes, the fields work in parallel, with both camps trying to get to the bottom of things.  "Bottom" in physics means understanding how hypothetical physical calculations (e.g., correlation functions, metrics) are affected.  "Bottom" in mathematics might mean a definition and/or a relation to existing structures.
A: John Baez addresses some of these questions on his page "Advice for the Young Scientist".  I think it is more oriented towards, for example, someone deciding between grad-schools or later appointments rather than choosing between majors at the university, but it may still be useful to you:

Math or Physics?
This is for people who are torn between a research career in math and one in physics.
Nobody can stop you from learning and thinking about both math and physics - you should go on doing both! The real issue is whether you want to work in a math department or a physics department. It's possible to switch from one discipline to another after grad school, but it's not easy, since departments prefer to hire people with an appropriate degree. So, it's wise to decide which job suits you best before you apply for grad school - if not sooner.
To decide, you need to know how these two jobs differ. For this, it's
  best to talk to as many mathematicians
  and physicists as you can, and find
  out what their jobs are like. Talk to
  your professors! It's also great to go
  to some conferences - there's often
  money for students to attend
  conferences. I can only summarize:
Mathematicians get promoted by
  publishing in math journals;
  physicists by publishing in physics
  journals. Read both kinds of journals
  and see which you can best imagine
  yourself publishing in. Spend time in
  a good library and browse. I spent my
  whole undergraduate career doing this!
  There are some journals at the
  boundary of math and physics, like
  Advances in Theoretical and
  Mathematical Physics and
  Communications in Mathematical
  Physics. Look at these. But also look
  at a bunch of journals that are
  full-fledged physics, like Physical
  Review A, B, C, D, and E, and Physical
  Review Letters, or solidly
  mathematical, like Annals of
  Mathematics or the Transactions of the
  American Mathematical Society. You'll
  see these come from different worlds!
  Mathematicians typically write in the
  definition/theorem/proof style, while
  physicists write shorter papers, and
  more of them, often packed with
  formulas, but usually leaving out all
  the details of calculations.
Physics is a faster, looser, more
  energetic discipline. You are also
  evaluated more heavily on how much
  grant money you can pull in.
Here's another way to put it: do you
  like things to be clearly stated in a
  rigorous way, or do you like to use
  your physical intuition to get to the
  answers?

He also has a page titled "How to Learn Math and Physics".
A: *

*This I don't know. I think that everyone have his own way of doing research... But you'll need to work hard anyway.

*I don't think there is a clear distinction now. There is some "gradient" between those.

*I think that there is a reason beyond teaching students experiment even if they would be theorticians. But matematicians keep coming to physics as if they don't care about my opinion. 

*Everyone I know who tried it -- made it. 

A: 1. How does one get to work as a theoretical physicist
Go to a good university and get a firm foundation in math and physics.  Go to graduate school in theoretical physics.   Then apply for post-doc and faculty positions; repeat until you get a faculty position.
3. Would it be advisable to do my bachelor in mathematics and try to get into theoretical physics later on?
Do whichever you find yourself liking better.  A background in math could be excellent preparation for theoretical physics.  You will either do "mainly physics with a lot of math" or "mainly math with a lot of physics." 
4. Is there a real chance of getting into research afterwards? 
Of course, if you're good.
A good way to start (right now) is to get yourself a copy of Landau and Lifshitz's Mechanics and work through it.  See where that takes you.
A: I can see theoretical physicists at work every day, and I've also seen experimentalists... My own field is computer simulations. To my understanding the answers to your questions can be like this:


*

*What theoretical physicists do all the day long? Calculate and think.

*Physics vs Math? Physics is more intuitive than mathematics, and there is more freedom in physics than in mathematics. There is also much less abstraction in physics.

*Maybe yes, maybe no. This strongly depends on what exactly kind of physics you want to do.

*Yes, of course. But one has to work really hard for this to happen. Usually everything depends on the level of one's publications. And - in my opinion - if one loves physics he has a REALLY good chance of having good achievements and, as a consequence, publications.

A: I was in the process of writing several comments all over the place, and realised it would be better structured to put them together as an answer. I answer specifically questions 3. and 4. of your question.
$\\$
"3. I have a great interest in physics, but I'm not really much interested in doing experiments: Would it be advisable to do my bachelor in mathematics and try to get into theoretical physics later on?"
$\\$
Firstly, my response to some of the answers above encouraging you do a physics degree, including the accpetd answer by ptomato, is do not listen!
$\\$
I realise that it may be hard to do before you've even gotten an undergraduate degree, but if you can try to decide which area of theoretical physics you think you might like to work in, it would help you greatly to decide how to specalise your undergraduate degree.
By that I mean that theoretical physics is a large area of research these days. I have some friends who switched into mathematics to go on to become string theorists, black hole physicists, cosmologists, particle theorists, quantum information researchers etc. as well as fields outside of theoretical phyiscs like meteorology, many kinds of maths biology (from cancer to nuroscience to informatics), genomics etc. On the other hand, some, such as myself, stayed with physics to go on and become cosmologists, astrophysicists, computational physicists, geophysicists, complex systems physicists etc. all doing theoretical work.
(I should say that we are only at PhD student age now, not university lecturers or anything like that, but this gives you an idea of the many fields that you can go into).
What you'll notice is that absolutely none of us went into condensed matter physics, a very large area of theoretical physics. And this is why I highlight ptomato's answer as terrible terrible advice; If you fore people to do things again and again that they are not enjoying, they will more likely learn to resent it, than grow to love it! His claim that 
$\\$
"These students run into trouble when they're looking for internships or final projects, because no professor will accept them" 
$\\$
is pure and utter rubbish, again, in my humble opinion, which is based on the experience of my fellow class-mates. In fact, I vividly remember one physics lecturer saying that he prefers to take on the maths-physics students for PhD research. Using a metric of "number of students who went on to further research", the maths-physics degree had the highest number, by a long shot!
The university I went to is quite big in condensed matter physics, and there was always a high focus on this in our lectures and labs (labs were compulsory in all years. We had no choice in our lecture courses. I believe this is not common in some other countries, so maybe it is not a worry for you?). Many of us did not like it, which resulted in over half switching into pure maths. The only reason I did not was because I was stubborn, proud and arragont (which, as a side piece of advice, don't be stubborn, or overly proud etc. I can tell you, you won't get anything out of it, and will loose a year playing catch-up to your more humble maths-physicists friends who took the leap!).
$\\$
From the sounds of your comments, it is obvious that you really like maths. So why not do a maths-physics degree? That way you'll see both sides, learn lots of interesting things from both departments, and keep your options open.
I just repeat that doing a straight physics degree, in my opinion, given that you want to become a theoretical physicists, would be a very bad idea.
$\\$
4.Is there a real chance of getting into research afterwards? (not that any kind of answer to this question would ever stop me from trying...)
$\\$
For your second question, again, I can only offer an answer from the point of view of a current graduate student.
To start, that attitude is definitely the right sort of one to have! People will tell you that it's very hard, which it is, and that you'll need a bit of luck, which you will, but most of all, you need to not give up before you've even begun!
As far as how likely it is to get into research, again, for starters, it depends on what you want to do. I have friends who switched into maths biology, and money is thrown at them. Others who want to be string theorists and algebraic geometrs, not so much!
Moreover, it depends on your country. I come from a country with very little funding in fundamental physics (theoretical particle physics, string theory, maths-physics etc.). Most of the funding goes to materials physics, experimental physics, computational physics etc. 
I now like in a country with more funding for maths-physics, which is great, but they are a little hostile to forigners, and funding for foreign PhD students and post-doc researchers is far harder to get than for nationals. But there is still some money out there, so it's not too bad.
In other countries, the US, Netherlands, Germany mainly, there is lots of funding, if you're good, and most PhD students once accepted get funding, and post-doc researchers even get a good wage (relative to where I am now, in case that angers anyone!).
$\\$
For now, my advice would be to hold on to your ambition, and work hard at your undergraduate degree. Assumedly you have several years before you need to think about these things.
$\\$
And with that, I wish you the best of luck!
