Theoretical Physics: Investigating the Fundamental Laws of Nature

[Hertz]

Hi,
So what exactly does somebody who is involved in "theoretical physics" do at their job? This is not the same thing as somebody who is involved in experimental physics, right?

If I understand right, theoretical physics is the stuff we learn in class. It's the theory to explain the phenomena, rather than the actual process of detecting the phenomena. Of course, you take lab classes to get experience in the experimental part of physics as well. I understand (somewhat) what somebody who is doing work in experimental physics does; they pick some very specific mysterious topic and try to collect data to learn more about it. So then does that mean that someone doing work in theoretical physics would then look at that data and try to develop a theory to explain the results? Does he get paid a salary, or hourly? What does he do at work? I can only imagine him sitting there staring at a blank piece of paper until he thinks of some awesome theory. (Yes I know that women can be theoretical physicists too :) ). What if he doesn't come up with any good theories?

Personally, I like studying theory. I wish that I could just keep studying theory continuously. And eventually I would like to be able to contribute somehow by coming up with theory of my own and publishing it. But what path do I need to take to follow these dreams? All I see right now is opportunities to work in a lab and collect data and stuff, but that's not what I'm passionate about. Maybe I'm underplaying the whole experimental route, I don't have any experience of my own so my opinion could turn out to be kind of ignorant. I appreciate any insight you can give me.

 

[Arsenic&Lace]

"Do they look at data and explain the results"
Normally. Sometimes they are ahead of the experimenters (anti-matter, general relativity, many other instances), sometimes the experimenters are ahead of them.

"How are they paid?"
I think it's a salary possibly augmented by grants but I'm not really sure.

"What if he sucks at coming up with theories?"
Many theorists spend their time computing the consequences of theories, and in fact few of them actually successfully theorize anything, at least as far as I can tell. It's pretty rare to get a law named after your or to develop a major new method in physics, especially the further into the fundamentals you get. So there are other ways to contribute than to make the major milestones.

"I don't think I like experiment"
Well very few physics majors even become physicists, and fewer of those become theorists, so I suggest you look into experiment, especially since some branches, such as materials physics, transition into industry fairly easily. A good middle ground I'm personally pursuing is computation, where the focus is exclusively on the utilization of powerful computers to do Step 2 of the theoretical physics process, which is to compute the consequences of somebody else's theory.

 

[Simon Bridge]

Theoretical physicists spend a lot of time with computers.
Generally, as a theoretical physicist, your job would be to identify unsolved problems in physics and attempt to solve them. In the private sector these problems would be given to you while in academia you may actually get to choose.

This can mean that you start out with a bunch of data and you have to figure out what sort of process the data came from. You can also come up with a novel theory and then go look for evidence for it in a lot of data that someone else has collected.

Studying existing theory is a necessary but small part of the job - you have to come up with new theories, which is actually quite easy, but they also have to be improvements on the existing ones - and useful - which is a lot harder. You can go some way, though, just finding holes in other people's theories.

You will fail a lot ... and I mean a LOT. This is normal and expected - in fact ruling out a particular theory "the following sort of model is untrue" is often very useful. In science, more is learned from failure than from success.

However - there is no getting away from the fact that at some stage someone will have to do an experiment.

 

[radium]

I'm a first year grad student in condensed matter theory so I can give you some example from what I've seen.
A lot of the analytical work done involves looking at new phenoma from modifying existing theories. For example, the field of topological insulators was strongly catalyzed do by a paper on the quantum spin Hall effect in graphene. Here the authors realized that when you include a spin orbit term, you can invert the bands which makes the system topologically nontrivial and causes you to have surface states in multiples of two in which the up and down spins travel in opposite directions (a spin current) while the system is insulating in the bulk of the material.This phenomena is a manifestation of time reversal symmetry and Kramer's theorem in a system where the spin orbit acts like magnetic fields for the spins but time reversal is preserved overall in the system. The physics in the bulk indicates that the surface states have to be there because of the topologically nontrivial wave function.

 

[bigfooted]

I'm involved in theoretical physics. These are some of the questions I'm thinking about:
How are charged particles formed in a flame?
How does vortex shedding influence heat transfer?
How does vapor coming from an evaporating fuel droplet influence turbulence?

I get paid to work on these problems because they will lead to better products and more money for the company. The last problem was part of my PhD research where I studied particle-turbulence interaction in spray flames. It led to a model that increases the accuracy of simulations of for instance Diesel sprays, so you can design better car engines.
All of these problems are investigated by doing experiments and simulations to create a better theoretical understanding.
The people I work with either call themselves computational physicists or experimentalists. They have mastered a powerful tool that will help them in building better theories and models. You need tools to build something...