# Qualities of a successful theorist

1. Dec 28, 2015

### spaghetti3451

What are the most important qualities that a budding theorist ought to have?

Two of the most important I know of are the ability to finish textbooks cover-to-cover in a short period of time and technical problem-solving skills.

Can you think of any others?

2. Dec 28, 2015

### Vagn

It depends on the field/sub-field. The requirements for a theorist doing density functional theory will likely be different to those of a theorist working on problems beyond the standard model.

3. Dec 28, 2015

### spaghetti3451

Can you elaborate?

How do the requirements for theorists in the two fields differ?

4. Dec 28, 2015

Staff Emeritus
Why do you think that's important?

5. Dec 28, 2015

### Vagn

DFT theorists primarily work in condensed matter physics and chemistry, in contrast a theorist working on models beyond the standard model is working with particle physics and cosmology as a guide. The knowledge and a lot of the mathematics required for each are completely different.
DFT is also more immediately testable in lab based experiments, so a DFT theorist shouldaim to have a decent working knowledge of the experimental techniques used in the fields (e.g. XPS, LEED, various optical techniques, diffraction techniques etc.) and have contact and collaborations with experimentalists.

6. Dec 28, 2015

### spaghetti3451

I think it's important because it shows that the student can absorb vast amounts of material in a very short period of time, which in turn can lead to greater productivity.

7. Dec 28, 2015

### spaghetti3451

I'm not talking about specific skills. I am talking about a very general set of skills that can serve a theorist well in his research career.

8. Dec 28, 2015

### Calaver

From what I've heard, textbooks are rarely read cover to cover, especially in advanced fields. The reasoning is that many textbooks include superfluous information that isn't necessarily important to the overall picture. A PhD in mathematics at my (high) school said she has never read all of a textbook (even for teaching a class). Although it should be noted that some historical or really well written textbooks and most papers are different.

Perhaps someone who has more experience can provide a deeper insight into this?

9. Dec 28, 2015

### spaghetti3451

I say 'cover-to-cover' because any budding string theorist, for example, ought to know Peskin and Schroeder, Muller-Kirsten and Wiedemann, Wess and Bagger, and Polchinski cover-to-cover. I'm assuming that's the bare minimum.

I've read somewhere once that Witten, for example, only became really good in advanced mathematics (that he used later in his research work) after he read a whole of bunch of yellow Springer textbooks (with guidance from Michael Atiyah on what to read) while at Oxford.

10. Dec 28, 2015

### bacte2013

I never understand how it is possible to not read the books (at least in mathematics) from cover-to-cover. I often found that the mathematical books are written in author's unique style and in pedagogical order such that skipping some chapters will present the problem of deciphering later chapters of a reader's interest.

11. Dec 28, 2015

### Choppy

Considering the probability of actually becoming a professional theorist is quite small, I think at least one important quality a "budding theorist" should have is flexibility.

12. Dec 28, 2015

### spaghetti3451

But, textbooks on mathematical methods for physicists tend to have self-contained chapters, or a self-contained chunk of chapters, at most.

The difficulty is mainly with the physics textbooks, I think, but even there, one never needs to read all of the first $n$ chapters to understand the $n+1$ chapter.

13. Dec 29, 2015

People usually don't read textbooks cover to cover since every book has it's strengths and weaknesses. For example, I find that Peskin and Schroeder is helpful for learning how to calculate cross sections per se but almost useless for getting a conceptual picture of renormalization. There are much better books for that.

As a theorist, it's more important to have a deep conceptual understanding of a topic and being able to see connections between different things than being able to cram information from a textbook into your head.

Also, different fields involve more brute force computations than others. If you are doing particle physics you are going to have to a lot of intensive QFT calculations. In areas of CMT involving topological phases, one barely has to calculate anything, it is much more conceptual .

14. Dec 29, 2015

### spaghetti3451

What kind of computations are usually done in string theory?

15. Dec 29, 2015

### JorisL

Hate to break it to you but the range of topics covered in e.g. Polchinski is far too broad for a single PhD.
You are supposed to gain mastery of a topic, in general this means you gain very specific knowledge.

An important part is being able to accept certain methods without deeply understanding them at first.
You need to progress in your research.
By doing this you can use some spare time/less productive time trying to understand the concepts better.

I'm still struggling with this myself but its great advice my thesis advisor gave me. (in my case the deep understanding requires lots of full string theory while calculations happen in supergravity)

Compactifications come to mind.

16. Dec 29, 2015

### spaghetti3451

Is it absolutely crucial to develop excellent technical (i.e. mathematical) and problem solving skills to be able to do computations in string theory?

Side question: How much of general relativity and QFT do you still need to remember/use when you're now doing string theory?

17. Dec 29, 2015

### JorisL

I like following a more mathematical approach to physics. Which means I exploit form notation to the fullest.
To validate that approach I rederived several identities I use time and again.
Another aspect of mathematics I needed is knowledge of linear algebra. That way I was able to restrict the moduli I have to look at.

All in all you can benefit from a broad mathematical foundation. Lie algebras are important too, more so depending on the specific part of string theory.

I benefited most from GR. In fact I won't use QFT in the forseeable future.

18. Dec 29, 2015

Staff Emeritus
Then that's the skill set you should have said you need. I would argue with that - I think depth is more important than speed - but it's not what you wrote. Besides, you gave three examples. Over a forty-year career, doing something three times isn't that often.

That said, I don't think this thread is about the skills a successful theorist needs. It seems to be more about your convincing yourself that the skills a successful theorist needs are exactly the skills you already have.

19. Dec 29, 2015

### spaghetti3451

I don't have any of those skills I talked about :P

20. Dec 29, 2015

### bacte2013

Do you take notes from the textbooks, by any chance? I know that question is off-topic of this thread, but I lately have been suffering from the weird belief that I must create a comprehensive notes for each subject (i.e. linear algebra and analysis) by using various textbooks and taking notes from them. I somehow could not get myself out from the belief that having no textbook-style notes will prevent me from recalling the understanding later on.

21. Dec 29, 2015

### spaghetti3451

Taking notes allows you to reshape the content of each topic you study from the textbook in your own personal way. And so taking notes is absolutely vital to good studying.

A better approach is to read a paragraph, close the textbook, and rewrite the essence of the paragraph in your own words, or better still using a mixture of pictorial symbols and verbal words. That way, you might retain the information for a longer period of time.

Also, it's a good idea to skim read an entire chunk of text that you feel encapsulates a single major concept of the topic you're studying, and then to read it again until you get the gist of it. Then you might want to read it carefully. Afterwards, you might wish to see if you could present the same thing in a way that seems more logical and natural to you, i.e. reshuffle the various chunks of text in such a way as to make you feel that the argument flows naturally. You know you've understood the concept when you find that if you're given the first and last sentence of your notes, you can reproduce all the intermediate sentences because your presentation of the same argument happens to flow just naturally and logically.

If you're in a hurry to finish a textbook, you might just wish to time yourself per page, say, 30 minutes per page. That way, when your 30 minutes are up, you're forced to move to the next page regardless of whether you understand all of that page or not. Usually, for me, I understand 80% of the page and I highlight (whatever of the page I don't understand) in the notes with a pencil. Once I revise my notes and its highlighted parts a week later, everything happens to fall in place (eventually). Deep understanding takes time, so no matter how much time you spend on a single page, you won't get better understanding in the short-term. So, speed reading is probably the best strategy because you know that when you come back to revise your notes, what you did not understand the first time will gradually begin to make sense. And anyway, it is not possible to understand/impractical to want to understand 100% of a textbook. Achieving a 90% understanding is sufficient and you can move on to the next textbook.

22. Dec 29, 2015

For advanced books you definitely have to read through chapters several times to really understand what is going on. For many books I learn something new every time I read it.

I have realized I retain a lot more taking notes. It's tempting to just copy a lot but I find the most helpful thing is to just copy the things that you find especially insightful conceptually. It's also good to read several things simultaneously in slightly different topics since then you will be able to see connections in between seemingly different things. It really helps you see how everything is motivated. For example, in CMT, spin liquids, the quantum ising model, superfluidity and the FQHE may seem very different at first, but there are things that they have in common.

I definitely agree speed is not the most important thing. Sometimes it may seem frustrating that it is taking you a long time to understand something, but then all of the sudden you will have a breakthrough and realized something very deep. For me, that was the case with RG and the motivation of CFTs.

23. Dec 30, 2015

### WannabeNewton

Have you had any experience with theory research yet?

24. Dec 30, 2015

### spaghetti3451

Yes. I learned some basic material in false vacuum decay and how to calculate the functional determinants that arise in the decay rate.

25. Dec 30, 2015

### Student100

I don't think you read WN question correctly.