Theoretical Physics PhD worthless nowadays?

some_dude

For me it's very difficult to learn stuff from an "applied" point of view. So if my goal were applied math or engineering I'd probably still be on the same path. I just can't stand when a class or subject is taught from a high-level perspective where you need to just take on faith a huge foundation of material.

Locrian

You never need to take it on faith. Just go and research the background yourself.

some_dude

That's rarely feasible. E.g., some undergrad (non-pure math) course invokes Fourier transforms: if you want to go and research the background you're going to need to go study Harmonic analysis, but first you'll need measure theory, and before that undergraduate analysis. In other words, you'd be probably still be researching the background long after said course ended.

DrRocket

I agree.

When you are asked to take a lot on faith, I take that as an indication that the person who asks that of you does not understand the material himself.

I don't call that a high-level perspective, I just call it hand waving. Sometimes that which is being described during the hand waving is correct, and sometimes it is not.

DrRocket

You picked an interesting example, a pet peeve. Some of the hand waving with regard to Fourier transforms and Fourier series is necessary, because what they are telling you is false, particularly with regard to convergence of Fourier series. They do not, in general, converge as nicely as many engineers and scientists have been led to believe. I have seen text book statements that are just plain false.

It is even more difficult to prove a false statement rigorously than it is to wave your hands and convince the naive but skeptical.

The only reliable way that I know of to handle the situation that you describe, is to do what I did. But leaving engineering to go get a PhD in pure mathematics is rather extreme, and more of a commitment of time than one might reasonably expect from the average student. It is, however, effective.

some_dude

Well, I'm glad to hear it's effective, because it's not far off from my motivation either.

Fourier analysis is interesting in that respect, because I've never found anything as tedious as having to do that stuff in a rote way without any clue what was going on. Trying to "play around" with those integrals in any sort of non-thought out way just seems to lead to a nightmare explosion of symbols. My friend is doing some research on that this summer, and I'm going to some seminars to hopefully cut through some of the fog that remains.

DrRocket

There are of course lots of books, good ones, on harmonic analysis, but I am particularly fond of the following two:

Fourier Analysis on Groups -- Rudin

An Introduction to Harmonic Analysis -- Katznelson

Both are graduate level books, and there is minimal overlap between the two.

If you can find an old Dover Edition of Katznelson's book it is MUCH cheaper than the very slightly updated new hardback edition.

arunma

Twofish-quant, I know it's been a couple months since you posted this. But perhaps you (or someone else) can help clear up some misconceptions on my part.

I too have always been told that there are far fewer academic positions available than trained PhDs. Here's a graph from the American Institute of Physics which says that in the past few years, we've been training between 1000 and 1400 PhDs per year in America:

http://www.aip.org/statistics/trends...ad/figure6.htm

Now this data says that in 2005, there were about 1100 new physics PhDs trained in America. However, this table says that in 2005, there were 324 faculty positions available:

http://www.aip.org/statistics/trends...f06/table8.htm

Obviously we're averaging over all majors here, such as condensed matter, high energy, space physics, astrophysics, etc. Nonetheless, in the class of 2005 this leaves 1100 people competing for 324 jobs. Now obviously I know that in the best physics tradition, I'm making some simplifying assumptions here. People applying to 4 year colleges will be able to compete seriously without doing postdocs, but people applying to state universities will have to do postdocs. Nonetheless, you can see why I'm worried.

My own informal polls (i.e. talking to my friends at the bar) suggests that most physics grad students want to pursue faculty positions rather than go into industry. Indeed, for virtually all grad students outside condensed matter, it's very difficult to find an industry job that involves doing the same research one did for his PhD. And every time my department has done a faculty search, it consideres quite a few candidates, for just one position.

These statistics make me worry about my prospects as a physicist. I really don't want to end up being a computer programmer or financial analyst, but since I'm an astrophysicist, I'm not really sure where I could get a permanent industry job that involves doing actual physics (I know, it was my mistake for not doing condensed matter when I had the chance). After glancing at these stats, I had all but given up hope on getting a tenure-track academic position. But now you're telling me that the general interpretation of this data is wrong. Could you elaborate? If I have a good chance of getting a faculty job, I'd really like to know.

twofish-quant

One mistake that people make is to think of a degree as some sort of meal ticket, in which you get the degree and that degree somehow turns into a meal ticket in which you get money from. It really doesn't work that way. The degree is a characteristic of yourself that can be sold if you can do some sales and marketing.

The thing about physics and math Ph.D.'s is that I don't know of anyone that doesn't have a middle class job, and I know of a many people with physics and math Ph.D.'s that make pretty large salaries.

some_dude

Well I might have a look, though someone's also recommended Elias Stein's advanced book to me, looks intimidating so we'll see. I'm strongly inclined to be a geometer and have found Mikhail Gromov has an enormous amount of geometry flavored analogues to the more analytical standard analysis material I have difficulty grasping. I've seen first hand many natural analysts have abilities with manipulating integrals I don't have (and lacking that makes studying Fourier analysis tough).

Oh, I know - been there, done that. From the ages of 18 to 22, I was just always looking for the right "credential" or line-on-the-resume to get the "big job". To some extent, I do think there are degrees that are "meal tickets": e.g., Harvard MBAs, medical degrees, dental degrees, CPAs, etc. But I've also come to believe those types of mass-produced jobs provide no intrinsic value - it's only the "prestigue" and money I could see as motivating me to go that direction, certainly not something I'd be doing if I weren't chasing a carrot. Math is the complete opposite: chicks are, at best, ambivalent it and you aren't really guaranteed to make a dime, but it's still worthwhile in and of itself.

arunma

Don't know anything about dental derees and CPAs, and know that the MBA isn't all it's cracked up to be. But from my career-related research, I do know that in America the medical degree is most certainly a safety net for the rest of your life. Don't get me wrong, you've still got to work long hours and stuff. But if you've got MD (or DO) after your name, you've basically got a guarantee that regardless of the state of the economy, you will get a job and not be laid off. In addition to the fact that there's always a steady supply of sick people, the AMA has taken steps to protect the medical profession, such as restricting the number of medical students it educates yearly, and not starting new medical schools. I wonder what the job outlook would be for physicists if departments accepted fewer grad students.

Anyway, if I knew in undergrad what I know now, I definitely would have stuck with my pre-med program instead of switching to physics. But hey, maybe I'm wrong. Twofish says that there's a shortage of candidates for tenure-track positions. Now if only someone will find a way to explain the AIP data to me...

Dr Transport

You are forgetting all the hundreds of Post-Docs who are looking for faculty positions, the ratio is more like 10:1, and departments are getting hundreds of applications for every position. The academic community needs to start training their graduate students to work in industry, not just as an academic someplace after working as a slave post-doc for 4-5 years. You are correct, condensed matter is most likely one of the few disciplines where you have many of the skills out of a PhD program to slide into industry, optical physics is another. Experimentalists have an easier path in because of the amount of lab work in many industrial disciplines, theoreticians have it harder but can hack it, if they learn how to work with the other areas.

Having worked in industry for the past 10 years with a PhD in Solid State, I can talk with some authority on this matter.

Andy Resnick

I agree 100% with Dr. Transport. I got a PhD in Optical Physics, worked in industry for 6 years and went back to academia, where I am now.

arunma

Good point Dr. Transport. I was making simplifying assumptions to show that getting an academic position with a PhD in physics is really hard. But I guess it's even harder than these assumptions would suggest.

In this economy I'll take any job I can get. But if I have any ability to choose, I'm really looking for the following criteria (in order of importance):

1. Minimal risk of being laid off.
2. Work that involves doing actual physics. Since I was trained to take data and make scientific conclusions, I would like to do this. And since I'm a physicist, I'd like to be doing physics and not computational biology (or whatever it is high energy PhDs do these days).
3. I don't want to be a programmer. These days it sounds like the transferrable skills of a physics PhD lands most people in programming jobs. Don't get me wrong, I can program so long as there's a non-computer end goal in mind. In my current research most of my time is spent programming, but that programming is done with some astrophysics objective. I don't want to end up working for Microsoft developing software.

If I can get these criteria met in industry, great. But it sounds like academia is the only way to go. How do you compete for an academic position when there are so many candidates? Again, Twofish said that there is a shortage of candidates for academic jobs. If so, I'd like to hear more about this. Or if anyone knows how I can get into an industry job (as a physicist, not a programmer), I'd really like to know how to do this too.

Choppy

Where did he say that? From what I've seen in this forum, he's a strong proponent of the "don't go into graduate study with the expectation of getting an academic job" school of thought.

I think he's saying that just because you can't find work as a professor, doesn't mean you're condemned to a career at McDonalds. There are other options for people who've completed graduate degrees and there are many cases of people who end up doing quite well financially because they've figured out how to parlay the scienctific skills they've acquired into marketable assets.

The trick, I suppose, is really figuring out how to do that. How does someone who's spent the last four years running stellar evolution simulations convince a financial company that she's worth $200k per year? And if she does managed to do that, how does she find enough personal fulfillment to remain committed to whatever they need her to do?

Dr Transport

In this day and age, lay-offs are always an issue, minimal risk means that you are indispensable, that takes time and energy on your part to get there.

All engineering work involves some physics and engineers know this, the lead in my group is a mechanical engineer and says all the time "Physics works, you just need to know how to apply it".

All jobs have programming as part of the duties, remember, you're analyzing data on a daily basis and you have to model the experiments.

twofish-quant

I'm curious what I said that gave people the idea that I think there a shortage of Ph.D.'s for academic jobs. There is a *VAST* oversupply of Ph.D.'s for tenure track academic positions. What I do believe is that the job prospects for Ph.D. physicists is extremely good, you just have to broaden your perspective a bit.

It's pretty easy to do that, since the hiring manager is someone that has done numerical relativity or something similar.

Depends on the environment, but it helps that the work environment is very much like graduate school.