PhD in Physics: What to Expect?

[Azelketh]

Hi

I thought that physics forums could be the place to ask you guys (others who are currently in or have already completed PhD programs) some questions.

I've recently started a PhD program in physics in the UK, and my general state of being is complete and utter confusion.
In my attempt to do the first thing that I am told any PhD student should do, and read around the subject area to find out what is known, what isn't, what is currently being worked on by others ect
I have found that it takes me ~ 80 to 100 hours to read each paper. At this rate i'll be a decade into my PhD before I finish the initial preliminary review! Did anyone else find that their comprehension of material was so low as to render background reading as useless as this?

 

[Travis_King]

I imagine it would be a better idea to review loads of abstracts and get an idea of what you're interested in looking deeper into first. Then, from there you can look through more abstracts and see what's a hot topic or a burgeoning field at the moment and see if there are any gaps or areas that aren't getting much attention and determine where you think you could make a contribution.

You don't need to read every single paper to get an understanding of the research currently being done.

 

[Vanadium 50]

100 hours per paper means you're doing something wrong. It also means you are only reading one or two papers per month, which is also a problem. Why do you think it's taking you 2 orders of magnitude longer than most to read a paper?

 

[SteamKing]

You could copy each paper in longhand in 80 or 100 hours of work.

I've read my share of papers dealing with hydrodynamics, which is never an easy subject, and I've got to tell you, there is little in the way of novel ideas presented in a typical paper. Much of the work is based on what has been done before, and most of the new ideas presented advance the subject incrementally at best. Unlike a murder mystery, it is OK to skim the paper until you reach the conclusions to determine if a more in depth study of the paper is warranted. Your time is a precious commodity; treat it as such.

IMO, most papers are written to give the author a publishing credit, nothing more.

 

[bp_psy]

Maybe he is talking about one of those 60-100 pages long review ''papers''. I am reading two of those at the moment and they do indeed take a lot of time.They are supposed to get you to the point where you can understand what the hell people talk about in the journals so it is understandable that they should take a lot of effort.Or at least that is what I hope.

 

[jesse73]

You could copy each paper in longhand in 80 or 100 hours of work.

I've read my share of papers dealing with hydrodynamics, which is never an easy subject, and I've got to tell you, there is little in the way of novel ideas presented in a typical paper. Much of the work is based on what has been done before, and most of the new ideas presented advance the subject incrementally at best. Unlike a murder mystery, it is OK to skim the paper until you reach the conclusions to determine if a more in depth study of the paper is warranted. Your time is a precious commodity; treat it as such.

IMO, most papers are written to give the author a publishing credit, nothing more.

This is true from my experience and is frustrating about academia.

 

[Azelketh]

100 hours per paper means you're doing something wrong. It also means you are only reading one or two papers per month, which is also a problem. Why do you think it's taking you 2 orders of magnitude longer than most to read a paper?

Well the biggest contributor to the time per paper is trying to understand the papers.

I can 'read' a paper in 10 minutes no problem, I'll just have absolutely no idea what any of it meant. I think that mostly its my lack of familiarity with the physics. I am looking at papers to do with chalcogen impurity donor states in 3-5 materials and I am just not up to scratch with the theory.

For example when a few weeks ago I first encountered the 1s (A1) --> 2p0 transition , I had never seen the A1 notation before, hence I did not understand what the transition was. Even after a few weeks spent trying to look it up, all I can say now is that the same A1 notation is used to represent the representation of a class of operations in the space group of the material. How that exactly effects the transition I still don't have any idea.

The biggest difference it seems to me, considering that 5 months ago I was finishing an undergrad degree, is that when you do not understand something as an undergrad you simply let it slide as you must continue with the course , you'll constantly have new (mostly admittedly basic) material to learn. Whereas that doesn't happen here, I don't understand something in the paper, so I try to understand it...months pass...

And I have thought that perhaps I should just continue without trying to understand thin gs, but then the whole paper doest really make much sense.


Do you guys ever have similar problems?


Also thanks for replying!

 

[Useful nucleus]

For example when a few weeks ago I first encountered the 1s (A1) --> 2p0 transition , I had never seen the A1 notation before, hence I did not understand what the transition was. Even after a few weeks spent trying to look it up, all I can say now is that the same A1 notation is used to represent the representation of a class of operations in the space group of the material. How that exactly effects the transition I still don't have any idea.

I faced a similar situation, actually exactly the same problem about symmetry in crystals and molecules). At that point I realized that I need to work on group representation application to molecules and crystals from a textbook. I tried to avoid "the details" of papers that rely a lot on group theoretic notions until I spend 2 or 3 month reading a text on the subject in which I have a weakness.
I'm not saying that one has to dedicate months of textbook reading for every single notion they do not understand in a paper (life is short). But by skimming over many papers, you will be able to determine the things that you really need to study from textbooks and possibly take graduate courses (if available in your school).
Do not get frustrated, move on to the next paper, take notes of the things that you potentially need to work on. Take an iterative approach to reading papers rather than a linear one.

 

[Azelketh]

...
Do not get frustrated, move on to the next paper, take notes of the things that you potentially need to work on. Take an iterative approach to reading papers rather than a linear one.

hmm I might have to do that.
It is just incredibly frustrating though, when after so much time reading a paper if asked to explain what it is about I can pretty much only say "they measured some stuff *shrug*" end of my understanding.

 

[ZapperZ]

Don't you have an academic advisor?

I don't know how they do it out there over in UK. Still, you should have someone who can guide you over such things. He/she should be able to tell you of a few papers that you must read thoroughly, and a few others that you can simply skim over and get the gist of the results.

You need to keep in mind the purpose of reading these papers. It is to get you up to speed on the development in your field! You need to know (i) what has already been known (ii) what are the major issues and problems in the area (iii) what are the hottest topics being worked on now, etc.. This is where your advisor can steer you in the proper direction so that you don't get bogged down in spending too much time on a paper that might not be that important or central to your work.

Zz.

 

[billiards]

Coming from geophysics I'd recommend you to read review papers (I assume you have those in physics). These will give you the overview and are an excellent source of references for more detailed reading should you need it. Obviously the more up-to-date the review the better, generally speaking.

Also don't despair. It takes time to learn your field and also to learn to read papers critically, but once you get over the initial few hills it gets easier.

 

[Choppy]

The biggest difference it seems to me, considering that 5 months ago I was finishing an undergrad degree, is that when you do not understand something as an undergrad you simply let it slide as you must continue with the course , you'll constantly have new (mostly admittedly basic) material to learn. Whereas that doesn't happen here, I don't understand something in the paper, so I try to understand it...months pass...

I think this really is an important point and one of the key differences between being a "student" and a "researcher."

When you're a student, you can, at the end of the day, hand in the assignment or exam without deeply understanding the material. Often, due to time constraints, you learn what you have to to get the grade you want, and move on.

In research, it's a lot harder to just move on because you've got a problem to solve and it will still be there tomorrow if you don't solve it today. And you won't make any progress on a problem until it's solved.

One piece of advice is that even with reading papers, you have to give yourself time to "digest" what your read. Not everything will make sense right up front, no matter how much reading you do. Sometimes you have to work with a problem on your own until you have your own meaningful questions about it. Then, what other people have written will make a lot more sense. Sometimes a month or so after reading something, you'll have an "ah-ha" moment and the material will click.

It doesn't help that papers are written across an whole spectrum of quality. Sometimes a very academically important paper can be a struggle to get through, not because of the material, but just because the authors can't communicate well. It frustrates me to no end when I'm working my way through something someone else has done and just can't see how they got a particular result and eventually email the author to discover that the paper has a typo.

 

[Azelketh]

I think this really is an important point and one of the key differences between being a "student" and a "researcher."

When you're a student, you can, at the end of the day, hand in the assignment or exam without deeply understanding the material. Often, due to time constraints, you learn what you have to to get the grade you want, and move on.

In research, it's a lot harder to just move on because you've got a problem to solve and it will still be there tomorrow if you don't solve it today. And you won't make any progress on a problem until it's solved.

One piece of advice is that even with reading papers, you have to give yourself time to "digest" what your read. Not everything will make sense right up front, no matter how much reading you do. Sometimes you have to work with a problem on your own until you have your own meaningful questions about it. Then, what other people have written will make a lot more sense. Sometimes a month or so after reading something, you'll have an "ah-ha" moment and the material will click.

It doesn't help that papers are written across an whole spectrum of quality. Sometimes a very academically important paper can be a struggle to get through, not because of the material, but just because the authors can't communicate well. It frustrates me to no end when I'm working my way through something someone else has done and just can't see how they got a particular result and eventually email the author to discover that the paper has a typo.

This difference between being a student and a researcher does seem to be key. I would go so far as to say that I never deeply understood any of the material I studied as an undergrad.

Thanks for your replies.
It seems that I'm not alone in my confusion.