14

I am a PhD student in computer science. My advisor keeps telling me I need to read papers, but I find most of them really long and boring, no matter what field they are in (but especially if there are a lot of math equations). Usually my mind starts wandering after one or two pages.

My advisor thinks I should read one or two papers a day (apparently he does this himself), but if I followed his instructions (and yes, I've tried doing that), I would literally spend the whole day reading papers, or 20% reading papers and 80% procrastinating.

I am perfectly capable of reading papers if there's something specific I want to gain from the paper (like if the paper is really relevant to my project, or if I need to implement an algorithm from the paper). But I have a hard time doing broad literature reviews and getting to know a field if I don't have a specific problem in mind. A lot of times I just end up not doing them.

Does this automatically make me a bad researcher, and what are some ways I can compensate for my short attention span?

  • 10
    You haven't yet found a need to read; keep looking. – Mad Jack Dec 15 '16 at 4:44
  • See if there are any seminars that cover related topics, or if you and a group of students could do one yourselves. Sometimes it's a lot easier to tackle material together, and thinking about how you would explain the material to someone else forces you to approach the material differently. – David Dec 15 '16 at 5:12
  • 3
    Reading a paper does not necessarily mean reading it thoroughly. Most of the times a short glance would suffice, because most of them are essentially similar. There are a few original papers in every field and the rest are just some minor modifications – polfosol Dec 15 '16 at 7:02
  • 6
    "Two months of research can save two hours in the library." (Frank Westheimer) – Captain Emacs Dec 15 '16 at 13:02
  • 11
    My advisor thinks I should read one or two papers a day — Then he doesn't mean "read"; he means "skim". Actually reading a paper can take days. – JeffE Dec 15 '16 at 13:17
5

Interesting question, I have thought about this alot both during my PhD studies and afterwards...

There are some answers and comments touching upon the aspects like why you should be reading, or what it means to read. I will try to give my two cents on the question title, which I assume is what's really on your mind, that is whether or not you're suitable for doing research if you don't like reading literature.

Short answer: It depends...

There's no way of knowing, or objectively guessing, whether or not it will work out fine for you. There are way too many parameters to consider

Long answer: It depends...

I'm sure you've heard, and will hear and see, people reading constantly. They will be telling you what you should read, how much you should read. I know I have...

While I don't doubt the sincerity or the good intentions behind these sentiments I am not sure how accurate or helpful they actually are. Hearing and reading those kind of comments have demotivated me at best, and pushed me further and further away from pursuing a career in academia at worst.

I was super ambitious in the beginning; I highlighted numerous journals that I deemed important in my field, signed up for their eTOC alerts. Read the short abstracts of anything that got published there. I had email alerts for a bunch of keywords, that were relevant for my work. Those articles that matched any one of those would be labeled in my inbox, I'd make sure to go through the whole article, albeit quickly...

About half way into my PhD studies (which is in bioinformatics, for the record) I've realized that there is quite literally no way for me to stay on top of literature, to a satisfactory degree. There is just too many people out there, doing way too many experiments that may be relevant somehow to my work. Also around the same time I started to look at publishing a bit more critically. The way funding and career options look these days, there is tremendous pressure on people to publish.

That has several significant implications, two really significant ones are:

  1. there's an increasing number of publications (and the rate of increase is also increasing, although I dare not say exponentially)
  2. a significant portion of the articles are written to secure research funds, deliver results for a received grant or secure the next position; as opposed to being written to be read and understood easily
  3. stuff that wouldn't normally be published gets published. this could happen for a number of different reasons like the group that wrote the paper is too influential to get rejected or objectively critiqued, or the reviewers are too busy to give a proper review of the work, or the in the worst case of all, there is data forgery committed

So at the end of the day, you start doubting what's actually on the paper, and start wondering about what's been omitted, or how the article was framed in order to get it published.

Added on top of that is the fact that a significant portion of the articles are not read and an overwhelming majority of the articles are never cited. The exact numbers have been a matter of debate (here's one link there are many more) however the fact remains that a significant portion of the literature is ignored at best, treated as garbage at worst.

All of these reasons give birth to a skepticism that is hard to ignore. A lot of my colleagues and friends who've been through my phd studies with me essentially agree. I have yet to meet a PhD student who get invigorated and excited by a pile of articles that s/he needs to read... (Interesting side note here is how majority of the people that will typically talk about how much one needs to read are senior scientists.)

Finally the matter of work-life separation. Many academics would go around saying that being a scientist is a lifestyle not a profession. I have this very close to heart where many PIs I personally know well (including my brother) spend a significant amount of their free time reading articles. For the life of me, I have no idea why... The last thing I want to do when I get home is to read research articles. I think it's healthy to separate work and life beyond work, but I haven't gotten so far into my career, so take that with a pinch of salt.

So, as you can see, I share the same worry (unless I misunderstood the origin of the question) that OP does. I have been worried about this for several years now, am I a bad scientist because I am not interested in reading literature?

I am not sure, I have been getting pretty much only positive feedback on my work and skills, so I am not entirely sure it's that bad as it might appear. While I agree that one cannot completely shut out reading altogether and ignore literature, I don't think one needs to obssess about reading constantly. I honestly think it's a matter of etiquette, or an ambition to strive towards, even though most sensible people realize that it's not really feasible to constantly stay on top of literature.

I say: do read, at least as much as you need to, about the subjects you need to know about, the rest is optional. Don't worry about keeping statistics or meeting thresholds, just try to stay well-informed about your field.

  • "So at the end of the day, you start doubting what's actually on the paper, and start wondering about what's been omitted, or how the article was framed in order to get it published." ... in other words, you develop a critical mind. Is it a bad thing? – T. Verron Dec 15 '16 at 13:40
  • @T.Verron not at all.. I am just trying to express that it's a process – posdef Dec 15 '16 at 14:40
4

My advice is mostly on why I think you should read, but I will also say why you might be having difficulty.

It sounds like you are in Computer Science or a related area. There are, in my mind, three reasons to read a lot of papers during your PhD.

The first is to understand the existing techniques that are available to you. For example, the different dominant formalisms, especially those often used in your area, and the differences between them. If there is an over-arching framework that links them, then it is also useful to know that. By reading a lot, when you actually do research you will have in your head a wide range of possibilities.

The second is simply so that you know what you should research. What has been done, what are open problems and what problems could be made more interesting. Imagine working on a problem that has been solved, but you didn't know (whilst everyone else does) - that will feel frustrating and I have seen it happen even when the problem was well-known.

The third, is simply so you can work with others in the area! You need to get a good idea of who you would like to collaborate with and also, when you are at conferences, if you don't know people's work well enough how will you build a potential working relationship? I have had colleagues who did not read much and it was detrimental to everyone. There was no common ground on which to discuss ideas, they might have an idea and describe it only for it to turn be a well-known problem already solved a decade ago (for example).

In terms of why you are having difficulty, it's not unusual. From what I have seen it is simply that the person is not interested in the area. You say it doesn't matter what area you are looking at, you find it uninteresting. Perhaps the area does not matter to you, but you really like to solve problems. In this case, maybe view each paper as a way to understand the problem and - as you read - also think of different solutions, how you would do it better and related problems. This would perhaps help maintain the excitement and it's something you should do anyway (to get an idea of who's work is actually great and who is just hiding an easy problem under a complex solution). Be an active reader is what I am saying (in contrast to those saying you should skim), don't just read but also - do - work out how a formalism or algorithm works, what could be improved, any properties that might be there the authors have not mentioned etc. More generally, perhaps try and think of why you like research and see if you can incorporate that into your paper reading.

1

You'll find that broadening your knowledge base to include material even tangentially related to your main research focus will help with the generation of novel research questions, novel methodologies, or insights that can help fill the gaps in your current area. Expertise isn't just about depth of knowledge, so try to incorporate related material into your reading list, even if you don't have a specific problem you're trying to solve. The background knowledge will be helpful when you sit down to write up your dissertation or other research papers.

0

My suggestion would be that for the time being, you take a different approach to looking at papers. Skim a couple per week without reading the whole thing carefully beginning to end. Make yourself a list of questions and have those in front of you or in mind while you are spending 30-45 minutes skimming a paper. Read the abstract and the conclusion. Form an idea what the paper is intended to accomplish. Skim the whole thing enough to be able to make yourself a mental table of contents of the paper. Pick one section of the core of the paper to read carefully. Pick one table or graph to read carefully. See if you are interested in taking a look at one of the references.

Try to strike a happy medium between your advisor's concept of a paper reading program and your own natural instinct.

One other idea -- you might find a thesis easier reading than a journal article. However, it will be longer. Take that into account. (You might spend a week gradually getting through a thesis.)

0

It seems like you are a "doer" rather than a philosopher. So in your shoes, I would start by "doing" computer science.

I would "consult" rather than "read" literature. Basically, you have to know the literature just well enough to use it to justify what you are doing, as opposed to "getting ideas" from the literature.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.