7

Disclaimer! I am a first year CS/Applied math master student so I do understand very little in how academia works. I am trying to fix it before I go into PhD.


I have discovered a scientist who works in the field I am interested in. This scientists is very successful career-wise. The scientists finished a prestigious university and had stays in very good places. Now the scientists is an assistant prof in a high-ranking US university.

However, literally >90 % of his papers were written in co-authorship (as a rule, not the first nor the last author) with other more experienced researcher who we recognized as leaders in the fields. In fact, the only solo paper the scientist has is ones thesis! All PhD students so far have been also co-advised.

I have very little research experience but it seems to me that coming up with a good idea one can further develop is the trickiest part. Implementation (or, sometimes, even writing down actual proofs) is a technical work. I guess that the coauthorship was earned for some implementation work. It seems to me that there are much more people capable of / actually doing these technical work both in academia and in industry who do not get such recognition as the aforementioned scientist.

I may be envy but I find very little proofs of that scientist actual skills. Instead I have a feeling that the scientist makes very wise political decisions or is good as self-advertising.

Also I did very limited empirical studies and found that researchers who have highly-cited solo papers tend to have higher (4K +) citation counts then those who have almost exclusively co-authored papers.

My question is many-fold:

  1. is what I have described normal? Update. Is it normal to have almost exclusively co-authored papers?

  2. what should I do to mimic that behavior. It does not seem to be noble at all but apparently this is how successful academia people work.

  3. Are indeed funding and promotions based on citation counts? If this is not the case, how those researchers without solo papers are assessed?

  4. Is it true that those researcher who has successful solo papers are much better supervisors?

Update. I am talking here about theoretical research in CS/Applied Math. I guess in other fields there are many more opportunities for equally valuable contributions of many sides.

  • 4
    Co-authorship is more common in many academic fields when compared to solo authorship. – user2768 Jun 24 at 10:09
  • 8
    IMO, ideas are cheap. Work long enough (and not even that long) on a problem / field, and you'll have dozens of idea to try every month. The hard part is going through the work of implementing them, see if they work, if not why not (and how long do I want to spend trying to figure it out, vs trying something else). There is no point in time when you have an idea and say "Ah, perfect, I solved the problem, now just need to code it up and I'm done". – Ant Jun 24 at 19:09
17

I should point out one of the most famous of co-authors in math: Paul Erdős. Erdős never won a fields medal or even wrote a seminal paper as a solo author. What he did do was "wander" around to different universities and help people solve problems. He was an extremely prolific writer (over 1500 published manuscripts) and is considered one of the greatest mathematicians of the 20th century. The ultimate mathematical Renaissance man who could seemingly work on any problem. I think you would have quite a bit of difficulty in arguing that because Erdős never wrote a classical mathematics paper as a solo author, he was not a tremendous mathematician. (Though to be fair, in the past decade, his paper "On Random Graphs" can now be considered a classical work of mathematics. It just took 50 years for the work to become highly applicable).

I relate this story because having lots of co-authors does not necessarily mean having bad skills. Now I am not saying your professor is the second coming of Paul Erdős (I can't think of anyone currently in academia who can claim that mantle) but they might have other skills you are not giving them their due credit for.

Never underestimate the power of being able to organize a group of smart people and get them to work together on the same problem.

To answer your questions specifically-

  1. It's neither normal or not normal. Tenure positions are rare enough that there are certain metrics that are highly predicative (e.g. individuals who are ahead of the publication curve for their field tend to get hired in tenure positions) but there is no conclusive metric. Individually, there is enough variance within each position that you cannot really say.

  2. Collaborate! Having only solo author publications is probably not wise (unless you are in a field that is big on monographs). Having a variety of publications where you are first author, second author, and a middling author show your ability to collaborate and be a team player.

  3. Yes, clout does increase likelihood of getting accepted. Being a big wig with 10,000 citations at an elite university means you are more likely to get a grant than someone with a handful of publications from an R2. But to be fair, the qualities that led that person to be a big wig probably translate into their writing and research caliber as well.

  4. I have never seen any research or evidence that suggests that.

  • +1 And to answer the question "how those researchers without solo papers are assessed?" — They are assessed crudely by counting their publications, and more effectively by actually reading their papers, just like researchers with solo papers. – JeffE Jun 25 at 5:08
  • 3
    I don't think Erdos is a good example here. Yes, he famously had lots of collaborators, but he also wrote over 400 solo papers. – Especially Lime Jun 25 at 15:36
8

I think you are misinterpreting the data. What you seem to think of as "wise political decisions" is really just collaboration. Collaboration is a good thing in general. I'm not sure why you think that the leaders of the field would put up with someone just "hanging onto their coattails". I think they would probably resent that implication.

My suggestion is that you give up trying to "stand above everyone" and find some people who are as good as you are and start to share ideas and start some collaborations. If you wind up in a place that has a small faculty, then having already established a circle of collaborators will help your career greatly.

The person (maybe people) you open with have good positions because they were judged to have good potential. Their "citation counts" and all of that were just evidence of that potential. Their vast network of collaborators was likewise.

As you say, you are in a field in which insight is (all) important, but also rare. Sharing ideas can help. Hoarding doesn't.

  • >Their "citation counts" and all of that were just evidence of that potential. I see here a little bit of a "chicken or the egg problem". If citations come exclusively from co-authred papers, how do we know about the potential? – GlossyRetirement Jun 24 at 11:17
  • Again, it may that I am biased and envy. However, imagine the following. There is a bright student who gets good grades. He gets a chance to do an internship in a good team under the supervision of the leader of the field. The student gets some work done say implements some algorithms or helps to write the paper. He becomes a co-author. Yet what was done is totally comparable with home assignments which do not even require being creative. – GlossyRetirement Jun 24 at 11:24
  • 3
    @GlossyRetirement Why don't you try working solo and see how far you get? (this is not meant sarcastically, it will really answer your question in by far most cases) – Captain Emacs Jun 24 at 11:26
  • I've tried and find it super hard. In fact I had written several "papers" which are unpublished because the progress I had was meager. So I guess that the answer is that these people do collaborative research because it is easier? – GlossyRetirement Jun 24 at 11:27
  • 5
    Not "easier". Just more productive. Don't expect to be an expert at something having only just begun the journey. There are many steps to be taken. – Buffy Jun 24 at 11:36
5

I'm not sure what part of theoretical CS you're considering. In the parts I'm familiar with, co-authored papers are very much the norm, and author lists are alphabetical, so there's no significance in being first or last author. Other areas use other author-ordering criteria but co-authored papers still seem to be the most common case.

Citation counts differ dramatically between areas. For example, I have just finished a 50-page (co-authored) paper that cites about 25 other papers. In contrast, a seminar announcement on our departmental mailing list linked to a 14-page paper in another area of CS that cited more than 80 papers (or, should I say a 9-page paper with a five-page bibliography?). If that paper is typical of its field, citation counts must be much higher there than in my area. Perhaps your citations-vs-single author comparison is comparing a field where a typical paper has one author and cites a lot with a field where a typical paper has several authors and cites less.

There are plenty of universities where co-advisorship of PhD theses is common. Think of it as a sort of insurance policy in case the student doesn't get along with one of their advisors, coupled with two heads being better than one. It's not a reflection on either advisor's skills as an advisor.

I guess that the coauthorship was earned for some implementation work. [...] I may be envy but I find very little proofs of that scientist actual skills. Instead I have a feeling that the scientist makes very wise political decisions or is good as self-advertising.

Wow, you're spectacularly dismissive. From reading your question, I get the impression that you believe that scientific publication is primarily to prove how awesome the author is, that co-authorship must necessarily dilute that awesomeness, and that the only reason to bring on a co-author is to get them to do the boring parts. That's not how it works. Furthermore, unless you're reading papers that explicitly state what the contribution of each author was, your guesses of who did what are exactly that: guesses, and guesses based on almost no information. Don't condemn people based only on your baseless guesses.

c) Are indeed funding and promotions based on citation counts? If this is not the case, how those researchers without solo papers are assessed?

No. Not applicable.

d) Is it true that those researcher who has successful solo papers are much better supervisors?

Why would it be? They seem completely unrelated concepts, to me. If I had to guess, I'd guess that the researcher with many co-authors is better at working with other people, so is better at working with students. Doesn't that make more sense?

I am talking here about theoretical research in CS/Applied Math. I guess in other fields there are many more opportunities for equally valuable contributions of many sides.

Your guesses about theoretical CS aren't something that I, a theoretical computer scientist, recognize as a description of my field.

  • co-authored papers are very much the norm, I am not questioning this. I am asking more about a researcher having only such papers. Citation counts differ dramatically between areas I am comparing counts in the same field. Moreover, I have seen the same authors listed in direct (X, Y) and in the reversed order (Y,X). scientific publication is primarily to prove how awesome the author is, sure, but isn't this part very important career-wise? If not that then what? guesses, and guesses based on almost no information Agree. – GlossyRetirement Jun 25 at 7:11
  • 1
    @GlossyRetirement Can you find researchers in your field with single-author papers? What's the ratio between such researchers and those with no single-author papers? – user2768 Jun 25 at 9:30
  • @GlossyRetirement When co-authored papers are completely normal, having only co-authored papers is also pretty normal. I don't understand why you think there's something wrong with co-authored papers. Let me be absolutely clear: co-authoring is not an indication of any kind of inadequacy. Career-wise, doing high-quality research is what matters, and collaborative research is fine. You're not going to get somebody looks good who consistently contributes almost nothing to papers, because people will only work with them once, and it's hard to keep finding new collaborators. – David Richerby Jun 25 at 9:57

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.