How independent should an academic be?

Question

I am a computational scientist and I often find myself much more productive when working with others with resonating mind. When working alone, my motivation level is lower, distracted more easily, and I feel the progress is slower. However, in academic jobs, I often hear that it is important to show that you are an independent researcher.

• Q1. What are the defining properties of an independent researcher?
• Q2. How can I show that I am independent?
• Q3. I like working in teams, is academia (esp faculty in research university) not a good career path for me? (I could think independently, but I could do better with others, so why be independent?)

Answer 1

Collaboration is generally seen as a major strength in computer science and computational science, not a weakness. One of the biggest questions we ask about any faculty candidate in my department is "Who in the department/college will they work with?" We do occasionally hire faculty that work alone in their cave and occasionally emit theorems, but they're rare exceptions (and even they still work with their students).

The most important thing is to establish a reputation as a leader in your collaborations, rather than a follower. You need to establish your own research agenda, instead of just following someone else's. More importantly, you need to be seen to have your own research agenda, instead of just following someone else's. Your agenda must be visible in your publication record, in your recommendation letters, and (eventually) in your funding record. To this end, it is very important to do a few things:

• Limit your collaborations with more senior researchers, especially your advisor, especially between getting your PhD and getting tenure. As I've suggested elsewhere, publishing at least one paper without your advisor before you graduate is a strong signal of independence, even if you have other coauthors. Once you have your PhD, DO NOT publish with your advisor for at least a few years.

• Do not always publish with the same set of coauthors. It's fine to have two or three different groups of people that you always work with, as long as you can wind a consistent story that ties most of your work in those different groups together.

• (Once you have a permanent job:) Get at least one grant as a principal investigator, not just as a co-PI. Definitely join other grants as a co-PI, too, but you must have at least one grant with your name on top to get tenure.

Answer 2

I think I have a similar mindset. I like independent work, but I find that I am more motivated when working with others. However, I don't think that will hinder your ability to show that you are an independent researcher. As mentioned by @aeismail, the defining point of independence is that you are able to generate and pursue your own research ideas and agenda. If you do it alone or with others is irrelevant.

In a lot of cases, like when you applying for positions, etc. You will have reference letters to accompany your CV. Thus, the author-list is not the only place to show your independence. In the letters your coauthors and mentors can attest to your independence.

A research statement or cover letter also is a great place to highlight your independence. If you work with lots of different people, then you can show how all these projects tie into a broader theme which is distinctly your own. In computational science it could very well be a theme like "I like developing information theoretic explanations for neural and multi-agent activity" and then work with an auditory group, a cognitive group, and a population biologists can be great supporting evidence of the fact that your ideas and approach are unique and independent, not to mention widely applicable.

Taking a leadership and organizational role is another clear way to show your independence. A personal example: I prefer to read and discuss papers with others, so I organized a reading group that meets weekly (ideally). A bit of my time gets consumed in organizing and managing this group, but it is more than made up for by the extra motivation I get to stay on top of the literature.

As for the atmosphere; group-work is not an antithesis to academia. It is very important to show that you 'play well with others' and frequent collaboration is a great way to do so. Especially with the large push for multi-disciplinary work, being able to work with people of various background is a great asset. Further, as a faculty in a research university, one of the things you will be expected to do is supervise and teach students. This means you need experience in sharing and developing your ideas with others, something that working in isolation does not nurture.

Answer 3

The issue of independence is an important one, but it doesn't obligate you to work in isolation. Instead, what it means is that you are capable of generating and pursuing your own research ideas. This can be done in collaboration with others, but there's still needs to be some evidence that you can lead the effort.

Some ways to show independence are to have some collaborative efforts where you are in the "leadership" positions—first or last author. If you can do this across multiple independent collaborations, preferably in multiple disciplinary areas, that will show that you have some lateral flexibility, which is among the hallmarks of an independent researcher.

I don't see, in this day and age, the desire to work in teams as ruling out a career in academia. I might not lean towards it as a first choice—instead, I'd probably steer such a person into a research institute like a Max-Planck-Institut here in Germany, or a DOE lab in the U.S. Those offer more collaborative environments which would be a better "fit" for someone who prefers to be a team player than a team leader.

Answer 4

At least in my field (mathematics), there are people who do almost all their work collaboratively. Academia can be a good place for such people, but:

1. It varies a lot, not only by field, but by subfield, so you'll want to take that into account when choosing a specialty. (Math has a lot of single authored papers, while the lab sciences have very few. I gather CS is somewhere in between, and that it depends on the area. Even within math, it depends on the particular subfield, sometimes for reasons that have more to do with culture than anything intrinsic to the subject.)

2. If you're in an area where single authored papers are common, you'll eventually be expected to produce some, just to demonstrate that you're not be carried along by your co-authors.