I have been reflecting on the roles that I took in academia over these last 10+ years and I finally decided to write them down. I believe I have passed through at least four different roles in the academic system. These roles are informal and they only exist in my mind. I may have spent 3–4 years in each role. The roles follow a very sequential process: you only pass to the next one when you pretty much understand the current one. Interestingly, there is no one that says that you should pass thought all these roles, in this order. They just happen to happen to be like this. At least it happened to me.

The roles are:

  • Understanding what is research
  • Learning how to do research
  • Conceiving research questions
  • Helping other to do research

Although I worked in several tech companies during my undergraduate studies, I did not have an opportunity to experience research. Therefore, I only started to scratch the research world in my masters. I started my masters in 2008 and finished in 2010, so I that I played my first academic role (understanding what is research) during this time window. In this particular role, my main obligation was to understand what research is all about. Since I had zero research experience, but reasonable industry experience, I had that feeling that research as about creating software (eg, if I create a software that does X, I would be able to graduate). I remember my former advisor telling me that I should work on something that the industry is not yet doing (and not doing the same thing that industry was doing). I also remember that it was quite difficult to understand what she meant by this (doing a thing that this great company is not doing? what? how could this be even possible?). By that time, my comprehension about research improved, but was still opaque. The problem that I worked during my masters was also a kind of challenge-free, so I knew that if I could implement the stuff that I was supposed to implement, I was good to go. For me, research was a lot of work and little to no reward. In this sense, research was pretty much what I experienced in industry, not to mention the hassle of being a student with little income.

I finished my masters in 2010 with this rough notion of research, and still in 2010 I get back to the industry. I nevertheless started my phd one year later, in 2011. I have to say that I started my phd still trying to understand what is research. I eventually figured out that research is about understanding the unknown. It is not about creating software solely, but to create software to test hypothesis. I do not remember how I sorted this out, but I believe it was due to the fact that my phd advisor often asked me to do more than one research work. These works were done in collaboration with other students and professors, and were targeting different research topics. All of this contribute to build my research understanding. The next logical step was to learn how to do research. This is not to say that I did not do research in my master. I did. However, I did not have the conscious understanding that what I did was research. For me, it was all about engineering. Moreover, I have to say that it was a lot of effort to understand how to do research. Although I have published some papers here and there in the first two years of my phd (most of them in collaboration with other students, as aforementioned), my phd thesis was pretty much empty. My first paper related to my thesis was accepted only in my third year. Why was that? Because it took me a really while to find a problem. My advisor was really open about anything, but I was in charge of looking for interesting problems to find answers to. This process of conceiving problems turned out to be incredibly difficult. During the first 1–2 months, you dedicate yourself to read the literature, and you figure out that you do not know anything. After many other months of intense reading, you start to comprehend the landscape of one very particular part of the literature. At this moment, it seems that every interesting question already had an answer. It seems that every interesting piece of work that worth doing is already done. This lack of comprehension of “what to do next” is a symptom of this second role: I understood what research was about, but I did not know how to do it. The solution now is clear: I have to ask new questions. Ask a new question is easy. Ask a good new question is hard. Ask an unique question is harder. Ask an unique question that could give you a phd title is even harder.

I eventually I ended up with some research questions during my phd (some of them borrowed from the literature, some other stated my by former advisor, etc), but I believe I only started to understand how to conceive research questions after my phd. We can go to a bar and some hours after come back with a list of research questions. But how do you know which one will not lead you to a dead end? Or how do you know that this research question was not already asked in that bedrock paper from the 80s? How do you know reviewers will care about this research question? Will readers care about this research question 10 years from now? Will you care to finish a research paper that has this particular research question? There are many other questions related to the importance of the research question. And to conceive good research questions, you should care about these other questions. Every single time. One may say: “yo, I can write a paper without thinking too much about research questions”. Yes, of course you can do. In particular if you are working on a subject that you know a lot (you know the open questions). You can also read the “future work” section of the most important papers in your area to easily grasp other research questions. However, one key aspect of conceiving research questions is to find something that you would love to do. More importantly, regardless if the question is unique, unanswered, timely, beautiful, etc., would you trade your precious time to find an answer to it? It is not about finishing a masters or a phd. It is about doing work that matter (at least to you). Ultimately: look for question that keep your eyes open, day and night.

Finally, the last role, which is the one that I am pursuing now, I believe, is about teaching other to do research. The goal now is not to have your paper accepted at that top conference, but, instead, helping others to have their papers accepted as well. Since I am just starting this role, I have little to share. But one thing that I am thinking is that: since these roles last for about 3–4 years each, which role comes next?

Categories:

Updated: