Where Does a Great Research Problem Begin?

Jun 30, 2023 9:13 PM
phd thoughts
July 12, 2023
Files & media

A PhD is a journey of many steps, but contrary to common belief, the difficulty is rarely a matter of the difficulty of any single step. The difficulty is not a hard math problem, a system too complex, an argument too arcane. The challenge lies in discovering what direction one should explore. If a PhD is about answering questions, then we can view this core challenge as the process of choosing the right questions to ask. I have personally wasted too much time looking for the right question, the question that achieves the perfect balance of whatever stuff makes a tasteful research question. Such tasteful questions do exist and it’s often easy to identify them once they appear. But they do not appear easily.

The mistake I have made—and have seen others make as well—is patiently waiting for problems to appear. True, it is possible to simply read a research paper or two and come up with a good research problem. But it is not easy. Many experienced researchers struggle with it, and if they can do it, it’s often because it reminds them of something from their previous research experiences.

Learning how to choose a research direction is not just something that new grad students need to figure out. At some point, every researcher—academic or otherwise—will need to explore an area in which they have no experience. A topic with which they have no background. And in this, there is only one choice: to start. I am reminded of a quote by Antonio Machado in Border of a Dream: Selected Poems.

Traveler, there is no path. The path is made by walking.

Too often we wait to take action because we don’t know what the most productive next step is. In the case of research: because we have not settled on the perfect problem. But Traveler, there is no perfect problem. A problem only appears perfect in retrospect, when you understand all the context that makes the problem interesting. This is in part because if the problem were so obvious that it could be reasoned out of thin air, it would have been asked and answered (it it is even answerable). It’s true that some research questions are considered low-hanging fruit, but it’s also true that if it’s hanging low enough it will either be picked immediately or is maybe not worth picking at all.

So, pick a direction, any direction. If something seems interesting, follow up on it. At some point, you’ll find that you have bushwhacked a path to an interesting problem, and eventually, maybe even an interesting body of work. Once more unto the breach.


Apart from its poetic language—even in translation from the original Spanish—this quote speaks to me.

The path from point A to B only exists as a retrospective concept. When we’re trying to go from A to B, it’s comforting to know a way to get there. But every context is different, we can never know exactly how the journey will unfold. The only reasonable choice is to start with whatever we think is the first action that will get us closer to the goal. What the next step is may remain unknown or it may become clear after taking the first step. This, we cannot know until we’ve taken that step.