Back when I was working at Nuance Communications, we had a yearly internal research conference, in which people would share what they had been working on with other teams. Usually there were about 200 or so people at the conference, which was probably about 1/4 of all the people working in research. I got the opportunity to go several times, and it was always very fruitful. One year I recall they did some interviews with people, asking them what they thought the research mindset meant to them. It has been a few years, and I don’t remember any answers specifically, but it is something I think about fairly frequently, and I have some of my own thoughts about the topic.
For me, the defining factor of the research mindset is the desire to understand something at a deep level, to ask not only what questions, but also how, and in some cases, also why. I worked with many people with PhDs at Nuance, and I feel that they all shared this mindset. I think it is one of the things that people learn in a PhD program. It is a unique experience to be able to spend several years deeply studying a particular subject, and that experience can really change the way you think. Of course, there are other ways to gain a research mindset, but that is the path I took.
I have been working in software development for over 15 years now, and I have seen lots of cases where people without this research mindset would have benefited from it. I would frequently see cases where we would have an issue, e.g. the website was too slow. Some engineer would start working on the problem and begin by searching the internet for possible solutions. I think this was the wrong approach. In my opinion, before one can start looking for solutions, one must first understand the root cause of the problem, and that requires research. In computer programming, that can involve adding additional logging or profiling. I recall one time reviewing some code for a colleague. He mentioned that it took minutes or hours to run, which seemed strange to me, because it seemed like a relatively simple task which should only take seconds. I used the python profiler to analyze which parts of the code were taking the longest, and I discovered that there was a regular expression being called thousands of times unnecessarily. This is actually a common issue in python code. I precompiled the regular expression, which resulted in a 10X reduction in latency. My colleague, who had studied computer science and was a quite good programmer, was not even aware of the python profiler. I was quite surprised by that, given that I studied Linguistics and am a self-taught programmer. I think the research mindset has helped me.
I actually started a draft of this post several years ago, and put in a link to an article about research and academia, and the pressure to publish. It recalls some stories from the excellent book, Surely you’re joking, Mr. Feynman. My friend Brian first introduced me to it when I was in high school. My wife was introduced to it by her dad, and we read it to our kids several years ago. One of the stories I frequently think of from it is how he fixed radios when he was a young child. There was one time in which he went to someone’s house to fix the radio. He took a look at it for a bit, then he started walking around the room. The man asked him what he was doing. He was supposed to be fixing the radio. Feynman replied that he was working on fixing the radio. He was fixing it by thinking. The man was flabbergasted – fix a radio by thinking, what nonsense! Of course, he did figure out what was wrong with the radio and fixed it. I had a professor in college for an Asian religion class, Harold Kasimov. Frequently when someone asked him a question, he would take up to thirty seconds to respond. You could tell that he was deep in thought. He was really thinking about the question and how he should answer, instead of immediately blurting something out, like so many people do. It was awkward, but I also really respected him for this. I also do this sometimes. Knowing that it is awkward, I often try to say out loud “I’m thinking”, so my interlocutor is not totally confused.
I am curious to know if any of my readers have ever thought about it. Do you also have a research mindset?