Sunday, May 13, 2012

How to ask questions

Mike Mitzenmacher said once that I ask lots of questions. In teaching I always ask lots of questions, because the students are the ones who build the lectures. When I meet some well-known computer scientists I sometimes take the chance to interview them on research and life. Recently I have also pursued a project outside work that involves interviewing academics. How to go about asking questions, especially from strangers? First, I am interested in what they have to say. Seeing what is in other people's minds is always a thrill. Second, I have no interest in boiler-plate answers. In class, I want to hear how students think about a problem, not a repetition of what the textbook says. In interviews, when I ask someone, for example, why he or she decided to become a researcher in theoretical computer science, what I'm hoping to hear is their story, not the answer that they think they're supposed to be saying. (That's why it works better with senior people: they have more freedom of speech, because they have no worries about jobs or tenure.) Third, if they seem open to questions, I try to see how far I can go, and it can sometimes go to the limits of people's comfort zone (or beyond if I misjudge). It is a way to try to learn about people, in a fashion similar to when trying to learn about a scientific problem: to understand it, ask a variety of questions about it, aiming not for the trivial nor for the unanswerable, but for the zone at the threshold. The downside is that it's not very polite. In British culture, it can come across as downright rude. The movie "A room with a view" starts with a scene where a man offers to trade his hotel room with a woman he does not know, after hearing her complain about the lack of view from her room. He means well and shows common sense, but is perceived as insufferable. No sense of proper reserve! How incredibly ill-mannered! I have heard that academics in general have the reputation of lacking good manners. Could that be related? Could the willingness to ask hard questions about open problems, if it is unchecked, correlate with a tendency towards bad manners in society?


  1. Regarding our last point : in scientific academia, it is commonplace to point out inaccuracies, oversights, or fuzzy reasoning. Of course, you can be more direct with close colleagues, in a meeting room, than if you are questioning a speaker in front of a conference audience (pointing out problems may then be humiliating for him) ; still the practice is to point out problems in reasoning or in facts.

    In common life, it is generally bad manners to point out problems in reasoning or in facts. Many "technical" people (including computing engineers) tend to want to correct factual inaccuracies or logical fallacies when talking to non-technical people. They are then perceived as abrasive, rude, obnoxious, nitpicking, arrogant, elitists, etc.

    You have to choose your fights. If somebody talks nonsense, even if you can disprove him or her, is it really worth it?

  2. To be clear, I think I would have meant when I said that you ask a lot of questions that it was a compliment. Hopefully, in whatever context I said it, it was.

  3. So... why did you decide to become a researcher in theoretical computer science?

  4. Do you ask questions just because you want to be claire?