Skip to content

Advice on soft skills for academics

Julia Hirschberg sent this along to the natural language processing mailing list at Columbia:

here are some slides from last spring’s CRA-W Grad Cohort and previous years that might be of interest. all sorts of topics such as interviewing, building confidence, finding a thesis topic, preparing your thesis proposal, publishing, entrepreneurialism, and a very interesting panel on human-human interaction skills.

I took a look at a couple of these and they look like useful advice for grad students. I wrote back to Julia that I used to have a hard time as a professor convincing students their evening would be better spent having dinner with the invited seminar speaker than revising their homework.

Advice to grad students is a longstanding, niche writing genre. There are even widely-cited classics, like the transciption of Richard Hamming’s talk You and Your Research, which has been recommended to me more times than I can count. There’s also a later YouTube presentation, which I haven’t seen (yet!).

Hamming’s advice on lunchtime behavior reminds me of one of the reasons I like places like Google—they still have active round table tech lunches. We try to do that once per week after the Stan meetings. If you’re in town and want to join us, drop me or Andrew a line.

P.S. I got the term “soft skills” from a friend of mine who’s going through soft skills training at Amazon before they’ll let him speak to customers—things like giving presentations and staying on topic.


  1. jrkrideau says:

    In some disciplines it seems to be difficult to get students to realize that writing coherently in whole sentences and paragraphs rather than in point form is a needed skill.

    • Torquemada in Training says:

      jrkrideau: Refer them to the previous thread where we rank on semi-literates who thought they could put those oppressive notions of coherence, clarity and grammar behind them after high school.

      Also, in my experience, both primary and second-hand, it’s not just some disciplines, it’s darned near all of them. (If you want to start an ecumenical rip-roarin’ bitchfest in the faculty lounge, just bring up the subject of poorly constructed reports. Names will be named, ivory tower milquetoasts will turn apoplectic and no slander will be off-base. Wading through incompetent writing is the second most aggravating grievance on everybody’s mind. The first, of course, is parking.)

      The reason for the problem is obvious: WRITING IS HARD! Just as thinking about and correctly performing STATISTICS IS HARD.

  2. Sander Greenland says:

    That talk became the final chapter in Hamming’s final book, which has many other interesting parts:

    See also his earlier book, a free preview of which can be downloaded here:

    • Sander,

      Such riveting reading on my patio under a full moon. Thank you, thank you thank you. lol

    • Shravan says:

      Related: Street-fighting mathematics, by Mahajan (MIT Press, free to download)

      edx course

    • Shravan says:

      I don’t agree with Hamming in some things he says to young researchers:

      1. He tells them to drop activities that do not directly lead to advances in one’s research, and to apply their mind single-mindedly to their work. I hope he didn’t practice what he preached.

      2. He says somewhere that one should only study successful people’s lives (Newton etc.) if one wants to become great (and when one isn’t naturally great, like Newton, Feynman, or Tukey etc.). One should ignore failures. I feel that the failures are at least as informative. Of course, he points out that it’s difficult to ask someone why they failed in their career, so he never did that (and probably nobody can explain why they failed, and those who succeed might come up with post-hoc explanations). But one can and should study one’s own failures, painful as it is. A failure is often more informative than a success.

      • Shravan,

        You are right. He has a view of success and failure that takes into account the life trajectories of a subsets scientists and engineers.

        I liked his comments on history writing which has been subject to different interpretations. In other words it is simply not helpful to privilege only a few as have tended to do in past.

  3. Shravan says:

    Hamming, Intro to The Art of Doing Science and Engineering: Learning to Learn (March 28, 1995)

    seems like a very inspiring course. take that, ted talks.

  4. Soft skills often come along with the substance of the thing, whatever it might be. I am sorry to see schools spend inordinate time on confidence-building, group work for group work’s sake, and such. It is difficult (and sometimes misleading) to teach such skills outside of the subject matter. Yes, these skills are important, but one can learn them through involvement in the work itself.

    In childhood, I had some stage anxiety and excitement; it took me a while to realize that preparation, not anxiety, was the main issue. If I hadn’t mastered a piece, I would play it worse than usual in performance; but if I had it down, I would be fine (and would enjoy performing it). The same held true for presentations, talks, readings, etc. In adulthood, I made it a point, when possible, to prepare more than thoroughly. The anxiety quickly dwindled.

    As for interviewing, there are some basic principles, but beyond that, a substantive conversation–where one shows genuine interest in the position and institution, asks pertinent questions, listens carefully, and conveys one’s qualifications–does most of the work. If the interview is serious–that is, not just a formality–and one does these four things well, I doubt that it will go terribly wrong. The peripherals–how to sit, what kind of eye contact to make, how to shake hands, etc.–should not distract from the rest.

    As for “networking,” having dinner with an invited seminar speaker can be great, not as much for the “soft skills” as for the hospitality and interesting conversation. If I respect this person’s work enough to invite him or her to my seminar, or if I am the student of someone who does, then I will probably want to welcome the person and learn more from him or her. That said, there are times when anyone–including the speaker–might want to bow out of a social occasion. That too has its merit.

    • Martha (Smith) says:

      Lots of good points here.

    • I spent more time lunching with my father’s colleagues than I did attending their classes. I learned more probably by that former process. I was however in a very specific niche market as it were. And the grooming for it was not going to result from being a typical grad student. This is what I have realized now.

      Then again professors would join their students for lunch and even invite students to their homes. Campuses were always my home away from home due to my father’s vocation. A professor.

      As I understand it, there were considerably more informal gatherings among academics before the 80’s. I heard that expressed by Lawrence Susskind at a Program on Negotiation conference at Harvard Law School some years back.

    • Anonymous says:

      When reading this post, i actually feared for this and checked a few of the slide-collections of topics just to check for this exact thing (i.c. any mention of “power posing”). I quickly wanted to stop reading them, and did not find anything about “power posing” in the few slide-collections i checked.

      I am not a fan of these type of “advise” sessions for students (which is why i wanted to stop reading as soon as possible), and I am not sure what it tells me about the usefulness and/or quality and/or validity of them if i expected to find some stuff on “power posing” which apparently can indeed be found (although technically i don’t know what was said about it)…

    • Andrew says:


      There’s a lot in these slides about how to make a memorable presentation, creating good presentations, etc. But one thing they never say is that one requirement for a good scientific presentation is that it be clear regarding the evidence for its claims A presentation could be the slickest, story-filled delight you ever will see, but if it’s not clear on the evidence, I don’t think it’s a good scientific presentation. I notice another line from these slides: “Gotta Handle Questions . . . YOU are the expert.” But what if you aren’t the expert? What if the questions being asked have no good answer (except, “Sorry, I screwed up”)? Then what? I think that any lessons to students on how to present should warn against overconfidence. Actually, various Ted talks would be excellent ways to make this point.

  5. Torquemada in Training says:

    Just as the secret to a successful marriage is the willingness to utter those three little words – “I was wrong” – so too is the key to a less bumpy professional career – “I don’t know,” and then shut your pie hole. You won’t get called out for trying to bluff in a room full of bluffers, and frequently they will fill the void with their own opinions, some of which you can use later.

    • Martha (Smith) says:

      “I don’t know” brings to mind a couple of memories of good teachers I’ve had:

      1. My high school offered a third semester of physics, consisting roughly of an introduction to relativity and quantum mechanics (thus only bright, eager students signed up for it). One day we were pestering the teacher with questions, beyond what he could answer. I remember him sitting cross-legged on top of the desk (he had been a high school and college football player, so this was quite a sight in itself), shaking his head, and saying, “I don’t know. I really wish I knew, but I don’t.”

      2. Once in graduate school, in an advanced topics course, the professor stated an if-and-only-if theorem, smoothly proved one direction, then got stuck in trying to prove the converse. After a couple of minutes, I came up with a counter-example, and shyly raised my hand to tell him. He abandoned trying to prove the converse and went on to the next topic. After class, he thanked me, saying that if I hadn’t spoken up, he would have stood there for the rest of the hour making a fool of himself.

      Writing this, I remembered that the prof in the second example had played semi-pro basketball. Makes me wonder if maybe having serious sports experience helps in not being overconfident?

Leave a Reply