‘Tis the season, both for holiday stuff and for reference/recommendation letters. Such letters are one of the unusual features of academia, not always easy to explain to people outside. (For students: I’m sorry to say that asking people to write letters for you doesn’t stop when you graduate or even when you get a tenure-track job. There are still letters to be written for tenure and/or promotion, awards, and even certain grants.)

Lots has been written about them, including:

I’m writing letters for several different people this season, and have done sufficiently many to date that I decided to write a script to make sure that I send the right letter to the right place. It’s not uncommon for someone graduating with a PhD in astronomy to apply for 20 different positions, or someone applying to graduate programs to apply for the same number. While there are services like Academic Jobs Online and Interfolio that help with the distribution of reference letters, there’s still a lot of time involved in requesting, writing, collecting, and reading them.

So why do we bother? What’s so special about academia that we need this time-consuming procedure that nobody else does? Presumably we think there is some information contained in letters that’s not transmitted in the transcripts, CVs, cover letters, research and/or teaching statements that a candidate submits. My experience with being on search and admissions committees is that at least some faculty use letters as a major screening tool; a good set of letters is a necessity to get to the next round.

One assumption about recommendation letters would be that they give some kind of neutral, independent assessment of the person and/or their work. But I don’t think it’s that simple. Many of these kinds of letters also reflect on the letter-writer: saying that someone who worked with me as a grad student is terrible reflects badly on me, since it implies that either I didn’t train them well or I was foolish to take them on in the first place. As supervisors we generally want to see our students succeed and so we write letters that emphasize their positive qualities. If anyone doubts that this is necessary, I can’t count the number of grad-school application forms that I’ve filled out recently that ask if a student is in the top 5, 10 or 25% of some cohort - it’s clear that anybody below that limit is disfavoured. At least in North America, a mostly-if-not-completely-positive letter is almost a requirement.

So if a recommendation letter is not an entirely neutral assessment, what else is it for? Sometimes letters are used to explain complex situations that may not be clear from a CV. Suppose a student changed programs or supervisors because of something out of the student’s control; a third-party explanation of the situation (and the fact that it wasn’t the student’s fault) could be very helpful. Sometimes letters are used to explain particular hardships or challenges that a candidate has overcome. It can be difficult for someone to write about such situations themselves without sounding either whiny or self-serving, whereas a letter from someone else can put the situation in context.

Emphasizing the positive and explaining the negative both seem like worthy goals, but I’m still not completely unconvinced that this makes recommendation letters worth the effort involved. It seems like there should be better ways to get information to the people who need it. Is the big difference between academia and industry the fact that it’s hard to fire people in academia, so we feel like we need additional security in finding the “right” person? Or is it that the competition for positions is so much more intense? I still don’t know, but will continue to think about this as I keep sending in my letters..