Dear [admissions / hiring / scholarship / fellowship / internship] committee. I am writing to you to support [student / colleague / supervisee / former student / prospective student]'s application ...
How many of these letters have you had written on your behalf? How many have you written for others? How many have you read, as a member of the 'dear committee' in question? Letters of recommendation pervade academic culture, and the world outside it. So take a minute, and go and read this: "Reference Letters Cost Women Jobs" at Futurity.org. Go. Then come back. (Thank you to reader Heidi for bringing this to our attention, btw.)
Are you agog? I am.
Reference letters are hard. They're hard to write and hard to read. They are consequential, and treated with some reverence and a lot of formality. A reference letter is held in strictest confidence, on the understanding that the letter writer is speaking some hard or secret truth to a committee of whatever sort, a truth whose utterance can only be assured by this guarantee of confidence. A graduate admissions committee is committing to support, say, an incoming PhD student for a minimum of four years: we want to know what we're getting into! A hiring committee is looking to make a huge commitment to a potential colleague--we hope a reference letter will help us pick those three candidates out of the pool of 50 that we might best interview.
But reference letters are also often written in code. They are increasingly prey to the cult of the superlative, where all the adverbs (fantastically! superbly! tremendously! extremely! immensely!) we've worked so hard to prune from our other writings magically reappear, because we need to make the applicants we write for stand out. It seems like every applicant is always ranked in the "top 2%" of everyone the referee has ever had contact with. This paradox requires careful interpretation of the narrative portions of the letter.
There is also a kind of reference letter that seems to be written under duress: where the referee has tried gently or not-so-gently to dissuade an applicant, who nonetheless insists on getting said letter. These are easily enough deciphered by committees, but they never come right out and say, "I don't want to write this letter, because I don't think this student is ready for grad school." Instead, they usually say something like "While I am happy to help out X with her application, I cannot be as thorough as I might be had I taught her in a course other than the 500 person introductory survey six years ago."
In my time at Waterloo, I have written more than 40 different reference letters for students. I have served on the university committee that ranks SSHRC graduate fellowship applications. I spent three years on the graduate committee readings SSHRC and OGS apps, as well as applications to the program. I have spent three years on the appointments committee. I have read hundreds and hundreds of reference letters, all written in some form of code.
So this article has knocked my socks off. It's not that I'm so terribly surprised that, generally, that gendered language is used to describe job candidates and students. It's that the code words the article notes are ones I very much recognize as keywords I look for when evaluating letters. I do tend to be impressed by some of the descriptors coded female, and less impressed by some of those coded male. But over all, I hadn't noticed these as gendered, and a lot of the ones from the feminized list of descriptors I have often understood to be code for "not that smart or accomplished."
Let me tell you, I've been digging through all the reference letters on my computer that I've written for others. And I find myself largely innocent of gender-torpedoing those I've written for. But I'm sure as hell going to be a lot more conscious of this as I read all the letters I use to evaluate candidates, and more careful still of how I write them.