“Letter writers who say applicants are cute, petite, lovely, sweet, appropriately dressed, has a great smile, has a supportive spouse–I’m judging you,” he wrote  “Perky,” and “bubbly” weren’t in the tweet, but were also used in the letters, he tells Quartz.

To be sure, as Shaw explained in follow-up tweets and a blog post, the offending language didn’t appear in every letter, but in about 10% of the nearly 200 he evaluated. They also didn’t read to him as dismissive or coyly hostile; rather, the phrases generally accompanied solid testimony to an applicant’s academic successes and capabilities. When they appeared, these inappropriate word choices were usually found toward the end of the letters, which is where reference writers tend to shift modes, describing the individual as a person, and answering the question, “Is this someone you and others would want to work with?”

It’s also possible, says Shaw, that people found themselves in vaguely gendered territory because of the discipline, which may have been part of the problem. Terms like “kind,” “thoughtful,” and “compassionate” are words you’d like to have associated with the professional who will hear out your neurotic thoughts and private confessions. They are borderline gendered words, he says, and once in that zone, “People went too far, to say things like ‘nurturing’ and ‘lovely,’ and words that are a little bit more gendered.”

He respects and is grateful to the letter writers who volunteered their time for this task, he emphasizes; he just wishes that some of them had avoided clumsy, stereotype-reinforcing language.

On Twitter, some suggested Shaw respond to the letter-writers to point out their error, and one person supplied a link to a bias calculator: Plug in your text, and it will tell you whether your implicit beliefs about women’s role and abilities are showing. And several women responded to Shaw with their own anecdotes about recommendation letters in which women were described as charming, attractive, stylish, great at balancing the needs of the job with that of the family, and, in one case, a star baker of cakes.

The anecdotes echoed studies of recommendation letters for PhD programs and faculty openings, which found, for instance, that men were often described as “trailblazers,” while women are “very knowledgeable” (but not set apart from their peers), and that women applying for professorships were more likely to be characterized with traits related to building relationships, like “caring.” 

Shareen Holly, one of Shaw’s former students, now a psychology professor herself, wisely reminded Shaw of the two questions he had taught her class to ask about the way they describe patients in psychological reports: “Is it applicable/appropriate for either gender?” and “Could you say the reverse?,” i.e., of the other gender.

After his tweet went viral, Shaw says that he was surprised by the number of professors who told him they’d never come across such minimizing language, and how many said they see it every day. He believes that it’s very possible he had read comments like this in previous years and was simply not attuned to it. This year, he had just completed equity training for a McGill hiring committee, which he says may have primed him for the reaction he had. But he was also reading this year’s batch extra carefully since he plans to hire only one student for his lab, and as “awesome” as having delicious baked goods would be, he says, “it’s not something I’m going to be too concerned about.”

Now Shaw intends to fine tune the instructions for future letter writers, so they can give praise and kudos without reverting to retrograde descriptors.

🖋 Sign up for The Memo from Quartz at Work

A dispatch from the world of modern work. Learn how you can help create a productive, creative, and compassionate work culture.