I dread the number of letters of reference that I’ll need to write for my trainees’ job applications in the coming months. How can I accurately describe their intelligence and capabilities without simply repeating the word “excellence”?
Dr. Editor’s response:
Dread is an appropriate emotion for reference letter writing. Letters of reference problematically reinforce the networks of prestige (Huemer, 2017; White, 2016) and reflect biases against women (Dutt et al., 2016; Lin et al., 2019; Schmader et al., 2007; Towaij et al., 2020; Turrentine et al., 2018). Who wouldn’t dread needing to feed such a beast?
If I had a magic wand that could remove time-wasting labour in academia, the required letter of reference in the first round of academic job applications would be one of the first things to go (after I waved away the CCV with my wand, that is).
As long as my magic wand remains in the realm of fantasy, though, you’ll continue to need to write letters of reference. To reduce the potential of biases finding their way into your letters, and thus unfairly disadvantaging trainees from marginalized and equity-deserving groups, I recommend you first draft your letter, and then edit it with three key lenses in mind.
1. Cut intensifying language
Intensifiers — the adverbs and adjectives that writers include to add force to their expression — don’t have the effect that some imagine they might.
Take these two examples:
- Dave is a trustworthy employee.
- Dave is a really trustworthy employee.
In which of these examples might a reader be left wondering if Dave will be pocketing staplers on the way out the door?
When we’re talking about Dave’s trustworthiness, adding the intensifier “really” turns “trustworthy” from a simple binary (either Dave can be trusted or he can’t) into a spectrum: if Dave is “really” trustworthy, does that mean there are other employees who might be “very” trustworthy or “quite” trustworthy or “outrageously” trustworthy? How does Dave compare to these other folks?
Being trustworthy thus moves from an absolute quality into a characteristic that varies by degrees. The intensifier introduces a shade of grey, and with it, the possibility of doubt.
That’s why calling your student’s recent coup a “prestigious award” or describing them as being “highly prolific” is not an effective strategy if you want to persuade your reader about the truth of your claims. As Colleen Flaherty has noted in reference to the Dutt et al. (2016) study, in letters of reference in the geosciences, “Phrases such as ‘scientific leader,’ ‘brilliant scientist,’ ‘role model’ and ‘trailblazer’ could put candidates into an ‘excellent’ category. Lower-key phrases such as ‘highly intelligent,’ ‘very productive’ and ‘very knowledgeable’ – in the absence of supporting information – tended to get candidates into the ‘good’ category” (2016). The “very”s and “highly”s aren’t helping your trainees.
Two options that are preferable to these grey-inducing intensifiers:
- cut the descriptive term; or
- replace the description with something quantifiable.
Report on accomplishments
So you now know not to tell your reader that your trainee is “highly prolific” — but what should you say instead? Describe their accomplishments: tell us that they have co-authored more publications than any trainee you have supervised in your 15 years at U of X, or that they submitted N manuscripts last year, or that they spoke at two international conferences, attracting standing-room-only crowds both times.
By putting the spotlight on what your trainees have achieved — their outputs, their impact, their efficiency — you’ll be providing the evidence necessary to persuade your reader of the validity of the claims you make about their “excellence.”
Being able to report on accomplishments means your trainees will need to be coached to keep detailed CVs, tracking not just grants, publications and presentations, but also participant numbers (e.g., “organized a three-day, 320-person conference, attracting speakers from eight countries”), media mentions (e.g., “interviewed for a 30-minute podcast episode that has since received over 10,200 downloads”) and public engagement (e.g. “delivered a one-hour presentation at the city library, attracting 73 members of the public”).
3. Trim the hedging
Many genres of academic writing contain examples of phrasing that might be described as “hedging.” Wait, sorry, what I actually meant was: you academics — when you write, you hedge.
Hedging blunts strong claims. When you say that a critic “suggests” or that some data “indicates,” you’ve pulled back from making a definitive statement. In contexts in which nuances are unpacked, such hedging language makes sense. To avoid having another scholar chip at the foundation of your argument, you protect your work with a defensive hedge. For a conference paper or a published article, such an approach makes sense.
In a letter of reference, though, you should only hedge if you want to sink your trainees’ chances of securing an academic job: “the inclusion of even a single doubt raiser — particularly negativity or hedging — was enough to lead to statistically lower evaluations of the applicant” (Madera et al., 2019). And, as Madera et al. (2019) show in their two-part study, women are more likely to be described in such application-sinking terms than their male colleagues, even when controlling for factors including productivity.
There are a few free hedge-detecting bots online — Readable’s “Hedge Word Detector”; Data yze’s “Difficult & Extraneous Word Finder” — but I can’t endorse either bot, as they both seem to me to be imperfect in their detecting and finding. For instance, in the first sentence in this paragraph, Readable flagged “can’t” in “I can’t endorse either bot”—I’d argue with the bot on that, as I feel the sentiment is unequivocal — whereas Data yze found no sentences with multiple hedging words.
Ultimately, to recognize and remove biased language from your own letters of recommendation, you’ll need to learn to locate and remove the inappropriate use of hedges and intensifiers, as well as regularly and objectively review your corpus of recommendation letters and look for patterns that might show a systemic cultural misogyny, along with other forms of bias. Consider working with your colleagues to track your own data, either among yourselves or at the departmental level, so that you can begin measuring and then begin changing what matters.