As it happens, I’ve now had the chance to sit on four or five search panels for academic positions. This makes me far from expert in the process, but I’ve been struck by the differences in how letter writers perceive their task in recommending a candidate (similar observations could be made about letter writers for PhD programs, I think, mutatis mutandis). In a way that has surprised me, on the whole I’ve noticed that the most prominent scholars often write the most detailed and involved letters of reference, and attempt to offer a documented case for their candidate. I had assumed that senior scholars would be too busy to write tailored letters of reference, but one has the sense that they take the task very seriously since seeing good people achieve the best positions is one of the most important means of guaranteeing the future quality of the field. But even when we are early in our careers, we are sometimes asked to write letters for others, and it’s not always easy to know what to include, particularly before we’ve had the benefit of seeing lots of letters of recommendations ourselves.
So what makes a good letter? A few brief, unscientific observations. Good letters are:
- Specific. The best letters are specific, and in multiple senses. They evidence a real knowledge of the individual in question, and can therefore supply not merely a list of abstract qualities, but can demonstrate those qualities in action by referring to particular illustrative events that make the case for them. Moreover, while sometimes candidates submit letters – often because their doctoral programs urge them to proceed in this way – that are simply general letters of recommendation meant to be deployed in any circumstances (‘I recommend X for a position in any college, seminary or university’), letters tailored for a specific context undeniably make a stronger impression. We all assume that everyone starts with a boilerplate, but tailoring the letter conveys that the senior academic is highly enough invested in a person to be bothered to write again and again and again, and so implicitly conveys that the applicant is worth someone else’s time as well.
- Long enough to say something. Variations in the length of letters also surprised me. While very short letters can of course be very positive, it seems to convey again something about the senior academic’s investment in the applicant, if a letter is hefty enough to be substantive. True, some letters can go on and on for pages, but most letters can say quite a bit in a solid 1.5-2.5 pages, which seems to be about the average length of good models I have in mind.
- Personal. Perhaps because they are major scholars, the best letters show the voice of their authors, and so convey something of the personal importance of the applicant to the academic. I don’t here have in mind letters laden with all the gory details of one’s personal life, but I think letters that offer selected anecdotes from personal interactions in a way that begin to suggest for the reader the personality and quality of the applicant ultimately help in making an applicant stand out.
- Comparative. When recommenders write for more than one person, it’s extremely helpful if they can offer, even implicitly, some way of assessing the relative merits of the candidates for whom they write, or at least to pinpoint more precisely the strengths and weaknesses of each as they appear to the writer. Even if a person writes for only one applicant, to place them in some notional rank (‘among the best three students in my thirty year career’, or ‘perhaps the strongest student I have ever had’) can be extremely helpful.
There is much else that might be said, I’m sure, but these are simply a few of the points that stick in the mind as I consider letters I’ve read (there are more authoritative sources here). Poor letters can convey that a writer wishes to damn the candidate with faint praise; that may be perfectly intentional in certain circumstances, but it’s worth reflecting on the process of writing recommendations enough to ensure that our letters have the persuasive effect that we do intend by crafting them appropriately.