Over on Twitter yesterday, Jonny (a philosophy of science postgrad) asked, "Suppose I have 10 letters of recommendation for PhDs, how do I determine whose to use?" Someone suggested we hold a thread on this here, so here goes!

First of all, congrats to Jonny on securing 10 letters. That might be an all-time record! In response, Craig Agule wrote: "Although this advice is not wholly uncontroversial, you should have some trusted advisor read all 10 letters, both to catch accidentally bad letters and to help you choose." Although I've never tried it, this is what I've been told before–and I've heard if you have your letters uploaded to Interfolio, you can have them sent to a trusted advisor (though I wonder if there are some ethical issues here, given that letters are intended to be confidential). 

In any case, I guess my own answer is a bit more complex. I think if it's possible to get a trusted person to vet your letters, that's a good way to go. But I also think some strategy is probably in order for selecting the best letters. First, when it comes to jobs at R1's and prestigious SLACs, I suspect name recognition may matter a lot. If you have some letters from really eminent people in the profession, my sense is that (all things being equal) those are probably ones to prioritize. Conversely, if you are applying to jobs at a "teaching" school, it is not at all clear to me that the same strategy is wise. In my experience, prestigious letter-writers tend to write research-heavy letters–ones that may make a candidate look like a bad fit for a teaching school (conveying in effect, "this person really belongs at an R1"). When applying to teaching schools, one should always include at least one "teaching letter", and I think one should try to figure out which of one's letters make one look balanced (i.e. both a good researcher and a good teacher). Finally, I think that whenever possible, it is good to include at least one "outside" letter by a person not from one's grad department, as everyone recognizes that faculty from your grad program have a vested interest in your success (and hence reasons to "oversell" you relative to someone who has no such vested interest.

Anyway, these are my thoughts. I have to confess that this is a problem I never really solved as a job-marketeer. I had seven letters, never had anyone vet them, and just decided on a case to case basis which ones seemed to best fit the jobs I was applying to (most of the time, my experience is that jobs have a three-letter cutoff). This probably wasn't the best solution, but it worked okay! I suspect this may be in part because (or so I've heard) search committee members these days often take recommendation letters with a real grain of salt, focusing more on judging your materials for themselves. As a four-time search-committee member myself, I can attest to this. When a candidate's letters don't match your own estimation of their materials at all (viz. "overselling" the candidate), I suspect the tendency is to go with one's own judgment. But again, this is just my guess, and I could be totally wrong!

What do you all think? It might be particularly good to hear from candidates who have grappled with this problem effectively, as well as from search committee members!

Posted in ,

6 responses to “Deciding which recommendation letters to use”

  1. Thanks very much for posting this. I’m in a masters degree now and applying for admission to a PhD program. Would your advice be the same?

  2. Marcus Arvan

    Ack, I’m so sorry – I (mis)interpreted your query as about letters for a new PhD heading onto the job-market!
    In any case, the short answer is: much of the advice I’d give is the same. Although I haven’t served on a PhD admissions committee myself, my experience in grad school and the profession suggest to me that two things matter for recommendation letters to PhD programs: (1) the quality of the letter itself (i.e. how complimentary and detailed it is), and (2) name recognition.
    In brief, my sense is that name recognition matters a lot, in the sense that admissions committees may be more likely to trust the judgment of someone whose work they know and recognize (i.e. a “name” in the field) than someone whose name they don’t recognize, whose work they don’t know, and whose judgment they therefore don’t know how to evaluate. On the other hand, I have heard stories of “name” people writing really perfunctory letters that don’t give committees much to go on, which can raise questions regarding how well the letter writer even knows the candidate or their work.
    Long story short, you want to find (A) the best letters (content-wise), (B) written by the best people. Hence, why it is so important to get someone (a faculty member you can trust) to read and vet them, if possible. Otherwise, if you can’t get them vetted, here’s what I would personally do. I would probably choose one or maybe two letters by people who know your work really well (i.e. faculty in your Masters program), regardless of how well-known they are, and then, if you have any letters by “name” people, include letters by one or two of the most prominent people.
    One final wrinkle, I think, is how well the letters fit your areas of focus. Suppose you focus primarily in Philosophy of Mind (or whatever), you’re applying to programs that have a very good reputation in that area, your writing sample is in Philosophy of Mind, and so on. In that case, I think it would probably behoove you to include at least one letter by a well-known people who works in that particular area. It’s fine if your other letters can speak to your general philosophical abilities–but if you’re “selling yourself” as a philosopher of mind in the application process, it’s almost certainly a good idea to have a specialist (especially a well-known) be able to attest to the quality of your work in that area.
    I hope this is helpful, and am curious whether anyone who has worked in PhD admissions disagrees. In any case, apologies once again for misinterpreting your original query!

  3. R

    I agree that it’s potentially beneficial to have letters from names that the people reading them are likely to know. I just wanted to add that that doesn’t necessarily require them to be famous. If you happen to know that people in the department you’re applying to personally know one of your letter writers I would make sure to include that letter in your application to that department, arguably even over someone more famous, as they’re more likely to trust their friend’s recommendation than some distant famous person. Especially when you have as many letters as the OP does this might end up meaning using different letters for each application (and asking your advisors some questions about their network).

  4. Thanks for that Marcus and R! It really does help a ton. No worries on interpretation. I think the initial interpretation is useful for many more people.

  5. SM

    I find the three letter cut off to be wildly inconsistent with the demand for both a teaching letter and an external letter. We often get advice that having a letter from someone external to your committee who can vouch purely for your scholarly chops is a good thing. But if a teaching letter is also a must, now we’re down to including just a single letter from our committee? What if you have co-chairs? This all seems silly. Hiring committees should either not implicitly be demanding external letters or they should allow you to submit more than three. Then again, I suppose one should just follow your advice, Marcus: don’t bother to submit the external letter to a teaching school.

  6. Marcus Arvan

    SM: I entirely agree. I think the three letter cutoff places candidates in an unnecessarily difficult situation. I don’t know why committees don’t just let candidates upload as many or as few as they wish. Sure, some candidates might upload too many, but in that case committee members can choose for themselves which ones to read, assign weight to, and so on. It seems to me much more fair to candidates to simply let them submit whichever batch of letters they think is best, rather than require them to make these kinds of very difficult decisions.

Leave a Reply to JonnyDCTIDCancel reply

Discover more from The Philosophers' Cocoon

Subscribe now to keep reading and get access to the full archive.

Continue reading