My first thought was that this proposal is beyond bizarre.  My second thought is it’s beyond bizarre, but worth kicking around a bit because it hits on some interesting issues about the profession.

So, there’s a perception that academic pedigree, i.e., where one did one’s Ph.D. and with whom one worked, particularly who wrote one letters of recommendation, matters disproportionately much to search committees, to the detriment of equally good but less prestigious candidates.  And there’s a stickiness problem, because it is commonly believed that one’s first job sets the course of the rest of one’s career, meaning that if it’s true that pedigree matters too much, there are plenty of people not getting good jobs their first time out, and having no real way to recover.

Portmore’s solution: candidates should submit blind dossiers, including blind letters of recommendation.

I think this is a bad suggestion on both practical and theoretical grounds.

First, the practical:

Schools get anywhere from 250-600 applications.  There has to be some way to make an initial cut.  Things that usually help make that cut:  one’s c.v., institution, letters of recommendation, area of research, publications, teaching, etc.  So, our question:  what makes the cut?

a) A truly blind dossier will have no names, no institution, no teaching (because the teaching portfolio of an R1 looks much different from the portfolio of someone who taught their own courses — dead giveaway), publications mentioned but not described.    And we can imagine the letters of recommendation:

To whom it may concern,

During my tenure at [redacted], I have taught [redacted] of [redacted], [redacted] of [redacted], and [redacted] of [redacted], and it is now my distinct pleasure to recommend to you [redacted].



b) Alright, completely blind provides no information (remember, the standard here is ‘enough to whittle this list of candidates down’) at all.  So let’s adopt a more reasonable approach (as Portmore seems to intend.)  Letters are blind.  The dossier has no name or institutional affiliations, but is mostly left intact otherwise.

The letters are still useless; what seems to be useful is not “X is a great scholar” but “X is a scholar comparable to others whom I have taught, and I back this up with my name and my own record.”   But worse, now that it’s not completely blind, it’s going to be relatively easy, especially for elite people at elite schools who tend to work the backchannels anyway, to figure out who is writing the letters for some candidates.

So our situation is not Susie Snowflake from Topranked R1 competing with Johnny Snowcone from Plucky University with blind recommendations, but rather that Susie, being at a top school and being known to the right people, probably is effectively competing with a full pedigree against Johnny Snowcone, whose letter is now completely meaningless because no one has any sense of who his allies are.   Making it blind is not going to hurt the R1s. But now we have a situation where Johnny Snowcone believes naively that his letters are weighted just as heavily as Snowflake’s.

So, onto the theoretical.

c) I think this position only looks attractive if one holds that pedigree and philosophical talent (whatever that means) have no relationship at all.  And I think that’s obviously false.  Pedigree isn’t a perfect predictor, and it might not even be a great one, but it’s also not as though everyone at the top programs gets equally good letters from all of the best people.   Over here, the question comes up: what could a letter tell you that isn’t just “pedigree”?  Here are some suggestions:

  1. A sense of the candidate’s relative strength.  “X is great” doesn’t mean much, but “X is the best to come through this program in ten years” is something that can only be evaluated if one knows the comparison class.  (This doesn’t help with hyper-inflated praise, but profs who praise everyone as the best ever get known for that, and people take their praise less seriously.)
  2. A sense of the candidate’s work.  Suppose one is at a small liberal arts college,and one wants to make a new hire, in a new line that’s supposed to round out the department.  The result is that no one in the department is even close to competent in the candidates’ research areas.   How might one evaluate it?  One can look at publications.  But one can also look at who wrote them letters, because those people, not because of pedigree exactly, but because of their own research, are more qualified to give an opinion.
  3. Departments have personalities.  It is true that one doesn’t hire trends, but individuals, but this isn’t the hiring decision, but the decision whether to continue reading the application.  It is useful to know, e.g., what school the person attends to know whether, e.g., the claim “X is ready to defend” is true.

So, that’s some suggestions.  It seems wrong to include all of that as ‘just pedigree.’  And there’s more!

d) The whole project buys into the myth that the purpose of the job market is to find the young philosopher who most approaches the Form of the Young Scholar, and bestow upon them the bestest R1 position ever.  (This is in the Protrepticus.)  But it doesn’t really seem to be like that at all.  It’s not even comparable to having a blind audition for an orchestra, because philosophical talent isn’t the only deciding factor.  And because that’s not the case, it seems that more information on whether the person is a good fit, the better.  And it would seem strange if that didn’t include, as a first pass, with whom the person studied.

e) Pedigree cuts both ways.  This is why it isn’t just a cheap proxy for finding the Form of the Young Scholar.   Here are things I have heard from friends at small colleges, about those pedigreed souls at R1s:

“He’s great, we have no chance at him.  We’d rather interview someone who would be likely to consider an offer.”
“She’s great, but we don’t want to put ourselves in a position where she uses us as leverage to get a better offer.”
“He’s great, but we want someone who is likely to stay here.”
“She’s great, but we don’t want to be used as a stepping stone, because I am so goddamned tired of going to the APA every year.”

In other words, there are some jobs where plucky Johnny Snowcone has a distinct advantage because of *his* pedigree.   Yet no one has a particular problem with the idea that Snowcone might beat out Snowflake for a 3/3 teaching college job for reasons that have mostly to do with pedigree.  (Moreover, it seems that knowing that Snowcone is likely to accept an offer should be something on the radar, because this is not about finding the Form of the Young Scholar.)

All that said, we might still think that knowing someone’s institution lends implicit bias to the initial review of an application.   But we still have to cut down these 300 apps.

So, a suggestion.  Implicit bias is the result of first impressions (subtle things — you know the applicant is female, so you unconsciously read the c.v. more critically): so perhaps the first impression should not be the C.V., or the cover letter, but the dissertation abstract.  That’s easy enough to make blind, and a quick read of that might establish the person as someone to consider before one gets onto the C.V.  If the initial impression is good, it might very well stick, and at the least, it would remove institutional affliation from its place of prominence.