Against Letters of Recommendation for Academic Jobs (updated)
“The practice of soliciting letters of recommendation for academic positions is both foolish and immoral.”
So begins a post by University of Colorado philosophy professor Michael Huemer at What’s Wrong?
Part of the problem, he says, stems from a tension in the role of the letter writer:
The theory behind this practice is that the committee doesn’t have time to familiarize itself with the work of all the hundreds of candidates who are applying, so they rely on an outside person who is already familiar with that work. Thus, the letter writer is, in theory, acting as the agent of the committee, with the task of helping the committee select the best person. In reality, most letter writers conceive themselves as agents of the candidate, with the task of helping their friend (the candidate) get the position that the friend wants.
The letters end up being useless, and the profession’s reliance on them has bad effects. Here are a few reasons why:
(1) The probative value of “letters of recommendation” is approximately zero… It’s basically like asking, “Tell us whom you’re friends with, and let’s see who can get their friends to say the most exaggerated things about them.”
(2) It contributes to a system in which the most coveted, scarce goods are distributed according to personal connections, rather than merit.
(3) It creates incentives for people to curry favor with the “big names” in the profession and thereby causes more of this to happen…
(4) It unfairly rewards people who are unashamed about bothering their friends for favors, while systematically disadvantaging people who are considerate of their friends’ time, or who tend to make friends with low-status people, or who tend to have unusually honest friends.
(5) It is manipulative. It recruits outside people to do work for the committee… but without paying them or giving them any benefit in return…
(6) It imposes an utterly unreasonable cost on the profession. [By requiring letters of recommendation, an institution offering one position] is imposing hundreds or thousands of man-hours of labor on the profession
See the original post for elaboration on these points.
Discussion welcome.
UPDATE: The following is from David Boonin (Colorado):
In case your readers are interested, here is the background on Mike’s blog post: our Center for Values and Social Policy is currently conducting a search for Visiting Fellows for the 2018-19 academic year. The ad is online here. We were originally going to require two letters of recommendation, but Mike sent me an e-mail suggesting that we drop that requirement and explaining why. I found his arguments pretty compelling, decided not to ask for letters this year as a result, and asked him to write his points up as a blog post. So a note to your readers: if you work in the values area and will be on paid leave (or retired) during part or all of the coming academic year, you can apply for one of our visiting fellowships without being unfairly disadvantaged for being someone who is considerate of your friends, or who tends to make friends with low-status people, or who tends to have unusually honest friends…
(photo of untitled sculpture by Tara Donovan)
“man-hours”? It’s 2017.
Seriously? The man is talking about a very serious issue that perpetuates the worst kinds of insider-ism, and this is your concern? It’s like a right wing parody of liberals.
All the more reason to be conscious of the very serious issue of sexism that pervades our language, I’d say. I mean, if the white male dominance in philosophy is not an insider-ism, then I don’t know what is.
Not sure if serious or trolling
Serious.
Goedele, white male “dominance” in philosophy is ambiguous between a totally benign tendency for more men to choose to go into the profession than women and something more sinister. Given your remarks about insider-ism, you seem to be alluding to the latter. But I haven’t seen particularly good evidence of it. On the contrary, I’d be willing to bet philosophy departments, given their left-leaning tendencies and general agreement with progressive causes, and the wide range of external incentives (funding, public perception, internal pressure from their universities, etc.), are both ideologically disposed to and externally pressured to recruit female philosophers at, if anything a disproportionately high rate from among the pool of applicants. I have seen efforts to encourage more women to enter philosophy. I’ve never seen efforts to encourage more men:
https://necpluribusimpar.net/women-underrepresented-philosophy-care/
If you think this isn’t the case, I’d be curious to hear why, and on what basis you believe there is insider-ism beyond the bald fact of the proportion itself:
Indeed it is, Daniel. I wish there were some way to test this, but I would honestly put money on there being an inverse correlation between comments like this and genuine concern for the relevant issues. This sort of thing comes across as about as genuine as David Brent’s efforts toward gender and racial equality. But not as funny. If you have nothing to say on the topic, could you please just say nothing?
He’s going to be one of my letter writers.
That is just too funny.
I agree entirely with this. Fortunately, our department has never been status obsessed, so letters from the current philosophical royalty never meant much to us. It does, however, mean a lot at a lot of places and thus has precisely the clubby effect the author rightly worries about.
Yet your department still requires letters. So it’s (or HR, or both) part of the problem, as Huemer sees it. If you’re going to bring up your department every chance you get, and continue to take shots at other people’s departments on top of it, try to at least be a bit more even handed, huh?
Thanks, E.
Actually E, it doesn’t any more (see David Boonin’s update above), at least not for the positions we are currently searching for. So, if you’re going to take shots at our department, try to use up-to-date information, eh? 🙂 To be fair (I know, I know, this is an internet discussion involving various people hiding behind anonymity, why would anyone be fair?), I only discovered that we changed our policy by reading David’s post above.
Alastair, I think E was taking shots at Daniel’s, not your department.
That’s right, N.
It’s Kaufman who constantly extols the virtues of not being a part of the “philosophical royalty.”
I think Alastair was confused because we (Alastair and I) have a colleague in our department named “Dan Kaufman”. But the Daniel Kaufman in this thread is a different person.
Mike is correct. I was confused by the Kaufman proliferation. Two (at least) Daniel Kaufmans in philosophy! Our profession is truly blessed.
Ha that’s funny! Whether we’re blessed is a different question.
A few years ago, some philosophers were arguing that screening interviews are worse than useless (“just read what’s on paper!”), and now this about letters. Yes, reading writing samples alone is better than reading letters alone. Reading entire dissertations would be better still.
Huemer’s arguments do seem persuasive; I myself was nodding along as I read them.
But I then remembered chairing every TT search in my department since 2000, and we are a small department, so guess who reviews the 200-450 applications each time? There is no divide-and-conquer here. Doing this without letters in a department as small as ours would just have magnified the pedigree effect (as others have noted). Colorado has 18 professors, and perhaps soon to be more? What works for a handful of large departments might not work generally.
I agree with the article. That you want to start fights and that JW wants to allow them is your problem and his. I would think it a good thing that people speak from their actual experience, rather than out of their rear ends and that they speak openly rather than anonymously. But I am more than happy not to comment here anymore. Cheers.
By implying that JW might want to disallow ‘fights’ with you, you’re getting dangerously close to contradicting your constant, Millian free-speech-grandstanding. By commenting here you are agreeing to be taken to account. As for anonymity, some of us are on the market and for reasons you should understand cannot risk alienating influential members of the profession like you.
JW is more than happy to call people out for various things and enforce certain rules of civility. I was surprised that he allowed this is all, given other times in which he has intervened.
I am not commenting on this anymore. But as you posed this directly to me, I thought it only right that I should answer you.
It’s quite rich of you to complain about civility when most of your activity here consists in verbally assaulting people (see your first comment).
As someone who has been active in publishing and public policy, but has not been connected to academia for many years, I’ve found obtaining these letters to be the most difficult part of job hunting. I’ve been forced to ask for request letters from professors I knew over 30 years ago rather than commissioners or agency heads I’ve worked with recently.
The points Huemer raises all strike me as important. I’m not sure how to weigh them against other factors, though. For example, for hiring committees without a specialist in the relevant area, a letter from a specialist in that area can be very helpful (though how helpful it is depends on how it’s written). If the profession were to ditch regular letters of recommendation, we would need something else to play that role.
I agree that the most helpful aspect of letters is that they indicate the nature and significance of the candidate’s dissertation. But, that valuable could be achieved by well-written dissertation summaries that normally accompany the CVs of junior candidates.
Let me make a qualified defense of reference letters. A caveat up-front: this is from the perspective of someone who’s been on search committees at Oxford and USC, where in both cases research quality and research potential is a major aspect of what we’re looking for. Daniel Kaufman is quite right to keep reminding us that this is not the typical hiring context.
The first thing to say is that in the overwhelming majority of cases, references are being written not for peers but for (current or recent) graduate students, normally your own or ones you have some other professional relation with. That dulls (it does not eliminate) nepotism issues. I am almost never asked for a letter where the request is supererogatory. It also qualifies Huemer’s concern about time costs. In practice letter-writers do not normally tailor their letter to the particular job (I don’t; most of the references I have read don’t). You have a standard letter, perhaps tailored a little bit towards different categories of jobs. The cost of writing references is real, but it’s on a per-student, not a per-application, basis, and it’s only a handful of hours per year.
So what are the positive benefits of references? I agree that sheer praise (“this person is marvelous”, “I wish we could hire them”) is of negligible use in most cases. There are two genuinely useful features:
1) Letter-writers will often give quantitative assessments of their students as compared to other students and junior academics they have known. These have to be read reasonably carefully but do convey information: there is a substantial difference between “this is one of the two best graduate students I’ve ever worked with” and “this is one of the three best graduate students I’ve worked with in the last five years. (Of course there is always a clear Gricean implication: “One of the N best” conversationally implicates “Not one of the (N-1) best”.)
2) More importantly, letter-writers explain a student’s research, its significance and originality, and how it fits into the broader debate in a way that is normally much better and clearer than any account the student themselves can give. That’s very helpful in initial assessments as well as later in the process when you’re reading a student’s writing sample (unless you’re a specialist in exactly the area the student’s working in, it can be difficult to assess the relevance of the student’s particular work to anything broader.
Finally, at some point these discussions have to engage with the need by search committees for *some* relatively-quickly-assessable way to sift the initial application pool down to a longlist.
The theoretically-ideal way to assess (the research part of) an application, no doubt, is to very carefully read the writing sample, follow up any of the cited literature to get a sense of the relevance and significance of the writing sample, and pay very little attention to anything else. This would take, at a conservative estimate, about three hours per application. But USC’s open-field Assistant Professorship search last year had *more than five hundred applications*. If a SC member were to spend even fifteen minutes on each application, that would be three full working weeks. Even if applications are initially divided up between many people, those people are still going to need to rely on things that can be assessed quickly.
What are the options?
– Pedigree. (I don’t need to rehearse people’s objections to *that*.)
– Awareness of candidates via networking. (Ditto.)
– Publication record. (Highly unreliable for early-career people.)
– Skimming the writing sample.
– Letters of reference.
None of these are perfect; all have flaws. But until and unless someone lets me step into a parallel timestream for the weeks it would take me to make an ideal assessment of a candidate, I’ve no desire to reduce the number of imperfect measures I have available to make a preliminary assessment.
“– Publication record. (Highly unreliable for early-career people.)”
Would you mind expanding a bit more on this, David? We’ve had positive experiences in our department with placing (significant) extra weight on publication record.
Well, the starting point is that I think all these indicators – references, pedigree, publications, the quick impression you get from skimming a writing sample – are imperfect-but-useful. (I pay plenty of attention to publication data.)
That said:
– we all know that the referee process can be capricious, especially given that people are incentivized to submit to very-oversubscribed generalist journals. For senior people that averages out, but at early career there’s still a lot of noise.
– there is a tendency for people to try to write up somewhat half baked ideas as soon as possible to get publications on their CV, and sometimes – refereeing is capricious! – it works. So publications have become less useful as people have realised that attention is paid to them.
– it advantages the student whose dissertation breaks into bite-sized chunks over the student who does something more unified.
“2) More importantly, letter-writers explain a student’s research, its significance and originality, and how it fits into the broader debate in a way that is normally much better and clearer than any account the student themselves can give. That’s very helpful in initial assessments as well as later in the process when you’re reading a student’s writing sample (unless you’re a specialist in exactly the area the student’s working in, it can be difficult to assess the relevance of the student’s particular work to anything broader.”
I agree that there is real value to this. But, instead of spending all the time writing them letters, why couldn’t faculty assist candidates in better expressing the nature and significance of their work in summary form?
with the best will in the world, you’re not going to be able to *teach* the student to be able to do this as well as you yourself can. You could write it for them – but that’s largely what the letter of reference is.
“with the best will in the world, you’re not going to be able to *teach* the student to be able to do this as well as you yourself can. ”
True, though one need not deny this in order to cogently reject the letter system. Imagine a system in which candidates write non-optimal summaries of their work on which committees must rely, but avoids the pernicious effects of the letter system, and a system in which faculty write nearer-to-optimal summaries of candidates work but involves the pernicious side effects of the letter system. Given that all candidates would be competing on the same playing field in the former (all are writing, with assistance, their own summaries of their own work), it seems clear that we should prefer it.
Not necessarily. If every student’s summary was non-optimal to the same degree, you’re probably right. But in practice, this just adds more noise to a situation where there is already much too much noise.
“But USC’s open-field Assistant Professorship search last year had *more than five hundred applications*”
When I chaired a search committee, it seemed like philosophy departments run much wider searches than other departments. It seemed to be the general expectation that departments would have written rather specific job advertisements and would be spending a lot of time scrutinizing applicants’ CVs and cover letters in order to assess how well they fit the job posting. Our broad postings that draw enormous pools that can’t be narrowed down by merely looking at their AOSes are exceptional.
There are definitely good things about our search procedure that I wouldn’t trade for other departments’ at all; it seems like we were much more likely to read applicants’ actual work than other departments, and that’s a good thing. But the open searches with 500 applicants do seem like they detract from the process; for one thing, by reducing the amount of time that people actually can give to reading the applicants’ work. (And by giving a greater emphasis to pedigree and publication record to narrow the pool with less effort.)
As for the options you mention, recommendation letters seem like they approximate pedigree enough to inherit most of the same objections (as Michael Huemer pointed out). Skimming the sample seems to me like the best quick way to go. (Yeah, it’s bad in some ways to reject someone on the basis of a skim, but maybe not as bad as to reject them on the basis of insufficiently enthusiastic recommendation letters.) It would definitely be a problem where there aren’t enough people in the department with relevant expertise, but otherwise I don’t think it’s a bad idea to emphasize looking at the sample more and recommendations less, at the very least.
Letters of reference contain a lot of information that you don’t get from pedigree.
As for your last suggestion: By all means shift the emphasis according to your assessment of what’s most useful. But let’s be clear what “skim” means, realistically, given time constraints. It can’t mean “read quickly”. It means something more like: read the abstract, skip to a random bit of the argument to get a quick sense of the applicant’s philosophical aptitude and writing style, skim the conclusion. And that’s already being optimistic.
I agree about skimming. Nevertheless, I wonder if the letters of recommendation would give me more useful information than I’d get by reading the introduction, skipping to what looks like the most important bit, skimming the conclusion, and if the paper looks good enough setting it aside for a bit more attention.
Part of the thought is this: Once a dossier has passed the initial cull, what do I do? I read the writing sample to see how good it is, because I’d like to judge the candidate by their work. Skimming the sample in this way seems to me like it gives a better indication of what the sample will be like than reading the recommendations.
I don’t know, though. If I really wanted to study this, during a search, I’d pick one batch of candidates according to recommendations, another according to a sample skim, and select another batch of the same size randomly; then I’d give all the samples a good read and see which pools produced the most finalists. But that would require me taking a lot of extra time during a search, and I’ve never done that.
Brief comment:
I don’t think “X is one of the N best” conversationally implicates “X is not one of the N-1 best”. It implicates merely that the speaker does not know how to rank X among the N best. Consider: Steph Curry is one of the three best NBA players. He might be 2nd, he might be 3rd. But he’s not 4th.
In general use, I agree of course. But in letters of reference, less so. (Though the comment was largely flippant.)
And this illustrates one of the problems with letters that I neglected to mention. People seem to have the illusion of a common understanding, where in fact there is none. There is not in fact a common agreement on what is implicated by a sentence in a letter, or what the ‘code’ meaning of a statement might be.
A reference letter doesn’t just state an opinion (“This person is terrific”); it gives an argument for it. And this argument can be persuasive or not. Search committees don’t (I trust) believe everything big-name philosophers tell them.
A feasible, arguably superior (though not perfect) alternative would be a system in which qualified letter-writers are randomly assigned (perhaps: anonymous) job candidates within their subfield whose dossier they are asked to judge carefully. Those letter-writers then write a letter of recommendation that will be made available (perhaps: anonymously) to hiring institutions to which that candidate applies. This accomplishes many of the theoretical goals of the current system, while avoiding the pitfalls of 1-4. I also think this reveals why 5 and 6 shouldn’t be independently worrisome. This system wouldn’t impose a new burden on the profession (6); it would be an attempt to more efficiently address a burden that already exists. And, arguably, this is not best thought of as people from one school’s being manipulated into doing free work for the hiring committees at others (5), but rather professionals being asked to play their part in the solution to a coordination problem.
I like this idea. I do have some worries about it introducing extra (broad) networking and luck. As it stands, getting around enough to know the people doing the hiring (or presenting to them in a memorable way or (…)) already presumably gives some sort of edge. This adds an additional edge in getting around in getting a nicer recommendation if a good impression was already made. Anonymity would mitigate this, but if someone does something memorable and the interactions last longer than getting names exchanged, the dossiers will give identities away, anyway.
That said, spending time interacting with others in the field seems like something positive and not so bad to have extra incentive in. There’s probably some worry for people without such opportunities, but that’s already bad. That already needs fixing. This would make it worse for them, though.
The anonymity might have an odd effect of incentivizing novelty. Seeing the n-th paper to make a small incremental move on the existing literature might not lead one to remember the author well. Seeing the stuff you can bust out at parties to keep someone’s eyes from glazing over would make a dossier more memorable and thus break the anonymity. I’d take the novelty as a plus, though.
That would absolutely impose a new burden on the profession. Familiarizing yourself with a candidate’s dossier is a pretty big ask. In the status quo, you’re usually only writing letters for people whose dossiers you’re already professionally obligated to be familiar with, since you’re mostly writing letters for students you’ve been advising or whose committees you’re on. In your system, students would still need advisors, the time you spend reading your own students’ work wouldn’t go away, it’s just now you’d also have to spend time reading the work of randomly assigned students for whom you write letters.
First, I didn’t say that it doesn’t impose a burden on *individual professionals*; I said that it doesn’t impose a new burden on *the profession.* Mike claimed in 6 that it does, because it “is imposing hundreds or thousands of man-hours of labor on the profession.” But the need to judge candidates already exists within the profession; a LOR system is (in part) meant to consolidate that work, eliminating some of the redundancy of people on each committee’s having to look as carefully at the work. That is, in theory, less of a burden on the profession.
It is also not at all clear to me that the system I mentioned would impose a greater *net* burden on individual professionals. Especially as more and more candidates are on the market for longer, many professionals find themselves writing for people other than their students (I personally have 4 writers who I never worked with as a student) or for students whose work has changed significantly since they worked together (my writers frequently ask me to send them new things so they can get up to speed). If there were no longer any expectation to do that work, but only to familiarize oneself with (say) one new dossier each year, that might very well be less work overall for many people.
My concerns with this system:
1) Regarding burdens: you would move from a system where some minority of letter writers are somewhat unfamiliar with a student’s work, to a system where all letter writers are completely unfamiliar. That can’t fail to be a big aggregate increase in work.
2) You would create a huge logistical difficulty. At the moment – and contra comments about “working for free” – most references are written for your own current or recent students, and doing it is part of your job if you work for a PhD- granting institution. By separating letter writing from PhD supervision you lose all mechanisms to make sure that letters actually are written and are written diligently.
I absolutely agree
People who come from smaller school ABSOLUTELY need letter writers, if they are to get a job at all. Getting rid of letters benefits the rich, who get richer. Unless committees actually read the work of an applicant (and this is rare at smaller schools), letter writers are absolutely needed.
This is entirely right. We have some evidence of what a system without letters is like; much of the UK uses it. And the result, at least to my casual observation, is that the bias in hiring towards the bigger name schools is even bigger than it is in the US.
If you don’t have letters in a file, and there are 500+ files to read, then every student from a smaller school who doesn’t have a large number of publications (e.g., more than 4) is going to find it very hard to get past an initial cull.
Repeating myself more concisely, perhaps the solution is in part to narrow down the search so there aren’t 500+ files.
I sort of see this – and maybe that’s what is needed. I definitely use letters much less in a targeted search with 80 candidates than I do in an open search with 500 candidates.
But that has downsides too. For one thing, the more narrowly targeted the search, the easier it will be to miss people whose work doesn’t fit traditional categories. And we are encouraged by the university to have as broad searches as possible for diversity reasons. I don’t know quite why they encourage this, but it could be because of this ‘falling between the cracks’ problem.
And we did what felt like a fairly targeted search for a PPE candidate a couple of years back and got 250 applicants. That’s more manageable than 500, but it’s still a lot.
Why not use a simple formula for the initial assessment of candidates? Something like the following maybe:
(#of publications in top journals)*10 + (# of conference presentations at top conferences)*3 + (# of publications in non-top journals)*2 + (#of conference presentations at non-top conferences) + (Some kind of normalized teaching score based on the eval scores provided by the candidate).
This formula would provide a ranking, and then search committee members could go down the list, starting with the top-ranked candidates. They’d stop once they found a suitable candidate.
If the formula turned out to give a sub-optimal ranking, the coefficients could be revised.
Why not do this? Because grad school is supposed to be about becoming an expert and writing an awesome dissertation, not about publishing in top journals.
Yeah, if you think that right now grad students are publishing too little and want a reform that encourages them to publish more, that could work. But I don’t think most people agree that grad students are (collectively) publishing too little.
Arthur Greeves, that may be what grad school is about, but the goal of a hiring committee is to choose the best candidate for the job. If someone’s published in a top journal that’s a decent (not perfect) and easy indicator that they’re an excellent candidate.
Brian Weatherson, would the proposed screening/ranking system incentivize students to try to publish even more than they already are? Maybe, but so what? That doesn’t mean it’s not a good way to screen candidates. Also, it’s not like you have to advertise publicly that this is how you evaluate candidates. So it’s not necessarily the case that implementing the system in your program will change anyone’s behavior.
Ben, you’re right that publishing in major journals is a good indicator. But there will be excellent candidates you will miss, if this is the main indicator you use. And scholars will be encouraged to publish before they’re ready, which is unfortunate.
The top journals — a curious construct , in its own right, with no official sanction in the profession, and deserving of more scrutiny than I’ve time to give it — tend to publish some sorts of scholarship more than others, some authors more than others, and some topics more than others, or at least so Kieran Healy’s data would suggest. I can imagine that weighing publications in them heavily will track some forms of excellence, but I also imagine that doing so would systematically work against diversity of applicants and scholarly endeavors in other ways.
I think there is another problem with references that hasn’t been mentioned in the article. I know of graduate students who are not keen on disagreeing with or confronting (senior) academics about certain issues (not academic ones, but departmental issues, for example) because they are the ones that will be writing their reference letters. Especially in the discipline of philosophy, I find that highly alarming.
Absolutely, I have received my PhD degree a few weeks ago. Part of the reason why I could not dare challenging my supervisor’s stone age beliefs was the dire knowledge that I would be relying on her recommendation letters to eventually find a job.
I propose that these positions be allocated via sortition, to avoid bias. Those chosen could then earn or fail to earn tenure in the ordinary way, with department members taking the time to learn enough about their work, teaching, and collegiality to make an informed decision. (Just Kidding! [Or am I? {Would it be an obviously worse system? What about a matching system like newly minted medical doctors in the US?}])
OK, the comments section is truly bizarre. What Daniel Kaufman is replying to keeps changing. What gives?
It seems as though the comment about the use of “man-hours” has disappeared, and (perhaps consequently?) some of the replies to comments are wandering into the wrong threads. Looking over it, it also seems like Spencer Case’s comment that Huemer is one of his letter-writers has disappeared too. I’m guessing this is some kind of database error rather than deletion?
On that topic, I’ve found that posts on individual topics often display very oddly on my computer in Firefox and Safari. (My Firefox setup is admittedly weird and things occasionally break on it, but my Safari setup should be pretty much unmodified.) If I click on a post title, everything is huge and the layout is messed up, and comments never appear. The only way I can get individual posts and comments to display properly is by clicking on a link in the “recent comments” sidebar. Has anyone else been having this issue? (There’s going to be some selection bias here, as anyone else who has this issue and doesn’t click on “recent comments” won’t see that question….)
Hi Matt. Thanks for your comment. As I note in the comments policy, if a sufficient number of readers report a comment, it automatically gets removed. When this happens I get notified by email; I check the comment and then I can manually re-approve it (or not). Sometimes this takes a while, as I have other things to do besides run Daily Nous (e.g., teach, write, eat, spend time with my kids, see friends, go to a show, watch a movie, sleep… you know, life).
As for the formatting issues, I will look into that. If others have the same problems, it would be useful to hear about.
Hi Justin. I didn’t mean to suggest you should be spending more time on the blog. I just assumed that it couldn’t be a deliberate deletion issue, because… well, framing this as an open letter to commenters:
To anyone who reported Tim Collins’s comment: Come on, even if you disagree with the comment about the use of “man-hours,” or find it unproductive, it doesn’t seem like the sort of abusive or trolling comment that should get reported.
To anyone who reported Spencer Case’s comment: What the heck, people? I can’t see what on earth could have led you to report that. Do better.
(Also it seems undesirable that when a comment gets deleted its replies get attached to other comments, but that’s probably a software thing that is hard to fix on your end.)
If we assume that commitees read dossiers credulously, unprofessionally, incompetently, or cynically when they’re looking at the letters of reference, then it’s hard to see why we’d expect better when they turn the page and read CVs, cover letters, and submitted research samples. But if we don’t assume this, none of the arguments adduced in the OP has much force against the practice of soliciting and considering letters of reference — among much other relevant evidence.
David Wallace wrote:
“Even if applications are initially divided up between many people, those people are still going to need to rely on things that can be assessed quickly.
What are the options?
– Pedigree. (I don’t need to rehearse people’s objections to *that*.)
– Awareness of candidates via networking. (Ditto.)
– Publication record. (Highly unreliable for early-career people.)
– Skimming the writing sample.
– Letters of reference. …”
I submit that there are additional options. I suggest a few below. These options, if made priorities, would, I think, go some distance to put into practice the boilerplate claims with respect to “equal opportunity” and “affirmative action” that many departments include in their job postings :
(1) identify and concentrate on applicants who work on emerging issues and in dynamic subfields that members of underrepresented groups in philosophy have introduced and gravitate toward,
(2) identify and concentrate on members of underrepresented groups,
(3) identify and concentrate on philosophers whom you know fit the criteria in both (1) and (2),
(4) develop other creative ways to hire for the benefit of the profession at large rather than merely to serve your own departmental interests, and
(5) think more progressively, expansively, and innovatively about what your department defines as “teaching needs” and departmental “needs” in general.
On (2) and (3): If you mean “look with particular care at applicants from underrepresented groups, to make sure you don’t overlook talent”, then all searches I’ve been on have already done that (rightly so in my view). But (whatever its other virtues) doing so does not solve the problem of how to sift all the other applications, and increases rather than decreases overall time costs.
If you instead mean “construct the longlist by discarding applications from non-underrepresented groups”, I agree that this would be quick and practical to implement. But it would also violate any university Equal-Opportunities policy I’ve ever come across; at least in the UK and (I think) in California, it would also be against the law.
Of the other suggestions, (1) is a special case of “define your search more narrowly”, (discussed by others above) which does help somewhat and has its advantages but is not a panacea. (4) and (5) don’t amount to concrete suggestions that I could use in a hiring process.
David,
I found your response to me curious. For example, you have listed criteria that departments can use to prioritize candidates. I suggested additional, different criteria that might be used. In response to me, you have suggested that I might be recommending something illegal. I don’t see how any of your recommended options couldn’t also be taken to mean, e.g. discard any applications that do not come on Oxbridge or Ivy League letterhead.
FYI, discarding applicants from low-status schools would not be illegal. Discarding white male applicants would be.
That’s what I would have guessed. (Though in the UK at least, I can theoretically imagine an indirect-discrimination case, if candidates from protected groups are disproportionately represented in low-status schools.)
I would be interested in learning more about the US legal framework here – I’ve found it much harder to get information than I did in the UK, presumably because it tends to be a state-level matter?
In general, anything that counts as unconstitutional discrimination will be ruled out in all states, but some states can, either legislatively, or through ballot initiative, or through state constitutional amendments, have more expansive definitions of discrimination than that used at the federal level. E.g., a number of states–California, Michigan, Florida, maybe others, links below–have ruled out race-based affirmative action in public schools, but that’s not mandated across all states.
https://en.wikipedia.org/wiki/California_Proposition_209
https://www.nytimes.com/2014/04/23/us/supreme-court-michigan-affirmative-action-ban.html?_r=0
From three relevant laws:
⋅ The 14th Amendment:
“. . . No state shall make or enforce any law which shall abridge the privileges or immunities of citizens of the United States; nor shall any state deprive any person of life, liberty, or property, without due process of law; nor deny to any person within its jurisdiction the equal protection of the laws. . . .”
⋅ The Civil Rights Act of 1964, Title VII:
“It shall be an unlawful employment practice for an employer – (1) to fail or refuse to hire or to discharge any individual, or otherwise to discriminate against any individual with respect to his compensation, terms, conditions, or privileges of employment, because of such individual’s race, color, religion, sex, or national origin . . .”
“Notwithstanding any other provision of this subchapter, (1) it shall not be an unlawful employment practice for an employer to hire and employ employees . . . on the basis of his religion, sex, or national origin in those certain instances where religion, sex, or national origin is a bona fide occupational qualification reasonably necessary to the normal operation of that particular business or enterprise . . .”
⋅ Title VI:
“No person in the United States shall, on the ground of race, color, or national origin, be excluded from participation in, be denied the benefits of, or be subjected to discrimination under any program or activity receiving Federal financial assistance.”
⋅ Education Amendments of 1972, title IX:
“No person in the United States shall, on the basis of sex, be excluded from participation in, be denied the benefits of, or be subjected to discrimination under any education program or activity receiving Federal financial assistance [. . .]” (followed by a series of exceptions)
In response to my initial comment, David wrote:
“(4) and (5) don’t amount to concrete suggestions that I could use in a hiring process.”
In fact, if one conducts a search of the PhilJobs site, one will find some job postings from departments that are using criteria of this kind to improve the representation of their departments, using phrases such as “The department is especially interested in candidates who…”
which means it encourages their applications and will examine them with particular care, not that it would select them over and above other criteria.
My point was/is that my list of criteria is a list of criteria *that can be used quickly to assess candidates in the longlisting phase of a job search*. Most of your suggestions (whatever their other merits) don’t provide such criteria. I identified one that was ambiguous and which on one disambiguation would provide such a criterion, but then noted that for legal reasons that criterion could not be used.
I think you may be reading my earlier post as broader than it is. It is not a discussion of all the various vicissitudes of hiring. It is aimed rather specifically at the problem of obtaining a longlist in a manageable amount of time given a very large number of applications.
David,
once again, I think that you have been too quick to dismiss what I have suggested. My suggestions in 4 and 5 concern areas of research and teaching, for instance. Surely, these criteria can be used to narrow the number of applications. They are already, aren’t they? Doesn’t a job advertised as “Open” get many more applications than a job advertised in a very specific subfield? I am suggesting this sort of criterion with the *bonus* that the areas of particular interest be emerging areas, underrepresented subfields, especially ones to which members of historically underrepresented philosophers gravitate.
You seem to be determined that the criteria you list are set in stone. Were all of the criteria you list, that is, were all of the criteria currently used to arrive at a limited number of candidates always used? No. That, to me, suggests that they need not always be and need not be the only ones used. Why don’t you give my suggestions a little more thought? You seem to have written them off in great haste.
Shelley, if you have a careful look at Philjobs this year, you’ll realize that underrepresented areas, or areas attracting underrepresented groups, are pretty heavily represented, perhaps even to the disadvantage of those doing more traditional stuff. Whether it’s for ill or good I don’t know, but at least you can’t complain that criteria like those you propose are not being used (1) to narrow down the pool of applicants and (2) increase the representation of particular sociological groups or philosophical areas. Granted, disability still doesn’t seem hot enough for search committees, but this may change, and the range of applied ethics/bioethics jobs advertised makes ample room for those interested in these questions.
I understand one purpose of letters may be to facilitate the hiring department’s job. But that shouldn’t be in lieu of furthering the candidate’s interests. That’s a false dichotomy. Assuming—perhaps contentiously—that a reference shouldn’t agree to write a letter if it would hurt the candidate’s prospects, then letters, if they’re not overly emphatic and insincere, are serving two functions, not one. For many references (advisors, mentors, former colleagues, collaborators) it’s part of their job to support candidates they think are worthy of their recommendation. There is no, or shouldn’t be, inconsistency between facilitating the committee’s work and supporting one’s candidates. Given the terrible market we’re in, if the purpose of letters was merely to be transparent about whatever flaws and strengths a candidate has, then we had better do without them until everyone agrees to use them in this way.
I would just add that everyone is never going to agree to use them in that way.
Exactly. And that’s why it would be unfair to expect references to be transparent, neutral purveyors of information. It would harm candidates whose references are primarily intending to help SCs while others are primarily concerned to further their applicant’s interests.
In the UK, I expect candidates to explain the interest/relevance/broader implications of their research in their cover letters. If they really need referees to do this for them – the most common argument so far for reference letters – then I really worry about their skills as a teacher. (Surely intro to philosophy courses require someone to explain why some apparently obscure topic is so important). I really do feel that this is one way in which the UK convention of a moderately detailed cover letter is preferable to the US convention of little more than a cover note.
This is not the US convention anymore, and has not been for some years.
Yes, sorry, I should have been clearer. I know it isn’t the US convention (at least anymore), but I think it is a good convention and one which should be in place. Surely the entire point of the op was to question a convention – seek cover letters. Some people objected to this on the grounds that important information- context for the PhD – would be lost. My point was that by reinstating the detailed cover letter convention, that loss would be mitigated (at least, for candidates who would also be good teachers). Sorry if I was obtuse.
In the institution I work in, hiring has two distinct phases. In the first academic departments identify the candidate(s) we are most interested in having as colleagues. In the second, we try to persuade our administrators (Deans, Provosts and the like) to hire them.
Even if we don’t use letters in the first phase, they’re pretty much indispensable in the second phase. (Which means in practice we can’t completely avoid considering them in the first phase, since ‘can we persuade the university administration to hire this candidate?’ Is a question we always have to ask ourselves before moving from phase one to phase two.)
N wrote:
Shelley, … Whether it’s for ill or good I don’t know, but at least you can’t complain that criteria like those you propose are not being used (1) to narrow down the pool of applicants and (2) increase the representation of particular sociological groups or philosophical areas. Granted, disability still doesn’t seem hot enough for search committees, but this may change, and the range of applied ethics/bioethics jobs advertised makes ample room for those interested in these questions.
N’s remarks articulate the sort of reductionist (mis)understanding of philosophical work on disability that I labour to undermine in my forthcoming book,, Foucault and Feminist Philosophy of Disability (University of Michigan Press). The book comes out December 1. I hope readers and listeners of Daily Nous will order it! You can find out a bit about the book here:
http://philosophycommons.typepad.com/disability_and_disadvanta/2017/10/more-about-foucault-and-feminist-philosophy-of-disability.html
i’d be curious to know why you think it’s a reductionist misunderstanding without having to purchase your book, though
N,
I know that the cost of books is prohibitive to many of us. So I hope that many university, college, and public libraries will add my book to their catalogues. I think you will derive an idea of my arguments in regard to the question of how philosophical work on disability/philosophy of disability gets classified if you read some of my work, in particular, my papers “Introducing Feminist Philosophy of Disability” and “Disabling Philosophy.” Both of these papers are available through PhilPapers and academia.edu.
I’ve got a really radical suggestions. How about screening through telephone interviews before contacting any references! What a thought!
In my department, we don’t even request letters of recommendation until candidates have been short- or long-listed.
I have just received my PhD degree and I am pretty certain that most academics would never give up writing recommendation letters, irrespective of how much work they put in it. This is a major source of power for them over their subordinates.
You must know some truly awful people Laertes. I am sorry for what has obviously been a most unpleasant graduate school experience for you. Not a single philosopher I know regards recommendation writing as an exercise of power. Everyone I know regards it as a serious, and seriously stressful, professional obligation. We are trying to do the best we can for our students. And again, none of the philosophers I know regard their students, or anyone else for whom they write letters, as subordinates (or at least if they do, they are careful to hide that attitude from me).
Just a few remarks (of course, I can’t comment on everything that has been said above):
– A practical proposal: if you agree with what I said about LoR’s, then, the next time your department advertises a position, propose that letters of recommendation be omitted from the application requirements. Either propose this as a motion at a dept meeting, or propose it to the person writing the ad.
Most of the things that people propose to improve the profession are things for which it isn’t clear what any given individual should do about them. This one isn’t like that; there is a clear, concrete step that you can take if you agree.
– Back to the debate: Are LOR’s sometimes useful? Of course. You can easily find cases in which a letter was helpful. But that has to be weighed against all the times that they are misleading and hence harmful, plus all the negative effects on the profession as a whole. I note that most of the negative consequences I cited have not been disputed.
– What should you do instead? FYI, I have been on many search committees, plus graduate admissions a few times. I don’t read letters. I rely on the rest of the file, which has plenty of information without the letters. I know there are other people who don’t read letters either. We manage. If you’re worried about how to cull 500 applications, that is a real problem, but it has little to do with the LOR debate, since the LOR’s for 500 applications would be over 2000 pages, and I doubt that one could or should read 2000 pages.
Moreover, I think that letters give you negative information — that is, they are worse than merely uninformative; they are, taken on the whole, positively misleading. Given that, I think “what else are we going to do?” is not a good argument.
Analogy: Patient P has an incurable disease. Doctor D1 proposes to beat P with a stick as a treatment. When D2 objects to the stick treatment, D1 says, “Well, how would *you* treat the patient?”, with the implication that unless D2 has a cure, we should go with the stick treatment. I would say D2 doesn’t need to propose a treatment, and then have a debate about the merits of that alternative; it’s enough to see that the stick-beating is harmful.
FYI, I would happily give up writing letters today, if I could. Professors don’t generally write letters because they want to, to exercise power. It’s not the letter-writer’s idea. It’s the schools *advertising* positions that coerce everyone into writing letters. They threaten to not hire or admit our students, if we don’t engage in this pernicious practice.
Just a few remarks (of course, I can’t comment on everything that has been said above):
– A practical proposal: if you agree with what I said about LoR’s, then, the next time your department advertises a position, propose that letters of recommendation be omitted from the application requirements. Either propose this as a motion at a dept meeting, or propose it to the person writing the ad.
Most of the things that people propose to improve the profession are things for which it isn’t clear what any given individual should do about them. This one isn’t like that; there is a clear, concrete step that you can take if you agree.
– Back to the debate: Are LOR’s sometimes useful? Of course. You can easily find cases in which a letter was helpful. But that has to be weighed against all the times that they are misleading and hence harmful, plus all the negative effects on the profession as a whole. I note that most of the negative consequences I cited have not been disputed.
– What should you do instead? FYI, I have been on many search committees, plus graduate admissions a few times. I don’t read letters. I rely on the rest of the file, which has plenty of information without the letters. I know there are other people who don’t read letters either. We manage. If you’re worried about how to cull 500 applications, that is a real problem, but it has little to do with the LOR debate, since the LOR’s for 500 applications would be over 2000 pages, and I doubt that one could or should read 2000 pages.
Moreover, I think that letters give you negative information — that is, they are worse than merely uninformative; they are, taken on the whole, positively misleading. Given that, I think “what else are we going to do?” is not a good argument.
Analogy: Patient P has an incurable disease. Doctor D1 proposes to beat P with a stick as a treatment. When D2 objects to the stick treatment, D1 says, “Well, how would *you* treat the patient?”, with the implication that unless D2 has a cure, we should go with the stick treatment. I would say D2 doesn’t need to propose a treatment, and then have a debate about the merits of that alternative; it’s enough to see that the stick-beating is harmful.
The social science data on LOR’s: https://orgtheory.wordpress.com/2014/10/29/letters-of-recommendation-still-garbage/
“The reliability problem is so severe that Baxter et al. (1981) found that there is more agreement between two recommendations written by the same person for two different applicants than there is between two people writing recommendations for the same person.”
The link to the job post directs to an email attachment no one outside that institution has access to. The proper link is: https://philjobs.org/job/show/8642