Most of the jobs I'm applying to are requesting three letters of recommendation, although one place is asking for four. Question: if I'm going to ask four people to write letters, should I submit four letters instead of three to the places that are asking for three? Is this standard practice or is it considered obnoxious? The advantage of four letters is that I could then submit a couple of letters from grad school and a couple of letters from my postdoc, which I think would give a more well-rounded picture of what I've done in both places. But I really don't want to annoy any SCs.
From what I've heard from members of our department who've been on committees, if the application says three, submit three. You don't want to come off like you are submitting blanket applications. Tailoring your materials to the particular school is probably the best and if you don't submit the proper application materials they may think you are prone to overlook directions. With that said, I don't think submitting 4 letters would ever keep you from getting an interview if you are at the top of a school's list. This would likely only ever come up if the department had 1 more invite to give and you and another candidate were neck and neck.
This is interesting - I've now asked my graduate mentor and my postdoc mentor about this and they both said, unequivocally, submit more than three - most candidates will. In fact my grad mentor wanted me to ask 5-6 people. I guess it goes to show that there's not necessarily a consensus on this among faculty.
A word of caution: If you are going to request extra letters, make sure they're going to be really good. Even a positive but less-than-compelling extra letter may hurt rather than help your application.
here's a risk: if a dept is strict, they might take "the first 3 letters they receive"… what if the first 3 they receive are not the strongest of your 6 letters?
when i was on the market, i was advised to only submit what each school asks for, not to leave anything out not to add anything in. for example, most schools ask for 3 publications; do you submit 3 or do you submit 6? but maybe letters are a different beast because anyone can go and get your pubs no matter how many they "ask" for in the ad… people who are in your area will likely go and look at more than the 3 pubs you send them.
Isn't the key to this the relevance? I could send 10 letters, but I don't need them. I need good letters from people who have worked closely with me and through key points of my career. I have one from someone who has seen me through all phases of my career, but with whom I did not work super closely. I have one from my grad school mentor, and letters from my two close post-doc/current mentors. I think that covers it. If I were most interested in teaching or clinical work, I might add one from a supervisor in that area.
As a faculty member who has gladly spent hours writing letters to help students get jobs, and as a member of many search committees who has had to read countless letters, I implore all of you to send only the number of letters requested. You may think that it is not a burden on your letter writers ("after all, they've already written it"), but keeping track of what letters got sent where does take time—and some of your letter writers will, like me, actually individualize letters for each job.
Moreover, I have found that letters usually serve more of a dissuading role than a convincing one, i.e., we will use red flags raised in letters as a reason to rule out a candidate, but I cannot think of a case where it was the letters that catapulted one candidate above another to make it to the short list. Most letters are uniformly glowing and thus relatively uninformative.
I personally wish the field would adopt the standard practice of requesting only names of references and then asking for letters only from those on a long list. But that's grist for another thread.
No one has mentioned yet that there is the additional issue of more letters if they come from well-known people in the field. I have been encouraged to ask letter writers simply because other mentors think that a letter coming from them would be helpful. More people might know the researcher and therefore his/her comments might be better received.
letters from big sharks out there are only helpful if they are genuinely and distinctively positive. If they are generic, I don't see how more helpful they are.
But, it can't hurt to have letters from people that everyone loves and respects. At least, it shows that you are well connected.