If I apply to a job, and it asks for 4 recommendation letters, is it good if I submit 5 or more letters? Why and why not?
In the US in math my understanding is:
- For the NSF fellowship follows the rules to a T.
- For other postdocs there's probably no harm in submitting an extra letter, but there's also rarely a compelling reason to do so.
- For tenure track positions it is common, and not frowned upon, to submit 4 or 5 research letters instead of the requested 3 research letters. If you have a compelling reason to submit more (for example, you work between two adjacent fields) then go ahead, but if you don't have a compelling reason you're also not going to be penalized for sticking with 3 research letters. It's better to have all the letters be great than to have more letters.
There's a lot of discussion about this in the comments on this thread.
The search committee, or whoever it is who is reviewing applications, asks for 4 letters because that's the number of letters they want. For a postdoc, four is logical because that is typically the size of a doctoral committee.
Unless you have a very good reason (which, off the top of my head, I can't think of), don't send more letters than they ask for. (1) They don't want to read them. (2) They will think you don't know how to follow directions and will toss your application (which is what I would do, if it wasn't clear why I was getting >4 letters). In my field of biology, there might be over 100 applicants for a position (particularly a faculty position; I've heard of 200-300 in some cases). I've been on numerous search committees, and I really can't fathom some of the materials that get sent in with applications.