Reference letters: How do you write them? How do you read them?
Experience indicates that lots of people have lots to say about this. So I’ll just get out a few ideas about writing to start a conversation.
I can’t think of any way to control gush, so I’ll move on. Here are some other things to consider.
1. In writing a letter, be careful that you do not inadvertently raise a flag. For example, if you say “X has the remarkable capacity to go off and master the relevant scholarship by herself.” You might think you’ve said something rather fine about X: you might even secretly envy her independence and diligence. But note that some may ask: Why is X doing this by herself? Is she unduly isolated? Is she non-collegial? Does she have no scholarly support? Other parts of the letter should therefore make it clear that X is a collegial person who is always ready to share and discuss her knowledge.
2. Contextualize the research, but do not repeat or summarize it. Many letters repeat at convoluted length what the applicant presumably says in her dissertation or writing sample. This is irritating, but it also raises a flag: Is this the letter-writer’s idea or the applicant’s? On the other hand, it is good to contextualize the research. What is the problem it addresses? Why is it an advance?
3. Don’t be naïvely anecdotal. “ When X gave a talk on campus, the Applicant asked a question that . . .” It’s never clear how these tales of victory and conquest are to be taken.
4. Placement officers ought to read all letters, and they should be quite bossy about editing them.
5. Maximum length: 2 ½ pages, 700 words. Minimum length depends on whether you are the supervisor or merely the third or fourth letter. But less than one page is too little.