I happened to be chatting a couple of days ago with a professor from Harvard about grade inflation. He is not worried by the fact that the scale’s been reduced so drastically. For him, A, A-, B+, and B (the only four viable grades at Harvard, as here) work perfectly well. They are a code, he says, and we can all decipher it. These four grades convey as much information as a broader scale would. The finessing can be done in letters of recommendation.
It took me a while to figure out why he and I differ on the issue. He teaches in a graduate school. For him, grades are primarily an admissions tool: he uses grades to work out whether a student should be allowed to enter his program. From this perspective, he is right. There is a code; we all do understand it; and the four-grade scale gives us as much information as we need. But I do not think of grades primarily as conveying information. I think of them as challenging a student to do her best. For this, I need sometimes to be able to give a student a lower grade without it destroying her career.
One possible solution is to give two grades, the code-grade that goes on the transcript and the grade you think the work actually deserves. This system is widely associated with Harvey Mansfield, but I’ve recently learned that Gayatri Spivak uses it too. The problem, though, as a colleague points out, is that our students aren’t likely to care much about the second grade. We aren’t Mansfield or Spivak, and work marked A-/B- is more likely to make our students resent us than to spur them to more effort.