A Gentleman's 'B+'

We're smart, we're talented, and doggone it, people like us (or at least we like ourselves). But every so often we members of the Harvard community feel the need to engage in a public confessional ritual. To say, in effect, that reports of our superiority have been greatly exaggerated.

I'm talking, of course, about the latest revival of the "grade inflation" debate, spurred recently by Prof. Harvey "C." Mansfield's short-tempered comment that grades became devalued because professors fear giving low grades to minority students. In addition to the familiar round of inter-ethnic name-calling, this infelicitous statements provoked additional discomfort among faculty and students by raising the question of whether grading policies reflect objective academic merit or subjective agendas and power relations.

Unfortunately, the entire discussion is predicated on the actual existence and undesirability of such a thing as "grade inflation" at Harvard. Having assumed such a problem, the debate has focused only on its causes and cures. This is the sort of collective mindset that prompts idle New York Times reporters to run up to Harvard and Yale to interview crusty old professors about the good old days when a C was a "gentleman's C" and an A was as rare as present-day Harvard professors who grade their own papers.

Some more productive and fundamental questions to ask about the current value of grades at Harvard would be as follows: Why are grades given in the first place? Who is their expected audience? What does a certain statistical distribution really mean, and is it proper to establish grading policies that aim at a specific statistical pattern? After examining these issues, we may find that "grade inflation" is a problematic if not a meaningless way to characterize Harvard students' academic situation.

To begin with, people go to Harvard (or any college) for three reasons: to edify themselves and seek truth, to learn useful skills, and to obtain a degree and a record which will have an impact on one's success as a job applicant. Grades are not relevant to the first two of these agendas.

Some have argued that a student will only work hard and truly learn the material if he knows he is being graded. Nonsense. This is not elementary school. By now we have learned some degree of self-discipline, and if we are seriously interested in edifying ourselves, we need no external incentives.

In fact, it would significantly improve the quality of student-professor relations if it were openly acknowledged that the threat of bad grades shape students' learning agendas such that they primarily seek not to learn what interests or uplifts them but what professors want them to believe.

We come now to the third purpose of a college education, namely getting a good record so one can get a good job. This is obviously the purpose of grades: to show how diligent and intelligent we all are, based on a series of evaluations of our performance here.

Yet the quantification of this student performance into grade points obscures the fact that the assignation of these objective-seeming scores is always relative to some context and to some philosophy of how best to compare students' worth. Should grades be calculated relative to others students in the class, to those in other classes, or to those in the whole school? And should they follow any statistical bell curve pattern, or is anomaly to be expected from such an unnaturally high-quality population of competitors? This is what the grade inflation debate should be about.

The purpose of grades is thus to winnow out the best students, the next best, and so on. How different, though, are most of the students here from each other? With the exception of the occasional genius or the handful of students who blow off a class due to laziness or lack of aptitude in that area, most people here work hard and do work that deserves a B+ or better.

Nonetheless, course instructors feel that they must ration the number of good grades they give, because if everyone got honors grades it would be "grade inflation." The result is large differences in letter grades between students whose actual test scores in class only differed by a point or two.

In humanities classes, these fine distinctions are even harder to make. Faced with a pile of papers which are all decently written, grammatical, properly researched and well-reasoned (not too much to expect from the average Harvard student), a grader who "must" fit the actual distribution of students' abilities into a bell-curve model of grade distribution will have no measurable criteria for winnowing students, and will pick the "best" papers based on such concerns as which ones coincided most with the grader's own views and methodology.

The outsiders for whose eyes these transcripts are ultimately intended--employers--have no way of knowing how minimal, according to real-world standards, the differences between Harvard students really are.

Just think: out of all the applicants to Harvard, only about 17% get in. Many more college-age people across the country have already been winnowed and tracked by their schooling up to then, with the result that they know better than to even apply here.

By the time a Harvard student body is assembled, its members have been sifted and evaluated for so long that most of the significant differences between us and our age-group peers are no longer an issue. Most of us really do high-quality work relative to our fellow teenagers--who are, after all, our real competitors in the job market after college.