Upper limit for grades?

I teach undergraduate math classes at a university. I have done so for around two decades now. I never grade on a curve. There is a good reason for this. One can not do well in a math class unless one knows how to do the previous math class’ subject matter pretty well (for those math classes that go in a specific order). Thus, it does no kindness to a student to give him an A in Calculus I if he doesn’t understand the most basic concepts since when he goes on to Calculus II, he won’t be able to have any success there either. And when he goes on to a science class that presupposes he knows some basic Calculus, he would run into trouble again.

It’s important that someone know how to add and subtract (with signed numbers and fractions and everything) before they start algebra. It’s important that someone understand algebra and trigonometry before they start Calculus. Giving everybody high grades is not the answer (though I’ve seen it done by colleagues) because it simply makes the students unprepared for later classes, where the professors then have to repeat the process of giving high grades to people who can’t actually succeed at anything, and what’s the point of that?

If someone passes a Calculus class, there is an expectation that they can actually successfully work Calculus problems.

And with relatively small class sizes (first place I taught might have 55 students in a class and the second place I taught might have 35), it is possible to get samples that differ greatly in ability. For example, at the worst (in terms of the preparedness of students to take math classes) place I taught, there was an algebra class added at the last minute. The only people who signed up for it were (apparently-- I can’t say with 100% accuracy) irresponsible people who waited until the last minute to sign up for classes. And having a class full of only irresponsible people was not a recipe for success. Tests that weren’t troublesome for other classes were huge disasters for this class, with often only one or two people passing a given test (and not always the same people). If I had graded on a curve, people who couldn’t work the simplest algebra problem would’ve gotten passing grades and then had to go on to more advanced math classes where they would’ve been hopelessly lost. Sometimes it is necessary to fail more than half of a class.

I have often taught the same exact class with the same exact assignments and tests during a given semester (or quarter), and the ability level of the different classes can be quite astounding. I can tell the difference. And even when the tests and assignments are different (as they would be in different semesters), they’re not that different, having only some numbers changed around for example, so I seriously doubt that there’s any significant difference in difficulty level.

And I remember when I was in college getting my bachelor’s degree, there was a particular class where I got the highest grade. And that grade was a “B”. And at the time, I was mildly perturbed that the highest grade in the class wasn’t an “A”, but looking back, I realize the professor was right to give me only a “B” since I had not done the work to deserve an “A”.

I taught an upper division Forensic Anthropology class, and I considered myself a tough grader. Since this was an intersession class (given during the Christmas break, meeting 8 hours a day), and wasn’t required for any major, most of the students were taking it because of their interest in the subject and were highly motivated. Almost everyone got either an A or a B (there might have been one or two C’s, I don’t remember). In a case like this, curving the grades would have the the height of unfairness.