Following up on a point you raised, I would agree that the results of grading on a curve may result in a non-normal distribution. This being said, there is difficulty in answering the question: what should a grade distribution look like? Would this seem somewhat arbitrary