Edmund Wilson (1895-1972), A Piece of My Mind: Reflections at Sixty
(New York: Farrar, Straus and Cudahy, 1956), pp. 139-140:
In preparation for writing this paper, I asked a Latin professor of my acquaintance—probably one of the most brilliant in the English-speaking world—for a professional explanation of the methods of teaching Latin that were followed in my own schooldays and that are still in practice today. He made no attempt to defend them. "It's just like the Court of Chancery at the beginning of Bleak House," he said. "Nobody has paid any attention to it or tried to do anything about it for ages." Reassured, I attack the problem. I had feared that my personal experience might have been exceptionally unfortunate—in Greek it was the other way—but I shall assume that it was more or less typical. I apologize for speaking to teachers exclusively in terms of this personal experience. They may well feel that I do not grasp their problems. But I should like to put certain things up to them.
My first drill in Latin—and it was merely drill—was designed to get me into prep school. I remember of it nothing but paradigms, which I endlessly wrote out in the evenings and which seemed to me very much the same kind of thing as algebraic equations; and a difficult progress through Caesar, who impressed me, as he did John Jay Chapman, as "not an author but a stone-crushing machine." It would be possible, no doubt, to make Caesar interesting even to schoolboys in their early teens, but I do not believe it is often done, and in any case the teacher starts in with a discouraging handicap. At prep school, I went on to Cicero, who was a good deal more complicated and, at my age, hardly more attractive. The worst feature of these two writers—and also of Virgil, whom I shall come to later—is that neither of them seems to a schoolboy to represent anything imaginable as actual human speech. Caesar appears impersonal to the point of not being human; Cicero, despite his invective, infinitely artificial. It was only in my freshman year at college that, arriving at Plautus and Terence, I was able to see that Latin had once been a spoken language—with colloquial contractions like other languages—in which people transacted business, gossiped, made love and quarreled. I was exhilarated at finding that a Latin author could be read rather than solved like a quadratic equation...