This is especially true in a domain that has no practical purpose. In computer science, the "exam" could be to go solve some hard problem, or write a program with a particular effect. In engineering, you can have students build a bridge and then stress test it. In athletics, the exam is competition. In business, it's making money. In dating, it's finding a mate (or whatever your personal goals are in that space). In farming, it's making food come out of the ground. For writers, it's a compelling or profitable story or book. For painters, it's a painting. Woodworkers can build a chair.

All those domains are testable. But what is the practical work product of a deep knowledge of medieval history? It's extremely difficult to test the past, so it's certainly not predictions about what already happened. Nobody needs knowledge of medieval history for any practical purpose in the modern day. There is no possible way to test for this knowledge in a practical scenario because there is no practical outlet for the knowledge. Literary criticism is the same, along with much that is called "liberal arts" today. So for subjects like this, exams and essays are the only possible work product, and it's true that at that point you're going to have to use contrived tests, since real tests don't exist.

Athletics and engineering have exactly same problem as medieval history. For engineering, just because it is practical does not mean you cant memorize it. If there is defined set of problems, you can memorize solutions. The same programmer can be able to solve one difficult problem and fail with other difficult problem.

The bridge test can be passed by memorizing one bridge.

The history course from American university required students to write few paragraphs as answers to around 5 questions that covered topics from lectures. Which is pretty close to what actual historians do.

Practical usage of medieval history: know when people are bullshiting for ideological reasons when they blable about how people in history behaved, what values they had and origins of certain traditions. Pretty much all "in the past people would" is a fantasy.

>But what is the practical work product of a deep knowledge of medieval history? It's extremely difficult to test the past, so it's certainly not predictions about what already happened.

This strikes me as a really sad way to see the world.

Would you then agree that say, archaeology or palaeontology are things that nobody needs knowledge of for any practical purpose in the modern day? If you disagree, why?

I didn't say we shouldn't study those things. I just said that there is not a much of a practical work product, and especially not one that a student could produce.

But, I also think that probably more people are studying them than need to be. If that is a student's interest, then well enough. Otherwise, I'm not sure that, e.g., a middle school medieval history class is a good use of a student's time.

There is a practical work product: research, formation of theories, etc. Edit: Possible new ways to understand the modern world. Just like archaeology or palaeontology.

In my opinion, the issue in liberal arts education is with grade inflation, along with the scalability of assessing learning discussed in the essay/these comments.

I suspect that really rigorous liberal arts education is much harder to do at scale than really rigorous science/engineering education because to be blunt, there is less room for argument in scientific concepts.

For example, properly grading student essays and research is more difficult at scale in the liberal arts and the level of the whole thing gets dumbed down as a result.