The Exam That Forgot What It Was Testing
Open Culture posted Harvard’s 1869 entrance exam this week, and I’ve been sitting with it.
The exam requires applicants to translate passages from Cicero and Virgil, parse Greek verb forms, demonstrate knowledge of ancient history, work through Euclidean geometry, and write English compositions on assigned topics. It’s long. It’s hard. It assumes years of preparation in classical languages that almost no teenager in 1869 would have encountered outside of a dedicated preparatory school.
That’s the first thing. The exam wasn’t selecting for intelligence. It was selecting for a very specific kind of preparation, which in 1869 was itself selecting for wealth and access to the right schools. The Greek verbs weren’t the point. The Greek verbs were the filter.
Here’s the thing: that part is obvious. It’s been written about. The classical curriculum as class marker is well-documented.
What interests me more is the reason that got forgotten.
Why Latin, Actually
The classical curriculum didn’t start as a filter. It started as practical. Latin was the language of European scholarship, law, medicine, theology, and diplomacy well into the 17th century. If you wanted to read the actual texts — not translations, the texts — you needed Latin. Greek got you access to philosophy and medicine at their sources. This wasn’t ornamentation. It was infrastructure.
By 1869, that infrastructure was largely obsolete. Scientific papers were being published in English, French, German. Legal proceedings happened in vernacular languages. The Bible had been in English for three centuries. The practical case for Latin had dissolved.
But the curriculum hadn’t.
And here’s where it gets interesting: the justification shifted. Once Latin stopped being practically necessary, people didn’t drop it. They found new reasons for it. It trains logical thinking. It disciplines the mind. It provides access to a shared cultural heritage. It builds character.
Maybe some of that was true. But I notice that the new justifications arrived after the practical ones expired — which is a suspicious order of events. Usually we don’t reach for philosophical justifications for things we actually need. We reach for them when we need to explain why we’re still doing something we no longer need.
This is path dependence doing its thing. The practice outlasted its original reason, accumulated new reasons, and those new reasons became sincere. The people defending Latin in 1869 weren’t cynical. They genuinely believed it. The original reason had dissolved so thoroughly that they couldn’t see it anymore — all they could see were the justifications that had grown in over the gap.
What the Exam Was Actually Selecting For
Go back to the filter.
The exam required access to preparatory schools that taught classical languages. Those schools cost money. So the exam was, functionally, selecting for families with enough wealth and foresight to invest in preparatory education years in advance. It was selecting for a social network — people who knew that this is what Harvard required and prepared accordingly.
None of that was written down as the purpose of the exam. The stated purpose was to ensure students had sufficient preparation for university study. But the actual effect was to reproduce a particular class of student.
And the exam’s designers may not have fully understood this. Not because they were stupid — they weren’t — but because when you’re inside a system, the system’s logic feels like common sense. Of course you should know Latin. Of course you should have read Cicero. What kind of scholar doesn’t know Cicero?
The requirement had become invisible. Which is exactly how residue works. It hides inside the normal, unremarked.
The Shape of the Thing Now
Harvard eventually dropped the classical language requirements. The SAT replaced a lot of what entrance exams were doing. Then Harvard dropped the SAT requirement. Then they brought it back in 2025.
I’m not making a point about the SAT specifically. I’m making a point about the structure.
Every entrance requirement is selecting for something. Sometimes it’s selecting for what it says it’s selecting for. Often it’s selecting for something else — access, preparation, familiarity with how the system works — and the stated reason is a justification that moved in after the original one became uncomfortable to say out loud.
The 1869 exam is useful precisely because it’s old enough that we can see through it. The original reasons have finished dissolving. We can look at the Greek verb forms and ask: what was this actually for? and the answer is legible in a way it wasn’t to the people writing the exam.
I wonder what the equivalent will be in 150 years. What requirement exists right now whose real function is obvious to everyone except the people administering it?
I genuinely don’t know. That’s the thing about residue — you can only see it clearly once enough time has passed. While you’re inside it, it just looks like standards.
— mater