I recently found a box of papers from high school and was shocked to see what I once knew. There, in my handwriting, was a multi-step geometric proof, a creditable essay on the United States’ involvement in the Philippine revolution, and other work that today is as incomprehensible to me as a Swedish newscast.

Chances are this is a common experience among adults like me who haven’t stepped foot in the classroom for ages—which might suggest there wasn’t much point in learning the stuff in the first place. But then again, maybe there is.

Research shows that people can often retain certain information long after they learned it in school. For example, in one 1998 study, 1,168 adults took an exam in developmental psychology, similar to the final exam they had taken for a college course between three and 16 years earlier. Yes, much had been forgotten, especially within the first three years of taking the course—but not everything. The study found that even after 16 years, participants had retained some knowledge from the college course, particularly facts (versus the application of mental skills). Psychologists in another psychology study, this one published in 1991, examined memory for high-school math content and had similar results.

These findings, among others, indicate that students forget less than they may think they do. And there’s value in what they remember. These conclusions carry important implications for the subject matter students study in school.

Naturally, knowledge sticks if it’s revisited. For example, one study of MIT students found that physics majors remembered material from a freshman course better than students who majored in subjects unrelated to physics. More striking, though, is that continued use can actually make knowledge indelible. In one rather remarkable study, researchers administered an algebra test among adults who had taken algebra anywhere from months to decades previously. Most of the adults struggled to remember how to do the equations, but those who’d studied math beyond calculus (subjects whose mastery requires an understanding of algebra) could still work basic algebra problems—even if they had not done so for decades. In other words, several years of practicing algebra in more advanced math courses made the former stick permanently.  

So why do adults remember some facts they learned in school but not others? For one, the context of a memory—where and when it’s learned—might be forgotten even if the content is recalled. That’s what happens when one recalls hearing a movie is good, but can’t remember who said so. Likewise, a student may remember a fact but not know she learned it at school. And if she hears the same fact many times, figuring out where she learned it first can be especially hard; who first told her that there are four quarters to a dollar? A parent? A teacher? Someone on Sesame Street?  

Other times a student remembers the context—he knows he studied French at school, for example—but falsely concludes that he’s forgotten everything. After all, it’s likely that some of the memory remains even if he recalls nothing. This invisible residue of old memories helps a person remember that same material again more quickly than before. Clever research studies on this phenomenon tested Mormon missionaries who learned a foreign language but didn’t use it again for decades; forgotten vocabulary was quickly relearned. Other research in more controlled laboratory situations showed comparable results.

Ultimately, this ability to retain some of that knowledge has practical benefits—and the reason for that has to do with the nature of intelligence.

Intelligence has two components. One is akin to mental horsepower—how many pieces of information a person can keep in mind simultaneously, and how efficiently that person can use it. Researchers measure this component with simple tasks like comparing the lengths of two lines as quickly as possible, or reciting a list of digits backwards. The other component of intelligence is like a database: It entails the facts someone knows and the skills he or she has acquired—skills like reading and calculating. That’s measured with tests of vocabulary and world knowledge.

Researchers have long known that going to school boosts IQ. The question is whether it makes people smarter by building mental horsepower, by adding to students’ database of knowledge and skills, or some of each component. Recent research published in Psychology and Aging shows that people who stay in school for a longer part of their lives are no faster at simple mental judgements (like line comparison) than their less-schooled counterparts. Other research published in Psychological Science shows that high-performing schools do little to boost kids’ mental horsepower. Instead, schooling makes students smarter largely by increasing what they know, both factual knowledge and specific mental skills like analyzing historical documents and learning procedures in mathematics.

This view of schooling carries two implications. If the benefit of schooling comes from the content learned, then it’s important to get a better understanding of what content will be most valuable to students later on in their lives. The answers may seem intuitive, but they’re also subjective and complex. A student may not use plane geometry, solid geometry, or trigonometry, but studying them may improve her ability to mentally visualize spatial relationships among objects, and that may prove useful for decades in a variety of tasks.

The aforementioned research also implies that the sequence of learning is as important as content. Revisiting subjects can protect against forgetting, and sustained study over several years can help make certain knowledge permanent. Thus, when thinking about what expect students to learn, it’s not enough that content be “covered.” Evidence suggests that a student must use such content in his or her thinking over several years in order to remember it for a lifetime.

Traditionally, some educators subscribed to the notion that it doesn’t much matter what students study, as long as it’s hard and they don’t like it. Through the early 20th century, educational theorists believed students should study Latin not because it was a useful language, but because studying it trained the brain to think logically. Although this theory was discarded in the 1920s, the 21st-century version arguably holds that technology has rendered memory unnecessary; what matters is learning to think. Both depict the mind as an all-purpose muscle that gains strength irrespective of the tasks used to exercise it. And both theories, it seems, are flawed.

Education-policy debates tend to focus on structural issues—things like teacher quality, licensure requirements, and laws governing charter schools. But research on human memory indicates that academic content and the way it is sequenced—i.e., curriculum—are vital determinants of educational outcomes, and they’re aspects that receive insufficient attention. In other words, perhaps what matters most after all isn’t mental exercise.