The Blog of Sam Zhang

When outsiders look at an university as prestigious at MIT, many might assume that everything about the education there is perfectly thought out, with every class (especially the large classes that many people need or want to take) taught by the best professors, in terms of both research and teaching ability, and every curricular choice carefully designed for the need of the student in both the present and the future.

The truth is that this is entirely not the case (nor is something like this feasible anyway, even at MIT). While I enjoyed my experience in many classes here, at the end of every semester, when it is time to fill out the course evaluation surveys, I find myself criticizing the teaching in several courses. Often, the most renowned researchers in their fields are nowhere nearly as capable of explaining these topics in the classroom. Both professors and graduate TAs have outside committments that limit the time they can spend on teaching; as a result, many classes are not well organized, or fail to offer meaningful feedback on students' work. I have experienced all of these things firsthand, and it is probably very difficult to make improvements on a large scale. Most course staff, as far as I can tell, are trying their best.

In addition to the question of what is taught in classes, there is also the question of which classes to take. In my free time, I tend to spend an unreasonable amount of time thinking about what I can take in future semesters, in such a way that maximizes (1) my enjoyment, (2) the usefulness of the content in the future, and (3) the number of graduation requirements satisfied. Obviously, it is (3) that has the least direct connection with my educational goals.

Certainly, graduation requirements are necessary. After all, a degree would entirely lose its value if students could obtain it by just taking the easiest classes they could find for four years. Each academic field has a body of knowledge that everyone working in it is expected to share, and that is reflected by the departmental requirements for each major. Since I am not familiar with the expectations of knowledge for many, if not most, majors, I will not say much about the structure of major curricula and the included degree requirements.

But, at MIT and many other universities, there is also a set of requirements that apply to everyone regardless of major. This complex system of General Institute Requirements (GIRs) includes 17 classes – around half of the courseload of a student taking four classes (48 units) per semester, assuming that they do not earn outside credit for some classes through AP credit (only available for the first semesters of calculus and physics) or advanced standing exams. Of course, many of these classes are, or at least can be, valuable and foundational educational experiences. Indeed, this type of class is, in most cases, the only type that is offered in high school and below. But sometimes, students end up taking classes that serve no purpose other than checking a box, a hurdle towards graduation.

My experience with the GIRs have been mixed. On the one hand, had it not been for the HASS and CI-H humanities requirements, I would not have taken Problems of Philosophy (24.00) in my first semester, a class that I enjoyed not only for the questions explored through in-class discussions, but also for helping me learn to write persuasively about complicated topics. (I'm currently planning a concentration in course 24.) On the other hand, 5.111, which was essentially the same as AP chemistry with a few harder problems and more things to memorize, did nothing to spark my interest in the subject. I knew going in that I won't ever take any class with chemistry as a prerequisite, and after the class, I remained certain that it should not be required of all students. In no way is any of the course staff responsible for this. Even if they taught the most engaging class possible, it would not have changed the fact that I, personally, could not benefit from the class, and that I would not have taken it if it was not a requirement. The same can be said about many other classes that are well-taught but that I chose not to take.

A few months ago, there was a call for white papers that propose changes to the GIR system, as well as a survey sent to students asking for their input. (For those of you on dormspam: see the email "Re: Do we need the GIRs?" sent on January 25, 2025. My response to the survey is in row 108 of the spreadsheet.) As a result, I decided to think a bit more about what the philosophy behind the GIRs should be, and what kind of curriculum should follow from that.

What purposes should the GIRs serve? One possible purpose is for the GIR to be used as a prerequisite for technical classes, when needed by students in essentially all majors. For example, I cannot think of any area of study in science and engineering that does not use single-variable calculus (18.01), and to a slightly lesser extent, this is true for multivariable calculus (18.02) as well. Another is to develop skills that are useful regardless of one's specialization. For example, CI classes (both humanities and in-major ones) are classes specifically designed to build communication skills.

Moreover, a requirement can also be worthwhile if it encourages students to explore areas that they would not have otherwise. The HASS requirement is a good example. Many students arrive at MIT having viewed themselves consistently as "the math and science kid" throughout middle and high school, finding their English classes the most boring parts of their day. As a result, without such a requirement, many would be inclined to take only technical classes at MIT. But are these students well-informed about what the humanities are like, given that their only experience of them comes from school, which only teaches a small subset of them (most high schools don't offer, say, anthropology or media studies), and often not very well? Probably not. Given that HASS classes tend to focus on ideas that are more or less universal to the human experience, unlike many technical subjects, they have the potential to appeal to all students. With a wide variety of departments and offerings, it is much more likely than not that for any given student, there is some, and most likely several, HASS departments whose classes they would enjoy. The excellent course evaluation ratings received by many HASS classes and instructors demonstrate this.

But when should we not require a class for students? In short: whenever students know going in that they neither need nor want to take the class, and the class does not help with changing that. I believe that this is exactly the case for the biology and chemistry part of the science GIRs. The large number of students who leave these GIRs to their senior year, and their low level of engagement, indicate that many students neither need these classes for their major nor enjoy them on their own terms. These students come in with little interest in the material and leave with little interest afterwards, forgetting everything they learned right after the final exam and wasting both student and instructor time. (I'm a bit more conflicted on the two semesters of physics, because it is foundational for a few more engineering majors, and my highly subjective personal opinion is that physics teaches more broadly applicable problem-solving skills than biology and chemistry. Moreover, I'm not aware of large numbers of students deferring the physics GIR to their senior year. Perhaps a compromise is to only require a single semester of physics.)

Another effect of the current GIR system is to make it more difficult to take in-major classes as a freshman, since students are typically expected to mostly use freshman year to complete GIRs while subject to a credit limit. I believe that this effect is detrimental because (1) the earlier students take in-major classes, the earlier will they realize if their chosen major is right for them, and in the event that students do choose to switch their major, they will be more likely to graduate on time, and (2) underclassmen are already at a disadvantage when seeking internship and UROP opportunities, and the lack of major-specific experiences in freshmen year only exacerbates the problem. In fact, a few years ago MIT actually recognized this problem and experimented with allowing taking some GIRs on a pass/no record basis after freshman year, which evolved into the current flexible P/NR policy. But doesn't encouraging GIRs to be deferred and taken pass/no record indicate that these classes are, for many students, not as important as classes in their major? And if these classes are not the foundational experiences they are supposed to be, why not remove them entirely? (Another change that could help is to eliminate the freshman spring credit limit, since students at that point already have some experience with the MIT workload. Although the credit limit helps discourage unhealthy academic competition, since there is no credit limit after freshman year, such competition would occur regardless of changes in the freshman credit limit.)

One might respond by saying that GIRs like biology and chemistry, even when they are not directly relevant to students' area of study, still introduce them to new ways of thinking that can be generalized. Another point might be that we must maintain rigorous academic standards. But by removing unneeded classes, students can reallocate time, and departments can reallocate requirements, towards advanced in-major electives, increasing the curriculum's rigor. Another point is that students taking the GIRs might realize that an area they might not know is one that they would enjoy. While that might have happened occasionally in the past, I believe that large lecture classes on introductory subjects are not an effective means of simulating student interest; furthermore, course evaluation data indicates that GIRs get lower ratings than more advanced classes. Finally, one might argue that all scientists and engineers, regardless of specialization, ought to have some experience, however limited, with these subjects. In that case, a potential compromise would be to accept AP Biology and Chemistry credit, possibly only for those students whose majors do not need these classes.

I would also make a point about the REST and laboratory requirements. It is unclear what function these requirements currently serve, when many, if not most, majors already include classes satisfying them as required classes in the department. For example, 18.03 is a REST required by many majors, and so is 6.100A/B, and taking these classes is enough to complete the REST requirement (with the "at most one REST in department" constraint not doing anything since these classes are in different departments). To me, it seems like the main effect that these requirements have is that students who take unusual academic tracks or are in certain majors would be required to take a few more classes that have no clear role in their academic program, or are much too basic for them (unless they try to petition another class, but there is no guarantee that the petition would be approved). I am currently in this situation with the REST requirement, since I did not take 18.03, 18.06, 6.100B, or 6.1200, which are classes in my area of study that typically satisfy it, opting instead for more advanced classes. My view is that major-specific requirements would better serve the purposes that these requirements originally had.

Ultimately, many of the GIRs seem like historical artifacts rather than a product of ongoing consideration of what MIT students truly need. I agree that it's time to start reconsidering them. Certainly something like the current HASS requirement should remain in some form, and a few foundational classes required for all students are probably appropriate (perhaps 6.100A is becoming such a class?). But we must keep in mind that what is "foundational" evolves over time.