The Task Force on the Undergraduate Academic Program (TFUAP) issued a call for whitepapers in December of 2024 requesting inputs regarding the undergraduate program (see ovc.mit.edu/tfuap). I am thankful for this invitation, and have sent a whitepaper, which I would be grateful to précis here. I am offering a heartfelt appeal to take extreme care with the maintenance and growth of what has, up until now, served as a robust yet flexible foundation for our entire curriculum: The General Institute Requirements (GIRs).
Training and education are complementary – but different – activities. Professor Woodie Flowers taught me this, and I’m stealing from him directly as I write. Learning to spell is training; learning to communicate effectively is education. Education leads to a habit of courage. Training alone does not. We have stumbled at critical times in our history when we have wavered in our focus on education combined with training. We made this mistake, for example, in the 1930s when we over-taught the use of manual tables and handbook calculations to satisfy industrial demands, instead of instilling the educational tools necessary to invent the solutions that helped to stop U-boats, free children from behind barb wire, cure disease, and elevate the human experience. The General Institute Requirements (GIRs) serve as our red-carpet entryway to an MIT education. The educational experience afforded by the GIRs is, by far, a most democratizing and egalitarian experience. The GIRs greet the newest members of our community and offer them a compact: “Master the ability to think with these languages, and you are welcome in any major, laboratory, studio, or pursuit on our campus.” The GIRs have served this role since MIT’s founding. At our best, when educating the mind and the hand, we capture the heart. Hundreds of hours of committee meetings and listening tours will not automatically produce changes that celebrate and enhance our purpose: to educate young minds to stand with courage in front of problems that we do not yet even know, and design creative solutions that justify society’s faith in and support of our Institution.
I am worried by the idea that, in our efforts to review and revise the GIRs to meet the needs of an evolving future, the 10 learning goals identified by the TFUAP in their call for whitepapers should serve as the North Star.
Learning goals 1 and 3-10 are laudable and superficially inarguable. What objection, in the context of any curriculum revision, could one make to the desire for “lifelong learners,” “big dreams,” “time management skills,” “self-care,” “collaborative teaming,” and the ability to “take on leadership roles?” But the timing of the development of many of these skills is a debatable subject even amongst the faculty. For example, during the process in EECS that led to seismic changes in the department’s structure and curriculum as well as the formation of the College of Computing, the Course 6 CS Excellence Committee asserted that “Leadership positions are unattractive to CS faculty.” Students and faculty alike are not ready to be functional team members, to intelligently exercise choice, to plan wisely for the future, until they have individual skills and mental perspective to use and share. We must take care not to buckle and break the GIRs through a poor redesign in the futile attempt to design a curriculum that attempts, in a scant four years, to impart personal attributes that are developed over a lifetime. I propose that the learning goals 1 and 3-10 should rather be viewed as a general guidance to the actual specific work of crafting an educational program that prepares new minds for energized participation in a vibrant MIT community.
We start our students on the road to contributory membership in their professional communities when we lead by example. The current SME core classes teach the ability to communicate ideas in the powerful languages of physics, chemistry, mathematics, and biology that do in fact underlie or “infuse … disciplines across MIT.” The wise decision, for example, made by the faculty in the early 1990s to add biology to the SME core recognized the advancement of biological sciences to a point where a wide-spread, modern framework for understanding life on earth was extant. Like all science, this framework is not “finished,” but it is sufficiently defined to be recognized as a foundational way of thinking about our world. We devalue the compact we offer to our new community members when we lose sight of the relatively unique educational opportunities that our best teachers and curriculum have provided in the GIRs. As a freshman advisor and undergraduate instructor, I have seen firsthand how often our Advanced Standing Exams, extended Pass/Fail system, and loose scheduling that allows GIRs to be back-burnered until senior year, make it easy for both students and faculty to lose sight of how the GIRs define our community and our Institution. During this curriculum redesign, we must not make changes to service a superficial notion of progress, which ultimately and predictably devalue the practice and intent of the SME core.
Learning goal 2 presents what could be interpreted as an a priori demand from the Task Force to alter the Science, Math, and Engineering (SME) core of the GIRs to include “computational thinking.” I’m not sure I agree. Computing is in a very different state compared to other disciplines in the SME core.
The importance of computing as a tool and an economic engine is undeniable and exciting. I am very grateful to be part of MIT and EECS and ME during this socioeconomic revolution, and I am excited about the future. I teach embedded control, and I use computing extensively in my teaching and research, and I do not think that computing is a bubble or a fad. At the present time, computing is largely a tool, one that primarily necessitates training. The education required to invent and model and create solutions may or may not require computing context, depending on where the evolving definition of computing goes next. We may surprisingly quickly, for example, enter a world where “programming” as an enterprise becomes a vastly smaller undertaking for humans as computers further develop the ability to code for us. The nature of a computer, and what can be computed, both appear to be on the verge of potentially radical change in the coming decades as we develop new computing models and biological interfaces for computing. Will computing ultimately become a discipline, like biology, such that all well-educated MIT graduates should learn to see the world through its lens? Or will it be more like the infrastructures of electrical lines and plumbing pipes that undergird our society: most people use these systems to great advantage, and at MIT’s inception it may well have been believed that all well-trained graduates should know the intimate secrets of these exciting new developments. But ultimately, neither plumbing nor power systems became a universal language. Physics was, is, and will be the palette with which these systems are created, and physics is something every MIT undergraduate did, does, and will continue to need to understand. I strongly believe that the jury is out on which way computing is headed, and that it is not yet the time to radically re-center our undergraduate curriculum and GIRs around it.
In recent conversations I have had or overheard, common opinion seems to hold that there should be “no science GIRs beyond the department pre-requisites.” This is a superficially reasonable notion. What GIR would make sense if no department “needed” it? Unfortunately, this seemingly reasonable statement is profoundly irresponsible, as it can be interpreted in two very different ways, one nurturing for our community and one corrosive. For better or worse, our departments operate as business units that receive budget and other resources based on enrollment. Departments experience an overwhelming pressure to be “popular.” We would be poorly served as an Institution if we allow the pressures on departments to spill over into our considered curation of the GIRs. We need to view the GIRs not with a departmental lens, but an Institutional one, with recognition that the GIRs define us as a community. The departments, of course, should use a carefully crafted GIR program as prerequisites. The reverse should not be true.
The Institute should retain the GIRs and the SME core as the ongoing vital “roots” of the “tree” that is the undergraduate experience. I respectfully recommend the following:
- The passage of time has made it clear that the 1964 Zacharias Committee Report, which observed that “we believe that flexibility, choice, and early branching are desirable within the framework of the core,” was largely incorrect. The REST subjects, Institute Laboratory, and other changes implemented at the time both were and also continue to be formless. Rethinking the allocation of these valuable curricular slots might make for interesting opportunities for a new GIR subject in design, or problem solving, or computing in some form. Recapturing the REST subjects and limiting them to a tighter list, perhaps focused on computing, may be a functional option.
- The SME core should not be further penalized or altered from the 1964 decisions in order to make a slot for a computing GIR. We effectively have a computing GIR now as a REST subject taken by a large number of freshmen.
- The training aspect of computing may best be combined with domain-specific education in solving different kinds of problems. The algorithms and data relevant to a Political Scientist, for example, may be very different in structure and use from those used by a Materials Scientist. A colleague suggested an idea I find compelling: create a computing-intensive “CIC” requirement, analogous in some respects to a CIM or CIH requirement, with course offerings from many or all departments that could satisfy the requirement.
- Other than point 3, I strongly encourage that we avoid the sorts of fashionable curriculum changes that our past efforts have again and again found wanting: six-unit classes are frequently an inadequate exposure to any serious technical material that we expect someone to learn. I’m suspicious of P/NR after the first term, and also of “take X of Y” curricular plans. The “X of Y” plans have consistently proven to be an “every person for themselves” plan at all levels, for both the students and the faculty. These plans use competition to replace leadership, and they waste resources.
- The GIRs are not a sinecure. Our highest expectations for educational caliber should be brought to bear on the teams and units offering these experiences. The most important aspect of our core should be committed and passionate instruction for every minute of these classes. These teachers must bring a love for the material they teach and for the intellectual growth of the students.
- The SME core instructors have met this challenge for decades. They have held the line on quality, innovation, and dedication. They created some of my most vivid memories from when I was an undergraduate, and they have earned our thanks and respect. I would prefer that our “GIR re-examination process” begin with asking the current stakeholders for their recommendations. How could the outstanding service our community has received for decades be further improved in the estimation of the field personnel currently leading the charge? Should the Institute have a central physical laboratory for GIR students to conduct experiments in physics, chemistry, etc.? What staffing would this require? What, in the estimation of the core instructors, could and should be updated? Can introductory physics as a combination move to introduce thermodynamics or quantum mechanics in a non-trivial way that fits in the allotted units? Could and would introductory chemistry classes benefit from a physical laboratory component? Is there a practical way to offer 24 unit “fusion” classes that connect a GIR with a writing requirement? Does the Math Department see a foundational change in either the preparation of incoming freshmen or the mathematics introduction required to participate in modern science and engineering? Do these changes suggest useful revisions or enhancements to our 18.0x entryway that would be applicable for the vast majority of freshmen?
I am grateful to be here, excited to participate in the future, and I appreciate your time and patience.
