This is the third in a series of opinion pieces about ways to begin reimagining education in the 21st and 22nd centuries. The first was on the need for incorporating financial education into the college’s graduation requirements. The second advocates the formation of a War Studies Department.
The purpose of this exercise is to jog the minds of those reading and to begin launching some questions for further study and review. These essays are the general reflections of an alum of this college and this newspaper. They, however, remain my views and mine alone. In some cases, the views offered in this series likely differ from policy I would enact if I myself were writing the rules. That seems natural to me. The desire, here, is for spirited inquiry, discussion, and debate.
The purpose of this article is to correct what I view as a deficiency of description. I contend that when we talk about academics in the modern undergraduate college, we are really talking about three distinct things, each of which should be thought of and categorized as such. Disambiguating these terms will lend insight into the operations of the learning process and consequently might have normative implications for the structuring of our curriculum.
Described simply, I contend that the way we teach often entangles three analytically distinct elements in the world of knowledge: methods, topics, and disciplinary histories. Methods are ways to analyze information. Topics are subject matters which can be studied through various methods in various disciplines. Disciplinary histories are the histories of how inquiry and knowledge have come to be pursued and produced in given disciplines.
To lump these distinct elements together without teasing out their intricacies is to impoverish one’s view of what it means to learn. Though each element is related to the other two, they run along different axes. Each element or axis contains a broad-though-discrete set of items, and for the most part, each item in a set can combine coherently with every item in each of the other sets. Perhaps examples will clarify.
If methods are ways to inquire, then we can think of the following as examples of methods: qualitative reasoning through the written word; quantitative reasoning through graphs and numerical argumentation; formalized, proof-based logical reasoning; idiographic analyses like “close reading”; three-dimensional puzzle solving; game-based inquiry; and supervised and unsupervised machine learning techniques, to name a few.
At first glance it may seem as though each of the above methods has a fairly confined use for the purpose of academic knowledge. Qualitative reasoning is thought of as existing in the realm of the humanities and law. Graphs and numerics are thought of as being for economics, demography, and business. Formalized, proof-based, logical reasoning is a method usually reserved for pure math, computer science, and philosophy. Idiographic reasoning is thought of as fairly unique to English literature, art history, and philosophy. Three-dimensional puzzles are methods typically used for architecture and engineering. Games are used in math and economics in the form of game theory. Supervised and unsupervised machine learning methods are thought of as methods used in computer science.
On second glance, however, it is clear that each method bursts well beyond the field it is most typically associated with.
Qualitative reasoning, of course, is crucial to higher-level thought in most fields. It is not sufficient in the sciences to simply have graphs, or empirical evidence: descriptive theory as to what is going on — and why — is also necessary.
Graphs and numerics have moved beyond the sciences and are now applied seriously to study literature or to create abstract paintings.
Formalized proof-based reasoning is crucial to contemporary debates in applied constitutional law.
Idiographic approaches (that is, those extracting the general from the specific) are most commonly thought of as being used in literary “close reading” or in the study of art. Yet the hard sciences have recently shifted toward this approach as we’ve come to understand more about the uniqueness of the human genome and the need to extract general treatments or methods from localized problems in the human body. Idiographic approaches are also central to the common law case approach used in legal teaching and the case study method employed in business schools.
Three-dimensional puzzle solving techniques have been crucial to cracking protein folding problems in biology.
Games are used routinely in political science and international relations in the form of scenario planning, war games, and campaign strategies.
Supervised and unsupervised machine learning methods are now regularly used to explore questions in fields as diverse as English literature, linguistics, and biology.
The point here isn’t that there aren’t real differences between the fields, or that one field is “better” than another by virtue of having more applicable methods. The point is that each of these methods is a distinct way of inquiring about the world. More importantly, each method can, and ought to be, taught as just that: a method.
The alternative — what we do now — is to implicitly lump in methods as by-products on the way to a field, often entangling the work of the method with the essence of the field itself. This is a mistake, for three reasons.
The first is that it is easier to learn a method when it is clear that you are learning a method. When a method is taught as its own subject, those teaching it will presumably be experts in the method, enabling you to get a clearer sense of the mechanics of the method, as well as how to structure or approach its various applications. This has large implications for the efficiency of teaching and learning, the organization of the curriculum, and the credentialing of students as trained in a particular set of broad-based, cross-applicable skills.
The second reason is that those “second-glance,” surprising applications I listed above likely do not exhaust the ways that methods and fields can be combined and recombined. Training students in isolated, particular methods might cause them to have very real breakthroughs as to the ways a particular method used in one field might answer questions in another. To use an example from my own life, various research ideas about how to apply machine learning to literary analysis only came to mind after I studied machine learning carefully. It was the training in the method that led to novel insights.
It might not be that it is necessary to train students in methods strictly in isolation for breakthroughs to occur (though, naturally, isolating them could yield efficiencies for the sake of instruction); rather, teaching them various methods would expose them to techniques of analysis which are not yet commonly used in their own discipline, for reasons of disciplinary internalism described more fully below. If one is never exposed, they will not think to try.
The third reason, closely related to the first two, is the implications this has for what we call “interdisciplinary” or “multidisciplinary learning.” Those terms themselves have always been loosely defined. They are, as they stand, intentionally meant to categorize loose or not-yet defined combinations of disciplines, but that does not mean the mechanics by which we expect this to happen need to also be loosely defined.
To do interdisciplinary or multidisciplinary learning can mean many things, but one of the things it definitely means is to take methods developed primarily in one set of fields and apply them to another. If we are serious about wanting to expand options for interdisciplinarity and cross-talk, then buying into the taxonomy proposed by this essay is one way to proceed.
If methods are what I described above, then what else is involved in the process of knowledge production and transmission? Necessarily, methods are applied to the study of topics.
It is worth being specific about what I mean by topic. For the purposes of the disambiguation and disentanglement that this article calls for, when I say “topic” here, I am referring loosely to a subject of inquiry which people are curious about. If methods are verbs, then topics are nouns. Again, illustration by example is useful.
An easy topic to think of is that of food. Food is something people are curious about which immensely affects human life. It is studied through various disciplines, and by various methods. The study of food occupies classes in biology and ecology, business, economics, literature, chemical engineering, history, geopolitics, farming, and likely many more. Food can usefully be thought of through all of these lenses, and it is indeed only when you combine the study of food through all of the disciplines that you get any true sense of what food truly is.
An exercise that I will leave to the reader is to consider the various methods I’ve outlined above and think about how each of them might be usefully applied to consider the topic of food.
A further example of a topic, one relevant to Swarthmore, is the studies in “color” organized by the faculty a few years ago under the catchy headline of “Chromatic Cabinet.” Pulled from their website, their description is as follows: “Chromatic Cabinet is a group of faculty and staff from across the college who meet regularly to discuss color. The interests are broad and touch on the physics, perception, linguistics, history, sociology, literature, and philosophy of color.” Perhaps this in a nutshell is a good look at one iteration of what it is like to conceptualize studies in topics.
Importantly, sometimes topics are also usefully thought of in terms of the challenges they represent. “Humanitarianism” is a topic which poses a problem that must be tackled through many disciplines and studied through various methods. To think usefully about solving humanitarian crises, one must gather biologists and ecologists, experts in energy systems and engineering, moral philosophers and historians, political scientists and language experts, communicators, actors, filmmakers, artists, economists, programmers, and system designers.
The implications of disambiguating topic studies are clear for curriculum design: if topics are indeed their own element of knowledge, running along an axis that intersects various disciplines and various methods, then teaching various topics as topics enables us to gather students from various disciplines to share knowledge with one another on an even playing field. This in turn leads to the pursuit of a particular goal (in the case of tackling a humanitarian crisis) or the acquisition of philosophical knowledge (in the case of “color studies”) which would be unachievable if each were to simply stay in their own lane.
If methods and topics are as described above, what else is left? What, for example, is the use of “disciplines” as we currently have them? Why even have majors?
Analytically, a discipline must be a set of topics of study, pursued historically along similar lines, using some subset of academic methods, to answer general broad questions of inquiry. In other words, disciplines are like topics, but high-level topics, broad ones for which there are myriad sub-topics. If this is the case, how do they fit on their own axis of thought, or qualify as a unique element in the course of knowledge production?
Simply put, in 2022, disciplines exist as a force in themselves, regardless of what one might conceptualize their abstract role to be. There are some reasons for the force of disciplines that are indeed arguably inherent: sets of topics tend naturally to cluster together with various methods, leading to recombinations and inertia which ultimately begin to transform the topics and methods themselves.
For example, various topics, including many problems in logistics and also anything to do with the Internet, as well as various methods, including machine learning, were developed wholly out of computer science, a discipline which itself evolved out of a clustering of psychology, optics, physics, art, engineering, lambda calculus, discrete math, linguistics, literature, demography, and war studies, among other things. Disciplines do usefully coalesce organically and act as forces in the world of knowledge upon such coalescence.
But there are of course more-human reasons to consider disciplines as a distinct element in the knowledge sphere. It is naive to think that the type, organization, and quality of student education in the university is immune from the departmental jockeying, funding pressures, and charismatic personalities that are characteristic of faculty life. These characteristics play an important role in forming and perpetuating disciplines at the undergraduate and graduate levels and then back again. Indeed, the oral cultures of seminars go on to influence subsequent generations of teaching and syllabi, as well as how students, scholars, and practitioners conceptualize their fields.
The presence and influence of disciplines exist even when the topics at hand are both exceptionally interdisciplinary and imminentally practical. For example, as Shyamkrishna Balganesh & Taisu Zhang describe in their recent law review article, “Legal Internalism In Modern Histories of Copyright,” in spite of the influence and relevance of various other fields, the history of copyright remains a legal history. This is interesting because analytically, there is no necessary reason why the rules governing the use of art and computer code, as copyright does, ought to be more influenced by the field it originated from than the fields it governs or other fields which influence it. Yet the nature of the way participants in the law are organized among one another makes it so, argue Balganesh and Zhang.
The implication here for knowledge organizing is serious. If the force of an individual discipline holds true for intellectual arenas with high interdisciplinary overlap and also with stakes that affect daily material transactions (the consequential nature of which one might imagine resulting in an override of academic theory), then it is clear that disciplines indeed have weight in and of themselves. Regardless of what one thinks of the referenced analysis of copyright, the authors’ verbiage is useful for the concept I am formulating here. There exists what we could usefully call “disciplinary internalism.”
Rather than lament the presence of such internalism as a barrier to interdisciplinarity or as an impediment to the learning process in general, I believe that this internalism is crucial to the effective continuation of the modern undergraduate institution.
It seems clear that having distinct disciplines yields numerous benefits. It enables students to learn a common language and code of speech which then permits them to efficiently communicate with scholars all over the world as well as with practitioners in their fields. It highlights a space for open, unsolved problems around which people are communing. It has efficiencies for the sake of allocating resources and structuring and scheduling classes. Such efficiencies are not trivial but rather central to the whole endeavor of providing education for students. Perhaps most importantly, having disciplines exposes students to the study of the past.
Therefore, for the sake of clarity, and also to emphasize the essential role that continual cultures play in the process of knowledge production, this third element of knowledge (the third thing we talk about when we talk about teaching) is not just the idea of a discipline as it presently stands but the disciplinary history as it stands and has stood.
The distinction is worth noting particularly because many courses and majors are actually classes in disciplinary history. These courses tend to cover a broad range of topics and use a set of methods that are sometimes but not always explained. Many introductory courses are structured this way. Being surveys of the field, they expose students to what goes on in the major and what has been tried in the past.
Knowing their history enables students to identify forks in the road of knowledge that turn out to be dead ends, or better yet, unexplored. It helps save them time in pursuing their various curiosities as well as in identifying theories or approaches which turned out to be false.
If methods can be thought of as particular tools to use, and topics as particular objects to tackle, then disciplinary histories are those fuzzy-yet-nuanced experiences which shape an individual’s reference set, forming the frameworks through which they view novel problems or experiences.
If we think about methods, topics, and disciplinary histories as three distinct things, some suggestive implications open up. The first is to think about training students explicitly in methods, the reasons for which are outlined in part above. The second is to place explicit consideration on methodological training when reviewing funding considerations. This may be politically unpopular, but so are many things which are both sensible and fair.
For example, I contend that one of the reasons for the overflow of students into the computer science major, aside from the obvious desirability of computer scientists in the job market, is that computer science courses give students training in quantitative, programmatic, data visualization, and machine learning methods of analysis, all of which are exceptionally useful in other disciplines. Students can then leverage their training in these methods, presently taught by the computer science faculty, to be better chemists, physicists, quantitative sociologists, quantitative political scientists, digital humanists, linguists, and biologists. In fact, I recently learned from a professional contemporary dancer (check her out, her stuff is quite cool) that experimental computational methods are even influencing the world of choreography.
These facts spur two competing lines of argumentation: The first is that other disciplines leveraging computational methods ought to internalize their own costs more by training such methods themselves, which, indeed, some of them do. The counterpose is the contention that students are actually better served by getting rigorous training in various computational methods by people who are full experts in the nuances of such methods.
The solution is to extract the notion of computational methods out of the department of computer science altogether and fund it as a separate meta-department with a distinct purpose. The computer science department could then focus more fully on all the pressing problems that fall squarely in the realm of computer science, and the department of computational methods could work with other departments to craft an accessible course of study for those who need training in various forms of analysis.
An even better solution would be to recognize that this method-subject divide, as I show above, is not unique to contemporary computational methods, but in fact inheres to the whole problem of teaching and learning altogether.
A strategy for the long run would be to gradually adapt the department of computational methods to include other methods as described above, thereby offloading the training of generalized methods from their usage all together. Perhaps some of these methods could be taught in workshop form instead of full classes for the sake of flexibility and concision. For the cases in which methods are unique to disciplines, or are too niche to justify outsourcing, disciplines and courses should of course still retain their training.
Another implication which juts out of the taxonomy I propose is the idea of organizing more working groups around topics or problems. One could even imagine extending these working groups in time and establishing cohorts of students from different disciplines to develop multidisciplinary approaches for tackling specific topic-based problems over the course of a full year or two. Doing this would confer upon students general knowledge of what goes on in other fields, accomplishing what the division requirements intend, but with more efficacy and without forcing unwanted (and often dislocated) specialization onto students. It would also simulate collaboration skills necessary for real-world endeavors.
Of course, the real world has always seemed to be anathema to Swarthmore. But, then again, that is precisely the problem.
Perhaps the last thing that comes to mind is to encourage departments to more explicitly describe some of their courses to students as histories of the discipline and therefore not entirely representative of what subsequent work in the field could or might entail. Conversely, in disciplines that tend to proceed not by history but by doing labs or retaining facts (think of some math courses here as an example), incorporating more disciplinary history might lend students genuine insights into the thought processes of some of the field’s greatest minds and perhaps even show them that even the greats were wrong about many things.
Placing disciplinary history front and center might also help students see how fields are constructed and provide them with a more high-level intuition about areas which can otherwise seem fairly technical (the fields of economics and computer science come to mind here). This in turn can engender excitement about topics, which will give students more incentive and horsepower to grit through areas of the fields they find challenging.
The added bonus here is opening up disciplines to many students who come from backgrounds or high schools that didn’t provide them with very good primers for the course material they now face at Swarthmore. It would show that everything and everyone starts from somewhere.
The technique thus far employed by my series of articles about reimagining education to suit the next 100 years has been to identify specific areas in the curriculum in need of targeted reform. Running through each of my previous two articles was a justification for each reform on two grounds. On one level, I advocate for those changes because of present circumstances which highlight a utility for each change. On the second, I attempt to show how each change fills a distinct gap within the college’s larger mission of transmitting and advancing types of knowledge.
This present article proceeds by a different tack. Rather than begin by advocating for any particular normative change — that is, any specific sense of the steps Swarthmore should take to ameliorate clear deficiencies — this article aims to clarify what I view as a deficiency of description in the hopes that clarified terms will open a larger space of discussion. The implications outlined above are a tentative attempt to describe some proposals which jut out of this taxonomy, but I imagine there are others, and that various readers would have bright ideas not outlined above.
Our curriculum was designed a long time ago with varying aims in mind. Some elements of it were experimental. Others, like the way the honors system now works, were compromises to transform older forms of teaching into something more palatable to various reformers. Still others, like the writing requirement (and, in my opinion, the division requirements), are obvious failures of curricular design, even if such failures were in no way foreseeable to those establishing them.
Whatever one’s opinion of those past choices, however, it is worth taking the lessons from the past seriously and using them to think about a future not strictly or necessarily confined by that past. My views in this series are one stab at it, but I too am a dead hand as far as Swarthmore goes, and thus of only limited use. You, the reader, will have to be the guide.
After all, a school, to paraphrase Jefferson, belongs always to the present generation.
This article owes a strong debt to the Aydelotte Foundation at Swarthmore College, to the course taught in its name, and to Tim Burke and Rachel Buurma, who co-chair the foundation and teach that course. Many of the ideas in this piece formed from my participation in the Spring 2018 iteration of the Aydelotte Seminar, though the opinions and quips expressed above do not necessarily reflect the beliefs of the Foundation or its leaders and affiliates. All errors, of course, remain my own.
An additional, historical debt is owed to Frank Aydelotte of Swarthmore and the Institute for Advanced Study. An old Swarthmore motto from a secret society long since gone was the Latin phrase Acta Non Verba, “Actions, not words.” If the methods of Frank Aydelotte can be neatly pinned down, they might be embodied by a similar, though modified, phrase: “Actions through words.”