Diving Deeper Into What 6 Million Syllabi Say About Higher Education
Which book is assigned most often in colleges around the world? That's something that hasn't really been knowable, since no one's been able to collect every syllabus out there and analyze the reading lists on them to see what pops as the top ranked. But this month, a new tool came out that starts to make it possible to answer questions like this.
It's called the Open Syllabus Project and it's run out of a public policy center at Columbia University. Its leaders have gathered more than six million syllabi from around the world and made it possible to do searches across them.
We wrote about the project the other day, and for this week’s EdSurge On Air podcast, we dive deeper—and talk to one of the one of the professors who emerged as a most-taught author.
We’ve provided the transcript below for accessibility, though it’s a better experience in audio.
David McClure: Basically, we just go out on the public web, get these documents, and then each kind of represents one professor's little view onto how to teach something, what matters, what's important, what's the sequence between the different readings, how they relate to each other. In the past, we encountered these documents in a one-off basis as students, but now we have this huge sea of about six million of them and it just makes it possible to start to analytically try to understand the whole teaching and learning system all at once.
EdSurge: That's David McClure, a grad student at the MIT Media Lab who's helping to build the Open Syllabus Project. McClure also created an interactive visualization called a galaxy view that shows almost every text in the collection as dots on an interconnected map of knowledge. The larger the dot on the map, the more often it's assigned in a course.
McClure: Basically, when two books are close together, it means they tend to show up in the same types of courses in the data set and yeah, I think a lot of fidelity is lost in that process and it's almost more of an intuition pump than a piece of evidence. But it does give us a broad sense of what books are similar and dissimilar and how the different big blocks of the academy interact with each other.
Basically, it gives a rough approximation of which individual books are similar and which sort of fields and subfields are more or less similar. For me, I think the really cool think is I think it suggest these structural differences in the way the disciplines are put together.
For example, there's an environmental studies cluster on the graph which is really tight and cohesive even though the individual documents come from lots of different fields. It's as if environmental studies has this really coherent reading program but it's drawn from lots of different disciplines whereas something like history or political science are much more scattered and broad in the layout. It's like they have a number of different distinct subfields more spread out across the sea of knowledge, right?
This really is a new kind of window into higher ed and one that its designers hope can be used as a kind of ranking of teaching materials. Each book in the system is assigned a core of a one to 100 that measures its influence in classrooms. The director of the Open Syllabus Project, Joe Karaganis, says he hopes it becomes a tool that professors use to prove their value to their bosses and help them get promotion or tenure.
Joe Karaganis: In a sense, you value what you can measure and we hope that one of the outcomes of this new metric will be a way of incentivizing the production of more and better teaching-facing materials.
EdSurge: But this plan may be better in theory than in practice because in my search through the system, most of the popular authors on syllabi are long dead or [inaudible 00:03:49] retired. It seems like it just takes a long time for a book to work its way onto a large number of syllabi.
Oh, right, so I never said what the most popular text in higher ed is. Turns out, it's the Elements of Style by Strunk and White. As a writer, I love this. I had to read that in college, and I still use it as a reference. Number two is another writing book, A Writer's Reference by the late Diana Hacker. Number three, a calculus textbook, its name is Calculus, by James Stewart, a Canadian mathematician who passed away in 2014 from what I can tell on Wikipedia.
Ranked fourth is Human Anatomy and Physiology, sounds like a good pre-med text, by Elaine Marieb and published back in 1989. And number five on the list is Plato's Republic. Couldn't get much further back then that. These are not exactly unknown authors waiting to be discovered—or hunting for acclaim.
Actually, some professors worry the collection could actually have harmful side effects. One potential problem is that the system could lead to less diversity in classroom assignments as people just look to the data set for guidance. John Becker, a professor at Virginia Commonwealth University, told me that maybe some marginalized texts would get even more lost because a department decides not to adopt it because it didn't get a high enough score.
What does Joe Karaganis say to that concern?
Karaganis: We've heard that from a number of people who've encountered the project for the first time and I guess I have two thoughts on that. One is that the main force for standardization of classes is really textbooks, so that is not something we'll have any impact on one way or the other.
With respect to teachers or faculty who are coming to look for materials for their classes in which they're composing a sort of complex syllabus with many texts, the tool I think is really what you bring to it. If you have a kind of lazy approach that you just want the quick answer to what the most frequently co-assigned texts are, our data can certainly show you that.
If you're interested in digging down into some of the lower reaches of co-assignments and looking for some of the more interesting and maybe less expected connections, then we can support that too. I'm somewhat agnostic on that question. I think it will be useful to creative faculty and lazy faculty alike.
EdSurge: So you’re saying you don't feel like the tool will create that either way, that people will just kind of do what they're doing anyway?
Karaganis: I think there is a powerful pressure across some parts of the university system to standardize curricula, but I think that takes the form of textbook teaching, not standardization around the kinds of custom or bespoke classes that I think that faculty member was referring to. Many faculty in especially the humanities and social sciences, treat it as sort of the norm.
EdSurge: A bigger downside of the system though could be in the political realm. When the Open Syllabus Project released an earlier version of this database a few years ago, it had about a million syllabi. Right wing commentators were quick to point out that the third most taught title was Communist Manifesto by Karl Marx, now ranked sixth. But they flagged it as evidence of extreme liberal bias in the academy.
Karaganis: Well, I learned last time around that there was no countering these arguments, that the data didn't matter at the end of the day. It was people looking for a way to plug our neutral tool into their argument about the left-wing leanings of the academy. We have much richer data now and if there were ever a need to debunk arguments about the role that Marx plays in the canon again, we could do so.
The Communist Manifesto is a very widely assigned text and as I've been arguing for a number of years now and can now show, it's widely assigned because it's widely assigned across many fields. That's a very unusual characteristic for a text. In most fields, most texts are assigned in one field or two maybe.
EdSurge: It sounds like your argument is it's not being taught as a go do this philosophy, it's being taught in other contexts.
Karaganis: Oh yeah, well, that's the first position to take in this debate is that people are not assigning works as endorsements of those works but often to criticize them as well.
EdSurge: I asked McClure about this idea as well. I guess I wonder if there's any potential downside from having this kind of collection of data.
McClure: Yeah, I think in the wrong hands, absolutely so this is why we're super careful to never actually publish the documents and you're totally right. There's a certain culture around the political right, a trend to kind of out faculty on the left that they think are teaching things that they don't like politically which for me is super troubling and that's why we don't actually publish the documents.
We don't put them out there in a way that you could go and search for some keyword and find the class that might be associated with one scholar. That I think could be problematic and we don't do that for exactly that reason. But we're kind of curious about these emergent overall connections that come out of huge collections of the documents.
But yeah, I think we really are careful to never make it possible to filter down too closely in the data where you could attach individual scholars or faculty members to different books and stuff like that.
EdSurge: The system lets you see top ranked books by campus and it's interesting to poke around and look at what's popular where. The most assigned text where I went to undergrad at Princeton University is Leviathan by Thomas Hobbes. It's kind of highfalutin. At Harvard, the most assigned text is Bowling Alone by Robert Putnam, a more policy oriented book.
At Berkeley, number one is The Communist Manifesto and second place is Intermediate Accounting by J. David Spiceland. At one of the largest community colleges, Northern Virginia Community College, the top assigned book according to the database is All Is Quiet on the Western Front. At Liberty University, a religious university, the top book assigned is Manual For Writers of Research Papers, Theses, and Dissertations by Kate Turabian followed in number two by God, Marriage, and Family: Rebuilding the Biblical Foundation by David Jones.
You know what, though? It turns out the first thing most professors do when they get to this collection is search for their own book and see how popular it is.
Karaganis: Oh, I think that's the first step for any academic coming to this thing. It'll be the vanity search and curiosity searches of friends, people looking to confirm their intuitions about what's frequently assigned.
EdSurge: One esteemed author who did a vanity search like that in the collection to see how he ranked was Greg Manikew, who is the author of the most-assigned economics textbook on the planet. Actually he has the top three econ textbooks. And on his blog this week he linked to our article about the syllabus collection and wrote: “What authors appear on the greatest number of syllabuses for college courses? Shakespeare is number 1, and Plato is number 2. I show up at number 22.” he writes of his place in the syllabus universe. Pretty impressive.
I was curious to hear from one of these scholars who came up at the top of the rankings, so I reached out to the author of the most-assigned history textbook out there, James Roark, who is an emeritus professor of history at Emory University. And I was able to reach him by cell phone. How does it feel to be that highly ranked?
James Roark: It feels good. You know textbook writing is a very different project from monographs and the kind of work that historians do most of the time. You’re not trying to impress your colleagues. You’re not writing for a small historical audience. You’re trying to help 18-year-olds who quite likely are in a course they wish they weren’t in learn American History and to do well and perhaps get a little excited by it.
EdSurge: It's important also somewhere in here to say some obvious caveats about this collection. All of these syllabi are in English, so obviously there's much of the world left out and more of higher ed out there that's not included in this selection. Also, some syllabi aren't on the public web which is what this is looking to because the professors haven't shared them, so there are gaps and blind spots in this data.
What has surprised the developers as they put this collection together?
McClure: I guess maybe [it’s] a little bit dispiriting is how dominant the traditional white male canon is in the list. For all the questions about whether the classics are still taught, they seem to be in this data set which was, again, maybe not surprising but you maybe want to see a little less of that overall.
In terms of the galaxy graph, I think I was struck by really how distinct the sciences are. The sciences in that way out form a very discreet separated off unit. I was curious and interested in the interstitial zone between the sciences and the social sciences, humanities and trying to see what fell in that middle zone there.
I think that's one of my interests here is what are the bridging texts or texts that form connections between fields that are generally far apart.
EdSurge: Have you looked yourself up?
McClure: I have. I'm definitely not important enough to show up there. I looked at all my old professors—it's always great to see where things land. My dad is a history professor, so I always look him up when I drop in the data.