Tuesday, July 15, 2014

More on Disruption and Liberal Education

Talcot Brewer, a philosophy professor at the University of Virginia, has published a long essay entitled "The Coup that Failed: How the Near-Sacking of a University President Exposed the Fault Lines in Higher Education." The president in question is Teresa Sullivan, who was removed and quickly re-installed by the governors of the University of Virginia, but the essay also reports on turmoil at a liberal arts college, Mary Baldwin College. But these are only case studies for an analysis of the bigger questions.

[Board members] Dragas and Kington explained their actions as being the result of “philosophical” differences with Sullivan, yet they declined to say which of the great questions of existence had divided them so irrevocably as to require her dismissal. Enterprising reporters from the student newspaper made use of the Freedom of Information Act to shed some light on this mysterious philosophical disagreement. It turned out that Dragas and Kington had come to believe that the rise of online learning would soon pose an existential threat to the university, and that it had to embrace the trend quickly or risk being left hopelessly behind. Sullivan had been reluctant to move in this direction with the boldness they thought necessary. She was threatened with imminent dismissal, and agreed under duress to step down. 
At first blush, this does not sound like a philosophical disagreement. It sounds like an ordinary empirical disagreement about whether, and under what conditions, the university would be able to attract enough qualified undergraduates to sustain itself. One party to the conflict, President Sullivan, was less impressed than her adversaries by Harvard Business School professor Clayton Christensen’s widely discussed prediction that online instruction would prove to be a “disruptive innovation,” one that would pose a threat to the very existence of traditional suppliers—not by providing a better product but by providing an inferior substitute that was either vastly cheaper or more convenient. Christensen had insisted that traditional providers of higher education could survive this disruption only by “changing their DNA”—that is, fundamentally changing their mode of instruction, partly by using online instruction to lower costs and reach more students. Sullivan seems to have thought that this was alarmist and that no program of online instruction would soon convince parents to forgo the rite of passage into adulthood that we call “going to college.” 
On the surface, then, the conflict between President Sullivan and the leaders of the Board of Visitors seems to have been a difference in market outlook, not a difference that could be termed philosophical, even in the loose and popular deployment of that term. Yet I believe there were important philosophical disagreements in the background. Discussions among members of the Board of Visitors touched not only on the importance of taking bold steps to deliver instruction via the Internet but also on the importance of taking bold steps to trim away departments with relatively few majors. German had been mentioned; so had classics. The guiding idea of the would-be reformers was that the university should be continuously reshaped to meet changes in student demand. Consumer sovereignty should be extended from the problem of determining which products should be displayed on the shelves of which stores, to the problem of determining the contents of the proper education of a young adult. 
A rather long analysis follows, with a few distractions, but the heart of it is an analysis of the question, "Why study the humanities?" Brewer proposes three answers: They are actually useful, even if they don't seem to be; they support citizenship in free societies; and they are just good in and of themselves. I think Brewer doesn't treat answer #2 well enough; I continue to associate the study of the humanities directly with burning into the souls of students an appreciation for freedom of thought and action, and Brewer seems to have a rather limited view of what citizens of a democracy can really do with that freedom anyway. But it is a thoughtful analysis, well worth reading.

And relevant to ongoing conversations about educational disruption. When I read this passage in particular, I thought some more about the Faust-Summers dialog at Aspen.

In the concluding section of another of Plato’s dialogues, the Phaedrus, Socrates argues that the written word cannot itself capture and deliver what needs to be understood; at best, it can incite readers to turn toward the phenomena themselves and secure understanding through a more immediate apprehension of them. Further, the written word is not ideally suited to play even this indirect role, since its author is not there to respond to successive attempts, on the part of the reader, to attain a firsthand discernment of the phenomena that inspire it. What Platonic philosophy hopes to deliver to students is no more amenable to summary statement in a treatise or textbook than what Freudian psychotherapy hopes to deliver to patients. The quest for understanding is irreducibly idiosyncratic, because the sources of blindness and delusion are irreducibly idiosyncratic. If the reader cannot speak to the author, the possibility of useful communication is greatly reduced. If this is right, then the spoken word taken in itself—delivered, say, in the form of a lecture rather than in the course of a conversation—is no better a vehicle for philosophical enlightenment than the written word. 
Philosophy, in short, lives in conversation. The student must be called on to speak, and to do so sincerely rather than strategically—e.g., with an eye to a grade. This is what puts the student himself or herself into play. If this does not happen, then philosophy does not happen. Thus, philosophy does not happen in the passive uptake of lectures—whether they are delivered in a large lecture hall or in a Massive Open Online Course (or MOOC, as such courses are quickly coming to be known). If university-style philosophy is in danger of being replaced by the “disruptive innovation” of online education, perhaps this is because university philosophy classes have assumed a deficient form, one suited to fields whose findings can be mastered in passive uptake. 
The passage that came to mind is Larry Summers's argument for why the humanities, not just introductory college mathematics, are due for a Christensen-inpsired, MOOC disruption. The same forces that acted on Kodak have to act on higher education, Summers explains.
It is just not right that Hamlet is exposited 15 or 25,000 times in high schools and in colleges. And so just as 100 years ago every professor wrote out their own notes for their students and then we realized that some professors could write the best textbooks and those textbooks came to be used and the role of what a professor did who hadn't write a textbook was changed fundamentally, something of that kind has to happen in higher education.
The same conservatism that croaked incumbents like Kodak is happening in higher education today. Two of the signs he sees are "the discussions on every campus about not to elevate too many people as expositors different from ordinary faculty in their teaching," and the tendency of universities "to be very protective of existing areas of excellence." Both of these are jarring examples given Brewer's thoughts about dialog and his fears about UVa closing down low-demand departments. Summers also proposes a figure of merit on how much a university has realigned its priorities as the ratio of dollars spent on libraries to dollars spent on online education. I'll take a pass on commenting further on that.

I actually think in one way that Summers does not go far enough in pressing the analogy he posits. MOOCs are indeed more like textbooks than courses. Of course universities have historically not been in the textbook writing business, though they have been accessories; nor have the textbook writers been the most respected professors, as Summers seems to suggest was the lesson of that analogy.

But the real question here is what, in alumni talks I have been giving, I have been calling the "hydraulic model of education," the theory that what educators do is pour their accumulated knowledge, like a fluid, into the empty vessels of the brains of their students. Of course that is not true, as everyone who has thought about it, from Plato on down, has known, but which, as Brewer observes, is one of the key issues at stake in the current higher education wars. I quote Plutarch on this, in his essay on listening to lectures: "The ind is not a vessel to be filled but a flame to be kindled." Which is pretty much the point Brewer was making.

I suppose the reason I am trying to move from lecturing to active learning, rather than going in the other direction to MOOCs, is that I still hold to an ideal like Plutarch's, and fear those who assert with confidence that mass production is the way to go. We can become more efficient, and MOOCs can help in the same way better textbooks can help. But we won't wind up with anything nearly as good if we think that Hamlet just needs fewer, more skilled expositors for us to improve students' understanding of the human condition.

Bonus link: An interesting story about Ecole 42, which teaches software engineering and nothing else, basically by putting a bunch of bright people together on a desert island with an Internet connection and each other. The journalist suggests that Harvard should be worried; and so we should, if this product is what we think colleges are supposed to be optimized for producing.


  1. This comment has been removed by the author.

  2. If "philosophy, in short, lives in conversation," then there is not much philosophy in community colleges, because essentially no conversation about any academic interest transpires in such venues. There is more in state colleges (Cal State, eg) and yet more in state universities (University of California at Berkeley, eg).

    It is perfectly arguable right now that all of the community colleges in California (at least) in the aggregate require perhaps a single calculus professor ... plus a larger set of teaching assistants to answer students' particular questions. An on-line substitute of that one professor's lectures may not be the best teaching, but it sure is cheaper than the alternative. And one could also argue that there is no need for any of the current brick and mortar now designated within any of California's community colleges for calculus instruction. The governor, legislature and state officials are aware of all that and considering what action to take given California's now-perpetual financial crisis. The Cal State and University of California situations are more nuanced, but profound changes are afoot there, too. Then there is the legislature's growing alienation from what they consider the esoteric research ventures of the University of California, coupled with the state's hi-tech businesses increasing confidence in their ability to employ foreign-trained talent at the expense of any need to pressure the state to generate and educate its own technologically trained workers. And there's plenty more.

    "Disrupted?" That is a specialized term coined by HBS nabob Christensen, and much abused and extended by others (one of Lepore's better observations, although she is hardly the first to make it). Higher education is, metaphorically speaking, being disrupted the way a city is disrupted by an incendiary bombing raid. Whether that corresponds to the Christensen definition is another matter.

    One obvious feature of American higher education is its fragmentation or lack of concentration. No institution of higher learning has more than a tiny fraction of anything that might be defined as a "market." There is also obvious segmentation by quality, or at least perceived quality, as represented by the community college/state college/state university example. Christensen's notion of "disruption" of an industry should play out very differently in such a fragmented market than it does in a more concentrated one, where the corrosive effects of the disruptor may be seen internally within the markets served by a single market participant. The disruption of the steel industry may be largely tracked by watching US Steel being disrupted as its lower end sales drop first. But a disruption of higher education can't look like that.

    1. Brewer himself offers a bit of the "compared to what?" question, not about community colleges but about lecturing in universities.

      Should SJSU ditch its philosophy professors for Michael Sandel's MOOC?

      It was Clay Christensen (and his coauthor Henry Eyring) who extended Christensen disruption theory to higher ed.

    2. Yes, Christensen and his co-author applied his theory to universities. I didn't mean to suggest otherwise.

      I meant only to point out that where a market is as highly fragmented along the lines of perceived quality as is higher education, one should not expect developments in any single institution to reflect the possible disruption (Christensen's definition) of the industry as a whole, where in the case of a highly concentrated industry that is more possible. For example, in evaluating whether Christensen is right about the American steel industry being disrupted by mini-mills one can watch US Steel serially lose its markets from the bottom up. But in the case of a highly fragmented industry such as higher education it is unlikely that one will find single institutions that reflect the overall disruption of the industry. Harvard, for example, exists within a narrow band of perceived quality. If higher education is being disrupted (again, in Christensen's sense) Harvard might exhibit no symptoms early on, while institutions further down the quality ladder completely fail. As the rising tide of disruption of the higher education industry reaches Harvard it's crisis may materialize all at once, everywhere, throughout the University. It will likely not be a question of one "low demand" department or another shrinking or closing down. When the hurricane hits Cambridge, it may hit all at once after a long period of calm.

      Lepore's New Yorker hit piece on Christensen takes special care to lash him on his application of his theories to higher education. She tries to make it seem like one more example of other cases she discusses in which his theories have supposedly failed. But there is a big difference that she seems to deliberately (and, in my opinion, perhaps disingenuously) tries to obscure: There is no data from which to mount an argument to show Christensen's theories have failed with respect to higher education. It's too early. That is not true of, say, the steel industry, where there is lots of data and the tussle is over construing it. Lepore's attack with respect to higher education seems pre-emptive and parochial. Almost as if it may be the whole point of the article.

      Once again, you provide and amazingly sophisticated, unique, civilized forum. Thank you.

    3. I wonder. You say there is no data to show Christensen's theories fail for higher ed. That suggests that his theories are falsifiable, in the Popperian sense. Are they? I noted in my other post on all this that Syre seemed surprised at the idea that anyone would think they should be used to make predictions -- he says they are useful only as explanations for what has already been observed. An unfalsifiable theory is an unreliable guide for action.

    4. I didn't mean to suggest Christensen's theories as applied to higher education aren't falsifiable as Popper meant that. Lepore and Christensen's other critics (and they are many, although Lepore denies that they exist) certainly believe that his theories are falsifiable as applied to commercial ventures (and, indeed, often demonstrably false). Christensen disputes that the data in the commercial cases really does undermine his theories, but he has been open to accepting some major criticisms and has re formulated his theories over time. One of his big complaints about Lepore is that all of her serious criticisms in the commercial case are addressed in his own reformulations and subsequent editions of his books. He frankly admits that he got the causation wrong in some cases and had to be shown the right answer by an economist at Tufts, for example. Such, he argues, is the nature of research.

      I just meant that Christensen's theories with respect to higher education haven't been around and applied enough for the relevant data to be produced yet. Certainly Lepore cited nothing in that area comparable to the data she adduced for the commercial cases.

      All levels of education are notorious for their difficulties in formulating appropriate standards of "success." Unlike a commercial venture one cannot just look at a bottom line of profitability. But up to that limitation, I see no reason Christensen's theories would not be falsifiable with respect to higher ed, but it will take some time to collect the data. For example, I have heard President Hennessy of Stanford note that drop out rates are something to watch carefully in any institution seriously introducing MOOC-like technology, which obviously can be seen as depersonalizing. That will happen somewhere and then drop-out rates can be tracked. But of course there is no a priori reason why one must accept drop-out rates as the ultimate standard of success (and Hennessy doesn't say that it is). Hennessy did say that he considers MOOCs and flipped classrooms to be just examples of thousands of ways this technology might penetrate higher education. Even if MOOCs don't disrupt the industry, some other application might. Time will tell. But it will take time.

  3. Harry, this is a DYNAMITE piece. Thank you for posting.

  4. Interesting piece. One quibble: The question of "closing down low-demand departments" sounds bad, I know, but it's inseparable from "how to find resources to invest in key areas." If you want to do the latter, you might have to consider the former -- if not close down, maybe shrink a bit.

    1. Actually that's an important point, and not a small quibble. It's partly true. Certainly departments with decreasing enrollments are shrinking everywhere, just as departments with rising enrollments are growing. (Whether institutions acknowledge doing this, as part of a rational and predictable plan, or just drag their feet about authorizing hires, is another question.) And some institutions (University of Phoenix, say) certainly can and should close down departments for which demand is so low that they are loss leaders. But research universities shouldn't do that lightly. For one thing, demand fluctuates in ways that are not easily predictable. 25 years ago Middle Eastern Studies was a niche area, with few enrollments. That changed dramatically after 9/11. Enrollments in Arabic courses zoomed, etc. And second, research universities have a role that other higher ed institutions don't have. They create and preserve learning, and that entails training scholars even in fields that are not sexy at the moment (either not yet, or not any more), and sustaining the scholarly collections that they require. Every now and then Harvard asks itself why it has a Celtic department, or whether historical linguistics are relevant. These questions often come from exactly the kinds of people Brewer is writing about. In fact there was a proposal at Harvard to shift resources from organismic to micro- biology and shut down the former department as just 19th century science. There is nothing wrong with talking about new and more efficient ways to organize ourselves administratively; fields do grow and evolve (we certainly are experiencing that in the applied sciences at Harvard). But as I noted in an earlier post, the entire discussion of disruption in higher education ignores the research role of universities. Their contributions are not just in teaching, and their teaching role is very much wrapped up with their roles in creating and preserving knowledge. Any discussion of efficiency and disruption that fails to mention that is shallow and likely to go off the rails.

  5. Interesting post, as always. I think a missing piece in the discussion, though (in addition to the one about resource allocation), is the sky-high cost of university education. I agree that an educational model based around MOOCs will be worse than the one that we are used to. But the one that we are used to is increasingly unaffordable for many people. It does seem possible to me that at some point, we will have the cheaper and worse model of MOOCs, just as we now have phone trees, ATMs, flight kiosks, and so on.

    1. Of course you are right, that is what makes this all so relevant, and why one can't talk about "higher education" as though it were a single thing. At Harvard and a handful of other shoals, nobody graduates with any debt (unless they have intentionally borrowed so they could avoid making money in the summer, etc., and even then the amounts are likely < $10K). The question of whether those who can afford to pay the high prices at Harvard will continue to think it is a good investment is on the table of course, but I don't think it's a major public policy issue in the way the student debt problem is.

      And I don't doubt that there will be interesting experiments such as Ecole 42 and some of them will succeed. Maybe Minerva will too. My main observation on all this is that it gets us back to talking about what college is actually for. That conversation needs to happen in something other than economic terms.

      But here is a puzzle about the economic argument. I can barely balance a checkbook so forgive me the following dumb analysis. This should be a subject for another day, but I am puzzled about the student debt crisis. There are certainly ways to get yourself in ridiculous amounts of debt going to college, and many people, like the poor woman in Ivory Tower, are in deep trouble. But the average student debt for a college graduate, according to the most recent figures I could find is $29,400, while the median starting salary of people with college degrees is $46,900, and the median starting salary of people with only a high school diploma is $30,000. So you "lose" four years of your life, but in less than two years, you earn back in increased salary an amount equal to your debt. If there is a problem, and I don't deny that there is, it sounds to me a lot more like a problem of personal financial management than a cost-of-college problem.

      Sorry for the quick and dirty analysis. I know I am matching means and medians, some people are unemployed, the data aren't for the same year, and so on. And none of this is to suggest that institutions with very high prices and very little scholarship aid aren't going to collapse. They will. And there may well be more people who get smart and if they fail in their effort to shoot high, go to a lower priced school rather than to the unaffordable "best" school they get into. But again, that sounds to me more a problem of counseling and educating people to make sound decisions, than like the death of Eastman Kodak.

  6. You may be interested to read an interview with Lawrence Bacow in Medium[dot]com, "Will EdEx Save GenX from Outrageous College Costs?" It doesn't actually address that question, but includes some informed commentary about the issues involved. More directly about what the financal-model representations of edEx are is an article in the Chronicle of Higher Education, "How edEx Plans to Earn, and Share, Revenue from Its Free Online Courses." The primary cash-flow is expected to come from not the massive online enrollments, but from licensing to other institutions! The whole project really seems to still be in the "dot-com" phase. Apparently (I have read elsewhere) it is expected that donors will cover the tens of millions which Harvard has pumped into edEx already.

    Similarly as with the student budgeting and cost-calculation decisions which can make the difference between a long-term financially precarious situation and a more comfortably predictable working model of planning, institutions can suffer the consequences of hubris in presumptions about returns from investments and donor capacity/generosity versus grandiose capital and program expenditures. At Harvard, aside from the many hundreds of millions spent on various buildings and remaking of public spaces in recent years, there are still some billions more in payments having to be made for renovations, new construction, new-building operations expenses, and payments on borrowings, etc., which were required to extract the university from obligations incurred during the financial crisis. Endowment returns are not what they used to be, and statements from outgoing CEO Jane Mendillo indicate that is likely to be a continuing reality. As Harvard Magazine editor John Rosenberg wrote in the magazine's spring issue, much of what comes in during the capital campaign will in fact be required to cover the crisis-recovery costs. Tuition assistance is much more generous than in the past, and thus requires a greater revenue stream. A recent tax return showed that Harvard has committed $75 million to the new green data center in Holyoke. Administrative ranks have burgeoned. So decision-makers have set themselves up for a scenario wherein certainly it will be necessary for a lot of alums to establish very lucrative careers for themselves, during which their enthusiasm about their own experiences at Harvard and the prospects for the future will inspire them to donate hugely, spurred by a worldwide coaxing campaign. Anything "disruptive" which might improve this picture no doubt will be seized upon eagerly. That's the context in which thinking about academic and other priorities takes place.

  7. I agree with much of what Tal Brewer has written, but any analysis of the U.Va. debacle of 2012 is far from complete if it does not address the powerful evidence that it was, at least in significant part, about politics, of the ugliest kind.

    About six months after the incident, when people were still interested in knowing what had been at the bottom of the whole thing, I found something significant and published it in The Guardian. In a nutshell: just a few weeks before Sullivan's "resignation" was announced, administrators mysteriously (and, as in the BOV's firing of Sullivan, refusing to explain why) vetoed the imminent offer of an endowed chair in climate science to none other than Michael Mann. It was clearly a violation of the academic freedom of the Env. Science department. The donor of the chair, lo and behold, was the vice-rector of the BOV, Mark Kington. There's less than a snowball's chance in hell that Cuccinelli's two main adversaries could be banished from U.Va. at roughly the same time (one of them not even being an employee!) and under similar conditions, yet the incidents unrelated.


    I can hardly blame Brewer, or anyone for that matter, for not being aware of this because it has been so vigorously suppressed. People who know the article may think the reason they didn't see the evidence I reported surface in other newspapers was that someone debunked what I wrote. Far from it. To this day, I've heard no remotely sensible argument against the plausibility of my hypothesis, yet the press, from the U.Va. student paper to the Washington Post, has adamantly refused to even let their readers know the article exists, let alone discuss the absolutely verifiable facts in it about the context of the U.Va. scandal that have hardly been reported anywhere else and that many people at U.Va. don't even know.

    There would be no need for the press's (and other organizations' such as the AAUP and SACS, U.Va.'s accrediting agency) desperate avoidance of this if I hadn't gotten things more or less right, and if the facts about Mann's candidacy for the Kington Chair weren't so powerful in and of themselves.

    It's chilling to see how effectively some powerful person or group has been at suppressing this information. Truly the coverup is as bad as the original "crime" -- we're talking about some of the country's leading newspapers very likely committing deliberate journalistic malpractice, as if they can just pick and choose the facts they take into account, and complain about "insufficiency of evidence" for a plausible hypothesis without lifting a finger to look for more. (Oh the e-mails I could show you from the Wash. Post's reporters!)

    It boggles the mind that the connection to the global-warming deniers, and particularly Ken Cuccinelli (who came pretty close to being elected governor of Va. last year, in spite of his possible involvement in this!) were not enough to act on the consciences of journalists and of people at U.Va. who are well aware of this but are for some reason committed to pretending they aren't (I"m looking at you, Faculty Senate).

    Sorry for the rant. I know it gets a little far afield from the legitimate philosophical and other issues that Brewer has raised. But it needs to be said, and it needs to be repeated until the press is forced to break its apparent vow of silence on it. If any of you are truly interested in this, have any questions or doubts that I can address, and/or would like to help raise greater awareness of it, whatever. . . I urge you to get in touch.

  8. This comment has been removed by a blog administrator.