On the very same day that the New Yorker uncovered an unflattering picture of his role in the unfolding of the US stimulus program, Larry Summers was back pontificating about the academy on the pages of the NY Times. This time he was advocating for a cost- and time-efficient new world order for post-secondary education. I had planned to ignore it, until his article was forwarded to me as a member of the UBC Board of Governors on the eve of a strategic session on the University’s current and future finances. As I had expected, the subject did come up in that meeting. After all, Summers’ article could pass for a user’s guide on financial efficiency in the servicing of a university’s undergraduate enterprise.
Fortunately, most Governors had quickly recognized how “banal” Summers’ article was. Why banal? Because all of his statements were either wrong or correct but well known and already practiced at many post-secondary institutions. So, I decided to ignore it again until Margaret Wente decided, a couple of days ago, to echo the banalities in another one of her attempts at “disrupting” higher education, Actually, she may have heard them first from Gwyn Morgan.
What Summers wants you to know is that the world is changing rapidly and that undergraduate education should be changing accordingly. Fair enough, but how? Well, the most charitable way of summarizing his insight is that education needs to take advantage of the economies of scale and of the availability of information made possible by the computer revolution. Another level of efficiency can be achieved by “investing” in no other language but English, since it is anyway the global language of trade.
And this is how a computer would summarize –or is it process?– his insights:
- Students don’t have to learn facts anymore because there is too much knowledge to absorb and they can look them up on the Internet anyway. “Education will be more about how to process and use information and less about imparting it”.
- Students don’t have to work anymore because someone else in their group will work for them. Oh! And they should move their desks into circles and groups so they can learn better.
- Teachers don’t have to be or to know anything because they can show videos and assign books of other teachers to their students.
- Students don’t need to learn any language besides English, since the latter is becoming the “global language”.
- Probability and Statistics are more important than Trigonometry and Calculus and “data analysis” is more important than History.
OK, it is a surrealistic version of his points, but I did say that’s how a computer would process them, didn’t I? Now, how would a human handle them? Well, speaking for myself, I find it hard to refute Summer’s arguments without running the risk of sounding equally banal. But I shall try anyway.
First, I have always thought computers are into “processing” information, while humans are into understanding and original thought, both of which require the hard work of mastering some material. The only alternative is to allow someone or something to do our thinking for us. But can our undergraduate students rely on online information regarding evolution, revolutions, climate change, wars, peace and other disputed matters?
Now I agree that collaboration is great, but I am well positioned to know that it has already been fostered in academia for several decades now. Still, I argue that much of the creative work is done alone, after having invested long and uninterrupted hours to study and master topics. Fine, here I am running the risk of being banal myself, but let me single out one of Summers’ points that I find dangerously misleading and simply not true. He says that, “For most people, school is the last time they will be evaluated on individual effort.” And I say, that in almost every job these days, performance evaluations are still conducted on individuals and much less on groups.
By promoting the concept of “video-teaching”, which –by the way– has also been used in universities for decades now, Summers is really pushing the notion that going to class to hear a lecture is not only outdated but also cost- and time-inefficient. This reminded one commenter of a Doonesbury cartoon from many years ago, where Uncle Duke is running for president and asked about his education policy. He says he would fire all of America’s teachers except for the best, who would present their classes on television to the whole nation. When the other candidates are asked to comment, they say ‘that’s a great idea, I wish I had thought of that!’
“Those who know no foreign language know nothing of their mother tongue”, wrote Goethe. I may add that they will also not know enough about other cultures to avoid the patronizing tone implicit in Summers’s comment about English speakers “treating patients in Africa or helping resolve conflicts in the Middle East.”
Most shocking is his musing about which scholarly content is more relevant to the workforce of the future. For one, it is a pity that he thinks that students should stop at the pre-calculus 17th century discrete probability theory of Pascal and Fermat and not learn the more calculus-empowered probability and statistics of Bernoulli, De Moivre, Laplace, Markov and Kolmogoroff.
Or did Summers mean that we should only teach students the latest probability and statistics software packages, because all of them will/should be joining the world of finance? Mr. Summers will be surprised to know how much trigonometric functions (and Fourier series) are relevant even in the world of finance. And, we do hope that the former secretary of the US treasury knows about the calculus-based “Black-Scholes model,” and the disastrous effect that the blind and universal use of its software packaged version has had on the 1980′s financial markets.
There is no doubt that Probability and Statistics are extremely important subjects, whose relevance was late to emerge in our educational system. But there is much more to mathematics. Here is a Table produced by the British Columbia Institute of Technology showing how various areas of mathematics –mostly calculus– are used in “real life”, from bio-medical engineering and Petroleum Technology, to Prosthetics and Wood Products. All are industries, where our undergrads could work some day. But let me stress that this Table only deals with elementary mathematics, very elementary mathematics, the one discovered and developed many centuries ago, and which is taught in undergraduate programs all over the world, as it should.
And one last point. Summers uses the following reasoning to make his point about video-learning. “In a 2008 survey of first- and second-year medical students at Harvard, those who used accelerated video lectures reported being more focused and learning more material faster than when they attended lectures in person.”
He should have paid more attention in his statistics courses because unless the group of students who used video lectures was randomized, this data is normally deemed to be meaningless.
It seems that online education is creating new opportunities for classical teaching. I can’t help posting this email I just received (I’ve anonymized it). Enjoy.
> From: ANON
> Date: Tue, Feb 7, 2012 at 11:36 AM
> Subject: Re: Machine Learning
> To: Nando de Freitas
>
> Hi Nando,
>
> I would like to request your help in finding a machine learning tutor.
> I am working through the Stanford course in my free time, and would
> really like to work every week or two with someone who has a good
> grasp of machine learning (both the tools and how to use them, and the
> larger context) and is good at explaining things. I think this will
> aid my learning and enjoyment.
>
> Thanks,
> ANON
On the other hand, it is true that there are very nice new online resources, e.g.
http://david-hu.com/2011/11/02/how-khan-academy-is-using-machine-learning-to-assess-student-mastery.html
I see these as complimentary to my teaching. In fact, my lectures are also videotaped and posted on the web, but most students still like to come to class.
What I see new here is: the vast marketing behind some of the new initiatives, online marking systems, the possibility of adapting the online lectures automatically to each individual, the ability to acquire massive amounts of data about the aptitude of a large group of individuals.
I like the online personalization aspect and I think it is worth improving further. My opinion on this is of course clouded by the fact that I’m a machine learning prof, but nonetheless I see the new media as nice multimedia books that adapt to your learning rate. I’d love to improve the offering of these to my students, though I suspect that many would still end up coming to class; like ANON above.
Concerning the financial aspects, I am surprised that the media has not realized that there is a lot of money behind the idea of identifying the top 20,000 individuals in profession X and referring these to recruiters or companies directly. The issues of privacy and fair competition are worth thinking about in this domain. Many business people have jumped into online education because of the value of trading people’s data, and not necessarily because of pure altruistic reasons. I am not saying this is a bad thing, but I am certainly saying that the press has been incredibly silent with respect to this (most likely because they haven’t thought about it as they are too busy paraphrasing each other with the same hype).