I love working in a University. Every day I encounter new ideas, and I get to chat to some of the smartest people on the planet. But I see signs, almost every day, that universities are poorly equipped to face the complex challenges of the 21st Century. Challenges like poverty, climate change, resource depletion, sustainable agriculture, and so on. The problem is that universities are organized into departments that correspond to disciplines like physics, computer science, sociology, geography, etc. Most of the strategic decision-making is made in these departments – which faculty to hire, which degree programs to offer, what research to focus on. But the grand challenges of the 21st Century are trans-disciplinary. To address them, we need people who can transcend their own disciplinary background; people who are not only comfortable working with a range of experts from many different fields, but who actively go out and seek such interactions. In marketing speak, we need T-shaped people:

Jelinski, who is vice chancellor for research and graduate studies at Louisiana State University, talked about a new “T-shaped” person with disciplinary depth, in biology for example, but with the ability, or arms, to reach out to other disciplines. “We need to encourage this new breed of scientist,” she said. [“Researchers Seek Basics Of Nano Scale,” The Scientist, August 21, 2000]

Universities don’t do this well because the entire reward structure for departments and professors is based purely on demonstrating disciplinary strength. Occasionally, we manage to establish inter-disciplinary centres and institutes, to act as places where faculty and students from different departments can come together and learn how to collaborate. A few of these prosper, but most of them get shut down fairly rapidly by the university. Here’s what happens. A new centre is set up with an initial research grant, perhaps for 3-5 years, which typically pays only for researchers’ salaries and equipment. The university agrees to provide space, administrative staff, and pay the utility bills for a limited time, because opening a new facility is good press, but then expects each centre to become “self-sufficient”. This is, of course, impossible, because no granting agency ever covers the full cost of running a research centre. The professors who want to make the centre succeed spend most of their time writing more grant proposals, most of which don’t get funded because competition for funding is tough. Nobody has much time to do the important inter-disciplinary work that the centre was established for. After five years, the university shuts it down because it didn’t become self-sufficient. A research centre at U of T that I’ve spent a lot of time at over the past few years is being shut down this month for these very reasons.

The same thing happens to inter-disciplinary graduate programs. While departments run graduate programs focusing on disciplinary strength, some enterprising professors do manage to set up “collaborative programs”, which students from a range of participating departments can sign up for. The collaborative programs are set up using seed money, some of which is donated by the participating departments, and some of which comes from the university teaching initiative funds, because they all agree the program is a good idea, and the students will benefit. However, after a few years, the seed money has been used up, and no unit within the university will kick in more, because the program is supposed to be “self-sufficient”. No such program can ever be self sufficient, because the students who participate are accounted for in their home departments. The collaborative program doesn’t generate any extra revenue, and the departments view it as ‘stealing their students’. Without funding, the program shuts down. A collaborative graduate program at U of T that I serve on the advisory board for is ending this month for these very reasons.

Not only does the university structure tend to squeeze out anything that does not fit into a neat disciplinary silo, it also generates rules that actively prevent students acquiring the skills needed to be “T-shaped”. For example, my own department has “breadth requirements” that graduate students have to meet when selecting a set of courses. “Breadth” here means breadth across the discipline. So students have to demonstrate they’ve taken courses that cover several different subfields of computer science, and several different research methodologies within the field. But this is the opposite of T-shaped! A T-shaped student has disciplinary *depth* and *inter-disciplinary* breadth. This would mean deep expertise in a particular subfield of computer science, and the ability to apply that expertise in many different contexts outside of computer science. Instead, we prevent students from getting the depth by forcing them to take more introductory courses within computer science, and we prevent them from getting inter-discipinary breadth for the same reason.

Working within a university sometimes feels like the intellectual equivalent of being at a lavish buffet but prevented from ever leaving the pasta section.


  1. Have you seen? http://worthwhile.typepad.com/worthwhile_canadian_initi/2010/10/confessions-of-a-central-planner.html
    It’s a good discussion about changing the university structure so the incentives line for departments to offer courses to non-majors.

  2. @Elizabeth: Yes, I remember that article. I didn’t like it when I first read it, and still don’t like it now. Sure, there’s a lot of good insight there about supply and demand within a university environment, and how to re-jig incentive structures to make things a little better. I think it might help reduce *one* major cause of unhappiness for students. But (as is typical for an economist), all the really important things are ignored as externalities.

    Universities don’t exist to optimize the supply and demand of seats in courses. They exist to educate. [They also exist to do leading edge research, although that’s usually achieved purely by overcharging for teaching, and diverting some of the funds to a research mission.] The mechanisms Nick talks about pay no attention to the quality of education, because his cost models cannot account for it. The closest he gets is to use popularity of a course as a proxy for quality. But it’s a very poor proxy, especially given the byzantine problem most students face in navigating through their course selection tasks. I will concede that if he’s successful in switching from an excess-demand to an excess-supply model, the quality of that proxy might improve a little. But I doubt it – there are too many other perverse incentives (e.g. grade inflation, workload pressure, student indebtedness, etc) that mess up students’ course selections.

    The core problem is that he’s attempting to optimize within a system that has deep structural problems. What’s wrong with modern universities is not to do with allocation of resources within the system. It’s the structure of the system itself. Partly it’s the structural issue of how society funds (or rather fails to fund) higher education. And partly, with respect to my desire for T-shaped students, it’s the structural problem that the very existence and identity of each department is based on it being very distinct from other departments. Changing how you shuffle tokens between existing departments does nothing to change this. If you improve resource allocation in this system in the way Nick suggests, you switch from having departments competing to avoid taking students to having departments compete to win more students. Nick thinks this is a good thing. Maybe it is. But it doesn’t break down the silos. It doesn’t reward faculty who innovate (innovations are rarely popular with heavily indebted, risk-averse students). It doesn’t incentivize intellectual adventurousness.

    You can’t make fundamental changes to a broken system merely by tinkering with the parameters. You have to change the structure.

  3. Adrian Schroeter

    Steve have you considered starting with ensuring that students within your department get depth by loosening the breadth requirement . That might eventually allow students to take more courses outside of cs and generate more interdisciplinary breadth.

  4. Pingback: The Need for T-shaped Students | Steve Easterbrook | Aug. 15, 2013 | Serendipity | In brief. David Ing.

  5. @Adrian: Every year I put forward arguments for loosening our breadth requirements. And every year I’m defeated by my colleagues. They equate breadth within the field as the hallmark of a great grad program. I say students are smart enough to make these decisions themselves, and need more freedom to choose.

Leave a Reply

Your email address will not be published. Required fields are marked *