Printed: 2020-07-02

Institute for Ethics and Emerging Technologies

IEET Link:

Interfaces and Education

David Eubanks

February 24, 2013

In my last article, I used a cartoon model of intelligence to examine different aspects of whatever that thing is we call critical thinking. The usefulness of the schematic goes well beyond that exercise, however. Specifically, there's the fascinating idea of a "unit of usefulness" often called an interface. It's worthwhile examining how it works in the context of education.

An interface allows a trip to be made all the way around the diagram, which I'll reproduce below.

An elevator is a good example. We start with a motivation:
  1. We want to be on a different floor of the building
  2. We observe the elevator controls, which are carefully constructed so as to be unambiguous, and clearly map reality to a specialized language.
  3. Our internal model (i.e. ontology) of the physical arrangements allow for a simple calculation to map what we want into the language of the elevator's button scheme, so that we can properly predict what's about to happen, and
  4. Act accordingly by pushing a button.
  5. Finally, we observe that our motivation is satisfied by arriving on the right floor, designated with language that lets us know that has happened.
This process is the main function of intelligence (predicting what actions will satisfy motivations), and we do this all the time with and without interfaces. The difference is that an interface makes it transparently easy, by means of a carefully designed language and apparatus for physical change that are congruent. 
Technology Produces Interfaces at an astonishing rate these days. The miraculous devices we carry around are obvious examples, but there's another influence that may not be as apparent: the evolution of societies into machine-like systems creates interfaces too. The Department of Motor Vehicles is a sometimes-reviled interface with government bureaucracy. It's unpleasantness isn't merely from the long wait, but from being treated like a pile of documents rather than a human being. But it's ubiquitous. The person working the check-out line at the grocery store is an interface too, and we can choose to limit our interaction to that mode of operation entirely, rather than acknowledging that this is a human being. The opening sequence of Sean of the Dead portrays this zombie-like element of our lives, even before the inevitable infections begin. It is typified by automatic behavior that characterize a problem already solved.
This is the result of any interface too--no matter how complex an airliner is, to the passenger it's primarily a way to get from one place to another. Everything is standardized. It may be stressful, but that's not because you don't know what the plan is. Much of modern life is developing facility with standard interfaces--like driving a car or using a phone, and the pace of technology creates stress because we can't keep up with all of the new ones. Perhaps that's why Apple products are so popular--they make the interfaces easy to use.
So much of our internal ontology--the way we understand the world--is now tied up in technological and social interfaces that would be very foreign to our forebears. 
Interfaces in Education abound. The whole bureaucracy is designed to partition reality into neat categories of "nominal reality." Unfortunately for this endeavor, humans are not very good subjects for this, and it leads to a lot of mischief. Take, for example, the "inoculation" folk theory of education that crops up continually in the academy. Students take a class on composition, and then are assumed to be able to write. Any subsequent problems with writing point fingers back to that class, which did not fulfill its role. The assumption is that Comp 101 is a reliable factory-like interface that takes raw material and produces good writers. This is a poor reflection of reality because writing is a tremendously complex endeavor that is more akin to kindling a flame than filling a vessel, to employ the ancient metaphor.
We parse learning into boxes called courses and majors and learning outcomes, and institutions certify these with their stamps of approval--theoretically providing an interface for consumers of the product. Of late, the advertised quality of that product (and the cost of producing it) has come into question, perhaps most infamously in Academically Adrift.
We could spend a lot of time dissecting why the interface model fails. Many of the articles I've written in this blog concern how assessments (particularly standardized tests) can create nominal realities that apparently create interfaces, but fail to reflect reality. The result is optimizing only the appearance of satisfying motivations, which is the central idea behind self-limiting intelligence.
Rather than rehashing the problems with the Reality-Language (i.e. measurement) detail of educational assessment, however, I'd like to comment on the expectations that students have. Particularly those who have come from a test-heavy public education, it seems (anecdotally) that they expect a college course to be a clearly-delineated interface, similar to the check-out counter. For example, they don't seem to take to open-ended problems naturally. An interface-centric attitude results in the following expectation: "show me exactly what I need to do to get an A," as if education were an algorithm. It's very easy to teach math courses that way (at least until creativity is required in later courses), but I think it does students a disservice. For example, they can learn how to take derivatives of functions without having any real intuition about what that means. This is well documented in an ongoing research project based on the Calculus Concept Inventory, which seeks to assess conceptual understanding. A quote from one paper (source):
the first term we did TEAL on term, the over all course evaluation was terrible, the lowest of any course I have been associated with at MIT, but I can plausibly argue that that term the students learned twice as much as under the lecture system, using assessment based on Hake normalized gains.
TEAL stands for Technologically Enhanced Active Learning, which de-emphasizes lectures in favor of more active approaches. The comment implies that it worked, but students didn't like it. I presume the reason is that emphasis on active learning of concepts is open-ended and less interface-like that what others have called the 'rent model': if I sit in class long enough, you pass me.

There's a reason why it took our species a couple million years to come up with calculus, and it's not because it's complex. It's NOT complex--anyone can learn the power law in a few minutes. It's the conceptual subtlety of thought and the precise language that expresses it that is the real value of the subject. And it's not just calculus, of course.

The parts of our environment that we can control with interfaces is the easy part--that's what interfaces do. Higher education (especially the liberal arts) should not just be a catalog of new interfaces to learn, but should cultivate the general ability to wrestle with problems that don't have interfaces. Questions of politics and ethics, and the expression of creativity cannot be reduced to a pre-packaged I/O device (despite the strident voices of ideologues who argue for just such a thing). 

If higher education is going to fulfill this role, it has to do as much work in unmaking minds as building them up, because many of our students are well-trained to expect (and demand) A-B-C-degree. The automated satisfaction of motivations by itself is a wonderful thing, but it can also make us dull with expectations that everything is push-button easy.

David Eubanks holds a doctorate in mathematics and works in higher education. His research on complex systems led to his writing Life Artificial, a novel from the point of view of an artificial intelligence.


Contact: Executive Director, Dr. James J. Hughes,
IEET, 35 Harbor Point Blvd, #404, Boston, MA 02125-3242 USA
phone: 860-428-1837