One of the things that taking Sam Freedman’s book writing class made me appreciate was that unwieldy genre called “the book proposal.” Somewhat akin to writing a dissertation prospectus, writing a book proposal helps you think about your project at several important levels: why it matters, why you should be the one to write it, what you’re going to say. That is, what’s the scope of the book? What’s the story?

I’ve struggled with this quite a bit, and no doubt it’s been one of the things that’s kept me from moving forward with Teaching Machines for years now. I wasn’t quite sure where the story should go. Initially inspired by my ride in a self-driving car with Sebastian Thrun, I thought that the story of the automation of education should include the push for AI and intelligent tutoring systems, for example. I knew I wanted to start with Sidney Pressey’s teaching machine, but I wasn’t sure where I wanted the book to go or where I wanted it to end.

When I first started shaping the book proposal for Sam’s class, I thought I’d start with Pressey and end with Alan Kay’s “Dynabook” – not a teaching machine, but a learning machine. The Dynabook was also, notably, not a device that was ever commercialized (or even really built) – that made it a nice book-end to Pressey’s failures to commercialize his in the 1930s. The Dynabook remains part of the education technology imaginary before the real advent of the personal computer – let alone today’s craze for internet-connected, personal, mobile teaching machines.

But Sam encouraged me to extend my history beyond Kay into the modern era of educational computing. He said that if I didn’t want to go up to the present, I should pick an important recent watershed moment, such as No Child Left Behind, to end the narrative. But NCLB didn’t feel like the right moment to close the book – so I decided I’d end with how Bill Gates’ vision of educational computing triumphed over Steve Jobs’ and how Gates has led efforts to put computers in classrooms in his role as the most powerful educational philanthropist of our time. That’s the book proposal I wrote for Sam’s class – the narrative would go from Pressey to the teaching machines of B. F. Skinner and Norman Crowder, then on to Patrick Suppes and the field of computer-assisted instruction. From there, it would look at Kay and then on to the Gates versus Jobs battle. Those men – yup, men – would be the main characters in the story.

When we workshopped our proposals, my peers in Sam’s class also wanted me to talk about education technology as it stands today, writing about the history of the field up to the present moment. And I get that – that’s what I’m known for too: criticism of today’s venture capital-infused teaching machines. But that wasn’t really what I wanted to do with Teaching Machines. I wanted to write a history. And all that history is too much for one book, I think.

Indeed, one of the major flaws with many other histories of education technology that I’ve read is that they start with the hornbook and run through the latest ed-tech trend. MOOCs. Khan Academy. VR. Whatever happens to be the “hot new thing” at the time the book goes to print. As a result, these books try to cram hundreds of years of history – of educational systems and instructional machines – into one story. It means that nothing is talked about with much depth or much context. It also means that often the books jump from technology to technology, as though these tools are themselves the driving force of history, as though these tools are what shapes policy and practices (rather than the other way around).

As I’ve undertaken my research for the book – a trip to Sidney Pressey’s archives at OSU, a trip to B. F. Skinner’s at Harvard, a trip to Patrick Suppes’ archives at Stanford, and a search for Norman Crowder’s – I’ve realized that what I need to do is not expand the scope of my book to include more and more about educational computing but to shrink it. In fact, I think I’m going to leave out computing altogether.

That means the book will go from Pressey to Skinner to Crowder. (No more Suppes; no more Kay; no more Gates – although certainly the billionaire Bill is likely to get a mention in the conclusion / epilogue.) This makes for a much, much better story. I promise.

Just as importantly, it means I’ll be able to look at that period – from about 1930 to 1970 – in much more detail. This will really be a book about teaching machines, not computers. I will really be able to dig into the ideas about individualization (personalization) and automation and behaviorism that were influential at the time (and to this day). I’ll also be able to focus on the work of IBM – its role in developing testing machines and later teaching machines. (No, tech reporters. IBM is not a “new entrant” into ed-tech. There’s a reason why students protesting the War in Vietnam used the punchcard as the system of a rotten educational and industrial war machine.) I’ll be able to write about the significance of the door-to-door encyclopedia sales (Grolier and Encyclopedia Britannica hawked teaching machines) but also about the commercial push for self-training and auto-instruction in the post-war years. I’ll be able to write about other professors who were working (often in Skinner’s shadow) on programmed instruction. I’ll be able to write about Lyle Spencer and Science Research Associates, which I think, as a Spencer Fellow, will be quite cool. I will be able to spend a lot more time on B. F. Skinner, who I believe had incredible influence on the 20th century with his promotion of a kind of psycho-technologism that is still with us today. I will be able to talk more about the role of the military in developing teaching machines – that’s what Norman Crowder did, of course, but it’s also what Pressey did in the 1960s. Contrary to many stories about Pressey, he didn’t give up on “ed-tech” after he failed to bring his early teaching machine to market.

A lot of stuff happened in the mid-twentieth century that really laid the groundwork for how we think about teaching and learning and technology today. That period shaped what we’ve expected computers to be able to do. I’d rather do a good job with three decades of pre-computer history than write another shoddy jaunt through every instructional tool that’s ever appeared in the classroom. That is precisely the kind of crap cataloging of ideas that venture capitalists and education technologists seem to want but that I refuse to give them.

Audrey Watters


Teaching Machines

A Hack Education Project

Back to Archives