Teaching Machines is a work of narrative non-fiction, a historical account of the development of auto-instructional devices in the mid-twentieth century. A historical account written by a journalist. A historical account written by someone who is much more comfortable with criticism than narrative history.

One of the challenges of this work – and I think this goes for any type of writing like this – is fleshing out the important historical figures as characters. I can only draw on the materials I have – I can’t invent characteristics or attribute actions or attitudes to people that I can’t back up with documentation. And I need to be careful about what I infer based on that documentation.

When workshopping the chapter I’ve written on Sidney Pressey, much of the feedback I received asked me to round him out a bit more. I’d based that chapter almost entirely on letters I found in his archives at Ohio State University, narrating Pressey’s attempt (and failure) in the 1930s to commercialize the teaching machine he’d invented. In the letters, Pressey becomes increasingly agitated. And he admits in several of them that he’s collapsed under the strain of trying to bring his device to market. But just how neurotic was Pressey? (Is “neurotic” even the right word?) I need to find some other materials – diaries perhaps, or maybe letters his wife (soon-to-be, at the time, ex-wife) penned about his state of mind. Pressey, as I’ve written him, is rather one-dimensional.

Crafting B. F. Skinner into a character is much easier. There’s much more material to work with, for starters – the boxes and boxes and boxes of his papers at Harvard. Skinner also wrote an autobiography – three of them actually, totaling almost 1000 pages. (It’s quite a different effort to narrate one’s own legacy than the short autobiographical entry that Pressey wrote about himself for A History of Psychology in Autobiography, Volume V.)

Norman Crowder remains a challenge here, although I also remain determined to include him as a main character in the book. I don’t know the location of letters or diaries. (That’s not to say there aren’t primary materials. There are plenty of mentions in newspapers and magazines.)

There are a couple of other people that I think will be worth developing into characters: Reynold Johnson, for example, a school teacher hired by IBM to develop one of its earliest testing machines; Leslie Briggs, a student of Pressey’s who went on to do a lot of teaching machine work with the Air Force; and Lloyd Homme, who worked with Skinner and founded Teaching Machines, Inc. I know where some materials are about Johnson – some are at the Computer History Museum and some, I think, are at the ETS archives in Princeton. I’m not sure about the other two – perhaps locating some of the military stuff for Briggs would lead to more on Crowder, I don’t know.

I have to balance the narrative and the characters here. That is, I think I know where this story should go – from Pressey to Johnson to Skinner to Homme to Crowder to Briggs (and back to Pressey again). But I have to make sure there’s enough so that the story is rich and exciting and not just a list of interesting machines and an anecdote or two about the men who invented them.