In 1960, the educational climate was so different from today’s that it is difficult to credit. Classes of more than 40 children, teacher shortages, no research money: but we were keen to try anything and everything, and several teachers in Aberdeen readily joined with me and a psychology lecturer (Leslie Reid) in the fun of trying to build a teaching machine.

The aim was not to replace the teacher, but to provide a teaching aid which pupils could use on their own for specific topics of difficulty. We covered a wide range of ability and subject: one topic I remember was teaching the correct tense to use after “depuis” in French; and another was “to teach recognition of letters and numerals to children whose ability to learn is subnormal”. (That is a quotation from our application for funding of Pounds 20 to cover a device for delivering “small token rewards”, jargon for Smarties) The general idea was one that we imported from the United States. We were the first in Britain to try to improve on the American use of behaviourist procedures which tended to rely on multiple-choice questions. We followed the American principles: breaking the topic into minimal steps in learning, checking at each stage, going back on failure and moving to larger steps on success.

Our first primitive machine was a box with two windows and two levers. The “software” (the term had not yet been invented) consisted of two rolls of paper, one on the left for the instructional unit and a key question, and the other on the right for the learner to write on. Moving the left lever displayed the answer to the question in the left window for the pupil to check. With rather poor imagination we called our machine a “comparator”.

Three prototypes were piloted in local schools in session 1960-61 with a programme on fractions for 10-year-olds, and a “cheatproof” version was developed (a shutter moved across the right window when the answer was revealed).

At the end of that session, we applied to Aberdeen Endowments Trust for a grant of Pounds 100: Pounds 40 to build five more devices, Pounds 15 for a mechanism to present cards, Pounds 20 for the “rewards device” mentioned above and Pounds 25 for “hardboard, wood, plastic, paperclips, screws and duplicating paper”. (It hardly seems credible: 20 years later, on a very different scale, I was a partner in a million-pound research project.) The file also contains sketches and detailed specifications for more sophisticated versions, with a projector and film strip linked to a tape-recorder with soundtrack and pulses to move each frame on. It was pointed out to us by cynics that we could have achieved the same end more simply by using cards with answers on the back, or with a “flip-pack” of paper strips stapled on the left side. (We tried both these methods.) The machine version did work: we sometimes felt it was too successful, in that pupils took only a few minutes to reach 100 per cent mastery of a programme that had taken us a week to write. But it was much too cumbersome. Our thinking was in terms of mechanical devices: cards, yes-no buttons, gear wheels, telephone dials (for numerals), 8mm film and filmstrips and tape-recorders (this was before the days of cassettes).

Now electronic developments have made that kind of thinking obsolete: the silicon chip, computers, VDU screens, photo-cells, edge-lit numerals, digital coding. But some of the programming principles from the 1960s are still valid for self-instructional units, and perhaps we could do more to develop the educational potential of today’s technology, combining the old with the new - though not on a Pounds 65 grant.

Between 1960 and today, of course, there has been another shift, not just a matter of electronics, which may be significant. In the sixties, we called these things teaching machines: today we would call them learning aids.

John Nisbet was professor of education at Aberdeen University until 1988 and is now an honorary research fellow.