Michael Tobin

Current attempts to produce a satisfactory definition of programmed learning are sometimes interpreted as a sign that programming has not lived up to its early promise. Why else, it is argued, should there be this uncertainty about its distinguishing characteristics? Ten years ago, there was none of this concern with what did or did not constitute a programme. It was a self‐in‐structional system, usually presented by some kind of teaching machine, and designed in accordance with one of two apparently opposing models of how human beings learned.