It’s contended the engineering discipline of human–computer interaction (HCI) is promoting during the last handful of decades within an random manner, driven by the necessity to solve real problems, instead of being informed by higher level, general theories from the discipline. A retrospective role for general HCI theories is recommended. A start at this type of general theory, which should be simple capable to encompass all of the discipline of HCI’s activities, is suggested, in line with the idea of tasks. Jobs are characterised because the strategies by which jobs are performed. An over-all systems modelling approach is introduced which divides the assumed world it models into work systems and also the application domains which are altered by work performance. The function of various work systems, based on their differing limitations and goals, to define different subtasks is introduced and highlighted with numerous simple examples.