- This was a Canadian conference, but reflects the growing interest, on a world-wide basis, in interaction between people and various kinds of machines. It is now a matter of history that the study of human factors really got started during the second world war, as the increasing complexity of the machines that man had to control (most notably the aeroplane, but also radar systems, weapons systems and similar systems involving non-trivial control-display relationships) challenged the human being's sensorimotor abilities. It was quickly realized that mistakes made by pilots and others in controlling these systems, or simple lack of required performance levels, could no longer be blamed on those operators but, rather, arose from the inherent impossibilities or difficulties designed into the tasks they were asked to perform, as a result of designers' failure to understand or take note of human limitations and characteristics when designing the system.
The subsequent invention of the computer, and the implementation of complex computer systems in which more subtle and heavily loaded communication between people and machines is required, has extended the scope of human factors engineering beyond the purely physical, beyond the psychology of sensorimotor performance, and into the deeper areas of interactive dialog. This is no longer a passive interchange of information, involving optimized control-display relationships and population stereotypes, with the bulk of the information processing allocated to the human. It involves active co-operation between people and machines at an intellectual level requiring the automation of those intellectual activities for which a machine may be suited; the best division of labour between human and machine components; and rather powerful channels of communication between the two in sensory modalities most appropriate to the nature of the interaction. In the ultimate analysis, the extent to which the machine may carry its share of the work will depend on the excellence of this communication, and the degree to which the machine is able to "understand" the task and the task objectives, and therefore safely take over some of the real intellectual load. Nilo Lindgren, writing in the IEEE Spectrum as long ago as March 1966, summed it up rather nicely in the following passage:
"And now the emergence of the computer sciences, in which the human characteristics must be matched with the machine at the intellectual and deeper neural levels threatens to place new burdens on what 'human factors' means as a name." Historically, communication by means of'graphic symbols and gestures seems to have preceded more formal and abstract communication by means of written language. We presume that speech fell intermediate between the two, and that songs and music based 221 0020-7373/78/030221