WillIAM HIRST , ULRIC NEISSER , AND ELIZABETH SPELKE

.Common wisdom has it that "you can't do two things at once," but people do combine activities all the time. They walk and talk, eat and listen, drive and think. When two things are done simultaneously, must one or both suffer? Do the tasks interfere with each other? Must one of them be done unconsciously? Are they somehow fused? Questions like these are often answered on the basis of assumptions about the nature of consciousness and the limits of mental capacity.Webelieve, however. that they deserve experimental study. We also believe that they have no single or simple answers. Every human activity is flexible; with practice, each becomes smoother and faster. Combining two skilled activities is itself a skill that will change as it is mastered. By stl,ldyingthe development of such skills. we have been able to test and refute some widely held assumptions about human intellectual limits and capacities. It is fashionable nowadays to treat the brain as if it were a computing machine and to equate the central processing unit of that computer with consciousness. Such an analogy clearly suggests that people should be limited to doing one thing at a time, at least when doing things that require central processing. We believe that the computer analogy is 'misleading. Mental processes change so fundamentally with practice that no assumption involving fixed structures and fixed capacities can be correct. To make this point. we trained people to combine two very complex activities. They learned to write from dictation while reading stories, and they learned to understand both what they were reading and what they were writing. The task is a difficult one and required a great deal of practice to master. We deliberately chose this level of complexity in order to push the question of divided attention to its limits. So long as activities are simple. almost anyone can learn to combine them. Indeed. every human activity involves the coordination of many different actions. In walking, for example. you must lift one foot while placing the other, swing your arms, correct your balance, and watch for obstacles. Most people have no trouble doing all these things together; they can even chew gum at the same time, and perhaps carry on a conversation as well. To reconcile such achievements with the notion of an indivisibleconsciousness. or of a limited central capacity, psychologists appeal to the concept of automatism. Automatic activities are said to by-pass the central processor, so that any number of them can be combined. If activities are automatic, they are not conscious; the individual is not aware of them. Such activities must also be relatively simple so that they require no consciousness or action by the central processor.The theory we are considering implies that complex actions, like understanding ideas or making decisions, can