THE SUBJECT of schizophrenia is responsible for a fast growing but increasingly indigestible mass of clinical descriptions and experiments. Many of these data are valuable, but some conceptual scheme for ordering and organizing these observations would be helpful. The obvious parallels between human thought processes and computer programs have inspired some useful models; thus, I was led to explore a computer analogy for schizophrenic thought-process disorder. Some years ago, before phenothiazines, a fire broke out on the back ward of a state hospital. Most of the patients were hallucinated, chronic, process schizophrenics. However, they quickly queued up and marched out as sane as you please. Mannerisms, responses to hallucinations, and other gross signs of disorder vanished until after they reached the safety of the yard; then things returned to normal, or, in this case, to abnormal. With such a clear