A concept from a book by J.W. Dunne: An Experiment with Time (1927)

J.W. Dunne (1875-1949) was a British aviation pioneer. As an engineer trained in the classical pragmatism at the end of the 19th century, he held a number of early British aeronautical patents. He was a successful, respected engineer with a comfortable income.
There was only one problem: He occasionally would have dreams about major disasters. A few days later he would open the paper, and there would be a report about the disaster he had dreamed.
As a good pragmatist, he set out to use the scientific method he had been trained in to try to figure out what was going on. His first subject of study was himself. (Later, he involved a few family members and acquaintances.) He began keeping a dream diary. After overcoming the common difficulties encountered when trying to remember dreams, he managed to maintain a highly detailed summary of each night's dream experiences. Next step: He began to examine the dreams for evidence of future events, not just disasters or other major occurrances, but of any magnitude. And he went a step further. He at the same time kept a list of past events in the same dreams.

After several months, he made a startling discovery (which was confirmed as time passed and the observation and analysis of his dreams stretched into years). He found that his dreams contained approximately the same number of past and future events.
From this, he began to think about time, and our experience of time.
Next conclusion: He came to believe, based on this research, that in sleep we escape the linear, one-direction time which we think of as reality while awake. He compared time to a piano keyboard. When we're awake, we're sitting at the keyboard striking on not at a time, in sequence, from left to right. We cannot normally strike keys to the left ("the past") of the one in front of us, nor keys to the right ("the future"). But when we sleep, we are freed from the rigorous linearity.
He concluded that in dreams we are in effect banging around on the keyboard, hence the peculiar combination of elements and time events.

All this is recounted in his first book, An Experiment with Time (1927). Several other books followed.

Steen Juhler   Comments are welcome: ny@tid.nu