I know that this is a subject of much debate among the fields of metaphysics, general philosophy, neuroscience, psychology, etc. Today I came up with an idea, I'm really not sure if others have, unbeknownst to me, thought of the same previously, or if I have come up with a new possibility. In any case, I shall present it to you here: As you will no doubt be aware, our minds work by taking input through a neural network, a sort of tree. For instance, IF you are thirsty, THEN look for a drink, OTHERWISE don't. IF you look for a drink, THEN see what kind you have. IF you have coke, THEN see if you have Dr. Pepper, OTHERWISE drink the coke. IF you have Dr. Pepper, THEN drink it. etc. We create new neural pathways as we experience things and learn. We then edit these pathways in order to make them more effecient. While these "programs" may run correctly, in order to change (improve) them we must know each step in the pathway. The only way this can be known is if either (A) each node on the web sends out a signal when activated, the signal reaches another web which analyzes it and adds it to a map of the former web. (B) diverse stimuli are fed into the network as input and the output is analyzed carefully and cross-referenced with output from different input. I propose that (B) is the way the brain does it, and that dreams are the brain mapping out the neural pathways aquired over the previous day. At night, when we are unconscious, we are isolated from any external stimuli- we are a controlled environment. I propose that the brain feeds itself specially selected stimuli in order to analyse its own reactions and use the information gathered to deduce any changes in its neural networks. This would explain why dreams contain elements of things from the previous day- any changes to the networks would be related to the stimuli we had experienced. Therefore similar stimuli would activate the same areas of the "program". Recurring dreams, perhaps, are the brain feeding itself the SAME stimuli to see how its response has improved or degenerated, and use that information to tell whether it has been improving its effeciency. Any comments, ideas, questions, etc would be greatly appreciated! Thank you.