The way in which perceptions, memories, intentions, etc., are represented in the nervous activity of the brain is not known. The main focus of this field of investigation is to discover these representations and understand the neuronal mechanism which generates them. This is a combined effort including recording the activities of several nerve cells in parallel from appropriate places in the brain of behaving monkeys, devising new mathematical methods for analyzing the data, and verifying the theory by constructing (by way of simulations) neural networks which mimic the postulated brain processes.
So far the results indicate that the prevailing view about coding by enhanced neural activity is only part of the truth. The exact timing of nerve cell activity contains much of the information about what is happening in the brain. This timing can be "read" by neurons due to their sensitivity to synchronized activation. This same property also serves as the basic mechanism by which exact timing is generated. A neural-network known as a synfire chain can both read and generate such well -timed activities. The experimental results support the synfire hypothesis and extend it by suggesting that activity reverberates in such synfire modes. We assume that reverberating synfire chains in different brain locations can resonate with each other, thereby generating a unified "mental" experience.