Hmm... I have a new hypothesis on how knowledge (or at least, scientific knowledge) works. Knowledge, imho is does not proceed from divine inspiration, or random spurts, or gradual building up, but in a combination of these processes via a system akin to evolution. 1. Mutation. This comes from random complexity in the human brain, contributing to what we consider creativity. From this, we have random thoughts occasionally to come up with completely new lines of thought. These provide our moments of inspiration, but few of these inspired thoughts are true. Sometimes, mutation of knowledge can create new ideas that are completely useless in the growth of understanding, and these grow of themselves as though a cancer. (such as many crackpot theories, group misconceptions and perhaps religion/superstition? Or maybe not. ) 2. Crossover. Rival ideas meet and recombine. When two people talk about two theories, this collision can create a new offspring theory that contains elements of both systems. Some selection is also involved at this point, where people choose which theory to listen to and which to discard. This breeding of theories creates a large population of people with slightly different world views, and helps to proliferate theories far and wide. PF can be considered a mating ground of knowledge. 3. Selection. Theories are each tested against evidence, and in competition with other theories. While a single negative does not kill the theory, it decreases it's frequency among the heads of the population, and the strength of conviction it's trueness. Positives do the reverse. Hence, predictive and improving theories are favoured over stagnant and unproven hypotheses. However, the evolutionary knowledge system reaches a natural equilibrium where each theory cannot acheive an absolute dominance. This can be equated to keeping a large gene pool, and this allows the sudden emergence of new theories over old stagnated ones. From the above system, a number of features may be noticed. A. The evolution of knowledge generally tends to local highs in correctness. It takes a difficult mutation/intuitive leap to reach new quantum shifts in understanding. B. The system is basically random, and is derived from the idea that (a) understanding varies across minds, (b) theories breed and recombine, (c) random additions of knowledge occur, (d) theories are judged on successfulness. With these 4 conditions, evolution of knowledge would naturally occur. Without any of these 4 (eg. if there is indoctrination, segragation of intellectuals, discouragement of originality or a faith based approach to knowledge), stagnation would occur. C. The process is infinite. Ie. There is no sign of a set ending to searching for knowledge, or an acheivable final goal of knowledge. D. Theories normally don't die. They don't have 100% dominance either. As a theories reachs very low levels of existence, factors come into play which gives it an advantage over high dominance theories. The limited lack of full rejection of ideas is useful as it allows continued testing and recombination of knowledge. Well? Comments/flames?