The bridge between the previous chapter and this one is anchored by the difficulties of solving NP-complete problems at that end and by the "No Free Lunch" theorem at this end. The No Free Lunch theorem is based on an idea that has been developing in mathematical search theory in recent years that is likely to have important implications for both theories of form recognition and of the brain. Indeed, it has already been identified as one of the key constraints on generalizing neural network theories from one domain to the next.