Arguments are given for the tenet that although the immune system has no long term goals, it does have short term partial goals - which are often contradictory. Simple mathematical models illustrate how feedbacks can (i) harmonize conflicting partial goals, (ii) improve the performance of a given type of effector cell, (iii) cause the preferential amplification of more potent effectors. In particular, it is shown that spatial organization can allow non-specific chemical signals to select specific immune elements that contribute to system goals. Comparison is made with other autonomous decentralized systems.

Back to Workshop Schedule