Measuring Evolvability as the Rate of Complexity Increase

Chrystopher L. Nehaniv

Facutly of Engineering & Information Sciences
University of Hertfordshire
Hatfield, Herts AL10 9AB U.K.
C.L.Nehaniv@herts.ac.uk
http://homepages.feis.herts.ac.uk/$\sim$nehaniv

Abstract:

Evolutionary systems that exhibit non-trivial evolvability are expected to show complexity increase over time, and open-ended evolution requires that such complexity increase be unbounded. Drawing on the mathematically rigorous measure of hierarchical complexity for biological systems (Nehaniv & Rhodes, Artificial Life, 6(1), 2000) one can (1) define measure evolvability as the rate of complexity increase, applicable to natural and artificial systems and (2) compute rigorous bounds on evolvability from mathematical principles. Moreover, whenever an approriate complexity measure can be defined on entities in evolving populations one can quantify and measure evolvability.