Myths and Legends of the Baldwin Effect

This position paper argues that the Baldwin effectis widely misunderstood by the evolutionarycomputation community. The misunderstandingsappear to fall into two general categories. Firstly,it is commonly believed that the Baldwin effect isconcerned with the synergy that results whenthere is an evolving population of learning indi-viduals. This is only half of the story. The fullstory is more complicated and more interesting.The Baldwin effect is concerned with the costsand benefits of lifetime learning by individuals inan evolving population. Several researchers havefocussed exclusively on the benefits, but there ismuch to be gained from attention to the costs.This paper explains the two sides of the story andenumerates ten of the costs and benefits of life-time learning by individuals in an evolving popu-lation. Secondly, there is a cluster ofmisunderstandings about the relationshipbetween the Baldwin effect and Lamarckianinheritance of acquired characteristics. The Bald-win effect is not Lamarckian. A Lamarckianalgorithm is not better for most evolutionarycomputing problems than a Baldwinian algo-rithm. Finally, Lamarckian inheritance is not abetter model of memetic (cultural) evolution thanthe Baldwin effect.