Posted by Sven on July 28, 2003 at 11:13:22:
In Reply to: Concerned About Selection posted by David on July 28, 2003 at 08:23:40:
this problem is associated with at least two things that have been discussed before:
speciation: a lenghtwise homogeneous population of poems will not breed poems of a completely different length.
cross-over (see my guest-book entry): in other beings (than poems) you have only homologous recombination. cross-over between non-homologous genes brings death or severe disease, becuase the hole genetic identity is jumbled. i suggest trying to overcome this by introducing a linkage variable, that holds long-travelled parts together and gives a bias for breakage at less established points. - possibly this can also be achived by introducing breakage points / snips as invible words as part of the genetic meterial, as somebody has suggested.