This is one of a series of posts on evolvability. It is based off my undergraduate thesis, which I wrote at the University of Puget Sound under advisors Dr. America Chambers and Dr. Adam Smith. The original thesis is available here.
Blog Series on Evolvability
I like digital evolution because it necessitates the examination of fundamental assumptions of what is necessary for evolution. Building a digital evolution system, a researcher must work out how the phenotypes that are being evolved should be genetically encoded. This decision raises an interesting question: how do genetic encodings for digital evolution systems influence the evolutionary process within these systems? I think this question is really interesting! Better understanding this question has practical implications for digital evolution, as well. So, I picked it up as the topic for my undergraduate thesis.
As I dug in, I realized that the question of genetic encoding was part of a larger topic: evolvability. Evolvability In the crudest terms, evolvability refers to the capacity of an evolving system for adaptive evolution [Pigliucci, 2008]. (Don’t worry! We will expand on this definition next week.) In addition to the question of how the genetic encoding scheme affects the evolutionary process, evolvability encompasses questions of how the actual content of the genotype can affect the evolutionary process. Consider, for example, a phenotype that, under a particular genetic encoding scheme, might be encoded by two distinct genotypes. How might the evolutionary implications of these different genotypes differ? Is one encoding more evolvable than the other?
Although my initial entrée was from the perspective of digital evolution, I found a biological perspective to be especially rich for discussion of many aspects of evolvability. The idea that the genetic architecture, physiology, developmental processes, etc. of biological organisms, in addition to facilitating immediate functionality over an organism’s lifespan, also has properties that facilitate the evolutionary process caught and held my attention. Because I had never directly thought along those lines before, I found it especially interesting (and cool!) to think about how, beyond its immediate functionality, biological structure affects the phenotypic outcomes of genetic perturbation. So, the scope of my thesis expanded to review and organize the ideas both biologists and digital evolution practitioners have developed to understand what evolvability is and where it comes from.
Over the course of a series of posts, I’ll be sharing an overview of evolvability, drawing from that thesis. I’ll define and illustrate examples of a number of concepts related to evolvability. Along the way, I’ll also discuss how these concepts are related to one another and share the conceptual framework that I use to organize them. The first step, though, is to rigorously define evolvability itself. So, let’s start!