Novelty of Approach

How novel is my approach? Seriously. I'm exploring the development of language using sexual selection, just like Werner and Dyer. I'm using an arrangement of "internal states and externally observable states" (Wagner et al.), just like every communication study. I guess I do the suitability metric stuff... Anybody done that? Anybody?

The more research I do, the less happy I am with my own! At least my project will be more relevant, the longer I stick to the drawing board. So what assumptions am I making, and what will improve my results?

Symbol Grounding: I'm assuming my agents can handle it. I should first review the problem, beginning with the beginning, Harnad (1990). Then I have Taddeo and Floridi (2005) for an overview of fifteen years of research. After that, I'll take a look at some of the nonsituated models that convey internal state, which I guess is every one.

Genetic Algorithms: Downloaded Holland (1975, 1992) today, maybe I'll figure out what schemas are about, again. After that back to my old Mitchell intro. I'll code something up, since my hands are twitching and it'll be fun. Should figure out what the genotype will look like for an agent, which I guess means figuring out what the agents will look like.