Virtual Worlds

Abstract

This article argues for using virtual worlds as experimental environments for social science questions at the macro level. The authors can foresee two major objections to this approach and will address them as to show why they do not prove to be significant. The first being that virtual worlds are not like the real world; therefore, one cannot generalize from events within them. The second of these foreseeable objections states that human society is too complex to be controlled in the way that controlled experimentation requires. Humans discover things by building environments suited for exploring the questions the authors have a rat maze is a very abstract environment, yet it is useful for exploring very general questions of mammalian cognition. The authors conclude that virtual worlds are no less valuable, on net, than other established experimental tools. The next stage in toolmaking, after Petri dishes, rat mazes, and supercolliders, should be virtual worlds.