A number of philosophers :171–187, 2011; Arnold 2011, in Ethics Politics XV:101–138, 2013) have argued that agent-based, evolutionary game theory models of the evolution of cooperation fail to provide satisfying explanations of cooperation because they are too disconnected from actual biology. I show how these criticisms can be answered by employing modeling approaches from the situated cognition research program that allow for more biologically detailed models. Using cases drawn from recent situated cognition modeling research, I show how agent-based models of (...) the evolution of cooperation can become more empirically-informed and relevant to explaining the evolution of cooperation in real populations. I argue that because situated cognition models allow for more detailed representations of not just agents’ brains but bodies and environments as well, they can be informed by a much wider breadth of empirical research than traditional agent-based models. Moreover, because the simulated bodies of these agents exhibit particular behaviors, they can be more readily compared to the behaviors of actual organisms. I conclude that employing situated cognition approaches to modeling the evolution of cooperation is a more promising way to investigate the evolution of cooperation than more traditional agent-based evolutionary game theory models. (shrink)

Questions surrounding the evolution of cooperation, especially human cooperation, have driven research in many disciplines. Two key methodologies used to research and explain the evolution of cooperation are modeling and narrative construction. A number of scientists and philosophers have suggested that advancing research on the evolution of cooperation will require integrating models and narratives. But, relatively little has been said about what challenges exist to integrating models and narratives, how to go about integrating models and narratives, and what particular benefits (...) might be gained by accomplishing such integration. In this dissertation project, I address these questions in order to show how models and narratives can be integrated to produce better research on and explanations of the evolution of cooperation. The major challenge to integrating models and narratives concerning the evolution of cooperation that I address in this dissertation stems from the criticisms that such models and narratives lack empirical support. I answer this challenge by defending models and narratives from these criticisms. In the case of models, I argue both that at least some models of the evolution of cooperation have stronger empirical support than supposed by critics and that evolutionary robotics modeling approaches can be used to create even better models. In the case of narratives, I argue that several means of defending narratives from evidential challenges already developed by philosophers can be used to defend narratives about the evolution of cooperation as well. To show how narratives and models can be integrated to better research and explain the evolution of cooperation, I first establish an account of integration by drawing on the work of other philosophers. I then argue that models and narratives can provide certain benefits to each other via integration. Models can benefit from integration with narratives by using the narratives as sources of information as well as hypotheses to test. Indeed, as models are used to test such hypotheses, they also gain explanatory power. Narratives can benefit from integration with models by using the models as sources of support and constraint. Moreover, models can help deepen the understanding of historical contingency provided by narratives. (shrink)

I show that Eric Winsberg’s principles of model-building given in Science in the Age of Computer Simulation are insufficient to argue for the external validation of simulation data in cases in which simulation results conflict, and that laboratory experiments have an advantage over simulations because conflicting experimental results can be decided between on the basis of reproducibility. I also argue that robustness of predictions serves the same function for simulations as repeatability does for laboratory experiments in either adjudicating between conflicting (...) results or allowing us to say that we do not have sufficient justification to validate the results. Finally, I argue for an interpretation of the argument from robustness that appeals to the convergence of many well-built and diverse models rather than the more common interpretation which appeals to the probability that one of a set of models is likely to be true. (shrink)