Search this site

Elliott Wagner

I am a graduate student in the Department of Logic and Philosophy of Science at the University of California, Irvine. 

A copy of my CV is here.  My dissertation abstract and research statement is available here.

I can be reached at elliottw@uci.edu.

Publications

“Evolving to divide the fruits of cooperation.” Philosophy of Science. Forthcoming.  [penultimate draft]
Social contracts are often modeled by either stag hunt games or bargaining games, but it is more realistic to use a combined game in which players first choose whether or not to cooperate and then choose how to distribute the benefits generated by cooperation.  In this paper I show that the combined game has different properties than either of the single games taken independently.  In particular, both cooperation and equal division are more likely outcomes of the combined game than of either individual games.

“Deterministic chaos and the evolution of meaning.” The British Journal for the Philosophy of Science. 2011. [preprint, official]
Communication is often thought to be impossible in signaling games in which the interests of the sender and receiver are opposed.  The justification for this claim is that no information is transferred at the equilibria of zero-sum signaling games.  But in this paper I show that a standard evolutionary dynamic does not lead to equilibria in these sorts of games.  Instead, the dynamic leads to persistent out-of-equilibrium play in which signals transfer partial information.

“Communication and structured correlation.” Erkenntnis. 71 (2009): 377–393.  [official]
Signaling games have many types of inefficient equilibria. These inefficient equilibria are sometimes likely outcomes of large-population evolutionary or learning dynamics.  But in this paper I show that if players are located in a network and play only with their neighbors, then these equilibria are destabilized.  Moreover, small-world networks are especially favorable for the evolution of information transfer.  Inefficient equilibria are destabilized by these networks, and Pareto optimal strategies quickly spread throughout the population.

Works in Progress

“The long-run stability of collective action.” [draft]
It is often thought that the inefficient hare hunting equilibrium is the plausible long-run outcome of stag hunt games.  It is the risk-dominant equilibrium, and in two-player games the risk-dominant equilibrium is the unique stochastically stable equilibrium.  In this paper I show that such reasoning does not extend to N-player stag hunts.  When N>2, the cooperative stag hunting equilibrium can be the unique long-run equilibrium of the stochastic replicator dynamic.  This demonstrates an important difference between two-player and $N$-player games and suggests an explanation of the evolution of large-scale cooperation seen in nature.

“The dynamics of costly signaling.” [draft]
In 1973 Michael Spence proposed what has become a canonical model of signaling.  He showed that certain costly signals can honestly reveal information even when the sender's and receiver's interests slightly diverge.  Spence's game has an infinite number of equilibria, but economists often claim that only one (the Riley equilibrium) is a plausible outcome of the game.  In this paper I show that this equilibrium is a very unlikely outcome of large-population learning dynamics.  Much more likely is convergence to a pooling equilibrium or to a type of mixed equilibrum that is largely ignored in discussions of Spence's model.

“Divergent interests and the evolution of inference.” [draft]
In models of human and animal signaling it is often assumed that the players receive identical payoffs and that there are only two players in the game.  In this paper I explore the possibilities if both of these assumptions are simultaneously weakened.  I show that in cases when the interests of the players diverge and there two or more senders, then a behavior that resembles rudimentary reasoning or inference emerges.  The receiver learns to put together the partial information she receives from each sender and infer what the true state of the world is.

Report Abuse|Remove Access|Powered By Google Sites