This paper considers contractarianism as a method of justification. The analysis accepts the key tenets of contractarianism: expected utility maximization, unanimity as the criteria of acceptance, and social-scientific uncertainty of modelled agents. In addition to these three features, however, the analysis introduces a fourth feature: a criteria of rational belief formation, viz. Bayesian belief updating. Using a formal model, this paper identifies a decisive objection to contractarian justification. Insofar as contractarian projects approximate the Agreement Model, therefore, they fail to justify their conclusions. Insofar as they fail to approximate the Agreement Model, they must explain which modelling assumption they reject.

Following the work of John Rawls, political theorists have fixated on the comparative stability of different equilibrium states of justice. This article identifies a crucial gap in this literature, namely, the lack of attention paid to nonequilibrating systems. Drawing on Kakutani's theorem, I present plausible cases in which society will fail to exhibit any equilibrium states of justice. An important implication for political theory is that, rather than focusing exclusively on stable equilibria, theorists should examine dynamic processes of justice, which need not exhibit equilibrating forces. To this end, a concept more useful than stability is that of robustness, or the ability of a system to maintain general desiderata in the face of an evolving conception of justice.

This paper contributes to the literature on how diversity impacts groups by exploring how communication mediates the ability of diverse individuals to work together. To do so we incorporate a communication channel into a representative model of problem-solving by teams of diverse agents that provides the foundations for one of the most widely cited analytical results in the literature on diversity and team performance: the “Diversity Trumps Ability Theorem”. We extend the model to account for the fact that communication between agents is a necessary feature of team problem-solving, and we introduce the possibility that this communication occurs with error, and that this error might sometimes be correlated with how different agents are from one another. Accounting for communication does not give us reason to reject the claim associated with the theorem, that functionally diverse teams tend to outperform more homogeneous teams (even when the homogeneous teams are comprised of individuals with more task relevant expertise). However, incorporating communication into our model clarifies the role that four factors play in moderating the extent to which teams capture the benefits of functional diversity: i) the complexity of the problem, ii) the number of available approaches to solving the problem, iii) the ways of encoding or conceptualizing a problem, and iv) institutional characteristics, such as how teams work together. Specifically, we find that whether (and to what extent) teams capture the benefits of functional diversity depends on how these four factors interact with one another. Particularly important is the role institutional dynamics (like search methods) play in moderating or amplifying interpersonal frictions (like miscommunication), and notably we find that institutions that work in one setting can be counterproductive in other settings.

A strength of liberal political institutions is their ability to accommodate pluralism, both allowing divergent comprehensive doctrines as well as constructing the common ground necessary for diverse people to live together. A pressing question is how far such pluralism extends. Which comprehensive doctrines are simply beyond the pale and need not be accommodated by a political consensus? Rawls attempted to keep the boundaries of reasonable disagreement quite broad by infamously denying that political liberalism need make reference to the concept of truth, a claim that has been criticized by Joseph Raz, Joshua Cohen, and David Estlund. In this paper, we argue that these criticisms fail due to the fact that political liberalism can remain non-committal on the nature of truth, leaving the concept of truth in the domain of comprehensive doctrines while still avoiding the issues raised by Raz, Cohen, and Estlund. Further substantiating this point is the fact that Rawls would, and should, include parties in the overlapping consensus whose views on truth may be incoherent. Once it is seen that political liberalism allows such incoherence to reasonable parties, it is clear that the inclusion of truth and the requirement of coherence urged by Raz, Cohen, and Estlund requires more of reasonable people than is necessary for a political consensus.

No abstract