A prominent explanation of the evolution of altruism is ‘indirect reciprocity’ where the tracking of reputations in a population promotes altruistic outcomes. This paper investigates the conditions under which the meaning of reputation-tracking signals can co-evolve with altruistic behaviours. Previous work on this question suggests that such a co-evolution is unlikely. In our model, we introduce a mixture of direct and indirect information: individuals directly observe the actions and signals of others with some probability, rather than individuals always relying only (...) upon information conveyed through signals. In this context, we see the co-evolution of altruism and the kind of ‘moral signals’ required for individuals to track reputation. Our interpretation of this model is that, although individuals can sometimes observe the behaviour of their fellows directly, they cannot always do so, and signals evolve to fill this gap. Although ‘can’ does not imply ‘must’—that our model can produce this co-evolution does not establish that we have found the only possible mechanism—we nonetheless consider this an interesting possibility result worthy of further investigation. (shrink)

In A Treatise of Human Nature, Hume argues that morality pertains primarily to character, and that actions have moral content only to the extent that they signal good or bad character. I formalize his signalling theory of moral/immoral actions using simple game-theoretic models. Conditions exist under which there is a separating equilibrium in which actions do indeed credibly signal character, but conditions also exist in which there is only a pooling or semi-separating equilibrium. A tradeoff is identified between the signalling (...) value of actions, and the consequentialist goal of incentivizing all character types to choose beneficial actions. (shrink)

In explaining the emergence of conventions surrounding human cooperation and helping of those in need, it seems as though honest communication of need is an essential part of the story. While previous results indicate that punishment promotes cooperation, this paper will argue that the story is more complicated. Namely, whether punishment promotes cooperation depends on what you punish. Punishment of those who lie about their need for a resource may instead impede cooperation, as the attempts to deceive that arise in (...) cooperative endeavors may be too costly to make cooperation worthwhile. (shrink)

Recognition of and responsiveness to the behavioral dispositions of others are key features of moral systems for facilitating social cooperation and the mediation of punishment. Here we investigate the coevolutionary possibilities of recognition and conditional social behavior with respect to both altruism and spite. Using two evolutionary models, we find that recognition abilities can support both spite and altruism but that some can only coevolve with spite. These results show that it is essential to consider harmful social behaviors as both (...) a product of and an influence on the core features of our moral systems. (shrink)

The objective of this work is to demonstrate how cooperative sharers and uncooperative free riders can be placed in different groups of an electronic society in a decentralised manner. We have simulated an agent-based open and decentralised P2P system which self-organises itself into different groups to avoid cooperative sharers being exploited by uncooperative free riders. This approach encourages sharers to move to better groups and restricts free riders into those groups of sharers without needing centralised control. Our approach is suitable (...) for current P2P systems that are open and distributed. Gossip is used as a social mechanism for information sharing which facilitates the formation of groups. Using multi-agent based simulations we demonstrate how the adaptive behaviour of agents lead to self-organisation. We have tested with varying the gossip level and checked its impact in the system’s behaviour. We have also investigated the impact of false gossip in this system where gossip is the medium for information sharing which leads to self-organisation. (shrink)

Robustness analysis is widespread in science, but philosophers have struggled to justify its confirmatory power. We provide a positive account of robustness by analysing some explicit and implicit uses of within and across-model robustness in evolutionary theory. We argue that appeals to robustness are usually difficult to justify because they aim to increase the likeliness that a phenomenon obtains. However, we show that robust results are necessary for explanations of phenomena with specific properties. Across-model robustness is necessary for how-possibly explanations (...) of multiply instantiated phenomena, while within-model robustness is necessary for explanations of phenomena with multiple evolutionary starting points. In such cases, the appeal of robustness is explanatory rather than confirmatory. (shrink)