一项新的研究 自然的人类行为 by 黄金城官网 Department of Network and Data Science 助理教授 费德里科•Battiston and a team of international colleagues sheds new light on the importance of social networks to understand prosocial behavior. Battiston’s research emerges from the inquiry: If a self-oriented option appears the most rewarding, 为什么要合作呢?
通过应用来自 new theory of 'networks beyond pairwise interactions', Battiston and colleagues generalize the framework of evolutionary game theory on networks beyond dyadic games, proposing a new theory to boost cooperation in real, large and structured human societies.
Battiston notes that large-scale human cooperation is a fascinating puzzle from day-to-day interactions to greater endeavors such as volunteering to enlist for war. In such cases, humans routinely put aside individual interest in favor of a common good. Past pioneering work has shown the importance of kinship relations to sustain prosocial behavior in small-scale societies, showing that one is willing to pay a cost not only for one’s own benefit, but also for that of a person who shares one’s genes. 这一机制, 然而, is not able to provide a satisfactory answer to cooperative phenomena in much larger societies, 例如组织或国家.
As one lives and cooperates in networks, a foundational pillar of large-scale cooperation is driven by the structure of one’s social relationships. Scientists have shown that repeated games between the same pairs of individuals support the creation of robust mutual interactions based on trust even among unrelated individuals, despite that the temptation to defect would prove to be more rewarding in a single individual round. The discovery of this mechanism almost 30 years ago, 被称为网络互惠, spurred significant further research at the boundary of evolutionary game theory and network science. Battiston expands on the idea that the exact architecture of one’s social network can significantly affect prosocial behavior. 例如, with a highly heterogenous distribution of social contacts, the increased clustering between individuals, or of shortcuts in the network can significantly boost cooperation in human populations.
到目前为止, most investigations have focused so far on dilemmas where interactions are limited to pairs of individuals connected by the links of a network, such as the “prisoner's dilemma”. Battiston and colleagues go further, generalizing the framework of evolutionary game theory on networks beyond dyadic games, therefore becoming relevant to cases which occur at the level of groups. 这是, 例如, the scenario of taxes for welfare state, which are beneficial from an individual perspective only if most individuals of a population are willing to contribute, and the numbers of free riders is limited. Such more complex dilemmas are typically described by public goods game. 传统网络, intrinsically limited to dyadic interactions among agents fall short to provide an adequate representation of these 'higher-order' interactions, where larger groups of people interact. 然而, the use of different mathematical representations such as hypergraphs successfully capture the richness of these systems. 以这种方式, Battiston and colleagues were able to formally describe the evolutionary dynamics of higher-order interactions, and unveil how going beyond the dyad can boost prosocial behavior in our society.
阅读这项新研究 在自然中人类行为, 1月4日, 2020, on boosting cooperation with higher-order interactions.