The Limits of Egoism Are Not the Limits of Individualism

Two agents sharing the intention to do something are interlocked in a web of individual intentions with special social contents. This individualistic approach to shared intentional activity has been originally proposed by Bratman (1993). I will suggest that Bratman’s ‘shared intentions’ are of little help when interdependency of agents’ reasons can threaten coordination as in common-interest games with multiple equilibria. In particular, in Hi-Lo games (coordination games with at least two equilibria one of which is Pareto superior), it has been argued that agents fail to coordinate when they reason from an individualistic perspective. In this article, an explanation of individualistic choice in Hi-Lo situations is proposed. The propensity of agents to choose Hi is interpreted as the output of both a preference and a reasoning effect. Choosing Hi presupposes both social preferences and a framing of the situation as a collaborative choice requiring usual individual practical reasoning (although social in its scope). Agents that can be motivated by an other-regarding cognitive process of goal adoption have rational grounds both to infer a shared intention in the first place and then to specify it by generating new shared intentions as time goes by. Sharing an intention then is intending together to do an action. When we intend together, we intend that we do an action in a collaborative way.