Ric Sims

Minimal joint action is action jointly undertaken without prior planning or contentful intentions. Imagine a person lifting a heavy table upstairs. Her companion, seeing her struggling, rushes to her aid and picks up one end, and they both perform the task together. Despite there being no prior plan nor shared intentions, traditionally understood, this seems to be a perfectly legitimate case of joint action, and one that is not much discussed in the literature. The paper examines cases like this and proposes that there are kinds of joint action that can be analysed in terms of a teleosemiotic approach where coordination of individual agents is achieved through an environmental variable that depends on the results of their actions – stigmergic coordination, rather than through the content of intentional states. We leverage the fact that the mechanism of stigmergic coordination of actions through environmental signs takes place on two levels: that of individual agents and the group itself – hence establishing a notion of joint action at the group level. Normativity is introduced through a set of long-term broadly evolutionary processes – whether biological or cultural, which are responsible for the ontogenesis of tasks, their coordinative signs, and the agents that consume them.

This paper outlines an approach to analysing minimal cognition that brings out its social and historical dimensions. It proposes a model, the coordinated systems approach (CSA), which understands cognition as a coordinated coalition of loosely autonomous processes responsible for goal-directedness in a system. On this view, even individual cognition has something of a social flavour to it. The central concept of the paper is stigmergy: a process where the material trace of actions of system elements in their environment is a sign that coordinates a group of semi-autonomous processes in future actions – this is the social dimension. The historical dimension refers to longer term processes which establish the coordinative power of the sign and endow it with normative force. According to this proposal, a full explanation of cognitive capabilities should reference both dimensions. In the second half of the paper the CSA is let loose on some puzzles in 4E cognition. Can the model deal with old problems such as that of cognitive bloat, or new problems such as the supposed external memory of the slime mould Physarum polycephalum? Potentially, the approach could be used to analyse minimal cognitive phenomena over a range of scales from bacteria to human beings.