Dr Katharina Helming (Department of Philosophy at Leipzig University)

Abstract

Common knowledge, that is the ability to know things together, underlies uniquely human forms of coordination. However, there is a puzzle in the existing literature about how to characterize this kind of knowledge. Traditionally philosophers propose that common knowledge requires the ability to compute higher-orders of embedded knowledge states (i.e. I know that you know that I know that you know and so forth at infinitum that X). However, more empirically oriented accounts suggest that recursive mindreading is cognitively too demanding and thus not a plausible basis for human coordination. Instead, common knowledge is said to be achieved either on the basis of short-cuts or an arbitrary cut-off within the spiral of embedded knowledge states. But on such accounts, it remains an open question if and when (after how many steps) certainty about knowing things together can be achieved. The aim of this talk is to propose a new solution to this debate. By drawing on empirical literature across fields such as developmental, social and comparative psychology as well as linguistics I will investigate how people actually solve coordination tasks. A review of the literature shows that there is a reoccurring structure of embedded knowledge states that people use to cooperate successfully in different areas. More specifically, empirical data suggests that people actively generate three levels of embedded knowledge coordinate. Such tertiary structures are known as a “three way handshake” in computer science and can be redescribed as a mutual acknowledgment about content in social interactions. It is concluded that certainty about knowing things together in coordination is achieved based on a non-arbitrary cut-off within embedded knowledge states. The broader implications for the evolution of social cognition are discussed on this basis.