Evolution of Cooperation[1]: Recursive Belief and Common Ground

This series is a set of reading notes on Michael Tomasello’s work on the evolutionary origins of human communication and cooperation. The original notes were written in Chinese; this English version was translated and lightly polished with the help of Claude.

Tomasello (hereafter, T) argues that human “communication” rests on two foundations: first, a mutually assumed conceptual ground shared by both parties; second, mutually assumed cooperative communicative motives. What does “mutually assumed” mean? It refers to a recursive belief: “I know A, I know that you know A, I know that you know that I know A, …” T illustrates both points with an everyday example:

You and your ex-boyfriend broke up on bad terms. You and a friend walk up to the library, and your friend points to a bicycle at the entrance — your ex-boyfriend’s bicycle. The two of you then decide not to go in.

Your friend merely pointed at a bicycle. The act of pointing at a bicycle carries no information in itself: if your friend had pointed at any random bicycle, you would have been confused and had no idea what they were doing. The gesture means something because (1) your friend knows that this is your ex-boyfriend’s bicycle, and your friend knows that you know this too; (2) your friend wants to help you, and your friend knows that you know they want to help you.

Note that, for the first point, “I know” alone is not enough — what is also needed is “I know / I assume that you know.” For the second point, this kind of motive is unique to humans: a great ape will not engage in communication “for the benefit of another.” For instance, even when a female ape knows where a separated infant’s mother is, she will not tell the infant. In Tomasello’s words, “human communicative motives are so uniquely cooperative” — they have no real parallel in the animal kingdom.

Note that none of this involves language. It only involves the simplest form of communication — pointing with a finger.