I’m particularly interested in the human propensity to copy behaviours that lack any kind of knowable causal structure. This is how we learn arbitrary conventions—and I think it originates in a distinctively human way of building group identities.