This is a preview. Log in through your library . Abstract This paper proposes a tractable model of Bayesian learning on large random networks where agents choose whether to adopt an innovation. We ...