In a semantic network, a context may be defined as the shared neighbourhoods of two related concepts. A concept therefore has many potential contexts, each one defined by a relationship to another concept and weighed by the overlap of their semantic neighbourhoods. We can use the growth of neighbourhood intersection around weak ties in semantic networks as a signifier of emergent context. While this novelty index may be sensitive to the embeddedness of the data set it goes some way to isolating the friends from the acquaintances, which hopefully brings us a little more coherence…