The drive to extend the Web by taking advantage of automated symbolic reasoning (the so-called Semantic Web) has been dominated by a traditional model of knowledge sharing, in which the focus is on task-independent standardisation of knowledge. It appears to be difficult, in practice, to standardise in this way because the way in which we represent knowledge is strongly influenced by the ways in which we expect to use it. We present a form of knowledge sharing that is based not on direct sharing of "true" statements about the world but, instead, is based on sharing descriptions of interactions. By making interaction specifications the currency of knowledge sharing we gain a context to interpreting knowledge that can be transmitted between peers, in a manner analogous to the use of electronic institutions in multi-agent systems. The narrower notion of semantic commitment we thus obtain requires peers only to commit to meanings of terms for the purposes and duration of the interactions in which they appear. This lightweight semantics allows networks of interaction to be formed between peers using comparatively simple means of tackling the perennial issues of query routing, service composition and ontology matching. A basic version of the system described in this paper has been built (via the OpenKnowledge project); all its components use established methods; many of these have been deployed in substantial applications; and we summarise a simple means of integration using the interaction specification language itself.