Long-distance communication has emerged as a particular focus for archaeologicalexploration using network theory, analysis, and modelling. The promise is apparentlyobvious: communication in the past doubtlessly had properties of complex, dynamicnetworks, and archaeological datasets almost certainly preserve patterns of thisinteraction. Formal network analysis and modelling holds the potential to identify anddemonstrate such patterns, where traditional methods often prove inadequate. Thearchaeological study of communication networks in the past, however, calls for radically different analytical methods from those employed by most other forms of social network analysis. The fragmentary archaeological evidence presentsresearchers with the task of reconstructing the broken links of a ruined network fromobservable distributions and patterns of association in the archaeological record. Informal terms this is not a problem of network analysis, but network synthesis: theclassic problem of cracking codes or reconstructing black-box circuits. It is proposedthat archaeological approaches to network synthesis must involve a contextualreading of network data: observations arising from individual contexts, morphologies,and use patterns. This point is demonstrated with reference to a study of Viking-period communication in the North Sea region