Modeling 19th century quadrilingual networks of publishers and publications as multidigraphs
Аннотация
This article advances a network-based approach to translation and book history by modeling nineteenth-century European and transatlantic publishing as a multilingual, relational system. Using the case of Alphonsus Liguori – a widely translated Catholic author – we construct two parallel complex networks: a quadrilingual publisher network (English, French, Spanish, German) and a title-centered network linking translated works through shared publishers, chronological precedence, and edition counts. Both are implemented as weighted multidigraphs that integrate previously separate monolingual datasets with newly encoded cross-lingual correspondences, allowing analysis of circulation patterns, centrality, and community structure. Network analysis reveals a strongly hierarchical yet multilingual structure. A small set of titles occupies the highest centrality positions – with Visits to the Most Holy Sacrament standing out particularly strongly in the modeled title network – functioning not only as bestsellers but as infrastructural hubs that organize publication and translation pathways across languages. Publisher-level analysis shows that monolingual prominence does not map directly onto international structural prominence: multilingual presses gain central positions once cross-lingual correspondences are introduced. Community detection likewise identifies linguistically heterogeneous clusters containing many of the highest-ranked publishers; it also shows that the most influential clusters are also the most linguistically heterogeneous. These results are suggestive of a persistent multilingual structure, while remaining conditional on the construction of the networks, the title-matching procedure, and the inevitably uneven coverage of the underlying bibliographic data. Methodologically, the study demonstrates the value of multidigraph modeling for digital translation history. The parallel title- and publisher-level networks reveal patterns of connectivity, clustering, and centrality that remain difficult to see in monolingual or bibliographic views. The findings are nevertheless interpreted as evidence about the modeled corpus rather than as a complete reconstruction of nineteenth-century publishing.