With the multiplication of excavations, archaeological research (and even rescue archaeology) increasingly faces sites previously explored and thereby has to deal with documentation made with different methods and of different scopes than those of today. Beyond the historiographical issues and differences in ways of describing the past as viewed from the present, such an archival documentation is generally fragmentary, often scattered between various institutions or in private archives. The aim of this paper is to contribute to a debate on an important issue: how might we best make use of previous material, from archival documents to publications?

Herculaneum can exemplify the main processes one might face while studying a site excavated in the past. Initially explored through tunnels, it was excavated in open-area excavations in four periods (1828-55; 1869-75; 1927-61; 1996-98). Most of the city as we know it was uncovered by A. Maiuri (1886-1963) while he was Soprintendente and director of the Museum of Naples (1924-61). He managed to pass himself off as the only excavator of Herculaneum, and such a view was consolidated by the (incomplete) publication of his excavations in 1958. Up until the late 1990s, anything written about Herculaneum relied on his synthesis, with only a few re-interpretations based on observations made on a site that had been restored and presented by him. S. Mols was the first to use other sources to understand better the context of the wooden furniture discovered throughout the town, namely the daybooks (Giornale degli scavi di Ercolano [henceforth GSE]) composed during the excavation. Since then, slowly but surely, many of those involved in the study of Herculaneum started using those diaries, hoping to have a more complete and less subjective perception of the excavation — at least, one less influenced by Maiuri.