The Matrix: Connecting Time and Space in archaeological stratigraphic records and archives

Abstract

The records of archaeological stratigraphic data and the relationships between stratigraphic units are fundamental to understanding the overall cohesiveness of the archaeological archive of an excavation. The information about individual units of excavation identified on sites with complex stratigraphy is most often held in the site database records and stratigraphic matrix diagrams, usually documenting relationships based on the laws of stratigraphic superposition and the Harris matrix conventions (Harris 1979). However, once the matrix diagram has been used to record the information during excavation, there is far less consistency in how those stratigraphic records, and any associated phasing information, are finally deposited in the archives. For that valuable data to be successfully identified and re-used (particularly if the rest of the data is in a database), the stratigraphic and phasing data needs to be in a format that can be interrogated as part of the database. In practice, often only a (paper) copy of the matrix diagrams make the archive. This means that the critical temporal and spatio-temporal relationships upon which the phasing of sites is built, cannot usually be interrogated or (re)used without lengthy and wasteful re-keying of that data into another version of the database. The stratigraphic, sequencing and temporal information held in a matrix is fundamental in further studies of the site records and in working out how the site may relate to other excavated sites of similar or related dates and phases. This article will suggest ways in which the stratigraphic data from excavations and the reasoning processes carried out in subsequent analysis could be better managed, to make matrices (re)useable as part of a more integrated digital archive. This article examines how conceptual reference modelling, particularly using temporal relationships, can be used to explore these issues and how associated technologies could enable semantically-enriched deductions about the spatio-temporal and purely temporal relationships that fundamentally link archaeological data together. It will also consider where further work is needed both to deal with analysis of spatial or temporal records and to enhance Bayesian chronological modelling and associated temporal reasoning, and how this may form the basis for new linkages between archaeological information across space-time.