Reconcilable differences: Using retrospective photogrammetry to bridge the divide between analogue and digital site data collected during long-term excavation projects.

Over the last 30 years, high-resolution site documentation has rapidly developed, with analogue drawings and film photography being replaced with high-precision digital recordings. Today, most archaeological field data sets are produced using digital tools that store spatial and visual information in various digital formats directly, i.e., born-digital. A fully digital workflow makes the process of combining, comparing, and integrating field datasets quicker, easier, and potentially more analytically powerful. However, at sites where both analogue and born-digital data sets have been produced, additional procedural digitization steps are required before full data interoperability is achieved. In cases where the archaeological sites have a long excavation history, multiple generations of analogue and digital site documentation techniques have often been used, making it particularly challenging to physically reconstruct an excavated site based on its archival material. The Middle Stone Age site of Blombos Cave, South Africa, is a prime example of this type of challenging situation. This site features a more than 3-meter-deep and well-preserved archaeological sequence dated to between 300 and 100 000 years ago. Since it was initially excavated in 1991, multiple archaeological campaigns have been carried out (>15), and the excavations are still ongoing. The field documentation from Blombos Cave has, over the years, produced varied but rich datasets that have never been integrated into a single, coherent, and accessible archive. In this paper we evaluate the changes in excavation protocol at Blombos Cave over time, and we use this knowledge to digitally integrate and map the various stages of excavation within a three-dimensional framework using digital photogrammetry and archival photographs. The archaeological and analytical value of this approach is exemplified through multiple case studies, in which we demonstrate how and why the merging of old and new archaeological field data can lead to new results, specifically by offering more complete mapping and more accurate and analytically dynamic visualisations. The research history at Blombos Cave is not unique or site-specific. Our approach would be applicable to a wide variety of sites and contexts where long-running excavations have produced a mix of analogue and digital field data.