Archaeological datasets are increasingly visible, digital, and openly shared, yet they remain difficult to use in large-scale analysis within and beyond archaeology. This matters because other research communities, especially in the climate and Earth system sciences, are seeking long-term human-impact baselines: signals of land use, settlement dynamics, and socio-ecological change that can complement palaeoenvironmental archives. In practice, much archaeological data does not find its way to these research communities. It is encoded in discipline-specific categories, unevenly documented, spatially and temporally heterogeneous, and accompanied by uncertainty that is hard to carry into downstream analyses. Archaeology therefore risks being simultaneously highly relevant and operationally unusable for interdisciplinary synthesis. Large-scale questions with broad relevance require something different: not just ingredients, but a meal that is ready to be served and eaten.

This paper argues that bridging the gap is not primarily a matter of more data, but of archaeological synthesis understood as translation across communities of practice. We must deliver the pizza, not the pantry. Making archaeological evidence reusable requires treating interoperability, uncertainty, and bias as first-class research products. Concretely, this means a three step workflow: 1. curate data with provenance and machine-readable metadata so others can assess fitness for use, 2. formalise preservation loss, sampling intensity, and infrastructural inequalities that shape what becomes visible, and 3. produce harmonised outputs with credible intervals and documented assumptions that can be integrated into interdisciplinary workflows. Also within the field such practices will facilitate large scale meta-analyses.

We illustrate these points with an ongoing demographic synthesis effort that consolidates multiple archaeological proxies within a transparent Bayesian framework and produces reusable estimates plus diagnostics, a nutrition label that collaborative networks can rely on. The take-home message: interdisciplinary relevance depends on whether we deliver results in forms that other communities can immediately use, test, and build upon.

First Author: Martin Hinz

Co-Authors: Sophie Schmidt

Keywords:
data interoperability
knowledge translation
uncertainty and bias modelling
human impact baselines
transdisciplinarity