From 31st March to 4th April 2026, the annual Computer Applications and Quantitative Methods in Archaeology conference (CAA) took place in Vienna. ESTER was represented in Session 49, “Bridging Micro and Macro Perspectives in the Modelling of Past Human Ecosystems,” where Martin Hinz and Sophie C. Schmidt presented the paper “Generative Inference and Reflexive Modelling: Bridging Scales in Socio-Ecological Archaeology.” The talk was given on Wednesday, 1 April 2026, from 10:30 to 10:50.
The CAA talk placed ESTER in a broader methodological and theoretical context. It addressed a central problem in contemporary archaeology: how to connect highly uneven local observations with larger regional and continental patterns without flattening the uncertainties, distortions, and scale effects that shape the archaeological record. 
Our contribution argued that generative inference offers a productive way to think across these scales. Rather than treating models as descriptive summaries alone, this perspective understands them as explicit hypotheses about the processes that produced the data we observe. In that sense, modelling becomes an iterative and reflexive practice: one that remains grounded in empirical evidence while making uncertainty, assumptions, and the structure of inference itself more transparent.
The paper also highlighted that bridging micro- and macro-perspectives does not mean choosing between inference and simulation. Instead, it suggested a broader model ecology across scales, in which different modelling approaches can complement one another. For ESTER, this is especially important, because the project seeks to integrate site-level and regional evidence within a common inferential framework while remaining sensitive to local variation, proxy-specific uncertainty, and the uneven visibility of the archaeological record.
We are very grateful to the organisers of the session for creating such a stimulating context for discussion, and to the audience for the thoughtful engagement with questions of scale, uncertainty, and epistemic transparency in archaeological modelling. For ESTER, the session offered an excellent opportunity to present the project not only as a demographic endeavour, but also as a contribution to wider debates on how archaeological knowledge is generated through models.