Running commentary · August 2026
Permitted, Not Yet Measured: Paradata in the Data Space
The paper
Management team and Julia Schellenberg, Europeana Foundation (2026). Deployment of a common European data space for cultural heritage: PM. Periodic report M45. CNECT/LUX/2021/OP/0070, European Commission.
A commentary on the M45 periodic report of the common European data space for cultural heritage, written from the perspective of a national aggregator.
The M45 report records something national aggregators have been waiting for. The EDM schema and the Metis ingestion pipeline now support fields allowing metadata and paradata to be included at source, published in the dataset, and served through the APIs. A workflow to publish provenance information has been established, stamping records with the activities applied to them. EUreka3D-XR has delivered its paradata and sustainability guidelines with Europeana Foundation involvement. Taken together with the scale of Europe's investment in 3D digitisation, this is real and welcome progress, and the Foundation deserves credit for building the carriage before it was formally required to.
I would put the underlying principle plainly. Without certified paradata, a 3D acquisition is a picture of a thing rather than evidence about it. It can be viewed, but it cannot be relied upon, compared, revisited or defended. Reusability without warrant is a weaker promise than it appears.
Which is why the next step matters. Paradata is now permitted by the infrastructure; it is not yet measured by it. The Europeana Publishing Framework tiers assess resolution, rights and descriptive completeness, and the KPIs follow accordingly. A record with full capture paradata and one with none currently score alike. Provenance stamping, valuable as it is, documents what the aggregator did; the chain begins downstream of the digitisation event. And with external API traffic averaging 36.8 million requests a month, a 59% year-on-year rise the report attributes to AI crawling, data is leaving the infrastructure at a scale that makes the question urgent rather than academic.
Several things follow: a measured paradata dimension within the quality frameworks; capture-time obligations written into national digitisation contracts, not merely ingestion-time permission; formal ontological alignment, which the emerging work with ECHOES on the Heritage Digital Twin Ontology makes newly tractable; provenance that accrues in both directions, so that reuse and correction have somewhere to go; a CARE instrument to match the CARE commitment; and eventually fields for interpretation and uncertainty, which do not yet exist in any form.
It is worth noting that the EU Horizon project HERITALISE appears in this reporting, with the Foundation confirming feedback given on its state-of-the-art survey which i contributed to, a sign that the evidence base for exactly these questions is being built in dialogue rather than in parallel.
None of this is a demand made of Europeana alone. Aggregators should not ask the data space to require what we are not yet reliably able to supply. The infrastructure has moved first. It falls to us to fill it.
This is a personal reading by Anthony Cassar, PhD research fellow on the Memory Twin Framework. The views expressed are my own.