The 2026 Modernization

In 2026 the Scottish Court of Session Digital Archive became something no reading room can offer: the dispersed corpus of the Session Papers — digitized by institutions in the United States and Scotland — brought together and searchable as a single collection.

A scattered archive

The Session Papers are the printed record of litigation before Scotland's supreme civil court in the eighteenth and early nineteenth centuries: petitions, answers, memorials, informations, proofs, and the maps and plans that accompanied them. Printed in small runs for the use of the judges, they survive today in bound sets and loose gatherings scattered across research libraries on two continents. The University of Virginia Law Library holds one of the most significant collections outside Scotland; the University of Edinburgh and its partner libraries hold the great Scottish sets; further volumes rest at the Library of Congress.

That dispersal has always shaped — and limited — what scholars could do with the papers. A case's petition might be shelved in Charlottesville and the answers to it in Edinburgh. No catalog spanned the holdings. To follow one dispute through its paper trail could mean correspondence with three institutions; to ask a question across the whole corpus — every case touching the tobacco trade, every paper naming a place in Jamaica — was simply not possible.

The 2026 modernization set out to remove that limit. The archive now unites the digitized Session Papers of the University of Virginia Law Library, the University of Edinburgh, and the Library of Congress in one system: 189,205 page images preserved in a versioned dark archive, described by a catalog of 5,512 documents across 7,498 cases, naming 4,835 people and 2,205 places — and growing as review proceeds.

What the modernization delivers

One collection. Documents digitized on both sides of the Atlantic appear in one catalog, under one search, with one set of stable identifiers. The archive treats the corpus as the court produced it: a single body of records.
Every printed page, readable by machine. Layout-aware machine transcription makes the collection searchable at the page level. A query finds the sentence, not just the shelfmark — a transformation for sources that were, until now, discoverable only through case-level catalog entries.
New ways in. Beyond full-text search: an interactive map that plots the world the papers describe, from Fife farm towns to Caribbean plantations, over historic map layers; a network explorer tracing which people, places, and subjects each cause connects; and browsing by Scotland's historic counties.
A machine-readable archive. For computational researchers, a Model Context Protocol endpoint lets research software query the collection by meaning. See Searching the Archive.
Standards-based image delivery. Page images are served through the University of Virginia Library's IIIF infrastructure — the same interoperable framework used by research libraries worldwide, meaning any IIIF-aware viewer or tool can work with the archive's images.
Open methods. The computational work runs on open-weight, open-source models, on infrastructure the University controls, with every machine proposal reviewed by human catalogers. The approach — and the reasons for it — are set out in Computational Methods.

A living catalog

The numbers above are a snapshot, not a boundary. Cataloging is a living process: the computational pipelines propose records — a case determination, a person's role, a place identification — and the Law Library's catalogers confirm, correct, or reject each one, working from queues that show the evidence behind every proposal. Newly confirmed cases, people, and places appear on the site as they are approved. Corrections do double duty, teaching us where the pipelines fall short and making the next round of proposals better.

The same is true of coverage. Machine transcription continues across the corpus; place references are resolved to Scotland's historic geography as curation proceeds; and the connective tissue — case-to-document links, similar-case suggestions, the network graph — thickens as the catalog grows.

Built to last

Everything above rests on a preservation-first foundation. Before any computational work began, every source image was secured in a versioned dark archive, independent of any processing system, model, or vendor. The catalog's identifiers are permanent: every document, case, person, and place has a stable address that will survive future redesigns of the site. The derived layers — transcription, extraction, indexes — can always be rebuilt from the preserved evidence; the evidence itself is kept safe.