One Library, 100,000 Rare Books, And A Very Long Scanner Queue

Mass digitisation sounds like a machine problem. It is mostly a judgement problem.

On 9 July 2026 the University of Delaware Library, Museums and Press announced a partnership with the Google Books Library Project to digitise more than 100,000 unique volumes from its collections over the next three years. From summer 2026, staff began sending physical books, in secure batches, to a US digitisation centre.

Every one of those volumes had to be chosen by somebody.

The selection is the story

The stated focus is on works that are rare, out of print, or otherwise unavailable online — and specifically on material that other partner institutions do not already hold. In Delaware's case that means two strengths you would not necessarily guess: "Delawareana", the deep local and regional record of a small state, and historical trade catalogues, those unglamorous listings of what companies made and sold.

Trade catalogues are a good example of why this matters. Nobody reads them for pleasure. Historians of design, labour, agriculture, advertising and material culture read them constantly, because they are one of the few records of ordinary manufactured life — prices, sizes, finishes, what a farm or a kitchen actually contained in a given decade. They are also fragile, unstandardised and rarely reprinted, which makes them exactly the kind of thing that quietly disappears.

Subjects across the wider selection span literature, history, political science, education, art and culture.

What you will and will not be able to read

The access rules are not a detail, so it is worth being precise:

• **Public-domain volumes** become fully readable online once digitised.
• **Books still in copyright** appear as limited preview snippets — searchable, not readable end to end.
• **Discovery works three ways:** through Google Books, through the university's own library catalogue, and through the HathiTrust Digital Library, a not-for-profit collaborative of academic and research libraries described as holding more than 19 million digitised items.

That last point is the preservation argument. A scan that lives in one company's product is a convenience. A scan deposited into a library-run collaborative is closer to an insurance policy.

Trevor A. Dawes, the university's Vice Provost for Libraries and Museums and May Morris University Librarian, said in the announcement that the partnership will make more than 100,000 works available in digital form — a scale of digitisation, he noted, that the library could not have achieved on its own.

Three years is the honest timescale

This is not a switch being flipped. Books travel physically, in batches, and get handled by people. The university had already been building in this direction: a new digital collection of Special Collections print materials launched on 17 June 2026, and its digital collections platform is mid-migration to a new system, with legacy material still hosted elsewhere during the transition.

We could not verify how many volumes have been shipped or scanned so far, and no progress percentage has been published; treat any specific count you see elsewhere with suspicion until the library states one.

Why a small state matters

There is a tidy assumption that everything important is already online, and that what remains unscanned is unimportant. The opposite is closer to true. What is missing from the internet skews local, specialist and administrative — the record of a place rather than a canon.

A university in a small state deciding that its regional record deserves to be globally searchable is a modest act with a long half-life. Somebody in twenty years will find a trade catalogue from a Delaware manufacturer, at two in the morning, from anywhere, and will not think once about the batch it travelled in.

Have a Beaux day.