When a manuscript from a medieval scriptorium is opened for the first time in several centuries, the ink is the object most people notice. Paper archaeologists pay equal attention to the page that carries it. The fibres, the surface finish, the watermarks, and even the microscopic patterns left by a drying rack can reveal where, when, and how the paper was made. Over the last three decades, this unassuming branch of history has grown into one of the most productive partnerships between museum laboratories and traditional humanities scholarship.

The oldest European papers, arriving from Islamic mills in Spain around the eleventh century, were made by pounding linen rags into a wet pulp, scooping the pulp onto a screen, pressing out excess water, and drying the resulting sheet against a flat surface. A paper archaeologist can often identify such a sheet by the grid of chain-lines left by the screen, a pattern invisible to the naked eye but clearly visible when the paper is held against a strong light. Chain-line spacing varied from mill to mill, and a careful comparison with reference samples can often narrow the date and origin of an undocumented sheet to within fifty years and a few hundred kilometres.

Watermarks offer even more precise evidence. From about 1280 onwards, Italian paper-makers began twisting thin wires into shapes - a hand, an ox head, a three-pointed crown - and attaching them to the screen. Each sheet made on that screen carried a faint impression of the design. Because the screens wore out quickly and were replaced with slightly different versions, the precise form of a watermark can often be matched against dated archives to fix the year of production. The Italian scholar Paola Conti has compiled a reference database of more than eighty thousand watermarks, and a sheet of previously unknown provenance can now be identified in minutes by computer comparison.

The materials themselves record a further layer of history. Medieval European paper was made almost entirely from cotton and linen rags, which had entered the paper mills as worn-out clothing. Because the global supply of rags was limited and the demand for paper grew steadily, the nineteenth century saw a long crisis during which mill owners experimented with alternative fibres - esparto grass, straw, cornstalks - before settling on chemically treated wood pulp. Wood-pulp paper browns and becomes brittle within a century because of its residual acidity, whereas a rag-based paper of 1500 is typically still strong and flexible today. The shift in raw materials is therefore of more than antiquarian interest: it explains why many books from the late nineteenth century are now in worse condition than manuscripts five times older.

The surface of a sheet carries its own record. Before paper could be written on with a quill pen it was sized with a thin layer of animal gelatine, which prevented the ink from spreading into the fibres. The kind of animal glue used, and the way it was applied, varied from one region to another. In a recent study of sheets from a fifteenth-century Florentine workshop, researchers at the Opificio delle Pietre Dure identified residues of four different gelatines on successive leaves of the same volume, suggesting that the binder had gathered his sheets from several suppliers. The observation in turn made possible a closer dating of the binding itself, which had previously been attributed to the wrong decade.

Even damage can be informative. Mould grows on paper in characteristic patterns determined by the humidity history of its storage, and archaeological foxing - the reddish-brown spots common in old books - turns out to be a record of traces of iron in the original water supply reacting with atmospheric sulphur over centuries. The Canadian conservator Emily Chang has argued that such damage should be documented rather than simply removed. A conservator who bleaches away the foxing of a book may be improving its appearance while destroying an irreproducible record of its chemical biography.

Scientific imaging has greatly expanded the archaeologist's toolkit. Multi-spectral photography can reveal watermarks too faint to see with a light sheet, and X-ray fluorescence maps the distribution of trace elements across a page without any contact. A particularly striking example involves a previously unknown poem by a medieval Arabic writer, recovered in 2021 from a sheet that had been reused as the lining of a later European binding. The text had been washed away in the sixteenth century, but the iron in the original ink had soaked into the paper and could be mapped by X-ray, allowing the lost lines to be reconstructed.

The combined evidence of fibres, watermarks, size, and ink places paper archaeology in an unusual position. It depends on close cooperation between chemists, art historians, and curators, and it treats its objects as physical records of the societies that made them. For a manuscript that has survived four hundred years, every sheet is both a document in its own right and a dated sample of an extinct technology. Conti has observed that a library examined with these tools is no longer a storehouse of texts alone but an archive of the very materials on which texts depend.