The majority of the fifth floor of the Van Pelt-Dietrich Library Center is currently closed for construction. Learn more about this renovation to the Zilberman Family Center for Global Collections, including how to request materials.

  • Lecture

Using Handwritten Text Recognition for Rich and Abundant Historical Source Collections: The Case of Medieval Italian Financial Records

Laura Morreale, Independent Scholar & 2026-2027 SIMS Visiting Research Fellow

 

calendar_month
Friday, December 4, 2026, 12:00 - 1:30 pm EDT
location_on
Virtual
group
Open to the Public

Hosted by: SIMS and Kislak Center

Opening of a handwritten account book from Italy.

On October 2, 1466, Alamanno di Medici bailed out Salvestro da Palazzuolo from jail, where he had been imprisoned by the Otto di Guardia, the policing arm of the Florentine state. Although we do not know what Salvestro did to land himself in jail, his infraction—whatever it may have been—came at the cost of one fiorino largho, a rather hefty sum. Record of his imprisonment, the cost to free him, and an assertion that Salvestro would ultimately pay Alamanno back all appear in the Medici’s household leger, now held at University of Pennsylvania’s Kislak Center.

This small anecdote and the world of Florentine justice, economy, and social interaction that it evokes fills but four lines of the ledger’s nearly 250 folios, each of which contain up to 50 lines of accounting. The recorded expenditures reveal much more than one story of crime and punishment – here we see the day-to-day realities of child-rearing as the wet-nurse is paid, the math (abacco) teacher compensated, and clothing purchases made for Alamanno’s children. Familiar characters such as Mariotto the miller or Caterina the maid reappear in Alamanno’s payments to them over the years, and Alamanno’s wife Lena stars in her role as an important household actor as she pays the flour tax, purchases flax, or buys firewood.

The richness of these historical sources is hard to overstate, yet the many household ledgers that survive from the late Middle Ages remain largely untapped, in part because the manuscripts can be difficult to decipher and the writing lacks a narrative structure. However, historians are now reaching for new computer-based tools, including handwritten text recognition (HTR) programs, to transcribe these documents for easier access and analysis. Following a month-long fellowship at SIMS working with these materials, Laura Morreale will describe how to train HTR models to read difficult sources like the Italian financial records and will outline the costs, potential benefits, and risks of undertaking computer-based transcription for historical research.

Event Series

Featured Image: Opening of household ledger from the Medici Archive (UPenn Ms. Codex 1321, fols. 7v-8r).

Maps and More

Campus Libraries Map

Staff Information

Resources for Staff Committees