Research

Archives, contested heritage and the anthropology of technology

The evidence at each of my sites is simultaneously archaeological, archival and curatorial. The work falls into three strands: the historical anthropology of excavation archives, the anthropology of technology, and museums and contested heritage.

1. Clerical excavation and contested heritage in Rome, 1857–1949

My thesis examines the interpretative and methodological processes involved in clerical excavations at three case study sites in Rome between 1857 and 1949: the potential biases and assumptions of the excavators, the events leading up to the excavations, the excavations themselves, and their reception to date. How, for example, have initial interpretations, such as Fra Germano di San Stanislao's 1892 discovery announcement beneath Santi Giovanni e Paolo and the 1894 site monograph, shaped the current view of the archaeological remains in Rome? The period from the opening of the lower church at San Clemente in 1857 to the close of the Vatican excavations in 1949 was chosen as the focal point. The interpretative framework that governed all three campaigns was formed in the nineteenth century; the Vatican excavations of the 1940s applied that framework, largely unchanged, to a new site.

Contested heritage became one of the most useful ways of approaching the history of early Christian basilicas in Rome. A monument's meaning is renegotiated by each generation that maintains, moves or removes it; at archaeological sites the negotiation is carried out not through public argument but through excavation, consolidation and display, and is therefore much harder to see. The archival turn has made a parallel point about the record itself: what is catalogued, indexed and made accessible shapes what can later be asked of it, so that an archive's silences are as consequential as its contents. Both processes are at work at the three sites, excavated by the institutions that owned them, recorded in archives those institutions still hold, and presented to visitors through lighting, labelling and routing those institutions still control. Whose past does a given site embody? Who holds the authority to inscribe and re-inscribe its meaning? Whose narratives are preserved, marginalised or erased in the process?

Rather than a single 'moment of discovery and interpretation', each site saw a constant, iterative process of excavation, preservation and destruction. Initial and subsequent excavators were often expected to confirm existing Church traditions, which were then maintained through careful curation of the historical and archaeological record. When a later excavator opened a new area, the previous interpretation was assumed to be true and needed to be upheld in the new results, even where this created contradictions with the material evidence. What often fell outside the 'sacred' and 'secular' narratives were quotidian structures such as housing, shops, baths and storage. The traditional site literature had given little weight to a shared bath suite within the residential complex below Santi Giovanni e Paolo, yet a shared bath implies several households using common facilities, including in the centuries after the introduction of Christianity.

The on-site archives of the three sites span their history from the early modern period to the present, with gaps; some are shared with international archives, such as the Passionist Historical Archives in New York. Analysis of these texts revealed previously unknown aspects of the site histories and made it possible to revise, and in some cases overturn, several pre-existing assumptions about each site. Work in progress from this strand includes a book, Excavating the Sacred in the Eternal City, several articles, and entries on the three basilicas for Oxford's Cult of Saints in Late Antiquity database.

2. Anthropology of technology: who gets to read the archive

In a storage room near the Basilica of Santi Giovanni e Paolo, archaeologists' field notebooks are stacked from floor to ceiling. Much of it has never been published, and what is uncatalogued, or hard to read, tends to go unasked. AI tools now promise to open archives like this one to anyone with a question. I want to ask not only what such tools do to the authority to interpret the past, but how they might be designed to keep the archive's silences visible rather than smooth them over.

ArchAI is the tool I built in response: a human-in-the-loop system for reading and linking handwritten excavation and museum records. It stores each claim with its author, year and source, treats every suggestion as a proposal until a person confirms it, and recommends related records. Three questions guide the work around it. How do archivists, historians, curators and visitors come to trust, defer to, or challenge an AI system? What does an AI reconstruction of the past preserve, and what does it silence? And how might such tools be designed so that the judgement of those who know the records best remains part of every answer? I come to this as one of ArchAI's builders, and I treat that role as an object of analysis rather than a neutral vantage point. A planned field study will have archivists, cataloguers and curators work with an answer-only assistant and with ArchAI's attributed design, shown identical suggestions, so that any difference in how they catch errors comes from the design itself.

The excavators whose notebooks fill that room in Rome wrote down what they saw, what they missed and what they hoped to find. Their records shaped a century of scholarship on early Christian Rome; the tools now being built to read them will shape the next.

3. Museums and contested heritage

Museums are where my questions about records and evidence meet a working institution, and where much of my practical work has been done.

The Art Institute of Chicago

As a Leadership Intern in the Department of Collections and Loans, I review newly scanned historic registration records and index accession numbers in CITI, the museum's collection database. Registration is the layer that makes ownership provable and movement lawful: deeds of gift, purchase records, accession numbers, loan agreements, condition reports. It sits at the point where an object's paper history becomes its data, and where an error in the record becomes an error in everything downstream, from curatorial research to the public website. It is also where the questions ArchAI asks about attribution are tested against a working institution.

The Eskenazi Museum of Art and the Kelsey Museum of Archaeology

The Resting Satyr in Context brought augmented reality to an ancient sculpture in the galleries of the Eskenazi Museum of Art at Indiana University and the Kelsey Museum of Archaeology at the University of Michigan, and was refined through visitor observation and user testing with the curators. Its evaluation was published in ACM Interactions, and the project was presented at the American Academy in Rome, the Getty Villa and the Archaeological Institute of America. For the Eskenazi I also led the Digital Roman Portrait Project, an online catalogue and 3D models of the museum's Roman portraits, and taught a course in augmented-reality exhibit design with the museum.

Sculpture collections in Florence, Rome and Atlanta

With the Virtual World Heritage Laboratory at Indiana University, I worked on the 3D digitisation of the Greek and Roman sculpture of the Uffizi Gallery, the Pitti Palace and the Boboli Gardens in Florence, presented at the International Congress of Classical Archaeology in Bonn; of the Casino dell'Aurora at the Villa Ludovisi and the Ludovisi sculpture in Palazzo Altemps, Rome; and of ancient sculpture at the Michael C. Carlos Museum, Emory University.

Teaching in collections

At Oxford I teach from objects in the Ashmolean Museum, including object-based tutorials and revision sessions among the Roman sculpture.

Exhibitions in development

An exhibit proposal, Excavating Voices, would let museum visitors question an AI built from an archaeologist's own papers, every answer grounded in a document on screen, while foregrounding the local workers whose words the archive barely preserves. It opens a line of work on AI, the dead and archival absence. Planned exhibitions with the Institute for the Study of Ancient Cultures, the Smart Museum and the Art Institute of Chicago form part of the next phase of ArchAI.

Methods

The framework assembled for the thesis was designed to function across different types of evidence, and to interpret each site against the traditional grain.

• Contested heritage: the question of whose past each site represents, read through Laurajane Smith's 'authorised heritage discourse', the dominant, expert-led account of what a heritage site is and who may speak for it.
• The longue durée: Braudel's three registers, adapted: cultural structures that persist over many centuries, such as the cult of saints; the conjoncture of clerical excavation between 1857 and the 1940s; and single interventions, such as the stair cut down to Area 21 beneath Santi Giovanni e Paolo in 1892, which fixed the interpretation of that space before it had been recorded.
• Assemblage and ensemble analysis: testing inherited functional categories against sealed deposits and surviving décor.
• Archival analysis and on-site observation: digitisation, transcription and translation of the on-site and institutional archives of Rome, alongside observation of how each site is toured and explained.
• Curatorial analysis: the modern presentation of each site: which areas have been renovated to emphasise their Christian aspects, which quotidian areas have been neglected, and how staging, the order of visuals and discovery narratives work together on a tour.
• Computational: Python, scikit-learn, PyTorch and TensorFlow; fine-tuning and evaluating Hugging Face and large language models; handwriting-recognition models in Transkribus; SQL, C#, C++; ArcGIS, Unity3D, photogrammetry and augmented reality.

Earlier projects

I have long applied computational methods to Roman material culture: houses, wall painting and sculpture.

Fieldwork and languages

Archaeological and archival fieldwork in Italy (Rome, Pompeii, Ostia), Spain and Greece. Latin (proficient); Greek (competent); Italian (competent, about two years in residence); French and German (reading).