Classical Corner: Crime and Punishment in Roman Corinth

PETER BROWN / ALAMY STOCK PHOTO

One of the most overlooked structures at Corinth’s Roman Forum has a lot to tell us about life—and punishment—at the margins of late Roman civic life, offering us a rare and haunting picture of what it was like inside an ancient prison.

Following several destructive earthquakes that shook Corinth near the end of the fourth century CE, a series of major renovations and restorations took place in the subsequent decades. Archaeological evidence suggests that a particular series of rooms in the northwestern part of the forum, known as the Northwest Shops, were part of a larger renovation program during this time. Among these rooms, a central vaulted chamber, referred to by locals as the Boudroumi (from a Turkish word for “dungeon”), is architecturally distinctive. It was originally built in the late first or early second century alongside 15 similarly sized adjoining chambers, all originally open to the forum via a shared colonnade.

Although we’ve known of the Boudroumi since the early 20th century, when the area was first excavated by Herbert F. de Cou, its function remained uncertain. What has changed is how we interpret the space in light of previously underexplored evidence—namely, the archaeological context of reused paving slabs that form the floor of the Boudroumi and surrounding rooms, and the graffiti incised upon them.

M.D.C. LARSEN, “A PRISON IN LATE ANTIQUE CORINTH,” HESPERIA 93 (2024), FIG. 9 / COURTESY OF THE AMERICAN SCHOOL OF CLASSICAL STUDIES AT ATHENS, CORINTH EXCAVATIONS

What makes the evidence from Corinth so important? Despite literary sources suggesting that most Roman cities had a civic prison, identifying such facilities archaeologically has proven difficult. Prisons rarely carried signage. And while we have some texts and inscriptions referencing carceral experiences, we’ve lacked examples where graffiti can be securely connected to a known architectural context. In Corinth’s Boudroumi, however, the graffiti we’ve found—etched directly into pavement stones still in their original position—provide that missing link between text, context, and structure. And from these connections, a new picture of incarceration in late antique Corinth emerges.

In 1901, de Cou noted the discovery of 24 large limestone paving slabs, many broken on their edges, all laid in a consistent level across the Boudroumi and adjacent rooms. These slabs, he wrote, were part of a single, intact pavement that sealed off earlier Roman strata beneath, laid in conjunction with the late fourth-century renovations.

What caught de Cou’s attention—and ours today—were the inscriptions and images scratched into their surfaces. One reads, “☧ Lord God who separated the darkness and made light to rise through the inhabited world, repay, Godbearer, repay (punishment) of Marinos, the one who threw us in here and made us spend winter (here?) Petronia †.” (The final words are challenging and require some conjecture.) The graffito writer used Christian imagery (both a cross and a christogram), used a special abbreviation for the divine epithet (as was common in Christian scribal practice), and made an allusion to scripture (Genesis 1:4–5). Another text reads, “May the fortune of those who suffer in this lawless place prevail. Lord, do not show mercy on the one who threw us in here.” These graffiti offer a rather unique proximity to the voices of lived experience of the incarcerated in the ancient world.

Other examples, including bits of text, gameboards, and various drawings, hint further at the lived experience, rage, boredom, and personal devotion expressed inside the prison. These are not elite inscriptions. The language is colloquial. Some texts feature spelling errors or grammatical oddities, perhaps reflecting the writers’ lack of formal education. But their emotional and theological clarity is striking: These are desperate voices seeking justice, deliverance, or divine vengeance.

How do we know that the Boudroumi was used as a prison? In his unusually precise records, de Cou carefully noted that the two dozen inscribed paving slabs belonged to a single, uniform pavement laid across the area, adding that as his team worked they were “tearing up [this] late pavement.” One of his associates observed that this pavement served as a clear stratigraphic marker, separating Roman layers below from Byzantine layers above. Nearly all the slabs had at least one broken edge, indicating they were reused—spolia—from earlier contexts. It is clear that the graffiti on these slabs were inscribed only after they were repurposed to serve as paving stones: Several of the inscriptions respect the broken edges, showing the damage came first and the graffiti later. This is how we know these buildings were renovated and functioning as a prison during this later period.

The renovations give us additional clues about the key features of a prison. The Boudroumi shows evidence of a wall installed to seal the room off from the forum, accessible only through a narrow, controlled opening that allowed for very limited light and the passing of small items, such as food. One small doorway, including a locking threshold, was cut to connect the Boudroumi and the room immediately to the west. Interestingly, the Boudroumi had two stories, with the upper floor apparently maintained during its use as a prison.

Finally, the presence of captive-themed imagery just outside the building, such as the Captives’ Facade, usually thought to have been erected in 197 CE to commemorate Septimius Severus’s defeat of the Parthians, may have visually reinforced the association of the northwest corner of the forum with incarceration. Additionally, several inscriptions from the late fourth century found in the forum—particularly in the northwest corner—speak to the renovations undertaken during that time. It is likely that the repurposing of the Boudroumi and Northwest Shops into a prison was part of this urban renewal. There is also good evidence that the city’s previous prison was located in the southern aisle of the basement of the Julian basilica, across the forum to the east, which appears to have been destroyed by earthquakes around the same period. If so, the need for a new prison would follow logically from that destruction.

Secure identifications of ancient prisons are still relatively uncommon, though this is now slowly changing, and the Boudroumi site will be important in developing a typology of Roman prisons. Both the building and its graffiti provide important and interesting insight in what prisons were like in the past and what it was like to be an ancient prisoner.