Archaeologists working at the ancient site of Apollonia-Arsuf along Israel’s Mediterranean coast found sculpted protomes, head-and-upper-torso figures, deliberately placed inside Byzantine-era winepress installations. The objects were not discarded or lost. They were positioned in a way that strongly suggests intentional concealment. No excavation record or institutional file identifies who buried them or explains why.
Why hidden protomes inside a Byzantine winepress demand attention
The find sits at the intersection of two questions that rarely overlap in archaeology: who operated small-scale wine production facilities in the Byzantine Levant, and what drove someone to cache ritual or decorative objects inside working industrial infrastructure? The protomes were recovered from installations that included treading floors, collection basins, and inscriptions, all features documented by the Apollonia-Arsuf project run through Tel Aviv University’s Sonia and Marco Nadler Institute of Archaeology. That project has studied Byzantine wine-production sites at Apollonia-Arsuf for years, cataloging the physical layout and material culture of presses that once supplied regional markets.
Protomes are not full statues. They are bust forms, typically depicting a head and upper torso, and they served as decorative attachments in ancient art and architecture. The Smithsonian Institution’s National Museum of Natural History holds a cataloged example titled a female protome, a record that confirms the standard definition of the form. The J. Paul Getty Museum similarly describes a griffin protome as an ornamental element attached to vessels, furniture, or architectural surfaces. These were not freestanding sculptures meant for display in open spaces. They were functional art objects, which makes their burial inside a winepress all the more unusual.
The tension is direct: if protomes were decorative attachments, they belonged on surfaces, not under sediment inside industrial basins. Their placement inside the winepress points to a deliberate act of hiding rather than accidental loss or gradual accumulation of debris. One working hypothesis holds that local operators cached the protomes during a single, short-term disruption of the winepress, perhaps a raid, a period of political instability, or a forced abandonment, rather than during a broader regional collapse. That hypothesis could be tested by comparing deposition dates and covering sediments across multiple presses at Apollonia-Arsuf, but no published data from the excavation project has yet confirmed or ruled out that scenario.
What the Apollonia-Arsuf excavation records actually show
The strongest available evidence comes from institutional sources rather than detailed field reports. Tel Aviv University’s excavation project page confirms that the Apollonia-Arsuf site contains Byzantine wine-production installations with treading floors, basins, and inscriptions. The project is an academic dig, not a salvage operation, which means its methods are expected to follow university-level documentation standards. But the publicly available project description does not include stratigraphic data for the protome deposit, does not name the covering materials used to conceal the objects, and does not provide radiocarbon or ceramic dating that would pin down when the burial took place.
The Smithsonian catalog entry establishes what a protome is, a bust or head-and-upper-torso form, but it contains no reference to a winepress context or to intentional deposition. The Getty record reinforces the understanding that protomes served ornamental and possibly symbolic functions in ancient settings. Neither museum record links any protome in its collection to the Apollonia-Arsuf site or to a concealment event.
This gap matters because the claim that the protomes were hidden on purpose rests on physical evidence from the dig itself, evidence that has not yet appeared in a published excavation report accessible to outside researchers. The institutional records confirm what protomes are and how they were used. They do not confirm the specific circumstances of burial at Apollonia-Arsuf, nor do they document the precise relationship between the protomes and the winepress installations.
The absence of published field notes or object registry entries from the Tel Aviv University project leaves a significant hole. Without stratigraphic profiles showing how the protomes were covered, and without dating evidence that would separate deliberate concealment from later disturbance, the claim of intentional hiding depends on the excavators’ interpretation of the deposit rather than on independently reviewable data. For now, the scenario of purposeful burial remains plausible but unproven, a working explanation rather than a demonstrated fact.
Unanswered questions about who buried the Apollonia-Arsuf protomes
Several threads remain open. No official statement from the Apollonia-Arsuf project has addressed who performed the burial. The identity of the person or group that placed the protomes inside the winepress is unknown. No published report supplies dating evidence that would confirm purposeful hiding as opposed to later disturbance, such as a collapse of the press structure that happened to cover the objects. Until such evidence appears, any reconstruction of the event must be treated as provisional.
The motive is equally unclear. Protomes could have been hidden to protect them from looting during a period of conflict. They could have been placed as a ritual closing act when wine production ended at that particular press. They could have been stored temporarily, with the intention of reusing them in a new decorative scheme, only for circumstances to prevent their recovery. Each of these scenarios fits some aspects of the find, but none can be firmly supported without contextual data from the excavation.
Another unresolved question concerns ownership. Were the protomes the property of the press operators, of a nearby household, or of a religious institution overseeing production? Winepresses in the Byzantine Levant could be tied to estates, villages, or monasteries, and each ownership model would carry different implications for why decorative elements were treated as valuables worth concealing. The current public documentation from the Apollonia-Arsuf project does not specify who controlled the installations where the protomes were found, leaving this dimension of the story open.
There is also the issue of quantity and selection. Without an object list or photographs from the dig, it is unclear whether the cache contained a single protome or a group of varied pieces. If multiple protomes were buried together, their styles, iconography, and degrees of wear could reveal whether they came from one decorative program or were assembled from different sources over time. A heterogeneous group might suggest opportunistic gathering and hiding of valuables, while a matched set would point toward the deliberate retirement of a single installation.
Finally, the relationship between the protomes and the broader occupational history of Apollonia-Arsuf remains to be clarified. Were these objects buried during a phase of prosperity, when wine production was active and the press was worth protecting, or during a downturn, when installations were being abandoned or repurposed? The answer would affect how archaeologists interpret the social and economic meaning of the cache. Without securely dated layers tying the protomes to a specific moment in the site’s chronology, that interpretive step cannot yet be taken with confidence.
For now, the hidden protomes of Apollonia-Arsuf function as a prompt rather than a conclusion. They highlight how much can hinge on the careful publication of excavation records and how easily striking finds can outpace the documentation needed to interpret them. As long as the stratigraphy, dating, and object catalog remain unpublished, the story of who buried these sculpted heads inside a Byzantine winepress, and why, will stay an open case in the archaeology of the eastern Mediterranean.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
