City Workers Planned a Tree in Belgium and Found a Medieval Cemetery Instead

City workers in Ghent, Belgium, planned to plant a tree on the Vlasmarkt. Before the tree went in, archaeologists opened a small test pit beside Sint-Jacobskerk.

They found four well-preserved burials at about 1.3 metres deep. Then a borehole changed the scale of the story: more burial levels below, roughly 20 people estimated in that tiny pit, and at least about 240 across the wider 23-square-metre investigation area.

That is why the Ghent medieval cemetery in Belgium story is worth slowing down. The public headline sounds like 240 skeletons were found during a tree project in Belgium. The source trail says something more specific: four exposed burials, then an estimate.

Quick answer

The City of Ghent notice says BAAC Vlaanderen archaeologists exposed four well-preserved burials at a reported depth of 1.3 metres in a two-square-metre test pit. A borehole indicated about five stacked burial levels. That led to an estimate of roughly 20 people in the pit, then at least about 240 extrapolated across the 23-square-metre investigation area.

So no, Ghent did not excavate 240 skeletons. VRT’s English report gives an accessible account of the same city-led story and notes that the tree plan was paused. The public record is clear about the estimate method, but it does not include a trench plan, borehole log, field report, or osteological catalogue.

What does the Ghent medieval cemetery evidence show?

Ghent was resurfacing Vlasmarkt and considering a tree at one location. Before planting, archaeologists from BAAC Vlaanderen examined the intended spot with a two-square-metre test pit. This was planned heritage assessment, not a tree root accidentally tearing through hundreds of graves.

The city’s August 27, 2026 notice says the test exposed four well-preserved burials at 1.3 metres below the surface. Archaeology Magazine later amplified the story internationally, but the city notice remains the best source for the estimate method.

The location matters. This is Vlasmarkt beside Sint-Jacobskerk. It is not the separate Ghent excavation at Spaanskasteelplein that also produced a large burial story. Mixing those sites would import evidence from one project into another.

Source trail at a glance

City of Ghent notice
Use it for the pit size, four exposed burials, 1.3-metre depth, borehole inference, five levels, roughly 20-in-pit estimate, 23-square-metre area, 240 extrapolation, and preservation decision. Be careful because it is a public announcement, not a full field report.

City press-room mirror
Use it as an official duplicate and media-asset source. Be careful because it is not independent corroboration. It repeats the same city record.

VRT English report
Use it for an accessible English summary and the current tree/preservation decision. Be careful because it does not provide a trench plan or independent count.

Church and place-context records
Use Monumental Churches Ghent and the Flemish heritage inventory for Vlasmarkt for broad place context. Be careful because church history does not date any individual burial.

Future BAAC technical record
This is the source gap. A field report, trench plan, borehole log, context register, burial register, osteological catalogue, or dating report could support stronger claims. Those records were not public in the checked packet.

For a record-by-record map, use the Ghent Vlasmarkt cemetery source guide, which separates available sources from technical archaeology records not yet public.

What did archaeologists directly observe?

Four burials were exposed in the opened pit. That is the direct count. The city describes them as well preserved and places them at 1.3 metres.

Those observations do not supply names, sex, age, health, kinship, cause of death, or individual dates. The public notice does not report DNA analysis, radiocarbon dating, pathology, grave goods, or an osteological inventory.

A headline that turns the four people into plague victims, soldiers, relatives, or victims of one event would invent a story around human remains. The most respectful and accurate language is also the clearest: four burials were exposed, then preserved in place.

How did four burials become roughly 20?

A deeper borehole indicated about five superimposed burial levels. The lower levels were not all opened and counted. The borehole supplied subsurface evidence that burials continued below the exposed layer.

The archaeologists used those stacked levels to estimate around 20 individuals within the two-square-metre test pit. The intuitive outline is four visible burials multiplied across about five levels.

But the public notice does not provide a technical calculation, layer-by-layer density table, or full borehole log. We should not reverse-engineer false precision beyond the method the city actually reported.

The first ladder is simple:

• 4 burials directly exposed.
• About 5 burial levels indicated by borehole.
• Roughly 20 individuals estimated in the test pit.

Each number is useful. Each has a different evidence status.

How did roughly 20 become at least about 240?

The city says archaeologists extrapolated from the test-pit estimate across the 23-square-metre investigation area. That produced wording translated as roughly 240, at least 240, or in some reporting up to or more than 240.

The safest English phrase is “at least about 240 estimated burials,” followed immediately by the method. The number is not the population of the entire historical churchyard, and it is not a count of excavated skeletons.

The extrapolation assumes the sampled density can inform the wider defined area. The public notice does not show whether density is uniform, how obstacles were handled, or what confidence range the archaeologists would assign. A technical field report could clarify those points.

The complete ladder is:

4 exposed burials → about 5 levels → roughly 20 estimated in 2 m² → at least about 240 extrapolated across 23 m²

That one line prevents the headline from outrunning the evidence.

Was the cemetery itself a surprise?

No. The city says the churchyard around Sint-Jacobskerk was already known and had been used from the eleventh through the eighteenth century. What the test clarified was the preservation, depth, and apparent density at this particular point.

Monumental Churches Ghent dates the parish foundation to 1093 and the oldest surviving physical fabric to the twelfth century. The Flemish heritage inventory provides stable place context for Vlasmarkt. Those sources help explain why burials beside the church make sense. They do not date any exposed individual.

This matters because “lost cemetery discovered” is too loose. Archaeologists confirmed material conditions at a known churchyard location. The ground still supplied new evidence, but the cemetery’s existence was not unknown to local history.

Are all the burials medieval?

That is not established. The city connects the remains to a churchyard used from the eleventh to the eighteenth century. That range includes medieval and early-modern centuries.

“Ghent medieval cemetery” is useful search language and broadly fits the site’s long history. It should not become “four definitively medieval individuals” or “240 medieval skeletons.” The evidence supports a multi-century churchyard context and a preliminary density estimate.

A future stratigraphic sequence, artifact record, coffin evidence, burial register, radiocarbon date, or osteology report could narrow particular graves. None was available in the public packet checked.

What happens to the remains and the tree?

The city says the remains will be preserved in situ, the tree will not be planted at that spot for now, and no wider excavation is currently planned. VRT reports the same current decision.

“Currently” matters. Municipal plans can change. The present source trail supports preservation in place and a paused tree location, not a permanent promise that no future work will happen.

Human remains should not become visual spectacle. If images are later used, the strongest choices are a respectful trench-context photograph cleared by the city or BAAC, an exterior image of Sint-Jacobskerk, and an editor-made estimate diagram. An invented trench or pile of bones would be both misleading and unnecessary.

What evidence is still missing?

The public notice is unusually clear about the numbers, but it is not a field report. A mature source page should look for:

• BAAC’s technical report and project identifier,
• a trench plan and section drawing,
• the borehole and stratigraphy log,
• context sheets and a burial register,
• an osteological catalogue,
• grave-by-grave dating evidence,
• the legal preservation decision,
• any future redevelopment archaeology plan.

The absence of those documents sets a boundary. It does not invalidate the city’s estimate. It prevents us from adding individual stories, exact chronology, or precision the announcement does not contain.

For a related example of how remote or subsurface evidence becomes an estimate rather than a direct count, read PSF’s Amazon LiDAR primary-source guide. The Herculaneum scrolls article also separates physical evidence from what technology and interpretation add.

The estimate should stay attached to its method

Ghent’s public account does not hide how four became 240. It distinguishes exposed burials, borehole evidence, a local estimate, and wider extrapolation. The misleading version appears only when those steps disappear.

The real lesson is simple: a headline number can be useful without being a direct count. Readers deserve to know which is which.

Check the records behind the Ghent cemetery estimate. Use the Ghent Vlasmarkt cemetery source guide, then create a free PrimarySourceFinder account, get 200 starting credits, and continue into source research. No credit card.

Sources

• City of Ghent, Vlasmarkt archaeology notice
• City of Ghent press-room mirror
• VRT NWS, English report on the Vlasmarkt cemetery estimate
• Archaeology Magazine, “Historic Cemetery Excavated in Ghent”
• Monumental Churches Ghent, St James Church
• Flemish heritage inventory, Vlasmarkt