Read the excavated town through streets, shop fronts, upper dwellings and coastal construction, while distinguishing ancient arrangements from later loss and rebuilding. Historical architecture, reviewed 19 November 2026.
Street in Herculaneum. Photograph: Larry from Charlottetown, PEI, Canada / Wikimedia Commons / CC BY 2.0 Generic. Resized from 4608x3456 to 4000x3000 and recompressed; no crop. A view of excavated fabric, not proof of a particular ancient dwelling arrangement.
The excavated streets are only part of the town
Herculaneum's streets can make the recovered buildings appear to form a small, complete town. The Archaeological Park's guide gives a different starting point: only part of the ancient settlement has been uncovered, and the excavated sector does not contain every kind of public building expected in a Roman town. What has been exposed is not the whole that once existed.
A ground-floor plan is incomplete in another way. Accommodation extended above the rooms and entrances it records. Nicolas Monteix's study of shops within urban properties compares changing ground-level arrangements with upper floors and mezzanines. In his selected sample, adding the upper accommodation changes the account of how much usable space developed. His sample is not a census of the town.
The street, house and upper dwelling therefore need to be read together. A doorway could lead into a large house, a commercial room or accommodation reached independently by stairs. A building's outline at ground level does not by itself identify all the units above, their connections or their occupants.
Preservation also has its own history. Some upper structures were recorded and then lost; others were rebuilt after excavation. The question is not simply how many storeys a building had, but how its levels worked together and what kind of evidence supports that explanation.
Shops and upper dwellings did not have one arrangement
Monteix follows commercial fronts through several excavated blocks rather than treating one house as a model for the town. His analysis distinguishes shops communicating with a larger dwelling from physically independent shops, some associated with accommodation. It also distinguishes the creation of more separate living units from the sale or legal subdivision of property.
That distinction matters above the street. An upper floor might enlarge a house without creating a separate dwelling. Elsewhere, a stair from the street could serve accommodation independent of the rooms below. Physical separation provides architectural evidence, but does not alone identify an owner, prove a tenancy agreement or name the people who lived there.
Alexandra Dardenay's study of Insula V makes the problem concrete. In the units conventionally numbered V17 and V18, she describes an upper apartment reached from the street by a staircase. A partition isolated that stair from the ground-floor accommodation in AD 79, although the room's surviving state does not convey that separation. An apparent connection is not necessarily an ancient route between dwellings.
The arrangement of a whole block is important too. Dardenay follows rooms separated from a larger house to form other units, while older floors and decoration remained. Adjoining buildings can preserve parts of an earlier shared history even when their last arrangements were distinct. A modern house name is not a boundary that remained unchanged throughout antiquity.
A sloping street front complicates the floor plan
The western front of Insula Orientalis IIa provides a different setting from a private house. Monteix describes commercial rooms and access to upper accommodation along Cardo V, with public spaces behind. Its building ground slopes, and the relationship between street alignment, rooms and the complex behind helped produce different depths and arrangements.
His reconstruction shows why counting floors from the street can mislead. Mezzanines and full floors accommodated changing ground levels. A level that was the first upper floor in one part could correspond to the second in another. The continuous upper accommodation and its access cannot be understood by repeating one identical shop-and-house unit along the frontage.
The complex is commonly called the Palaestra, but its principal ancient function is disputed in the sources used here. The official guide describes a sporting complex; Monteix argues for a principal cultic role associated with Magna Mater. Neither identification needs to be settled to understand the western front's mixture of commercial rooms, upper access and spaces behind.
Nor should this unusually large complex stand for every street. Monteix treats it separately from private properties and warns that its particular development needs comparison with other buildings. Shops and accommodation shared the town without all belonging to one architectural or economic arrangement.
The coastal edge adds another vertical setting
Miko Flohr's discussion of Cardo IV connects the street to the seaward side of the settlement. Drawing on earlier architectural research, he describes modest plots, later combinations of neighbouring houses and upper apartments reached directly from the street. At the coastal ends of the blocks, some houses expanded over the earlier wall into much larger seaside residences.
Their small street-facing beginnings did not measure their later extent. The same street could relate to comparatively modest accommodation and to houses that developed substantial gardens and rooms towards the sea. This is a history of changing properties, not a reliable way to assign a social identity to every doorway.
The Archaeological Park's guide adds a bounded example on descending coastal ground: the House of the Relief of Telephus, described as built on three levels. Its account also places vaulted rooms beneath the terrace above the ancient beach. Buildings and supporting structures used the difference in level between the town and its coastal edge.
These relationships should remain distinct from the central shop fronts. A house extending towards the sea, a terrace carried on vaulted structures and an apartment reached by a street stair are different ways of occupying height. They do not combine into a standard cross-section of Herculaneum or a complete reconstruction of its ancient shoreline.
Lost upper rooms can survive in conflicting records
The Casa d'Argo presents a different problem: much of the upper fabric known to its excavators did not survive. James Andrews reconstructs its arrangement from excavation accounts, plans and drawings, distinguishing at least two upper complexes with separate access. He also records contradictions between those sources rather than treating every illustrated detail as an observation.
His analysis distinguishes elaborately finished upper rooms towards the street from more utilitarian areas elsewhere. Access and visibility matter alongside decoration and finds. But those relationships do not establish every room's function. In particular, Andrews rejects an unsupported identification of niches as slave sleeping places.
Andrews considers reception, storage and potentially seasonal use, while acknowledging uncertain connections and the incompletely excavated western part of the house. Those are interpretations of unequal evidence, not a recovered timetable of domestic life. A drawing can preserve an otherwise lost relationship without proving all the furniture, decoration or activities it depicts.
This one house cannot explain the whole town. It shows why the wider comparison needs more than surviving ground plans: one must distinguish recorded upper structures, reconstructions of their arrangement and interpretations of their use. The absence of a floor today does not mean there was no accommodation above.
Rebuilding changed the evidence, not only the appearance
Excavation and restoration also affected how surviving dwellings were understood. Monteix compares the Casa a graticcio's excavation records with its later presentation. He identifies a reduced street balcony and changed positions of furnishings, then proposes a different division of its accommodation. The contrast concerns the building's spatial interpretation, not merely its visual finish.
That example does not mean every displayed object is misplaced or every wall is modern. It means that a restored room cannot automatically be treated as an untouched record of where things were found. Monteix's criticism rests on particular comparisons between records and later presentation; the original archives have not been independently checked for this account.
A 2019 paper by the Park and Herculaneum Conservation Project practitioners describes another part of the problem. The excavated town contains ancient material alongside modern roofs, lintels and supports for upper levels. These different elements have different conservation needs. The paper is a dated explanation of the mixed fabric, not a present-condition or safety assessment.
Read Herculaneum, then, through several connected relationships: streets and entrances, shops and differently reached accommodation, central blocks and coastal construction, surviving fabric and lost upper rooms. Keep excavation and rebuilding within that explanation. The town is more extensive than its ground-floor plan, and the recovered buildings are evidence with histories of their own.
Sources
- Monteix, Les lieux de metier (2010): selected urban properties and upper accommodation
- Dardenay (2021): adjoining buildings and the V17/V18 stair partition
- Flohr (2023): Cardo IV and coastal property development
- Andrews (2017): the lost upper floors of Casa d Argo
- Herculaneum Conservation Project practitioners (2019): ancient and modern fabric
- Archaeological Park, Brief Guide to Herculaneum: historical town and coastal buildings
- Monteix: excavation and rebuilding
- Monteix: Insula Orientalis IIa
- Monteix: sampled urban blocks
- Monteix: conclusion
Questions people ask at the gate
Is the excavated area the whole ancient town?
No. The official guide describes a partly uncovered settlement. The exposed buildings and streets are not a complete sample of its public and private life. Monteix's commercial/property study is a further selected sample within that recovered evidence.
Did an upper floor always belong to the rooms below?
No single arrangement applies. An upper level could extend a house or contain independently reached accommodation. Dardenay's V17/V18 example includes a partition isolating the street stair from the ground-floor dwelling in AD 79, despite what the surviving room might suggest.
Can a separate entrance identify a tenant or owner?
Not by itself. Access and physical independence help interpret architecture, but do not supply a person's identity or a property deed. Monteix explicitly discusses the difficulty of establishing ownership and rental relationships from buildings.
Was Insula Orientalis IIa certainly a sports complex?
The sources used here disagree about its principal function. The official guide uses the Palaestra identification, while Monteix argues for a principal cultic role. The western commercial/apartment front can be discussed without resolving that disagreement.
Are all the Casa d'Argo's upper rooms still preserved?
No. Andrews uses records of upper structures subsequently lost. The documents preserve important relationships but also disagree, and his account separates observations from proposed uses. It does not establish a complete domestic routine.
Do reconstructed rooms show exactly what excavators found?
Not automatically. Monteix identifies particular differences between excavation records and later presentation. The conservation practitioners also describe ancient and modern fabric together. Neither finding establishes that every room was altered in the same way or gives a current condition assessment.
Next walk starts at a new gate.
Was this helpful?
one tap, no sign-up


