Was the Venice lagoon inhabited before Venice existed? Around its edges, yes. On the islands, there is still nothing to show. The Holocene lagoon and barrier system began forming most probably around 6000 BP, and the lagoon deposits under the historic centre are 2 to 6 metres thick with a maximum age of 4670 plus or minus 70 radiocarbon years (Mozzi et al., 2003). Bronze Age settlement in this landscape sits on the mainland fringe, at Altino and further north at San Gaetano di Caorle as far as Concordia Sagittaria, while for the Venetian area itself no archaeologically controlled sign of settlement, or even of visiting, has emerged (Vagnetti, 2006, p. 275).
That is not the story most guides tell. Three Mycenaean sherds published in 2000 as picked up on a beach on Torcello were greeted as proof that Aegean traders worked the lagoon in the thirteenth century BC. The pottery is real and well dated. The findspot is the problem, and the same specialist who dated the sherds set out why (Vagnetti, 2006, pp. 273, 275). Systematic excavation on Torcello has produced no Bronze Age layer at all: the first certain remains of human occupation there are no earlier than the Roman period and rest directly on lagoon silt, and survey work on the earliest inhabited nuclei of the future city has found no trace of settlement, or better of habitability, before the fifth century AD (Vagnetti, 2006, p. 275).
What the lagoon does have is a Roman past with dates and measurements attached. Altino, on the mainland at the northern edge, was founded by the ancient Veneti between the eighth and seventh centuries BC and became a municipium in 49 BC (Paiano and Delpozzo, 2024, p. 20; Musei archeologici nazionali di Venezia e della Laguna, 2026). A brick well-cistern in the San Felice channel, dated to the first century CE, still holds mortar made with volcanic pozzolan shipped from the Bay of Naples (Dilaria et al., 2024). The bishop's seat moved from Altino to Torcello in 639, a date fixed by an inscription found in the crypt in 1895 (Paiano and Delpozzo, 2024, pp. 30-31).
Was the Venice lagoon inhabited in the Bronze Age?
The honest answer separates the lagoon margin from the lagoon itself. Altino, the Roman town whose ruins sit at Quarto d'Altino, owed its position to an abundance of waterways that drew people to the site from the Mesolithic onwards, and traces of more stable settlement there date to the Bronze Age, shown by finds made across the Altino countryside (Paiano and Delpozzo, 2024, p. 20). By at least the eighth century BC the inhabited core lay near the Santa Maria canal to the south. In the seventh century BC that sector was given over to sacred use and housing moved instead onto the raised ground that would later carry the Roman public centre (Paiano and Delpozzo, 2024, p. 20). The state museum service dates the town's foundation by the ancient Veneti to between the eighth and seventh centuries BC, its peak to the Roman period between the first century BC and the second century AD, and its abandonment to the seventh century AD (Musei archeologici nazionali di Venezia e della Laguna, 2026).
Beyond Altino the same picture holds along the northern edge: significant Bronze Age settlement traces are being brought to light around the lagoon, at Altino and further north at San Gaetano di Caorle and as far as Concordia Sagittaria. For the properly Venetian area, by contrast, no archaeologically controlled indication of settlement, or even of visiting, had emerged when the question was reviewed (Vagnetti, 2006, p. 275).
When did the lagoon form, and what is the caranto?
The lagoon was already there in the Bronze Age, though the ground under Venice barely was. The Holocene lagoon and barrier complex started forming most probably around 6000 BP. In the central lagoon the subtidal mud flats average around 5 metres thick, thinning towards the mainland, while tidal channel scours can exceed 10 metres. Under the city itself the lagoon deposits are between 2 and 6 metres thick, with a maximum age of 4670 plus or minus 70 radiocarbon years, and the barrier island sands reach about 15 metres (Mozzi et al., 2003).
Underneath all of it lies the caranto, the firm mottled layer Venetian builders drive their piles into. Coring shows it to be a buried soil, formed on the silty clay overbank deposits of the Late Pleistocene Brenta river system, whose distal aggradation continued until about 14,500 BP. It is in physical continuity with the Calcisols still farmed on the plain immediately behind the lagoon, and it corresponds to the surface of that plain, drowned by Holocene sea level rise together with local subsidence. The word itself is dialect, used by mainland farmers for the hard cemented crusts that plough blades turn up in the fields (Mozzi et al., 2003). The lagoon those figures describe now covers 550 square kilometres at an average depth of about 1 metre, cut by channels deeper than 15 metres (Madricardo et al., 2019).
The Mycenaean pottery said to come from Torcello
In 2000 three Mycenaean sherds were published as found by chance on a small beach on Torcello. They had spent time in a private Venetian collection before being handed to the heritage authority, and one carries the pencilled note "Torcello 1949" on the inside. On what they are there is no dispute: good quality Mycenaean ware, comparable with pottery from the main centres of the Peloponnese, most probably Late Helladic IIIB, roughly the thirteenth century BC (Vagnetti, 2006, p. 273).
On where they came from the objections are physical. The fragments are in excellent condition, with fractures neither worn nor carrying the signs of water transport a tidal beach would leave, and the spot could no longer be inspected because a quay had been built over it (Vagnetti, 2006, p. 275). The older museum material fares no better. Four intact Mycenaean vases and two Cypriot vases of the Cypriot Middle Bronze Age were kept in the Torcello museum, and because they are whole they most probably come from tombs, which across Italy is almost unheard of. A Bronze Age cemetery there would imply a stable community in regular contact with the Aegean, and excavation has produced no sign of one. Doubts were recorded as early as 1904, on the ground that the museum's director of the day bought objects widely without noting provenance; in his catalogue only one of the four vases can be identified, as no. 760, noted as excavated at Mazzorbo in 1881 (Vagnetti, 2006, p. 274).
Frattesina and the Bronze Age routes behind the lagoon
The Aegean material is not a Venetian phenomenon at all. Across the central Mediterranean, Aegean and Mycenaean finds are attested at over a hundred sites, mostly pottery from settlements rather than cemeteries, spread over a span running roughly from the seventeenth to the eleventh century BC. Imports are practically the only material until the late fourteenth century, after which local imitations conventionally called Italo-Mycenaean appear from the thirteenth century alongside imports that keep circulating (Vagnetti, 2006, p. 276).
In the Po plain the distribution map counts eight findspots: Fondo Paviani, Fabbrica dei Soci, Castello del Tartaro, Lovara, Bovolone, Terranegra, Frattesina and Montagnana, all in the plain between the Adige and the Tartaro apart from Montagnana, which lies further north along an old branch of the Adige (Vagnetti, 2006, p. 276). Archaeometric analysis of that whole corpus assigned most compositions to Italo-Mycenaean workshops locatable in Puglia or elsewhere in central and southern Italy. Only one fragment from Fondo Paviani and two from Fabbrica dei Soci are good candidates for import from central Greece or the Peloponnese, and even those could have travelled the well attested south to north routes overland (Vagnetti, 2006, p. 277). What the upper Adriatic really had in the late second millennium was a dense system of settlements with a high concentration of craft production, the Valli Grandi Veronesi group first and then the large centres of the Polesine, among which Frattesina stands out for its scale and complexity (Vagnetti, 2006, p. 278).
Altino, the Roman town the lagoon grew around
Altino is where the pre-Venetian lagoon has measurements. With the grant of municipium status in 49 BC the town was rebuilt on Roman lines, and the central imperial centuries down to the third century AD read in both texts and archaeology as its most prosperous period (Paiano and Delpozzo, 2024, pp. 20-21). Aerial photography and geophysical survey show a canal on average 25 metres wide crossing the monumental quarter for at least 360 metres, with at least two bridges over it, one on the line of the via Annia (Paiano and Delpozzo, 2024, p. 28).
In the sector known as Area Est excavation found the water's edge itself: a quay of squared sandstone blocks stepping down eastwards, four steps resting on a timber pile platform, and a row of large mooring posts set at regular intervals. Tree-ring analysis of those posts gives a date between 80 and 70 BC, which also dates the digging of the canal, and the same canal was silted up and finally covered over there by the end of the first century BC (Paiano and Delpozzo, 2024, p. 26). Magnetometry at Valle Rossa has defined the harbour basin as an L shaped structure: a smaller southern arm of 134 by 20 metres and a main arm, oriented northeast to southwest, of about 139 by 41 metres, large enough to take both river craft and the smaller vessels coming in from the sea (Paiano and Delpozzo, 2024, pp. 28-29).
A submerged Roman well, and how far the sea has risen
The best dated Roman object in the lagoon proper sits underwater in the San Felice channel in the north, near Lio Piccolo. A brick structure long read as a defensive tower was reinterpreted after underwater survey as a well-cistern: a flat, almost square box measuring roughly 6.7 by 7.7 metres, built entirely of local bricks of the Po Valley sesquipedale size, with the gap around the inner shaft packed with river sand that filtered rainwater down to the bottom where the well drew it. Associated evidence dates it to the first century CE or slightly later (Dilaria et al., 2024).
Its mortar carries the surprise. Analysis identified volcanic pumice that does not match any Veneto geology and instead points to the Phlegraean Fields near Naples, the pulvis puteolanus that Roman writers recommended for building in seawater, and this is its first attestation anywhere in the lagoon of Venice. The structure doubles as a measuring stick for the sea. It now lies at a depth ranging from 4.5 to 3.0 metres below mean sea level, with its top standing about 1.5 metres above the channel floor, and with a Roman relative sea level in this part of the northern lagoon about 2.15 metres lower than today, the base of the well-cistern then stood at roughly 2.0 metres below the surface (Dilaria et al., 2024). Nothing of it is visitable: it is a submerged archaeological site in a working channel.
Did refugees fleeing Attila settle the lagoon?
The familiar account has the lagoon settled in a single panic, by late Roman inhabitants of Altino fleeing barbarians on horseback. Those reconstructions derive from historiographical topoi that are widely diffused and well recognised as such, and they survived in the literature because they summarise a complicated process neatly (Paiano and Delpozzo, 2024, p. 20). The often repeated resemblance between the street plans of Altino and Venice, which is used to make Altino the ancestor of Venice, appears on closer analysis to be entirely fortuitous (Paiano and Delpozzo, 2024, p. 20).
What the archaeology shows is slower and better dated. The number and the size of settlements in the northern lagoon increase from the fourth century AD, with San Lorenzo di Ammiana, San Francesco del Deserto and Torcello the best known cases, and written sources follow that shift of weight from the mainland to the islands (Paiano and Delpozzo, 2024, p. 30). The transfer of the episcopal seat from Altino to Torcello is dated to 639 by an inscription found in the crypt in 1895 during restoration work (Paiano and Delpozzo, 2024, pp. 30-31), and it followed the Lombard conquest (Vagnetti, 2006, p. 275). At Altino the structures of power fade progressively from the seventh century AD, in a long depopulation ending in almost total abandonment (Paiano and Delpozzo, 2024, p. 21).
