Now published! A database of higher-order settlements on the Italian peninsula (350 BCE to 300 CE)

Followers of this blog!

The database I have been working on since 2012 is now freely available online at the  Archaeology Data Service:

Also available at this link are the database’s explanatory notes, a bibliography, a list of sites excluded from the study and a list of ancient settlement names without archaeological locations.

An article explaining the database’s research design, methodological development, format and functionality has also been published Open Access by Internet Archaeology: (authors: J. Sewell and R.E. Witcher).

The Roman reuse of Etruscan town-planning at Pyrgi

Perhaps the earliest of Rome’s colonies that was inhabited exclusively by its own citizens was Ostia, established at some point after the mid-4th century BC. Originally the colony was tiny, estimated to have covered 2.4ha. It was enclosed by rectilinear fortifications and it is also believed to have had an orthogonal street system with the same orientation as the walls. It was located not far from Rome on the Tyrrhenian coast near the mouth of the Tiber, and its function is believed to have been to defend this stretch of coast and the entrance to the river.

After the foundation of Ostia, ancient literary sources indicate that Rome founded a further 21 coastal colonies of Roman citizens down to the year 181 BC and the archaeological sites of all but two of them have been found with relative certainty. Archaeologists and historians cannot claim to know all the reasons why these colonies were founded. Ancient literary sources contain reports of seaborne raiders active in the waters of the Italian coast, however, and Rome considered the maritime power of Carthage to be a threat. Tensions between Rome the Punic city grew during the 3rd century BC, leading to open conflict for the first time in 264 BC.

Roman citizen colonies founded down to 177 BC, including those located on the coast.

Roman citizen colonies founded down to 177 BC, including those located on the coast.

Similar to Ostia, many of the Roman citizen colonies of this period were small, although not of uniform size, with rectilinear fortifications. The construction of straight fortifications with right angled corners is, above all, a response to building on level terrain, and many of these centres were located on coastal plains. To a certain degree, the colonies can be considered as permanently manned forts. Perhaps as many as 13 (of 21) of them were founded as completely new settlements, with the remainder being set up within or very close to pre-existing centres.

One of those of the latter category was Pyrgi, believed to have been the port-town of the great Etruscan centre of Caere (Cerveteri). It is located about 45km to the NW of Rome and has archaeology dating back to the 8th century BC. In fact, it is probably best known for its pre-Roman history. There are strong indications that it was frequented not only by Etruscans, but also by Greeks and Phoenicians. It possessed two great sanctuaries, in one of which temples of the 6th and 5th centuries BC have been excavated.

The location of Pyrgi

The location of Pyrgi

Google Earth image of Pyrgi with the outline of the rectilinear walls of the Roman colony still visible. Note the Etruscan  sanctuary to the south-east.

Google Earth image of Pyrgi with the outline of the rectilinear walls of the Roman colony still visible. Note the Etruscan sanctuary to the south-east.

At some point, probably in the first half of the 3rd century BC, the Romans’ dominion over the local area is reflected in their founding of a colony at the site of the Etruscan port. Rectilinear walls were constructed, enclosing an area of 5.7 hectares, that seem to have been orientated on the line of the coast. It has been presumed for many years that the other rectlinear citizen colonies were planned in a similar way to Ostia, with orthogonal street-systems that were orientated according to the line of the walls.

Yet recent excavations at Pyrgi by Flavio Enei and his team have produced results that have shattered the simple logic of this presumption. It seems as though the Roman colony was built directly on top of the Etruscan settlement, although the Roman centre only occupied about half of the area of its Etruscan predecessor. Roman Pyrgi had a rectilinear street-system, but it turns out that it was not orientated on the Roman walls. It seems that Etruscan Pyrgi already had an orthogonal grid of streets in the Archaic period, and the Romans adapted it for their colony’s street system. But the orientation of the Etruscan streets was close to being diagonal to the line of the walls, although it is not yet clear whether the whole of the Roman urban area was laid according to this alignment. The reasons why the Romans chose to maintain the Archaic planning is not clear, especially because a destruction layer was found associated with the period of the Roman takeover, suggesting that some or all of the former Etruscan structures were destroyed before the colony was founded. This destruction layer makes it less likely that the reason for the maintenance of the Etruscan orientation was related to the desire to preserve pre-colonial buildings. A fascinating riddle.

At some point during the first two decades of the 2nd century BC, the character of citizen colonies changed. New ones, such as Luni, were founded as much larger urban settlements than their 3rd century-BC predecessors. Several of the older small colonies were enlarged during the 2nd century and took on much more recognizably urban characters. This reflects the fact that the Roman concept of the colony changed over the centuries of Roman rule.

At Pyrgi, a castle was built during the 14th century within the walls of the former colony: the castello di Santa Severa. It still stands (see image) and is often open for visitors during the summer.

Further reading:

Enei, F. 2012 ‘Pyrgi e le sue mura poligonali: recenti scoperte nel castrum e nell’area portuale’, in L. Attenni and D. Baldassarre (eds.) Quarto seminario internazionale di studi sulle mura poligonali. [Alatri] Palazzo Conti Gentili, 7-10 ottobre 2009. Roma: Aracne, 313–325.

Ancient cities on the move: changing names and locations

Some of the ancient cities in Italy were less easy to track down due to them having been relocated at various points in history. Their names could move with them. Very many of the names of Italy’s ancient cities have survived through to today, with ‘Roma’ being the most obvious example. Often the ancient names have mutated over the centuries (Caere Vetus to Cerveteri), but the locations of ancient cities were also subject to change. The Romans themselves were responsible for relocating some settlements: Falerii Veteres, a Faliscan settlement, was forcibly abandoned in c.241 BC and the community was moved to a new location, known as Falerii Novi. Another classic example is the destruction of Etruscan Velzna in 264 BC (modern Orvieto), and the transferral of the population to the new centre of Volsinii Nova (Bolsena).

Roman Sipontum was founded as a Roman colony in 194 BC, and located on the coast of Daunia. Its archaeology seems not to date back earlier than the founding of the colony, but it is mentioned by Livy as existing in 335 BC, when it was captured by Alexander the Molossian, the uncle of Alexander the Great. In fact, archaeologists believe that pre-Roman Sipontum has been found, located nearby in an area with the modern name of Cupola-Beccarini. The earlier site is very large and seems to have been a dispersed settlement with a history spanning from the early first millennium through to about the moment the Roman colony was founded.

It seems that cities could also move on health grounds. The relocation of Salapia, a Daunian settlement, is described by the Roman architect, Vitruvius. It was allegedly transferred to a new location in the 1st century BC due to an outbreak of malaria in the old town. There are still some disagreements among archaeologists in relation to individual sites regarding the process of re-location. For example, the modern town called Ginosa, located in the Province of Taranto in Puglia, is built on the ancient Peucetian settlement called Ginusia. This suggests continuity from the founding of the settlement (c.800 BC), to the present. Not so. As far as I can make out, there has been no systematic archaeological study of the area of the modern town, but tombs are known dating from the 8th to the end of the 4th centuries BC – not that this necessarily means that the city was abandoned at the end of the 4th century BC – and late antique ceramic is also known. But the Roman writer Pliny the Elder, stated that the town existed in his day, the 1st century AD, as Latinised, Genusium. Some have argued that Roman Genusium was to be found at the village of Montecaglioso, based on epigraphic evidence, while others prefer Contrada della Madonna Dattoli, where excavations have revealed a nucleation of villa type structures. Either way, it seems that the site of modern Ginosa was reoccupied in late antiquity because it was easily defensible.

This is another cause of wandering cities, the insecurity of the late Roman and early medieval periods on the Italian peninsula led some communities to relocate to higher ground where they could defend themselves more easily. Falvaterra is a modern village on a hilltop in the Province of Frosinone in Lazio. It moved to its current location, it seems, as a result of incursions by the Lombards in the 6th / 7th centuries AD. Prior to that, it had been located on a plain and was called Fabrateria Nova, a settlement founded in the last quarter of the 2nd century BC by the Romans, which itself was a replacement of an earlier Volscian settlement called Fabrateria Vetus!

Sometimes the location and name of original antique settlements can be reconstructed based on names surviving today.  A good example is ancient Forum Cassii, located on the route of via Cassia in Etruria, the name of which has survived in that of a church: Santa Maria di Forcassi. But because some settlements were liable to move over time, surviving names might not relate to the original location of the settlements. Capua, for example, is located today where the ancient port-town of Casilinum once stood. Ancient Capua was destroyed in the 9th century AD, and its population eventually moved to the city’s current location a few years later.

Settlements moved to new locations for various reasons, but it has to be emphasised that it was actually communities of people that moved. The names of their settlements moved with them because they provided historical and cultural identity to those communities. Modern settlements with ancient names conjure up a sense of ancient tradition, longevity and permanence in the landscape. Most recently, in the 20th century, there have been cases of medieval and later settlements changing their name to those of antique cities, most notably under fascism. For example, in 1928, Monteleone di Calabria changed its name to Vibo Valentia, being the title of the Latin colony of the third century BC that was buried under its streets.

The afterlife of archaeological sites: the example of Monterano

The site of Monterano, looking north

The site of Monterano, looking north

The physical condition in which the database’s archaeological sites find themselves today is highly varied. Since classical antiquity all but the most inaccessible of them have been plundered for reusable building materials. This is even the case at Pompeii, where material was robbed from right after the eruption of AD 79, and then periodically through to the time archaeological excavations began. Sites on exposed hilltops have been subject to the ravages of natural weathering processes. Many low lying sites that were not protected by the authorities have been damaged by deep mechanised ploughing since the mid-twentieth century. At the time of writing, the quantity of sites in the database is 590. More than half of them are covered today with substantial areas of modern habitation, either because they were never abandoned or because abandoned sites were reoccupied at some point from the medieval period onwards. Of these, 224 have totally or almost completely disappeared beneath townscapes that have developed from a constant process of urban renewal over centuries. Often the only indication of their antiquity are stretches of ancient city wall or other forms of monumental architecture that have survived through their incorporation into later structures. Searching for the remnants of ancient architecture along the narrow medieval streets of Italy’s numerous hilltop towns is always fun.

An fascinating example of the unpredictable fortunes that can befall ancient sites after they were abandoned is represented by Monterano, located on a tapering spur of a hill, c. 8 ha in size, in the territory of Canale Monterano in northern Lazio. Prehistoric activity has been documented in the area, and the site is believed to have become an Etruscan settlement in the 7th century BC, perhaps surviving through to the mid-4th century BC. Hardly anything is known of the settlement from this period, but it is suspected to have possessed fortifications (although there is no direct evidence for this); the Etruscan presence is represented by tombs located at the base of the hill. Its fate during the Roman period is unclear, but during the early medieval period it had become the seat of a bishop, and a castle was built on the site. During the later 17th century, under the aegis of the powerful Altieri family, the entire settlement underwent a drastic makeover. The castle, which had undergone numerous architectural changes since it was constructed, was converted into a palace. A very Roman-looking aqueduct was constructed to supply the settlement, and a convent was founded, dedicated to Saint Bonaventure. Much of the monumental architecture from this period survives, but after the settlement was sacked by French troops at the end of the 18th century, it went into decline and was abandoned. Today, it is a spectacular ‘ghost-village’, dominated by the ruins of the castle, the convent church and the aqueduct. Inevitably, it became a popular filming location during the 20th century; a scene for Ben Hur was shot nearby in 1959. Well worth a visit!


Seeking the impact of the Roman conquest

One of the big changes in the pattern of major settlements across the Italian peninsula is apparent from the 4th to the 3rd centuries BC. The project has documented 84 fortified settlements of at least 2 ha in size, founded either in the 4th century BC or at a point between the 4th and 3rd centuries BC that were located in highly defensible locations on hilltops and which did not go on to become Roman towns. Only three such sites appear to have been founded during the course of the 3rd century BC, and the basis for the dating of two of them is questionable. What brought about this change?

It is tempting to see the consequences of the Roman conquest as being somehow directly or indirectly related. Based on information from ancient literary sources, most of the work undertaken by the Romans to secure military and political domination over the peninsula occurred during the years 343 to 266/265 BC. Based on this chronology, the construction of new fortified sites in each region generally seems to have stopped after it was subjugated by Rome. Yet the evidence that Rome was responsible for this change is merely circumstantial. Although the Romans directly appropriated some land from Italy’s peoples for various purposes, prior to the 1st century BC it is believed that most of the peninsula’s surviving communities were only bound to Rome by way of a treaty (foedus). Rome manifests itself as the dominant partner in these treaties because, through them, subject communities were legally obliged to provide troops to fight alongside Rome’s legions. Because very little is known about the remaining content of the treaties, we have no details on how or whether Rome might have limited the independence of subject communities in other ways. Although it is understandable why Rome would have wanted to stop newly subjugated communities constructing new fortified settlements, there is no direct ancient textual evidence that Rome did or could have. Moreover, the archaeological record in many of the peninsular regions does not reveal much sign of dramatic cultural change in the immediate wake of the conquest.

The result of all this is an intriguing puzzle. Based on the stark reduction in the number of new sites being founded in the peninsula’s regions, the conquest appears to have had an impact. Yet knowledge is lacking on the processes by which such change was brought about, to such an extent that responsibility for it cannot be pinned on the Romans with any certainty.

The old adage ‘the exception proves the rule’ is relevant in this case. One of the sites that might date to the 3rd century BC is La Romana di Castelromano, a very large Samnite fortress of c.56 ha, in the region of Molise, although Archaic pottery has been found there as well. Another is found in Etruria: Piazza di Siena near Petroio. The walls of Piazza di Siena enclosed an area of c.3.8 ha and have been dated to the 3rd century based on stratigraphic excavations undertaken there by the local archaeological superintendency in the 1990s. As can be seen from the photo, a limestone quarry has subsequently caused much destruction of the archaeological area, highlighting one of the many modern threats ancient sites are subject to.

The site of Piazza di Siena in norther Etruria

The site of Piazza di Siena in norther Etruria


Casalecchio di Reno

How the site of Casalecchio appears today.

Another site with both urban and non-urban characteristics was excavated in the area of modern Casalecchio di Reno. It is located c.4 km to the south-west of Bologna on the southern margins of the Po plain, on the left bank of Reno river. The site has been known since the 19th century from finds found in the area. Systematic excavations were undertaken by the French School at Rome during the 1960s, and taken up again in the late 1980s by the archaeological superintendency of Emilia-Romagna. As can be seen from the above photo, the area is now more or less beyond the reach of archaeologists due to the sprawl of modern conurbation. An Italian fighter-jet notoriously crashed into a school here in 1990, tragically killing 12 of its pupils.


The ancient site has revealed occupation covering a huge timespan, with periods of settlement dated from the Neolithic through to the Late Roman periods. During the 6th century BC the area was penetrated by Etruscan territorial expansion. Bologna itself had an Etruscan phase, when it is believed to have been known as Felsina. To the south of Casalecchio, further down the Reno valley, the remains of the famous Etruscan ‘colony’ at Marzabotto are to be found. Marzabotto is particularly noteworthy due to its precise rectilinear street system dated to the early 5th century BC. It is argued that this example of Etruscan town-planning reflects the early diffusion of surveying technology developed by the Greeks. Consequently, Marzabotto is one of the most important sites for the scholarly discussion on ancient urbanism in Italy.

At Casalecchio, a rectangular “city block” of c.100 m x 200 m was excavated by the French, similar in form and date to those known from Marzabotto, containing rectilinear structures with cobble foundations and tiled roofs. Yet in contrast to Marzabotto, this block appears to have stood in isolation rather than having been part of a more extensive urban system – the only such case I am aware of from my research on ancient Italy. Although the whole settlement at Casalecchio spread over c.60 ha (twice the size of Marzabotto), elsewhere within this area, the various archaeological features of this phase seem to be more typical of a dispersed rural site. Yet a planned block of structures would seem to reflect the actions of a centralised authority, further indications of which might be a rectilinear series of ditches excavated in another part of the site. These were interpreted as land reclamation works to both supply and drain the water of the nearby river.

The success of this site might be related to its location between the level area of the Po plain and the mountainous area to the south, accessed by way of the Reno valley. Although agriculture appears to have been a key element of Casalecchio’s economy, it also seems to have been involved in commerce and manufacturing.

This strange site was, like Marzabotto, affected by an influx of Gallic tribes during the 4th century BC. These tribes settled a very large area in the north-east of the peninsula, but initially they swept southwards and even sacked Rome, an event that left an indelible scar on the Roman psyche. Numerous tombs from Casalecchio indicate a Gallic presence down to the mid-3rd century BC.  A further phase of occupation is marked by the founding of a villa in the late Roman republican period, after nearby Felsina had been re-founded as Bononia, a Roman colony (the name from which modern ‘Bologna’ is derived).

Further reading, see: Ortalli, J. 1998 ‘Archeologia topografica: la ricostruzione dell’ambiente e dell’insediamento antico nell’esperienza di Casalecchio di Reno’, in R. Farioli Campanati (ed.) XLIII Corso di Cultura sull’Arte Ravennate e Bizantina, Seminario Internazionale di Studi sul Tema “Ricerche di Archeologia e Topografia”. Ravenna, 22 – 26 marzo 1997. (Ravenna: Ed. del Girasole), pp. 565–606.


A view of the site of attributed as Anxa. The lake once covered the level area in the foreground where there are now fields.

A view of the site attributed as Anxa / Angitiae Lucus. The lake once covered the level area in the foreground where there are now fields.

One of the sites that has fascinated me most is the strange settlement/sanctuary site on the western side of what was once Lake Fucino. This area is associated with a pre-Roman people known as the Marsi. Because of its location, it is believed to be the site of Anxa, otherwise known as Angitiae Lucus. Both of these names are known from ancient literary and epigraphic sources relevant to this location. The original Iron Age settlement appears to have been located at the top of the slope, which at some point had a fortification wall built around it. Subsequently, a new and impressive polygonal wall was built around a much larger area, stretching from the lake shore, up the steep slopes so that it enclosed the crest of the hill 200 metres above. Three separate trenches were dug against this wall by archaeologists. The recovered artefacts indicated that it was constructed in the second half of the fourth century BC.

Several features of this site aroused my curiosity. It is located only 11km to the south of Alba Fucens, a Latin colony founded by the Romans in 303 BC. If the new fortification of Anxa was built in the period 350-300 BC, as proposed, was it built before or after the Romans took control of the area? I don’t know the answer and, either way, it is difficult to explain. Based on the area enclosed by the walls, Anxa is c.30 ha in size, and Alba Fucens 32 ha. I can’t think of a reason why the Romans would allow (or instigate) the construction of such a large fortification in a recently subdued area so close to one of their colonies. Superficially, it would seem to be giving potential enemies a means to defend themselves. If it pre-dates the Roman subjugation of the area, the wall seems not to be typical of those from other Marsic sites.

The line taken by the fortifications meant that much of the area within the walls comprised a slope so steep that it is difficult to imagine it was suitable for habitation. The site had a strong religious association, and at least two temples have been excavated inside the walled area. ‘Angitiae Lucus’ can be translated as ‘the sacred grove of Angitia’, a local Marsic goddess, and various artefacts have been found nearby that are believed to have been deposited during rituals associated with the cult (votives). Did the wall then just enclose a sanctuary? Certainly, walled sanctuaries are known elsewhere in the central Appenines, built by the Samnites.

If the identification with Anxa is correct, then it became a municipium at some point from the 1st century BC onwards, and thus became a proper settlement if it had not been one before. It has also been suggested that it had a port which, given its location on the lake shore, seems likely. Yet the Romans attempted to drain the lake, and indications of ancient land-divisions marking out plots of cultivated land have been identified from aerial photographs in the area directly to the east of the site, where the lake once stood. So maybe Anxa lost its harbour facilities at some point, or they were moved.

If you are interested in where it is, just copy and paste the following coordinates into the ‘search’ field of Google Earth, and press return: 41°58’13.87″N, 13°27’37.19″E  You will see the excavated temples directly to the east of the yellow pin, and the large area of where the lake stood is easy to make out by the modern rectilinear field systems that cover it.