Roman Villas and Rural and Coastal Infrastructures

Amenities such as roads, aqueducts, bridges, warehouses (horrea, sing. horreum), jetties, and port facilities on rivers or the sea passed through rural landscapes and along coasts to connect urban centers with one another: They also connected villas with other villas and villas with towns and markets overseas. Shipbuilding itself in various shapes and tonnage (from rowboats to sea-going vessels) was strongly developed in antiquity.¹ These physical expressions of Roman infrastructure were accompanied by spatial organization: regions in which land was surveyed and subdivided into agricultural units (centuriation) in different ranges of size or local practices, often with standardized definitions of property types (farm, forest, pasturage) and valuation for taxation or other purposes, had been effected by Greek colonial poleis but became a distinctively Roman practice.² In its origins, the intention of such building activity and spatial arrangements may well have been military, namely as means to establish a sustainable Roman hegemony over great distances and to provide viable plots of agricultural land for veteran soldiers, the citizens of newly established towns, or newly colonized existing urban centers, some in Italy, many overseas. The new infrastructures ultimately allowed several cities to grow beyond the constraints of their immediate hinterland; facilitated travel of all kinds; bolstered private, official, and commercial communications; enhanced transport of agricultural and manufactured goods, including raw or partially prepared materials such as timber and marble; and, ideally, fostered stable rural and urban populations. As a result, cities flourished wherever Rome laid down its solid and spatial manifestations, in a seemingly never-ending process of conquest and, later, of continual expansion of colonial and municipal rights and responsibilities to new provinces. Conquest, construction, and coordination of urban centers with rural space and production were the means by which the Republican empire grew beyond Italy in the second century BCE, and these were the premier devices of the expanding imperial system later on. Proud cities and towns came to be no less a Roman phenomenon than they had been for the Greeks, and the success of urban centers in turn prompted further expansion of infrastructures by municipal, provincial, and imperial authorities.

Villas, Urban Architectural Typologies, and Roman Economic Expansion

Roman architectural forms – the most famous ones, at least – are associated with towns and cities. Their components were for specific institutions: fana or temples, big and small, in various architectural orders and plans, for local or Latin gods and later for the imperial cult; fora for commerce and assembly; basilicae for law-courts and places for the exchange of commercial information; curiae for the meetings of city-councils; macella (sing. macellum) or market-buildings for shopping and goods-exchange; fountains, sewers, and latrines for urban hygiene; large or small (public or private) baths (balnea, thermae) for sociable relaxation and personal hygiene in grand or intimate venues; theaters, amphitheaters, and music buildings (odea, sing. odeon) for entertainment; paved roadways for urban convenience; ceremonial streets for processions and semi-sheltered pedestrian traffic; arches for commemoration; and fortifications and camps for military deployment. The architectural forms could be very specific (basilicae, for example, usually required a dais for the magistrate to set up his chair and macella might have a fountain for washing goods for sale, building clean-up, and hygiene). However, the architectural language used – columns in various orders, porticoes, pilasters, ceremoniously decorated doors, and so on – could be readily swapped among different kinds of buildings. Such buildings and the institutional entities they housed and promoted were essential to Roman presence: places of assembly, points of trade and commerce, government officialdom, bureaucracy, some standardization of languages and laws, permanent or periodic military presence, universal taxation, and civic and business organizations. These physical and institutional components were among the distinctive contributions of the later Roman Republic when, starting in the second century BCE, Rome established an empire in the Mediterranean basin in Hispania, North Africa, Greece, and Asia Minor and, later, in Gaul, Germany, and Britain. All were intensified, multiplied, and expanded geographically and architecturally with the unification of powers in the imperial period, which began with the principate of Augustus (after 29 BCE) and continued for some five centuries through the late antique period (later third through the fifth century) until the ultimate and definitive “loss” of the Latin-speaking Western provinces of the empire, namely in the late sixth century CE.

Urban houses – domus – were also essential components of Roman architectural forms, in part because houses themselves became bearers of meaning and movement, a signifier of and for the family and its inhabitants. Their importance was not reserved for the elites. The urban houses of Pompeii, Herculaneum, Cosa, and Ostia could be grand or very simple, but even the simplest ones had touches of amenities: one room with a pretty mosaic floor, a decorated door, a small garden. For the upper classes, the domus (pl. domus) was, throughout the empire, the main shelter for private citizens within a familia (an immediate family and its extensions together with servants, slaves, and liberti, these last being manumitted slaves under continuing obligation to their former masters). As such, domus could be very grand affairs, sometimes in sober old-fashioned décors and honorific objects (wreaths, statues, inscriptions) proclaiming the antiquity of a family, at other times in up-to-date modern styles advertising wealth and fashionability. Domus were venues of public representation and political transaction between an incumbent head of the family (paterfamilias) and/or his wife (matrona, domina) and the incumbent’s supporters and adherents (clientes) in a strong system of social relations known as clientage as it developed in the mid- and late Republic and was maintained, in varying ways, during the empire.³

Villas were also bearers of distinctively Roman meaning. They were venues of social interaction as well as centers of profitable activity – in all its meanings – in the countryside or the littorals. Like the domus, they were sites of personal and political hospitality of all kinds, places of retreat, pleasure, and even learning, but most often justified by their productive capacity in the way of agriculture or other lucrative endeavors. Pliny the Younger commented that he had a full granary (horreum plenum) at one villa and a full bookcase (scrinium) at another, the villas being profitable agriculturally and productive intellectually.⁴ The expansion of villas into the Mediterranean was an important aspect of Roman economic presence and organization to supply food for cities, government, and armies, but with their proliferation came a social and cultural sharing of values that bound the elites of the empire into a recognizably Roman society. And while villas could be expressions of sophisticated wealth and cultural achievement, they could also provide a backdrop of rustic wholesomeness (either real or affected) and traditional values based in ancient morality.

With these assets and advantages, villas unsurprisingly came to dominate the Roman countryside and coasts. Strabo (late first century BCE) noticed that villas were increasing in number and changing the aspect of the landscape: of the coastline on the Bay of Naples, he said that their solid walls made it look like a city.⁵ A century later, Pliny remarked on the view of villas from his own villa at Laurentum, just south of Ostia, that it had “the aspect of many cities.”⁶ This repletion of the coasts and the countryside of Italy with villas corresponded to the fully domesticated environment of the peninsula by Varro’s day in the 30s BCE: Varro comments that all Italy resembled a huge single orchard.⁷

What was good for urban centers was good for villas. The Roman Mediterranean was largely an agricultural entity of villas and their estates with strong physical, transportation, and monetary infrastructures that facilitated seaborne and land transport commerce for several centuries and over fairly long distances.⁸ Villas depended on this. Consistent, diverse, quickly transportable production of food on well-capitalized and prudently developed farms is the normal expectation of all manuals describing villas and their agricultural estates, from Cato in the second century BCE to that of Palladius in the late fourth to early fifth century CE. Producing raw agricultural materials (mainly wheat or other grains) to satisfy merely local subsistence needs had long been bypassed: Instead, villas shipped ready-to-store grains for local milling or processing elsewhere as well as other produce, in addition to providing commercial brokers with processed agricultural goods with a longer shelf-life – notably wine, olive oil, table-ready olives, various preserved or pickled fruits and vegetables, salted fish and fish sauces (garum, allec/hallec, liquamen) – which could be graded and traded, often through middlemen, off the agricultural estates themselves and away from the villas. Pliny the Younger, for instance, sold his grapes on the vine in advance of harvest time, thus leaving to the buyers (negotiatores) the harvesting, processing, transport, and the risk of possible crop failure due to bad weather or pests.⁹

Villa owners needed a place to sell their produce, either in fixed town or urban locations equipped, in some cases, with convenient, sometimes impressive permanent market buildings or in periodic market days organized regionally called nundinae.¹⁰ Known cases of villa owners organizing periodic markets on their villa estates included the future emperor Claudius himself, the ex-praetor Bellicius Sollers (a contemporary of Pliny the Younger), and the senator Lucilius Africanus, this last authorized by decree of the senate, in 138 CE, to have a market on his African estate.¹¹ Significantly, Lucilius Africanus’ property was in regione / Beguensi territorio Musulumianorum in Byzacena (el-Bejar), that is to say, in the modern Kasserine region of Tunisia, where abundant archaeological evidence for intensive olive cultivation, and possibly wine production, exists in the form of many Roman rural sites equipped with multiple presses.¹² This was one of the regions whence olive oil was exported to Rome and other areas of the empire, and the trade in these staples was important to the food supply in both the open market and the government-subsidized distribution systems. Earlier, Sollers may have been organizing a market on his estate to attract brokers or government agents and thereby establish a price for his product.¹³ In any case, owners of villas and farms were at all times in competition with one another and anxious to get their produce off the villa estate and on its way to get the best possible prices at markets, auctions, and/or from brokers.

Fresh and longer shelf-life produce were also concerns. Assuring food supplies both locally and in Mediterranean-wide markets animated the productivity of rural estates for fresh comestibles of all kinds, storable commodities such as grains and flax, meat (mammalian, avian, and piscatorial, fresh and salted) from husbandry, hunting, fishing, and finished products like olive oil, wine, fish/fish-sauce, and tanned or untanned hides. If they were in a good locations near cities or towns, villas produced seasonal goods with a short shelf-life called pastiones villaticae (sing. pastio villatica) for choosy urban gourmets or great private or public feasts, but also for ordinary but demanding urban consumers. Pastio villatica included perishable fruits, meats, game-birds, fish and shellfish, honey, even flowers and plants. Such goods could be a source of considerable revenue from villa agriculture.¹⁴

Other urban needs were fulfilled by villas and estates. Depending on their location and what the estate offered (clay beds, stone quarries, woods), villas often produced bricks and roof-tiles, amphorae, timber, charcoal, barrels for transporting goods, bins, and other useful things essential to the local economies and the new Roman economy of the Mediterranean. These products required all the infrastructures that cities and towns – and for that matter, the military – also required: roads, bridges, and so on, as well as such special interventions (often imperial) as irrigation systems and draining of marshes. Assuring, implanting, and expanding the productivity and animation of inland and coastal landscapes were important economic functions in the Roman Mediterranean, and villas were one of their means. Villas and their estates also dominated and came to clarify the countryside, making it comprehensible and meaningful – these are issues to which we will return in several different ways.

Villas in Official Landscapes: Centuriation, Distribution, and the Alimenta

Villas existed in landscapes, and Roman landscapes were conditioned by being subjected to official organization. Such supervision most often took two forms: the first was centuriation, a series of physical and legal acts that defined agricultural land in terms of measurable standard units and in view of its character (marginal, hilly, marsh-land, woods, and so on) and catalogued it in recordable form for administration by local authorities. A second specific measure was the alimenta Italiae, a scheme apparently devised by the emperor Nerva but implemented by his successor Trajan in the second century CE. This was a means of local charity, paid for with a perpetual obligation incurred by (or on) local owners of estates in Italy. The scheme was as follows: to local municipal authorities, the imperial fiscus sent a certain sum to be disbursed as loans to Italian proprietors, with the annual interest payments on the loan principals used for the monthly maintenance of local children, especially boys. Various similar private schemes, which took their cue from this imperial initiative, are attested epigraphically in Italy and the provinces.¹⁵ One such famous case is the scheme devised by Pliny the Younger in support of freeborn children in his hometown, Comum.¹⁶ Because the alimenta were limited to Italy (perhaps not in all regions of the peninsula) and the length of their existence is uncertain, we omit discussion here.

A third condition of the landscape affecting villas and estates was certainly the distribution of settlement and habitation in relation to the local topography, proximity to rivers, roads, sources of water and irrigation, towns and cities, and many other variables. Modern exploration in the form of aerial and field surveys – for the most part noninvasive – have contributed mightily to our knowledge of villas and the rural landscape, and below we give a very brief account of ongoing Mediterranean-wide projects.

Centuriation

Agricultural production of all kinds was the principal generator of wealth – small, middling, and large – in Greek and Roman antiquity. In consequence, the structure of the productive landscape came to be a matter of official cognizance and technical intervention, associated with the ager publicus or “public land” made available to Roman citizens in annexed regions beyond the City itself or its immediate territory.¹⁷ These lands were reorganized with new surveying, along Roman lines, of territories adjacent to, and administratively dependent on, newly conquered old towns or else newly founded urban centers. In part, the purpose of such surveying was to define parcels to be awarded to soldiers upon demobilization or to attract settlers to newly formed colonies or new drafts of citizens to existing towns, both in Italy and elsewhere. The award of agricultural land became an important aspect of late Republican politics.

The acts of mensuration were very well developed technically, capable of both universal uniformity of application and flexibility vis-à-vis special circumstances of topography and local land practices. It was applied in Italy, Gaul, Hispania, Greece, and Africa; its application elsewhere is less certain, and it may have been flexibly applied, or not applied at all, to respect and accommodate local practices. Trained personnel called agrimensores (“land-measurers”) were developed together with specialized handbooks and mechanical devices for sighting, finding true north, and estimating heights.¹⁸ The result was a system of physical divisions of agricultural land in more-or-less equal, measurable, cataloguable, and registerable units reducible to valuation for tax estimates (for land outside Italy) or commercial sale prices. Once established, the divisions could be assembled in municipal lists of properties called cadasters for administrative purposes. The act of divisioning itself was termed centuratio (sometimes also limitatio), from an original delimiting into square “centuries” corresponding to an old assemblage of 100 individual properties (heredia, sing. heredium), each heredium (consisting of 2 iugera or c. ½ ha or 1.3 acres) capable, in theory and antique tradition, of sustaining a nuclear farming family. In practice even quite small farms were much larger (18–20 iugera or more = 4½ -5 ha = 12–14 acres), and medium-sized farms at some 100–250 iugera (50–125 ha = 130–325 acres).¹⁹ The centuriation units developed in size and layout according to local needs, but usually toward larger defined plots of agricultural land. Ultimately, by the late second century BCE, official grants in units of 500 iugera (250 ha or 650 acres, quite ample at any historical period) were offered to Roman citizens willing to take up residence overseas in the environs of Carthage.²⁰

How centuriation related to villas is not certain, because by the second century BCE or later, when villas clearly surface in the archaeological and written historical record, the amalgamation of originally separate smaller properties may already have occurred, and the neat divisions of agricultural land intended for distribution to farming families had been bypassed. It may well be that the proliferation of villas actually disturbed or overthrew the original intention and purpose of the centuriatio. For that reason, the technical issues of centuriation need not occupy us here. Rather, their discovery in quite disparate parts of the empire is part of the recordable expansion of Roman hegemony and thus of the villas that eventually occupied the surveyed territories.

Archaeologists become enthusiastic when traditional and new disciplines coincide and mutually confirm one another: the discovery of centuriation is an example. Its existence and processes had been known from the texts of the agrimensores and from certain inscriptions that seemed to be lists of local properties; historical land divisions could also be inferred as the underpinnings of existing divisions from modern large-scale maps that seemed to show a ghostly divisioning underlying visible land features and delimitations of more recent date.²¹

However, the discovery of centuriation is one of the most exciting stories of modern archaeological investigation by technical means. In the late 1940s and 1950s, aerial photography from low-flying military, map-making, or tourist aircrafts began to register what seemed to be man-made interventions on natural or agriculturally developed land in rural or suburban areas.²² Among many others, these could be in the form of topographical discontinuities (buried remains of walls and terraces invisible at ground level), hydrological anomalies (e.g., ruined cisterns retaining some water), and color or height variations in modern monocultural planting due to subsurface structures. For centuriation, the foot-worn tracks between surveyed fields could appear strikingly visible, as could the low earth ridges and/or ditches defining units of the divisions. Substantial structures including villas, but also rural temples and other buildings, could be revealed. The results of aerial photography were clinched early on by epigraphic evidence, most specifically in the cadastral lists, inscribed in 77 CE on marble slabs, of properties in the region of Arausio (mod. Orange) in Gallia Narbonensis.²³ This had been a colonia or newly founded town set up for veterans of the Legio II Gallica, and the regular divisioning and apportionment described in the inscriptions coincided with the design of field boundaries revealed by aerial photography.

The discovery of centuriation patterns has resulted in significant new understandings of Roman territorial expansion – and therefore proliferation of villas – in the western provinces and Africa. The analysis of rural space as Roman space has proved fruitful for both local phenomena and for imperial history; Monique Clavel-Lévêque is a significant exponent, as is Emilio Gabba.²⁴ Methodological considerations began early and are ongoing.²⁵ Other noninvasive geophysical methods for archaeology were developed in early form around the same time as aerial photography, also to the advantage of documenting Roman villas and with ongoing refinements of technique and results.²⁶ In recent years, the most notable new noninvasive techniques that are greatly contributing to archaeological mapping and planning of field campaigns are LIDAR (= light detection and ranging) and remotely operated cameras on drones.²⁷ The former has the ability to detect man-made structures (or micro-topographical features) even under very dense forest canopy; the latter allows the capture of low altitude images (even using a heat-sensing camera) much more quickly and cheaply than in traditional aerial photography.

Field Surveys: Distribution of Villas in Landscapes

Where are villas in their landscapes? What kind were they? How were they related geographically to other rural structures such as villas (clustering for reasons beyond the impetus for social proximity), means of irrigation, and means of transport and access (roads, rivers, and coastlines)? What did they replace, and what replaced them? These are the questions that can often be addressed (if not completely answered) by modern techniques of field survey. Like centuriation in aerial photographs, field surveys can be large scale. They can be reduced to maps keyed with symbols for the various man-made and natural phenomena to reveal patterns not fully visible on the ground. However, centuriation maps and field survey maps differ: Centuriation maps show the traces of administrative structures over large areas, while field-survey maps show how ancient societies distributed themselves – people, structures, and resources – in the landscape, sometimes in territories, sometimes in minute areas. Field surveys have contributed greatly to the knowledge of the ancient territory and gathered a large amount of archaeological data on rural settlement patterns and chronological trends. Over large areas, the landscape presents itself as a palimpsest, natural, then by man-made elements written on it, erased, and overwritten many times with interventions of all kinds. In consequence, field-survey results can be presented both in historical layers and as types of presences (farms, villas, cisterns, roads, and so on) when the visible remains are datable or identifiable. The effectiveness of field survey as an investigative tool depends on the type of terrain: Relying on teams of people walking through transects at regular intervals and noting what is visible on the ground, it is most effective in plowed fields and less so in areas with dense vegetation. The ability to date the use of a given location with some degree of precision depends on the recovery of diagnostic finds, most notably pottery sherds, which are both durable and common in the archaeological record.

Of course, villas are by no means the principal topic of field surveys, because the landscape is a chronological overlay of phenomena: Surveys can go from prehistoric times to the present. Still, the first case of an intensive and large-scale survey conducted in the Mediterranean regions did, in fact, prioritize villas in its results, mainly because it was undertaken in the ager Veientanus north and east of Rome, a region rife with ancient farms and villas. The South Etruria Survey, directed by J.B. Ward-Perkins from the 1950s to the 1970s, was and became a milestone in Mediterranean archaeology.²⁸ In field surveys, tell-tale signs can both reveal villa sites and differentiate them as to type, date, and even social status. For example, a spread of tesserae for mosaic floors in association with certain types of fine pottery can denote a villa of some decorative pretension and therefore higher status, whereas fragments of lesser-quality floors and crude pottery may indicate a farm and even a date.²⁹ In combination, a residential part (pars urbana), namely a residence with some claim to comfort or even luxury, and a service quarter (pars rustica), namely an agricultural section devoted to farm processes, can be inferred. In turn, even small traces of masonry are datable and give evidence of Roman rural presences at various times; sporadic finds of inscriptions or hardened tracks and road pavements can make connections among settlements both chronologically and geographically. The reach of such material in relation to the hydrography and proximity of villas to landscape and habitation centers can establish when and why villas were built and what their sustainable existence might have been. The processes of field survey are, in principle, noninvasive, but they can also locate important sites for excavation or preservation.³⁰

Villas in Agricultural Treatises and in Vitruvius

In discussing rural domestic architecture, writers on agriculture as well as Vitruvius had something to say about villas, but only something. Agricultural treatises were an important literary and practical genre in Latin literature, and its practitioners (except for Cato) combined the authority of previous Greek and Punic writers as well as earlier or contemporary authors with their own knowledge. For that reason, Latin agricultural treatises preserve, in ways that often seem adapted to Italy or, at least, to the southern parts of the northern Mediterranean coasts, an international culture of agriculture from at least the time of Theophrastus, the successor to Aristotle in the philosophical school of the Lyceum at Athens in the fourth century BCE. Most of the very numerous works on agriculture (some very specialized as to geography and topic) from Greek, Punic, and even Roman antiquity have not survived, but four Latin treatises spanning some six centuries are extant, namely:³¹

1. 1. The De Agricultura of Cato (Marcus Porcius Cato, 234–149 BCE, whose treatise appeared in the mid-second century BCE),

2. 2. The De Re Rustica of Varro (Marcus Terentius Varro, 116–27 BCE, who published his treatise in 37 BCE when he was in his 81st year),

3. 3. The De Re Rustica of Columella (Lucius Junius Moderatus Columella, c. 4–70 CE, whose book was written mid-first century CE and published with an earlier treatise on trees, De Arboribus), and

4. 4. The De Re Rustica (also called the Opus Agriculturae) of Palladius (Rutilius Taurus Aemilianus Palladius, fl. Late fourth–early fifth century CE, with a similar date for his treatise; the treatise includes a book on veterinary medicine in sheep and a book [De Insitione], in verse, on grafting for stable product and optimum harvest).

   To the agricultural treatises can be added:

5. 5. The De Architectura (book 6.6–7) of Vitruvius (Vitruvius Pollio, fl. late first century BCE, with a dedication to Augustus as emperor, therefore after 27 BCE), in which the author, a military and civil architect, discussed some characteristics of rural dwellings in relation to urban domus in very abbreviated form, with a longer section on the villa rustica and its utilitarian buildings.

The authors of the Latin agricultural treatises borrowed from earlier ones or from other sources, except for Cato, who cited no earlier or contemporary authorities and appears to have relied on his own experience. Still, Cato became the touchstone for later agricultural writers and apparently an impetus for others whose works have not survived.³² The advice he gave to an owner (whom he calls pater familias and whom he assumes is an absentee landlord but one intimately involved with the operation of the property) was continuously cited for six centuries by later authors despite his haphazard and disorganized presentation in twelve books and some 162 sections. The authority of Cato’s text may come from the fact that his is a very early example of a prose book in Latin as well as his reputation, long after his death, as the cynosure of traditional values and antique achievement, the embodiment of the mos maiorum. Cato speaks at some length about crops, vine- and olive-cultivation and processing, as well as giving contract language, prayers, and recipes. Other topics – specifically about the villa itself – are telegraphically conveyed: selection of soil types, general location of the farm (at the foot of a hill and facing south, near a town for markets, nearby water for irrigation, available manpower, and close to a river, the sea, or a reliable road for transport).³³ A tabulation of building costs (whether for the residential part or the farm is not certain) is also included.³⁴ That’s about it. Almost nothing is said about the villa itself, neither the dwelling for the pater familias, his relatives, and guests, nor the facilities for the personnel, beyond two economical but cursory prescriptions: “When you build, seek not [to build] the villa less than the farm, nor the farm less than the villa,”³⁵ and a vague recommendation to build a good villa urbana so the owner will want to visit it and keep an eye on the farm, thereby mitigating absenteeism.³⁶ The recommendation may be sarcastically intended, a rebuke to town-dwelling owners of rural properties.

In fact, despite the generosity of their insights, the clarity of their writing, and their evident experience, agricultural writers are quite laconic as regards the architecture of residential villas, and apart from moralistic judgments, there is nothing specific about decoration or contents, either physical or in terms of meaning. Their silence must have been intentional, because by contrast they wrote at length on how horrea, stalls for livestock, and presses of various kinds are to be built, and Vitruvius – similarly brief on villas – included detailed instructions on the construction of foundations, cryptoporticus (covered or below-grade passages and rooms), terraces, and terrace walls, the architectural elements of the platforms of basis villae and other villa structures, but not much on the villas themselves.³⁷

The silence is understandable: Treatises on agriculture focus on agriculture, agricultural processes, and nonresidential farm buildings. Still, Cato, Varro, Columella, and Palladius all assumed that a residential villa for the owner was an essential part of the estate even though their comments on the topic are very limited. Indeed, there may have been a commonplace in elite morality that regarded villas – and especially luxury in villas – as indecent and typical of vile modern attitudes: Tremellius Scrofa, who wrote on agriculture and husbandry in the late first century BCE, remarked sarcastically that present-day owners were more concerned with the orientation and temperature of their summer and winter dining rooms than with the best placement of windows in their wine and oil warehouses.³⁸ Scrofa’s remark at least shows that owners of villas were concerned – if obtusely and epicenely – with the architecture and comfort of their residences. However, because the agricultural activity at villas was held in the respect that tradition deserved and that recalled ancient simplicity, describing modern grandeur or even comforts such as baths or rooms with pretentious Greek names was avoided, except to mock or excoriate them.

In addition, with very few exceptions, actual financial return on capital or agricultural investment is lacking in much detail in these authors,³⁹ possibly because talk of money was regarded as vulgar among the elites and thought of as the distinctive mark of freedmen (liberti, former slaves liberated by their masters but still considered socially inferior no matter how wealthy or high up in administrative hierarchies).

Agricultural writers and Roman or Italian “antiquities”

The success of Roman military ventures and the expansion of Rome’s Mediterranean power had an effect on both Hellenistic Greek and Latin writers: they began, with some complacency, to define Rome’s identity. In the aftermath of the victory over Carthage in the Second Punic War (218–201 BCE), Romans began to ruminate their own history and cultural phenomena (both ancient and contemporary), in part prompted by Greek historians – notably Timaeus of Tauromenium and Polybius – who had included what we might call anthropological or ethnographic material in their accounts of the expansion of Roman power. Cato, besides his De Agricultura and other writings, wrote a long account called Origines on the antiquities and traditions of Italian cities with special regard for Rome itself.⁴⁰ Cato’s treatise on farming was a self-contained compendium combining his experience with what he claimed was traditional practice as well as good tips and recipes for his fellow villa owners, but it came from the same impulse that prompted his recording of antique lore.

A little more than a century later, Varro’s Antiquitates dealt mainly with Roman lore and ancient curiosities. Its publication was followed by his three books on contemporary agriculture, De Re Rustica, written, he tells us, in his 81st year of life (c. 34 BCE). The association between picturesque antiquities of Rome and Italy and current Italian agriculture formed a strong bond. According to these authors, the frugal and moral exemplars of the past found their prolongation into the present by means of farming methods and the morality of farmers themselves, Romans of the old stripe. The connection between farming and the nearly mythical social and moral vigor of the past – in Latin, the mos maiorum (literally, the custom of the ancestors) – seemed obvious to them. Cato gives a list: men who are good farmers are brave men and eager soldiers, they are held in respect and make a good livelihood; they are rarely dispirited. The list is, for all practical purposes, the personal characteristics of those who embodied the mores maiorum (including Cato himself).

Latin itself – its locutions associated with farming that were still in use – could also be cited as living instance of antique tradition. Varro wrote extensively on the Latin language, the etymology of its words and the derivations of its grammar and syntax. Not a few of his topics concerned the derivation of words from agriculture, and it was no accident that for the word villa he claimed an origin in the movement of produce from and to the production-site, as follows: the word for the manager of an estate was vilicus, related whence and to the place agricultural goods were transported (convehuntur fructus et evehuntur). Simple workers (rustici) called cartage roads for produce veha, and the place whence and to which produce was conveyed vella rather than villa, and “to haul produce” was “to do vela” (facere velaturam). In another passage, Varro bases the family names of the most ancient Romans in the quaint traditions of husbandry (Porcius [pig], Equitius [horse]) and the branch designations of families from the same: He cites the Annii Caprae [goats], Statilii Tauri [bulls], Pomponii Vituli [oxen].⁴¹ History, agricultural language (ancient and in current usage), and famous ancient names were proof of the antique prestige of agricultural pursuits and, by implication, villas. The priority and superiority of life in the country over city dwelling was a commonplace in Latin literature, and villas were venues to smoothly recreate antique mores in modern times.⁴²

Villas and Theory in the Agricultural Writers and Vitruvius

Cato’s lead in agricultural writing had assumed that owning a villa was a sine qua non, a necessity at a certain social status: any discussion of its theory and/or morality except the fact of proprietorship was so obvious as to be otiose. However, later authors targeting different, perhaps more educated but more dispersed readers at other social levels sought to provide a mild theoretical and moral framework with which to give meaning to the agricultural and villa-owning activity.⁴³

Setting aside the specifics of its discourses on agriculture, Varro’s agricultural manual De Re Rustica urges a balanced approach: the usefulness of a villa and its estate (utilitas) is primary, but its pleasurable aspect (voluptas) is important as well, as much for its own sake as for discouraging absenteeism by the owner (the anxiety about owners’ indifference and absenteeism was universally present in Latin agricultural treatises). Then again, frugality in management (diligentia) trumps mere luxury (luxuria) in villas.⁴⁴ Indeed, Varro is critical of excessively lavish villas (villae urbanae, mostly located in the immediate environs of Rome, also called villae perpolitae) whose owners disdained agriculture: He considered that their negligence drove up prices of food staples.⁴⁵ In his work, which takes the form of an imaginary dialogue between various characters, Varro juxtaposes the “unproductive” (inutilis in Latin) villa of Appius Claudius Pulcher, crammed with works of art and not much else, with Quintus Axius’ villa, which, while elegant, retained its original function as a unit of agricultural production, including breeding of pigs and horses.⁴⁶

Varro’s treatise had the advantage of clarity in describing the deployment of the discrete parts of villas: He distinguishes the categories of space and use, though he does not fully describe them. Thus, his term, like Cato before him, for the residential villa is pars urbana). Other buildings devoted to agricultural processes of all kinds are termed the pars rustica, and storage facilities, larders, grain bins, and so on are the pars fructuaria. These designations entered the language of agricultural manuals and became permanent.

Of course, there were exceptions to frugality in villas: interest by owners in the architecture, decoration, and pleasure of villas was endemic. Freed from the constraints of confined urban real estate and the traditions of domus, space and imagination could be fully deployed at villas. Varro himself – the proponent of utilitas – is his own prime example. In the third book of his agricultural treatise, he describes a special aviary-dining room connected to the peristyle of his own villa at Casinum (mod. Cassino), south of Rome. Here, the owner and author combined literary ekphrasis with design ingenuity and simplicity: With some delight, he describes a garden with two long pools framed by wooden columns on three sides. Netting enclosed the spaces both between the columns and facing the garden to form cages for his collection of birds, and the colonnade continued in a semicircle (also netted for more birds) at the end of the garden. The semicircle framed a circular pool surrounding a ring of columns enclosing a round triclinium, the whole complex having devices for convenience and diversion without pretension.⁴⁷ Villas could make for originality within a tradition of plain elements (the wooden structures) and arrangements for hospitality, and the avian collection itself was a sign of both otium and the learned pursuits (in this case, zoological) appropriate to villas and their owners.

Columella, picking up and further developing the concepts of Varro’s work and repeating the partes urbanae, rusticae, and fructuariae division, wrote in his agricultural treatise De Re Rustica of the importance of equipping the villa with comfortable and well-appointed residential quarters in order to lure the absent landlord to visit his estates on a regular basis. Like Cato and Varro before him, the estate’s operations could be entrusted to a trustworthy manager (either a freedman or a slave) to assure the villa’s productivity and relieve the absent owner of irksome duties.⁴⁸ That is virtually the whole of Columella’s remarks on the architecture of rural residences. Such generalities also characterized Palladius’ treatise in the late fourth or early fifth century, though his names and analogies for villas and their parts are partially borrowed from military terminology or civic architecture: The pars urbana of an estate is called a praetorium as if it were the chief officer’s headquarters and residence in an army camp, and the disposition of the cella vinaria is described by reference to a civil basilica or law court.⁴⁹

Vitruvius, in his De Architectura, urged that elegant residential quarters ought not to impede the functionality of the service and production quarters of the agricultural estates themselves.⁵⁰ As for all houses, the Vitruvian principle of decor (suitability) was applied: The size, decoration, and architectural deployment (including the number and character of rooms) was to be congruent with the social status of their owners and inhabitants. Otherwise, he refers his readers to his preceding account of urban domus as a clue to the organization of houses in the country (he calls them aedificia pseudourbana or pseudo-urban buildings). The only variation is that, unlike town houses wherein atria are in the front of the dwelling, peristyles were the first main space encountered in country houses, with the atrium set to the back of the house amid porticoes and walkways in or overlooking gardens.

The writers on agriculture as well as Vitruvius were providing a framework for thinking about villa ownership: Cato’s insistence on its traditions, Varro’s on its useful antimonies (utilitas/voluptas), and all the writers offering vague prescriptions for balance and sobriety in architecture and some disdain for modern luxury and pretension. Actual architecture hardly appears at all in these writers, so the Roman account of villas must depend on other sources.

Latin Agricultural Treatises in Later Manifestations

The persistence of the agricultural treatises (see above, pp. 7–9) in Roman readership through late antiquity presaged their continuation and revival in later times, and while the tenth-century encyclopedia known as Geoponika⁵⁸ is a compilation of both known and now lost agricultural treatises without specific citation, it attests the continuity of interest in earlier practices. Later, the European rediscovery of the agricultural treatises came to be published in printed form in the later fifteenth and early sixteenth centuries (e.g., Venice 1472, Bologna 1494, Paris 1512, Venice 1515), in editions more Latinate than practical. Some texts were published singly, but most were groupings of Cato, Varro, Columella, and Palladius with added material from Virgil and other writers, sometimes also with modern treatises on special subjects.⁵⁹ Translations followed in the eighteenth century, though these were in French and English, so to whom they were addressed (in view of their impracticality in northern agricultural lands) is not certain.⁶⁰

Instrumenta: Equipment and Personnel

Varro, in his book on agriculture, outlines the three categories of instrumenta or essential equipment of agricultural estates as follows: instrumentum mutum (“non-speaking” tools, by which he means rope, carts, hoes, ploughs, and so on), instrumentum semivocalis (“sound-making” tools, meaning animals, both beasts of burden and herded animals), and instrumentum genus vocale (“talking hardware” or tools capable of speech, namely slaves or other personnel).⁶¹ His categorization is useful in our discussion of villas because, quite obviously, villas and farms need on-hand inventory and good husbandry techniques. Of course, on-site personnel of all kinds – whether free and hired, bonded in some way, or enslaved – were important as well. Such social differentiation may have inflected the plan of villas, but slavery in agriculture has also inflected the modern interpretation of evidence – archaeological and literary – about them (see pp. 51–2).

To be viable, villas required substantial capital outlay for specialized equipment, animals and their maintenance, and servants and slaves, including managers and household slaves, not just agricultural laborers.⁶² Even ships (in the case of maritime villas), together with wagons, mules, skins, and barrels, and all the items needed to transport the produce from the villa by land or sea, were categorized as belonging to an estate; in cases of the sale of properties, they were by law and precedent included in the transaction.⁶³ Although agrarian legislation was enacted throughout the late Republic and empire,⁶⁴ villas are not addressed directly in the late Roman compilations of laws and legal glosses (the Theodosian Code, issued in 438 CE and the Digest of Justinian of 530–3 CE) beyond definition of the types of property and ownership conditions to which land was subject. However, the Digest contains an impressively long list of legal opinions in the matter of farm equipment (including slaves) at villas and houses undergoing probate for inheritance as well as lists of “servitudes” (obligations permanently attached to certain properties). There was evidently lively litigation about farm equipment and personnel, implying that villas were a matter of jealous attention by litigants and serious concern for judges and legal experts.⁶⁵

Because wealth was, for the most part throughout antiquity, generated from growing and processing agricultural goods, supervision of the villa and its estate was essential – whether by the owner, members of his (or her) family, and associates (including freedmen or slaves).⁶⁶ The architectural result of this need for oversight was that some of the social differentiation of urban dwellings came to be reproduced in villas. The superior status of an owner could be reinforced by the villa’s pars urbana – the residential sector with traditional or innovative plans, fine views and gardens, ancient or up-to-date decorations on floors and walls, and so on – and the inferior status of agricultural workers could be architecturally underscored in the same villa by the partes rusticae – separate courtyards, storage facilities, work areas, and simplicity in the spaces and surfaces of the living quarters. In turn, within the villa building itself, the superior and subordinate distinction among the owner, his or her family members, and the familia – the assemblage of minor relatives, clients, hangers-on, servants with special functions, outside personnel, and slaves – might well be made. Roman designers were good at making such differentiations both spatially and decoratively, to give order and clear social signage in the house.⁶⁷ For very luxurious villas, the transposition of plan and decorative environment from city to country was easy. For lesser ones, the suavities of the great villas might be harder to achieve. For villas whose owners were ambitious but did not have the means or the command of sophisticated craftsmen to produce the ideal of the perfect villa, their proprietors did as well as they could.

In addition, villas and their economic impact gave structure to the countryside and clarified its political meanings. In 138 or 137 BCE, Tiberius Sempronius Gracchus, a young nobleman in his twenties at the beginning of his career, passed through Etruria north of Rome on his way to take up a military appointment in Hispania. His journey most probably took him along the via Aurelia Vetus, the main coastal road begun in 241 BCE to assure military movement north of Rome along the Tyrrhenian Sea. The road had also opened the district to Roman agricultural investors. Perhaps near Cosa in the ager Cosanus, where there were many large villas within eyeshot of the road,⁶⁸ Tiberius Gracchus observed a relatively new phenomenon: a territory supposedly depopulated of local farmers and free workers, with “barbarian” slaves replacing them to till the soil or tend the flocks on estates perhaps of a size to be called “big farms” (latifundia, sing. latifundium).⁶⁹ Estates like these, it was alleged, had dispossessed and driven out normal farmers (agricolae), the owners of small holdings of land settled in the territory as colonists when the town was founded in 273 BCE, some 130 years before. Such men were, it was said, the sturdy backbone of Italian agriculture and also supplied the army with strong recruits.⁷⁰ This vision of harsh, rich, arrogant villa owners replacing hardy, independent agricolae with slaves – in other words, of innovative but cruel economic efficiency in estate management versus old-fashioned family-based farming by citizen-soldiers – gave graphic clarity to Tiberius Gracchus’s later political views, a clear example of how to combine a simplifying social narrative with nostalgia for an agrarian past of “simple” farmers, which, though well bypassed, remained dear to the Roman elite. Two centuries later, Pliny the Elder was repeating the same view: “latifundia ruined Italy and are doing so now in the provinces.”⁷¹ The events of Tiberius Gracchus’s career as a politician upholding the rights of poor citizens strongly shaped the later history of Rome.⁷² His political populism was prompted by seeing newly profitable slavery replacing traditional farmers on agricultural estates, but his insight was part of a larger Roman theme: the antinomy between the new and the traditional.

Instrumenta: “Talking Hardware” in Villas

Agricultural labor and its conditions are complex, delicate topics that elicit strong and sharply differing opinions. In Western history and intellectual development, not to say social and political conflict and even hatred, the structures of work in agriculture are matters of very long standing.⁷³ Discussion had begun already in Greece,⁷⁴ and for Roman agricultural personnel, the modern discussions have been large and wide-ranging as well as disputatious (see pp. 15–17).

Slavery, both agricultural and nonagricultural, and the relationship of enslaved personnel to nonslave servants were at all times issues of discussion and contention in Roman history. While general histories of Roman slavery in modern scholarship have not been lacking, and contemporary discussions are ongoing, slavery as an institution is subject to competing theories of society and economic structure that seek to explain important questions of outcomes. None of the questions about slavery in the Roman economy are abstract ones: the issues they address must have manifested themselves tangibly and even day-by-day in Roman villas. We can formulate some of these questions (among many others and more subtle ones) as follows:

Was slavery inevitable in the expanding Roman economy, especially in the agricultural economy in the second century BCE? Was it optimally efficient? Who were its main beneficiaries? Was it the cause of economic vulnerability and thus later decline? Is slavery a form of capital and thus subject to Marxist interpretations? What are the reciprocal effects of slavery on slaves themselves and their owners?

To be sure, different but often overlapping answers to these questions have been proposed in a scholarly literature that is very extensive; discussions begun already in the 1890s, 1960s, and 1970s are under lively revision and analysis.

Roman Slavery in Modern and Contemporary Scholarship

Simplistic contrasts between ideologies and views – for example, Marxist or non-Marxist, or emphasis on cruel exploitation versus melioristic interpretation of the evidence – are not adequate to summarize the contemporary scholarship on Roman slavery.⁷⁵ However, within a broad spectrum, antinomy between, on the one hand, the work of Joseph Vogt and that of his students and colleagues and, on the other, the insights of Moses I. Finley have resulted in somewhat clearer classification of some of the issues.⁷⁶ The focus of this book is on material evidence of Roman villas, not on how slaves were incorporated in them. While the means by which agricultural work was effected on estates remains an important topic, there are continuing developments in interpretation that are outlined in this section. Identifying slaves both spatially in architecture and in figural representation remains a task for ongoing scholarship.

An internationally based and periodically extended bibliography on slavery begun by Joseph Vogt in the Akademie der Wissenschaften und der Literatur of Mainz University is ongoing,⁷⁷ and an image archive of slaves with protocols for identifying them has also been established.⁷⁸ The more recent publication edited by H. Heinen of a glossary – literary, legal, epigraphic – on slavery is now essential to the topic.⁷⁹

Following Rostovtzeff’s lead in illustrating what might well have been the activities of slaves (see pp. 15–18), many scholars have bravely attempted to identify slaves, in some cases as types in terms of figural representation, in other cases in terms of their spatial presence in domestic and/or urban architecture. A bank of images has been put in train. However, on the topic of identifying images of slaves and servile spaces, few scholars have been as assiduous or successful as Michele George, Sandra Joshel, and Lauren Petersen, who have undertaken an archaeological detection of slave spaces in Roman villas and houses.⁸⁰

Roman Agricultural Personnel in Modern Interpretation

Villas were also farms, and farming is work. Work is subject to theories in larger social and economic contexts. The conditions of agricultural work were and are a real, persistent problem as well as a fundamental element in Western social thought from Greek times to the present. Agrarian legislation has been exhaustively studied.⁸¹ The issues are numerous and important, both in real situations and in interpretive contexts, perhaps none so fraught as that of the relation between slavery and free labor, as the example of Tiberius Gracchus and his observation of slave gangs and the dispossession of traditional farmers in the ager Cosanus illustrates (see p. 13). Even the account of what Tiberius Gracchus actually saw and how he interpreted it has been challenged.⁸² Though the historiography of this topic began much earlier, modern discussion of slave and free personnel in Roman agriculture began about a century ago, in the 1920s.

For our purposes regarding villas, there are numerous themes. Respected traditions of work and social usefulness were enshrined in Roman consciousness about the citizen-farmer, but there was another tradition: that of concern about the asymmetries of deployment of manpower in its many forms (slavery, free citizen farmers, various types of tenancy). Modern interpretations of these asymmetries have been even more subject to discussion. In addition, issues of ownership – in its crudest form, attributing a specific property to a specific owner, but in its more refined form establishing the class, family affiliation and history, and income-level of a villa’s owner and thus his relation to the workers on his properties – have also come into play. In these matters, the evidence about villas, either in literary or historical forms or from archaeological investigation, has colored interpretations in ways that have been congruent or conflicting.⁸³

As in the decades before and after the French Revolution, problems of agricultural exploitation and personnel became a concern in history writing just before and after the First World War. Why this should have been so may have been due to various pressures: The causes and effects of the 1914–1918 conflicts in Europe and its political repercussions were certainly a factor, in that a new understanding of historical causation (beyond dynastic and national analyses or ideological analyses on the political right or left) was needed. Socialist/Marxist or socially based analyses of historical phenomena came into play for Roman history. Inevitably, such studies revolved around the working personnel of rural estates, slave and citizen, and the discussion continues to this day.

Decisive in this area of study is that of M.I. Rostovtzeff, in his article of 1911 but especially in his Social and Economic History of the Roman Empire of Reference Rostovtzeff and Fraser1926. This study dramatized the coexistence and conflict of social classes (including slaves) from the late Republic to late antiquity, using a language that specified the competitive and incompatible interests of the classes as an impetus for change in Roman history. Rostovtzeff’s achievement was decisive in another way for villas: He illustrated, and later editions of his work illustrated even more abundantly, in attractive black-and-white photographs that gave authenticity and actuality to the text. His was also among the first books to show pictures of ordinary and “daily life” objects (admittedly only a few from villa sites) to enhance and extend the discussion of historical texts and inscriptions and thus to make a marriage of literary, epigraphical, and archaeological resources. His lead gave rise to a whole genre of later “daily life” presentations that have become commonplace in history writing both professional and popular.⁸⁴ The distinctly contemporary language and lively presentation in Rostovtzeff’s study gave it authority and immediate appeal, attested by its numerous translations and updated revisions. Previous to Rostovtzeff’s work, the issue of personnel (in agricultural enterprise, in the army, and elsewhere) had been a subject (in Reference Lot1927) of Ferdinand Lot’s speculation on the “decline” of the Roman Empire.⁸⁵ Both Lot and Rostovtzeff were following, specifying, and questioning the tradition of documenting “decline” begun by Edward Gibbon’s The History of the Decline and Fall of the Roman Empire (1776–1788).

W.E. Heitland’s Agricola of Reference Heitland1921 benefitted from Rostovtzeff’s earlier study (Reference Rostovtzeff1911) of the colonate or agricultural tenants (coloni, sing. colonus, which could mean farmer in a general way or tenant-farmer) who enjoyed free status but might be bound by various obligations and restrictions and whose fortunes and conditions changed considerably from the late Republic through late antiquity.⁸⁶ Heitland set up the issue of slave and free personnel in agriculture directly, noting that, despite the contradictory claims in Latin written sources, the parallel and overlapping existence of both forms of working personnel were the norm, though the relative proportions of free to slave labor could vary. To what degree they varied at specific times and places, he could not accurately tell except to note that there must have been adaptation of the proportion of working personnel (free to slave) to different landscapes and different historical conditions (competition from overseas, concentration of estates, new forms of tenancy).⁸⁷

The sensitivities of scholars about the use of slaves in agriculture was perhaps exacerbated by eighteenth- and nineteenth-century anti-slavery movements, namely when, especially in the Americas, slavery had become in large part agricultural and breeding slaves to produce more slaves had been normalized as a social practice and in legislation upheld by the courts. For the ancient world, slavery was a far more nuanced proposition, and M.I. Finley’s essay outlining the individuals, notable for their variety of occupations and status, who were slaves, had recent slave ancestors, or were freedmen is but one example of scholars emphasizing these subtleties.⁸⁸ For villas and agricultural estates, the extremely varied status and training of specialized slaves has been usefully tabulated with literary and legal references by K.D. White as well as the complex issues surrounding the money that slaves could amass toward paying for their freedom (peculium; there were other paths toward manumission).⁸⁹

Slaves in Villas

Issues about slavery have entered the interpretation of archaeologically excavated villas, for the most part Italian villas because they are best known, and villas of the late Republic and early empire because they are abundant and, in some cases, well-preserved. While slaves, as we have seen in the Latin agricultural treatises, were essential components in the operation of agricultural estates, slavery has sometimes been used as a lens by which to interpret archaeological data, thus reinforcing images of social oppression and economic cruelty (including the dispossession of free farmers) in excavated villa sites.

The tradition of slave-based agricultural work on latifundia and the consequent dispossession of free farmers has a distinguished and nuanced historiography; it forms the basis of analyses by L. Perelli (Reference Perelli1982) who discounts an extreme view, and earlier by A.H. Bernstein (Reference Bernstein1978) and D. Stockton (Reference Stockton1979), with the issue of army recruitment from age-pools [17–25 years] in the free rural population studied by Y. Shochat (Reference Shochat1980).⁹⁰ These studies in general assume that big slave-based farms with villas came to dominate Italian agriculture in the second century BCE and after. As we will see (pp. 17–18), the tradition colored the interpretation of the Settefinestre villa by Carandini.⁹¹

However, the assumptions of these studies have been modified toward a much less drastic interpretation (which also drastically minimizes the physical incidence of latifundia and emphasizes the continuation of small farms owned and operated by free farmers). Among others, K. Bringmann (Reference Bringmann1985) challenged the assumption that small farms operated by free personnel had been ousted by big enterprises; his view has been taken up most recently by A. Marzano (Reference Marzano2007), S.T. Roselaar (Reference Roselaar2010), and A. Launaro (Reference Launaro2011), in part following the view taken by D.W. Rathbone (Reference Rathbone1981 and Reference Rathbone1983) favoring a parallel existence of smaller/free and larger/slave agricultural enterprises in central Italy rather than a violent social opposition between them.⁹² In that respect, the interpretation of free and slave labor in villas has returned to Heitland’s views of 1921, of their parallel and overlapping structures.⁹³

Slaves in Villa Spaces

As regards the interpretation of villas, agricultural workers (no matter what their status) who lived outside the architectural confines of villa compounds are, of course, hard to discern. In luxurious villas with large spaces specifically allocated to household slaves or servants, servile quarters and passageways can sometimes be marked visually with certain patterns and colors to provide social signage, and such signage may have been for nonagricultural domestic personnel (janitors, cooks, personal servants, in-house and visiting slaves, and so on) or for outside suppliers or menials (messengers and deliverymen, estate managers, suppliers, and the like).⁹⁴

A problem of some long standing since the 1980s concerns the interpretation of certain architectural spaces in the villas at Settefinestre and Lucus Feroniae. The Anglo-Italian excavators of the Settefinestre villa, understandably excited by the achievement at discovering and publishing, in very favorable circumstances, the first completely excavated villa in Italy, regarded the site as having an exemplary character and “test case” for finding the strangely invisible presence of agricultural slaves, so often referred to in the treatises of the Latin agronomists. Settefinestre is famous, as much for its importance in the history of Roman villas as for the exhaustive detail of its publication, which set a standard for the presentation of such projects.

By contrast, and quite unlike Settefinestre, the investigation of the Lucus Feroniae villa was a rescue excavation, which had to deal with definite time constraints, after a good portion of the villa had been bulldozed away in the construction works of the Rome-Florence highway Autostrada del Sole. The part of the villa that could be saved had to be excavated in haste, often with the use of mechanical machinery, and consequently little stratigraphic data, particularly for the later phases, exist. However, this villa, which had belonged to the senatorial family of the Volusii Saturnini, presented to the excavators an architectural feature, a courtyard with modular rooms on at least two, but probably three, sides, which was distinctly comparable to the two courtyards with modular rooms excavated at Settefinestre. At Settefinestre, these courtyards had clearly been part of the service or utilitarian part of the villa, separated from the elegant residential quarter and the access routes reserved to the owners. The interpretation given by the excavators of Settefinestre was that these courtyards with rooms (cellae) were the slave quarters for the agricultural manpower needed on the large estate, thus finding archaeological confirmation that slave labor, especially the fettered slaves (vincti) referred to in the context of villas by the agricultural writers, was the key element of the “villa system” that had displaced the small and medium free farmers of Republican central Italy. The sociopolitical history of Rome in the second century BCE reinforced such interpretation, as the appearance of large villas employing slave labor in the area of Cosa was linked to the story of Tiberius Gracchus who, while traveling to Spain along the coastal via Aurelia Vetus, would have seen only imported slaves in the fields (see pp. 13–14). Both the villa of Settefinestre and that of Lucus Feroniae became instances of a larger interpretation of Roman villas in Italy, offering a model by which a specific architectural feature (a courtyard with modular rooms) became the signifier of the type of labor and exploitation of the villa, a model that would then have been exported to some of the provinces.⁹⁵ While at Settefinestre the excavators had reached the conclusion that the rooms were all slave-cellae by excluding other functions (such as storerooms or stables), at Lucus Feroniae the interpretation relied heavily on the architectural parallel with Settefinestre.

However, the two villas were actually quite different. The Settefinestre villa was relatively far from its nearest town and port (Cosa), while the Lucus Feroniae villa was only 500 m from the forum of the nearby small town. They were also quite different architecturally. Although about half of the Lucus Feroniae villa is missing due to the circumstances of its discovery, the courtyard with its cellae was not clearly separated from the pars urbana, as in the case of Settefinestre. Rather, it was part of it, and it included a central room (the so-called Lararium) with a monumental inscription detailing the political career of a member of the family.⁹⁶ Even in the case of the Settefinestre courtyards, attributing all the rooms to slaves is a very grandiose hypothesis, as it involves attempting a calculation of the number of hands, and from this datum the size of the estate. Roman historians and archaeologists working at other sites in central Italy have taken Settefinestre as the undisputable archaeological proof that slave quarters housing up to hundreds of slaves were typical of rural elite villas. However, other interpretations of the rooms are possible, as some of the claims made by the excavators do not stand up to close scrutiny.⁹⁷ The courtyard rooms at Settefinestre could have been multipurpose, with some used for slaves, others for storage (of tools and equipment, for instance), others used as stables for the pack animals (mules and donkeys) that were part of the instrumentum of a villa and that a site like this must have needed in order to transport the agricultural produce to the port of Cosa.⁹⁸ Therefore, both these villas, which have greatly influenced villa studies, cannot be taken as general models to guide the interpretation of other sites in Tyrrhenian Italy whenever a courtyard surrounded by modular rooms is present. Recent studies, most notably that of Launaro (Reference Launaro2011), have shown that, with few, localized exceptions, there is no correlation in Italy between the appearance of villas and the disappearance of farms; on the contrary, the number of villas and farms increase and decrease in tandem. The very concept of extremely large, contiguous, estates is not really applicable to most of peninsular Italy. Wealthy landowners as a matter of habit amassed a number of geographically scattered properties, and this had a clear effect on the management, organization, and size of agricultural processing facilities of the villa estates.⁹⁹

Origins of Roman Villas

The origins of villas, both the villas as architectural form and villas as units of agricultural exploitation, are matters of continuing debate. Three main hypotheses or lines of interpretation have emerged from the many discussions.

The first hypothesis is based on external historical origins: because farms and larger rural mansions that could be called “villas” were present in the Greek classical and Hellenistic world, the villa came to the Roman world from Greece and was naturalized in Italy via the Greek colonies of southern Italy in the third century BCE.¹⁰⁰ In this context, both in ancient Greece and Rome, social and political structures were based on census: landed wealth was the criterion by which membership in one or another class was determined. For this reason, ownership of land, i.e., farms and villas, was linked with the social and political history of ancient states. The Greek classical period and, to a greater extent, the Hellenistic period were certainly defining moments in terms of the delineation of clearer architectural typologies and a more regular diffusion of dispersed settlements in the countryside. In the Hellenistic period, the number of villas and farms populating the countryside of areas such as mainland Greece, peninsular Italy, and Sicily greatly increased. In certain geographic areas there was a clear connection with Greek cities’ territorial expansion. Rome’s expansion in Italy during the third century BCE was also accompanied by the appearance, in the conquered territories, of farms in greater densities than heretofore, as revealed by the thousands of rural settlements comprising farmsteads, farms, and villas identified in the South Etruria Survey. A similar process has been noted in northern Greece, where the growth of the Macedonian kingdom in the Hellenistic period prompted the appearance of large estates that seem to have promoted the introduction of new technologies in order to process crops more efficiently.¹⁰¹ However, farms and what some have labeled “proto-villas” did not appear out of nowhere in the Greek classical period. They were present in the archaic period too, and in the Etruscan and Carthaginian worlds.

The second hypothesis about the origin of the Roman villa is in a link with the Punic world of North Africa. Carthage, before its destruction in 146 BCE, was surrounded by many farms and villas, her wealth deriving as much from her agricultural hinterland as from her extensive Mediterranean commercial networks. Although archaeological excavations of Hellenistic Punic farms around Carthage are scant, it is known that courtyard villas fortified by corner towers were in use both in Sicily and in North Africa as early as the third century BCE.¹⁰² Discoveries on the island of Djerba, off the cost of Tunisia, have highlighted the great diffusion in the Hellenistic period of villas producing sweet wine, which was widely exported around the Mediterranean.¹⁰³ In addition, arboriculture and certain cultivations that were later taken up in Roman villas seem to have been developed in the farms of Carthage. The agricultural treatise by the Carthaginian Mago was judged by the Roman Senate in 146 BCE to be the only noteworthy Carthaginian writing deserving of translation into Latin, and agricultural manuals written later by Latin authors, in particular Varro’s and Columella’s, often quote Mago’s work as an authoritative source.¹⁰⁴

The third hypothesis as to the origins of Roman villas breaks away from the evolutionist model (from small farm to large villa) that both the “Greek origins” and the “Carthaginian origins” hypotheses imply. Whether positing an origin from the Greek or Carthaginian worlds, the “standard history” of the evolution of farms and villas in Roman central Italy put forward in the historiography of the 1970s and 1980s has been characterized by an evolutionist model inflected by external phenomena.¹⁰⁵ Small farms appeared in the third century BCE as the result of the Roman annexations in the peninsula, and in the aftermath of the Hannibalic war, which opened the way to Rome’s interventions in the East after the mid-second century BCE, some developed into villas. The availability of manpower (slaves) and riches that the eastern conquests brought about allowed the elite to create slave-staffed villas producing cash crops. The landed estates of the wealthy grew in size, incorporating former small- and medium-sized properties and large tracts of ager publicus.

However, both this evolutionist model and the two “external” origins proposed for the Roman villa (Greek and Punic) have been questioned in consequence of an important archaeological discovery made in Rome in the 1990s: the Auditorium Villa. The site, now engulfed in the modern city, was well outside the north limits of the ancient City, at least as it had developed before the second century BCE. A substantial villa existed on the site already in the years 500–300 BCE; further enlargement of the complex occurred in 300–225 BCE. The name “villa” for this archaic and very early Republican complex is justified primarily by its size, so far unparalleled in these more than 250 years; by its décor, which included figurative architectural terracottas similar to those known from Etruscan religious buildings; by its plan, featuring a central open court; and by the presence of an oil press.¹⁰⁶ The complex had a long occupation history (nearly 650 years), going out of use only after the mid-second century CE. The discovery of the Auditorium Villa has led to the reconsideration of finds at other later villa sites, where the recovery of archaic architectural terracottas had been explained by hypothesizing the presence of nearby archaic rural sanctuaries obliterated by later farms and villas rather than a well-developed decorative culture in early villas like those found in the Auditorium Villa. In fact, the Roman villa could well have evolved from palatial rural structures of the Etruscan world.¹⁰⁷

The Auditorium Villa has had reverberations throughout the discussion of “origins” of the Roman villa. While emphasizing its “Italic” character and archaic date, one of its excavators, N. Terrenato, has also highlighted the discrepancy that exists between the villa described in Cato’s mid-second-century treatise (a modest, functional farm, with no signs of grandeur or luxury) and the villas in the archeological record of this time, which are larger and much more extravagant than the “Catonian villa.”¹⁰⁸ In his agricultural treatise, the striving homo novus Cato was not describing the “new” capitalistic farm as a relatively recent phenomenon of Latium, namely the large terraced villas in locales (Tusculum, Tibur) at easy distance from Rome. Instead, as a novus homo, he was promoting himself by opportunistically idealizing the villa, ignoring the realities of the villas of the elite of his time in order to appeal to the core values (e.g., frugality) of Rome’s aristocracy with the aim of being fully accepted by it socially. Villas could certainly be part of self-promotion and social upward mobility in the case of luxurious rural dwellings, but in Cato’s case, the ostentatiously simple villa of his agricultural treatise was at once a rebuke to foolish, showy, contemporary mores and a claim to prestigious old-fashioned morality.

Villas and cities go together. As a city, the growth of Rome as a large-scale consuming entity did not wait for Rome’s territorial expansion in the third century BCE. There were clear and earlier indications of intensifying agricultural exploitation in its environs. At various sites in the south-eastern suburbium near modern Centocelle, where in the late Republic and empire luxurious villas were built, archaeological investigation has revealed numerous trenches dug in the tufa plateau in order to plant vines, dated on the basis of associated ceramic material to the fourth and third centuries BCE.¹⁰⁹ R. Volpe’s work in the suburbium of Rome places the discoveries at the Auditorium Villa within the wider context of land use at the time. She reconstructs a far more extensive landscape of villas within a highly productive network of agricultural systems, a landscape that developed much earlier than has been postulated up to now. The evolutionist model for villas, namely the traditional narrative of an evolution from small farms to large villas in the second century BCE, needs reassessment in the light of the new information on the growth of Rome and agricultural exploitation of its territory at much earlier dates than have heretofore been known.¹¹⁰

Demise and Decline of Roman Villas

Decline and demise of villas can be observed in various ways, but the definition of these terms is difficult. Indeed, the processes may not be fully separable, and while the words themselves may prevent nuanced analysis, still, how phenomena and institutions change in a downward direction is a permanent topic in historical studies. With villas, their decline – or, perhaps better, the variations of their successful sustainability – within the Roman hegemony has become a topic of some dispute regarding the interpretation of the material remains. The demise of villas, its processes and causes were not uniform: villas in many cases outlasted, even flourished, in the face of severe social change, invasions, usurpations, changes of régime, interrupted communications, and abandonment by central authorities. The resilience of some villas as sustainable entities can be discerned, while others were more fragile, easily subject to immediate or long-term pressures. The passage from success to another, perhaps lesser, manifestation is an element in our understanding of the trajectories of the Roman Empire.

For Italy, the widespread rise of villas in the Republic, namely as of the mid- to late second century BCE and intensifying through the first century CE, cannot be doubted. In addition, the villas that were built in those two and a half centuries – with special exceptions and variations for maritime villas¹¹¹ – were usually both agricultural affairs in productive rural areas and residences with some degree of comfort both in the countryside and on coasts. Agriculture for expanded profit and agreeable, newly sophisticated lifestyle were not incompatible in Rome and Italy of the late Republic and early empire, and the association came to be a permanent and successful feature of villa life.

For modern historians and archaeologists, there is a problem with the villas’ success: How can the decline of any one villa or a group of villas be interpreted? Foolish individual owners and idiosyncratic management may not be sufficient to explain simultaneous negative changes in any conglomeration of dwellings, be they domus in cities or villas in the countryside. In addition, decline described as a general downward trend over several centuries coarsens and may well distort both local situations and preclude instances of variations against the perceived norm, because seeming decline of one villa site may actually be an advantage to another.

The demise and decline of Roman rule has been a perennial topic of historical research and speculation since late antiquity itself.¹¹² However, for Roman villas, their demise is not in doubt. By the late sixth century CE, villas as a rural settlement type had, with very rare exceptions, lost much semblance to their original economic and residential functioning, even though the incidence of the end of villas was unequal in different areas of the western empire.¹¹³ Villa sites were gradually abandoned, suddenly abandoned, or bear traces of only seasonal or sporadic occupation. The economies that had sustained villas may have substantially changed and perhaps become merely local rather than regional or Mediterranean-wide, though a newly local agricultural economy may be successful and viable no matter the seeming degradation of residential standards.¹¹⁴

Easier to observe, but perhaps more difficult to interpret, is decline. How to define it, and where to draw the line between mere change of use and actual degradation, is hard. Sometimes grand rooms could be subdivided or changed for new uses when the villas themselves were still dignified residences.¹¹⁵ In certain villas in their late phases, beaten earth floors appeared in rooms once floored in mosaic, kilns were built in formerly residential rooms, bath suites went out of use, often new agricultural activities replaced earlier ones, and burials appear in rooms no longer conventionally habitable. While the evidence has been principally collected from Italian villa sites, similar patterns of “ignoble” re-use can be registered in the historical record elsewhere in the empire: in Greece, Gaul, and Hispania as well as North Africa. As in Italy, it is difficult to draw the line between change of use and actual abandonment.

A letter written by Pliny the Younger in the late first or early second century CE is instructive in this respect.¹¹⁶ Pliny needed the advice of a friend about a business decision concerning villas; the letter itself outlines the pros, cons, and variables about how to come to a decision about real estate. A property adjacent to his estate and villa at Tifernum Tiberinum had come up for sale.¹¹⁷ The proposition was tempting because its current asking price was HS 3,000,000, down from a previous asking price of HS 5,000,000, a 40 percent drop.¹¹⁸ Pliny outlines three considerations, two on the positive side, another on the negative:

1. 1. On the positive side, buying the property would attractively aggrandize his own, make for less traveling time, and would concentrate moveable goods (sumptus supellectilis, meaning furniture and household equipment) as well as human personnel (gardeners, blacksmiths, gamekeepers). This would lower the maintenance costs of the amalgamated estates and increase the productivity of his existing personnel (their workload would be increased with the purchase of the adjacent property).¹¹⁹ Greater efficiency and streamlining would result.

2. 2. A second positive was that he could focus on enhancing and improving only one villa (unam villam colere et ornare), while the other villa, the one on the adjacent property, could be left in “as is” state (alteram tantum tueri). This would make everything as convenient as it would be agreeable (non minus utile quam voluptuosum).

3. 3. The downside was the risk to expose so large a property to the same uncertainties of weather (and possible crop failure) and general risks as the other estate (sit incautum, rem tam magnam isdem tempestatibus isdem casibus subdere).

Other considerations in the letter concern the manpower and cost, involving both slaves and what he (Pliny) would have to do for the existing tenants on the other property. This was a real difficulty. To re-establish and increase the tenants’ productivity, he would have to buy new slaves for them, as he did not have any chain-gangs (vincti). The tenants’ lack of resources and the costs associated with restoring the productivity to its earlier levels were the key elements that had depreciated the value of the property.¹²⁰

We do not know what Pliny’s final decision might have been or if he went through with the deal and effectively let the other villa stay “as is.” However, the alternatives that Pliny outlined to his friend raise two issues about how to perceive the historical trajectory of Roman villas in Italy and elsewhere: decline and ultimate demise. The first issue is as follows: If modern archaeologists visited Pliny’s amalgamated estates and their villas, one enhanced and improved, the other left “as is” and perhaps converted to practical farm uses as workshop, warehouse, or even cemetery, they might well conclude that half the villas in the area had fallen on hard times, had become degraded, and were in decline. In fact, such an interpretation would be incorrect as describing the process: Pliny was making a business decision aimed at greater efficiency in use and productivity of personnel, and no actual decline occurred. The villa on his estate would be enhanced and improved (as far as the residential part was concerned), the “other” villa not. Were the latter to have been altogether abandoned in archaeological perception, that would not mean any end to villas in general or to the agricultural exploitation of the land. Instead, it could point to concentration of resources and wealth and a streamlining of agricultural production rather than to any decline in the viability of villas as such.¹²¹

The second issue is one of demise. Pliny’s difficulty in allocating his own slaves to his potential new tenants and not having chain-gangs of slaves of his own to work the newly acquired estate were potentially deal-breaking. This raises an important historical issue in Roman history, in Italy and elsewhere: manpower. As we have seen (pp. 15–17 and 19–20), slaves and free labor were already viewed as a problem in the late Republic and later. The positing of a chronic manpower dearth throughout the Roman imperial hegemony began in the 1950s and is a continuing topic of discussion. In the most extreme interpretation, shortage of manpower has been seen as an ultimate cause of the fall of Rome;¹²² more fruitfully, sporadic labor shortages have been seen as the origins of intermittent ups and downs of villa development (especially in Italy).¹²³ However, manpower shortage – whether long term or occasional – has not, as far as we are aware, been satisfactorily recordable for Roman villas, so we have omitted its consideration in this book.

Social Groups and Clusters of Villas: Fashions in Dining and Decoration

If identifying slaves as a social group in the spaces of actual villas is hard, identifying where members of Roman elites (of status or wealth or both) distributed themselves is easier: they clustered in order to congregate, interact socially, and identify themselves, both in urban domus in Rome (e.g., on the Palatine Hill) and in villas in the Roman environs and elsewhere in Italy. The political life of the Roman Republic in the late second and during the first century BCE and into the early empire prompted members of the elites (with some exceptions) to congregate spatially, and the results were nodes of Republican villas, first in the suburbium of Rome and in its southern proximate area of Latium, then in Campania and the Bay of Naples, then elsewhere. Latin authors noticed these conglomerations and commented on them as a phenomenon in the history of Roman social landscape and habits.¹²⁴ The hilly countryside to the south and east of the City was noted for groups of villas. The locales of Lanuvium, Tusculum, Tibur (mod. Tivoli), and Praeneste (mod. Palestrina) had, by the second half of the second century BCE, been thoroughly taken over as venues for villas, as had the coast around Antium (mod. Anzio) to the south (about a day’s journey) and Laurentum (where Pliny the Younger had a villa that he could reach in a few hours ride from downtown Rome). Roman high society – be it of the senatorial nobility, the equestrian order, or the elite of money that might include freedmen and their descendants – tended to cluster and thus to homogenize in the way of architectural and decorative taste as well as lifestyle. Cicero tells us that the great general, triumphator, and consul L. Licinius Lucullus (c. 115–50 BCE, consul in 74 BCE), the scion of several ancient noble families, when living in retirement at his luxurious villa at Tusculum, answered a charge of excessive luxuria by remarking that he had to live up to the standards of his next-door neighbors, one a knight, the other a mere freedman.¹²⁵ Villas smoothed the differences among disparate social classes where the urban domus might not.

The clustering of villas and the proximity of a sociable elite continued down the generations: Cicero, who built his villa at Tusculum, sets an affable chat about books in the library of a villa there. The villa was owned by Lucius Licinius Lucullus, the son of the triumphator and consul mentioned above, and the villa and its books were presumably inherited from his father. Cicero was looking to borrow some volumes from Lucullus but found the young M. Porcius Cato, son of the famous Cato the Younger (Cato “of Utica,” a descendant of Cato the Elder) surrounded with piles of books he was perusing. Cato exclaims that, had he known Cicero was at Tusculum, he would have dropped in on him at his villa nearby. The elder homo novus (Cicero) was chatting about philosophy with the young descendants of high nobiles of the patrician and plebeian orders: Their backgrounds, ages, and even political affiliations were different, but they all had villas at Tusculum.¹²⁶

Such spatial proximity of the elite despite the social differences among them prompted a certain uniformity of architectural choices: throughout the late Republic and early empire, hillside suburban villas around Rome most often were built on basis villae, that is, on substructures creating one or more levels of artificial terraces to dramatize the views and catch the breezes. In the ager Cosanus, where the Settefinestre villa discussed on p. 17 was located, there was a fashion for terraced villas to look like fortified towns with amusing little towers along their garden terrace walls – the villas referenced military architecture without really resembling it. More impressive, apparently, in the way of military aspect, were the villas of the great political figures of the late Republic: Marius, Pompey the Great, and Julius Caesar, grouped on the promontory of Misenum in the Bay of Naples.¹²⁷ Villa owners could also cluster in affinity groups: Tibur came to be favored by villa owners of senatorial rank from Hispania.¹²⁸ Outside of central Italy and the Bay of Naples, villas also tended to cluster: the Larian lake (Lacus Larius, mod. Lake Como) was replete with them, as was the Istrian peninsula, and major towns and cities throughout the empire often developed villas in close proximity.

The relative homogeneity of architectural ideas in villas was matched by modifications in social arrangement for dining and reception. Beginning in the second century CE, the older form of triclinium, in which a square U-shaped arrangement of separate dining couches (each holding two diners) set on three sides, went somewhat, but not completely, out of fashion. It was replaced, in both domus and villas, with the stibadium, a continuous semicircular platform for reclining which had the advantage of greater visibility, both for the diners among themselves and for the feasting group as a whole to be on view for observers, social inferiors who were invited to observe but not to dine.¹²⁹ Of course, to be observed while feasting implies social superiority for the feasters, so the stibadium enhanced social difference. This difference would have been all the more marked when the stibadium had a splendid decorated architectural backdrop (an apse or exedra, quite often expanded to a tri-lobed absidal room for three stibadia) and/or was set off and above the other spaces with steps. These new arrangements, which became virtually obligatory in villas throughout the empire into late antiquity, denote the increasing social prestige of villa owners.

Owners of estates may well have been apt to display themselves as more powerful socially in late antiquity than they had been in earlier times. In addition, ceremonies of hospitality may have become less affable, more formal and ritualized in view of increasingly greater claims of status and self-regard by owners. Such changes in villas were marked by the frequent construction of domestic basilicae, that is, square or (most often) oblong rooms with an apse (semicircular, partially semi-circular, or angular) on axis at one end, sometimes with the floor of the apsed area raised with a low platform or shallow steps. The architectural arrangement marks and enhances the villa owner, giving him spatial authority and presenting him as the principal actor in situations of reception and adjudication. Because the new dining arrangements and the construction of basilicae can be registered in villas from Gaul to Greece and from Britannia to Africa, it is clear that their architectural prevalence came from a shared uniformity of social habit among the elites of the empire.

This uniformity can also be seen in religion: in the fourth century CE and subsequently, the spread of Christian iconography evidenced in the building of chapels and/or mausolea with Christian content. Themes of decoration in floors and wall paintings in villas throughout the empire likewise attest the fraternities of taste in decorative and ultimately pious practice among the elites.

With very few exceptions, notably in the Bay of Naples and some literary instances that describe real or imaginary mural decorations, most of the decorated plaster, mural mosaic, or painted walls and ceilings of villas have not survived. However, what has survived in some abundance are their floors, from very humble (beaten earth) to utilitarian in brick (opus spicatum in various patterns) or terrazzo (opus signinum, sometimes decorated with tesserae), to mosaic (opus tesselatum, opus musivum) made of grouted quadrangular elements (tesserae, sing. tessera, commonly in a c. 1 cm² cube), to very grand floors in multicolored cut marble or other stones (opus sectile, mainly geometric but also floral and even figural). Mosaic floors could have geometric, floral, and/or figural compositions, often rather prosaic but others enhanced with separately manufactured special elements (emblemata), some very detailed in tiny cubes (opus vermiculatum). The different techniques could coexist within houses to differentiate the social importance of their rooms, with the simpler techniques applied to ordinary rooms such as bedrooms (cubicula) or corridors and the grander, special-order jobs for tablina and triclinia. The techniques and different patterns could also be used in combination in a single floor. The business of mosaics for villas was evidently a lively one for craftsmen as well as a source of cultural and aesthetic engagement for their commissioners.

While stibadia and basilicae were frequent, mosaic floors were virtually universal in domus and villas from the first century CE through late antiquity, and they were virtually universal geographically as well. Mosaic workshops worked both locally and much further afield, in particular the African companies that supplied the craftsmen and decorative ideas for villas in Sicily and elsewhere. The phenomenon itself is not surprising: good workmanship is, after all, its own advertisement in a socially homogenized elite. Roman owners, no matter where they lived, regarded mosaic floors as the premier and virtually obligatory embellishment of their villas, fully worth the trouble and expense of hiring local or mobile companies of craftsmen to come and work in often isolated places. This devotion to mosaic floors resulted in both a certain uniformity of decorative and iconographic ideas shared among several villas (the Four Seasons, Provinces, Dionysiac themes) but also in certain splendid, clearly original, and creative ensembles (particularly figural but also stylistic) that were quite obviously custom-made and may denote competition among elite villa owners in parading their literary culture and often abstruse mythological knowledge.¹³⁰

Villas and Biographies

As early as the second century BCE, investigation of their own origins and how the past conditioned their present came to be normalized in Roman self-consciousness. Families used the actions and achievements of ancestors to enhance or revive their prestige, and political factions resorted to historical exemplars to justify present policies. Historical allusion was firmly incorporated into Latin rhetoric, inscriptions, and literary production, and the presence of the past in Roman mental life constituted an impressive historiographical tradition. The past and its ancestral moral habits and traditions (mos maiorum, pl. mores maiorum) were irrefutably prestigious.¹³¹ Villas often articulated the themes of families and recalled their history.

Among the Roman elites, competition of all kinds was the norm, in cultural achievements as well as in politics and military prowess. In that competitive milieu, owning villas became obligatory at the upper levels of society, and biographical notices of political grandees often included some facts of their owning a villa or (most often) many villas, and where they were situated: Scipio Africanus (Publius Cornelius Scipio Africanus Maior, 236–184/3 BCE) retired from military and civil life to a villa at Liternum, using its simplicity as a standing rebuke to the malice and arrogance of his political enemies.¹³² Coquettishly or grimly retiring to villas came to be a biographical theme: The future emperor Tiberius, during his short self-exile to Rhodes (6 BCE–2 CE), coyly and virtuously made do with a small house in town and another in the suburbs (a villa) not much better. Later, as emperor, he grimly self-exiled himself, from 26 CE to his death in 37, to a villa or villas on Capreae (mod. Capri), of which the depraved luxury were proverbial.¹³³

As we have seen, the villas of Marius, Pompey, and Julius Caesar at Misenum were worthy of mention,¹³⁴ and the habits of the Republican grandees were smoothly transmitted to imperial ownership of multiple villas, from Augustus to emperors in late antiquity. In addition, villas became actors in emperors’ biographies: The emperor Augustus was brought up on a family estate near Velitrae in the Alban Hills south of Rome, and in its villa the room that had been his nursery was still shown almost a century later.¹³⁵ In addition, Augustus died (14 CE) in the same room of a villa at Nola in which his father had passed away seventy-two years before: such magical and fated coincidences about houses included villas, and the rooms in which these events took place were held in reverence.¹³⁶ Augustus’s and his successors’ villas became part of their biographies,¹³⁷ and Augustus’s wife Livia had a villa at Primaporta north of Rome that derived its name ad gallinas albas or ad gallinas (“At the White Hens” or “At the Hens”) from a famous, highly numinous event in her life.¹³⁸ Accounts of the emperor Galba’s villa birthplace and his subsequent installation of a statue of Fortuna in his villa at Tusculum were incorporated into his biography.¹³⁹ Even invitations to villas reflected notably on guests and hosts: Pliny the Younger, summoned on legal business by the emperor Trajan to the villa at Centum Cellae (mod. Civitavecchia) on the coast north of Rome, described (a little smugly) both his duties as imperial official and the perfection of the emperor’s hospitality as well as the architecture and arrangements of the villa itself.¹⁴⁰ Imperial patron and host, honored guest, and the hospitality at the villa and its aspect all went together. Such anecdotes emphasized villas as extensions of the persona or public personality of famous individuals: biography and villas went hand-in-hand.¹⁴¹

At lesser (nonimperial but still elite) social levels, villas could also express persona, sometimes negatively, sometimes positively. The villa of Publius Servilius Vatia (c. 65–25 BCE) near Cumae was, according to Seneca the Younger (c. 4 BCE–65 CE), superbly beautiful as to its site, planting, natural and artificial elements, and architecture, but its useless beauty was merely an epitaph on the man himself. The son and grandson of consuls and a descendant of triumphatores, Servilius Vatia himself had advanced to the relatively low elective post of praetor, then retired to a cowardly life of ease, gluttony, and lust and was called a “rich praetorian” (praetorius dives, an insult). The villa was the man’s epitaph: “Here lies Vatia!”¹⁴² Later, Pliny the Younger infused villas with emotion: When he visited his aunt’s villa at Alsium (on the coast north of Rome), which had once been the retirement home of his patron, Lucius Verginius Rufus (15–97 CE), he was sad because the great man’s cenotaph had never been finished.¹⁴³ The theme of self-expression through villas in the matter of family pride or personal conviction persisted throughout the presence of villas into late antiquity: in the fifth century CE, the name of the founder of a villa was inscribed in stone at the entrance to the dwelling in the case of the “castle” (burgus) of the Pontius family in Aquitania.¹⁴⁴ An entire family estate could be given over to new, Christian purposes, as Sulpicius Severus did with his estate called Primulacium, also in Aquitania, soon after 400 CE.¹⁴⁵

Descriptions of Villas in Prose and Poetry

Villas were subjects of ancient prose writing (often in letters) and in verse, and to the extent that such literary sources indicate how country dwellings impinged on Roman consciousness of the phenomenon, a brief account of them can be included here, though it is by no means exhaustive.¹⁶⁰ Villas appeared in Latin literature as the objects of satire, praise, verbal description (ekphrasis), moral uplift, and as centers of agricultural work. The Latin origins of such manifestations were Greek: Hesiod’s poem on farm activities, the early seventh century BCE Works and Days, written like the Homeric poems in heroic meter (dactylic hexameter), gave authority to both agriculture and the traditional morality – claimed or real – of rural life as the basis of the good life.¹⁶¹ Ekphrasis – encompassing the description of architecture and works of art in words – originated in Greek literature as early as Homeric descriptions of palaces and places in the Shield of Achilles and Telemachos’ praise of the palace of Menelaos. The continuum from the description of a house, its estate, and its agricultural activities as well as the life and morality of its owners and inhabitants finds an early example in Xenophon’s account, in dialogue form, of the rural property and country oikos of a certain Ischomachos in the fourth century BCE.¹⁶² Even though Ischomachos owned a house in Athens, the account of his country house formed the core of Xenophon’s beliefs concerning the sobriety, utility, and superiority of rural pursuits. The theme quickly became an important, versatile, and multifaceted commonplace of Western thought and literature and has continued ever since in many forms.

In Latin literature, the celebration of a villa, often with its setting, inhabitants, and activities, in ekphrasis form (either versified or in prose) was most often the celebration of the writer’s own property, in a continuation of the theme of a dwelling as expression of self. In the late Republic and early empire, Horace (Quintus Horatius Flaccus, 65–8 BCE) wrote of his favorite country house in verse, sometimes as graceful invitations to his friends to come and visit.¹⁶³ Some fifty years later, Seneca the Younger wrote three letters framing villas he had seen or visited in moral terms.¹⁶⁴ Toward the end of the first century CE, themes of quiet repose and learned reflection infused the very famous letters written by Pliny the Younger on two of the many villas that he owned or had inherited, the Tuscan or Tuscanum villa (villa in Tuscis at Tifernum Tiberinum, mod. Città del Castello) and the Laurentinum on the coast at Laurentum near Rome.¹⁶⁵

Perhaps the most elevated and elaborate description of a villa, its site and contents, is that contained in a long poem by Statius (Publius Papinius Statius, c. 45–96 CE). The villa belonged to a certain Pollius Felix, a banker, and his wife Polla; the couple owned two other villas.¹⁶⁶ The villa, of which the site is known, was a villa maritima near Surrentum (mod. Sorrento) in the Bay of Naples; the poem is a description of the successive terraces, varied views, gardens, and grand rooms that the poet, a protégé of the owner, enjoyed after arriving by boat on the villa’s jetty (he had embarked from the north coast of the Bay after a dusty trip from Rome down the via Appia). From afar, the villa’s silhouette beckoned the visitor with steam rising from the double-domed bath building (hospitality often began with a bath).¹⁶⁷ The villa was crammed with works of art (vague attributions are made to the Greek artists Phidias, Myron, and Polyclitus) and rich surfaces (marbles from Italy, Greece, Asia, and Africa are laboriously mentioned), and they, together with the site itself, were ornamented by the poet with glittering mythological allusions, both Greek and Latin. The beauty of the villa and its contents offset the questionable occupation that its owner, a money lender, may have exercised, and the mythological allusions as well as the authority of its art validated the villa’s luxury.¹⁶⁸ Natural beauty, art, and learned allusion gave cultured glamour and possibly topics of table-talk to Pollius’ villa and its feasts. The subjects of wall decoration, statues, and floor mosaics in domus and villas may well embody the experience of conversation – either in reality or in aspiration – in Roman dwellings.

Statius’ poem had some successors in Latin literature, particularly (but less vigorously) in the way of classical allusions, but the forms of ekphrasis adopted by Pliny strongly influenced later letter-writers and poets who described their villas, among them Sidonius Apollinaris (Gaius Sollius Modestus Apollinaris Sidonius, c. 430–89 CE) who described his Avitacum near Augustonemetum (mod. Clermond-Ferrand in the Auvergne) and wrote a letter-poem about it intended to persuade a friend to visit to escape the summer heat.¹⁶⁹ Like Pliny (whom he sought to emulate), Sidonius in his letters and poems about his Avitacum and other villas combined description of architecture, amenities (the bathing facilities are important to emphasize the hospitable promise of villas), unusual features (special waterworks, the wall paintings in a lady’s boudoir), pleasant inhabitants (familia and servants), and excursions to and from the villa. In his long poem called Burgus Pontii Leontii, written about the castellated villa of an Aquitanian friend and fellow nobleman, classical allusions abound, supplemented with descriptions of wall paintings showing events in Roman and Old Testament history, the whole confection supposedly described in prophesy by the god Apollo, no less, to a drunken Dionysus, no less, as the residence of the gods and the Leontii when Rome will have expanded to the far west!¹⁷⁰ The expansive spread of villas was not only a historical phenomenon in Roman consciousness: It was an export that came freighted with godly presence and historical inevitability, profitable agricultural work, beauty, art and mythology, hospitality and family history, pleasant friendship, and strong morality. The continuity of values is the more notable in that, by the time of composition of the poem, the Leontii and Sidonius himself were Christian. Nonetheless, for Sidonius, the Roman villa could be loaded up with the entire baggage of antiquity and readily shipped from Italy and recreated to wherever Rome ruled.

Roman Villas as a Visual Ideal

Villas and their reputation had a radiance well beyond themselves, their owners, and their locations. Seneca the Younger, who always noticed the logic or illogic by which people conducted themselves, comments on the topic: Wealth without moral principles, he tells us, begets foolish luxury in individuals. Then such individuals develop fads: among others, preoccupations with personal appearance, obsessions with furniture, and alimentary crazes in the way of cooking and menus. Conspicuous among the fads is the desire to make town houses (domus) look like the size of country dwellings, presumably extended physically and/or fictively to suggest the environment and perspectives of villas.¹⁷¹ In Seneca’s day, and to his certain eyewitness, the greatest of all imperial urban dwellings was the apex of such a fad: the Domus Aurea. This palace was built next to the Roman Forum by Seneca’s pupil and patron, the emperor Nero; it was the culmination of the ideal of rus in urbe (the country in the city). From the villa-like porticoes of the palace, there was a view toward an artificial lake surrounded by buildings resembling cities seen from afar as well as an entire pastoral landscape with herds of wild and domestic animals and an agricultural setting with ploughed fields and vineyards.¹⁷² The palace itself was set in gardens as if in the countryside. At about the same time, real villas (notably that of Villa A at Oplontis, which may have been owned by Nero’s wife) developed painted extensions in perspectival wall paintings depicting spaces even grander, in the way of gardens and architecture, than their already grand real ones.¹⁷³

The architectural and decorative influence of villas on domus in Pompeii and Herculaneum has been studied in detail by Paul Zanker; he has shown that, both in terms of architectural adaptations and arrangements, as well as decorative motifs, owners of urban houses delighted in villa imagery.¹⁷⁴ Villas were prestigious no matter how or where they were evoked.

For owners, villas may have been a normal part of their visual vocabulary. For others, possibly those who did not own villas but merely aspired to do so, the glamor of villas in visual representation was a topic of frequent choice among other genre subjects (still-lives, theater scenes, and so on). It was not unusual, even in quite modest urban houses in Pompeii and Herculaneum, for visions of villas to be depicted. Villas were sometimes part of landscape subjects that included visual shibboleths such as shepherds, cows and sheep, ruined buildings, vaguely-perceived rustic-looking passersby, distant mountains, trees, and so on: these were the universal ingredients of pastorales in Roman art. Sometimes villas in landscapes might be the sole theme of such paintings: Suggestions of porticoes, towers, entrances, and the other baggage of villas in landscapes or by the sea were the subject, nothing clearly defined architecturally but everything suggested. With the addition of a ruined building suggesting a temple or a cenotaph or mausoleum, the scene became numinous and evocative of death and memory, a type often called “sacral-idyllic.”¹⁷⁵ Vignettes such as these were usually presented, with exceptions, in minor parts of larger and more robustly depicted decorative structures, most often represented as small moveable pictures (pinakes) at eye level or else as framed drawings below formal, classically painted images with elevated subject-matter. The style of these vignettes is invariably fresh and sketchy with high contrast of light and dark, atmospheric effects suggesting moist, smoky air, and unusual, often aerial, points of view.¹⁷⁶ Other depictions of villas in Pompeian wall painting are more decidedly perspectival, with sharp delineation of architectural elements and articulation of parts, as if the rigid formal elements were noble in themselves.

The visual genre of such representations of villas may well have been a taste among patrons who owned a villa or merely aspired to doing so. However, decorators or interior design companies may well have had such images as part of their usual baggage of choices for their customers: Villas in lovely landscapes would, with certain other mythological scenes and others in a standardized repertory, always be saleable merchandise and very charming.

Later, in North Africa, in the very late second and through the fourth centuries CE, another genre of representation developed: grand villas in the countryside or on the sea depicted in mosaic floors.¹⁷⁷ These are usually (with exceptions) found in domus in Carthage and other urban venues, as decorations in houses tightly woven into confined spaces. For city dwellers in their domus, the vision of the villa (and its activities) was often combined with images of deities or, occasionally and significantly, with images of their owners, the dominus and domina, sometimes as themselves, sometimes equipped with divine attributes.¹⁷⁸ Agricultural work and seasonal activity such as hunting at villas were also shown. Architecturally, some of the villas in mosaic representation have colonnades (often on the upper story) or towers or both. The colonnades may have struck an imperial theme, as in the upper-level colonnade of the Palace of Diocletian at Spalatum, and the towers certainly alluded to both real and described towers in actual villas of the second through the fifth centuries. For the North African villas represented in mosaics, domed buildings within the perimeter signify bath buildings: They convey the promise of the hospitality of bathing that was a part of noble reception at villas, very much as Statius described the steam rising from his host’s villa at Surrentum (see p. 28).

The representation of villas worked visually on several different levels. Domus could be slightly transformed into villas with allusions to them, and the representation of villas in landscapes, either for their architectural effects or as part of numinous scenes, was one of the many types of paintings in the repertory preserved at Campanian sites such as Pompeii. In the North African representations in mosaics, the villa and the activities around it appear to have been freighted with a grander iconography: The frequent presence of the dominus and domina clinch the relationship between the personae of the owners and the villa and estate they owned.

Roman Villas: Post-Antique Manifestations

In modern consciousness of Roman architectural forms since the fifteenth century, villas are latecomers. Their components were often more fragile and certainly more physically dispersed than solid urban constructions, and the distribution of villas in the Roman landscape made them differently vulnerable to the depredations on the marble, masonry, brick, and concrete structures of towns and cities. While late antique,¹⁷⁹ medieval, and later transformations of both urban centers and rural organization were profound, the Roman infrastructures of the countryside – which would have included villas – were especially subject to the influx of populations with different methods of work and different expectations of dwelling conditions and social life. Isolation in the countryside had its dangers. While some villas – and the notion of villa life and some of the words and even ideas associated with such dwellings – certainly continued after the disappearance of the Roman hegemony, villas were for the most part abandoned, destroyed, reused for purposes other than residential and/or agricultural production, and quarried for building materials.¹⁸⁰ Such equipment, artifacts, and works of art as happened to have remained in them were taken away, either to be reused or for collecting elsewhere. Villas became invisible.

The literary sources about villas were much less fragile, or, rather, their rediscovery and publication made villas visible once more. Despite the loss of continuity of Roman villas in modern consciousness, their elements from literary sources and from primitive archaeology (certainly since the eighteenth century, in some cases even earlier) have constituted a lively source of information and interpretation for archaeologists and historians of ancient culture, society, and economics. Roman rural residences have been objects of thought for many different kinds of public: owners and builders of rural property, antiquarians searching for agricultural lore, practical farmers, professional architects, as well as classically inflected poets, letter writers, and even biographers, architects or architectural amateurs, and writers of fiction. Villas have served as exemplars of what is possible in the way of a Roman form with contemporary relevance; the ideals of educated and elite otium were part of their attraction.¹⁸¹

In choices for contemporary cultural information, learned tourism, and entertainment, and very much like extant country houses of later date and place that are “open to the public,” Roman villas in France, Italy, Spain, Portugal, and elsewhere are among the most visited sites by an educated and interested public, because they give immediacy and meaning to ordinary experience that temples, amphitheaters, and palaces do not always do. Villas provide contemporary visitors a voyeur’s view into the homely life of a supposedly glamorous past that is profoundly compelling.

Latin agricultural treatises were consulted, in the Renaissance and later, for their practical information, but such antiquarianism was as nothing compared to the architectural and cultural emulation of Roman villas. They were a living font of inspiration in Western culture and architectural history, as potent as archetypes from the suburbia of Renaissance Rome and Florence, the terraferma of Venice, and many other places beginning in the fifteenth century to the suburbia of contemporary cities. Country houses on the supposed Roman model were built almost everywhere in Europe and ultimately in the Americas: They are, in their Lilliputian, emulative, or grandiose versions, an almost universal phenomenon in the built environment. That many, if not most, of these later versions are imaginary does not detract from the power of the Roman prototypes that they emulated.

In this history of emulation, the rediscovery of Vitruvius’ text (in 1416 by Poggio Bracciolini)¹⁸² and Pliny’s letters (in 1419 by Guarino Veronese)¹⁸³ were crucial to the modern historiography of Roman villas. In the case of Vitruvius, the dissemination and illustration of his treatise on architecture were as influential – if obtusely so – on domestic architecture as on the correct design of the columnar orders and their proportions (Doric, Tuscan, Ionic, and so forth) as on public building and domus, though not on villas. Editions of Vitruvius followed those of the authors on agriculture, even though his De Architectura had been discovered earlier. Printings of his Latin text were numerous, and one of the earliest was an illustrated one; translations followed.¹⁸⁴

No less important for the history of villas in the Renaissance was the almost simultaneous discovery of the letters of Pliny the Younger. The descriptions of the Laurentinum and Tuscanum as well as other letters about villas strongly inflected owners and architects of rural residences, first in Italy, then throughout Europe. While local traditions of farmsteads with residences, some quite substantial, continued,¹⁸⁵ the combination of Vitruvian classical form in architectural orders, proportions, and details with a new and utterly up-to-date appreciation of nature and leisure as well as the authority of Plinian otium gave villas and country houses the allure of antique justification. The results were notable in actual buildings and on drawn reconstructions as well as in numerous books of “how to” live the villa life correctly. The tradition begat its own ambiguities: The double themes of pride of house and natural and architectural beauty, as well as the house as a symbol of oppressive privilege, form the bitter-sweet bases of William Butler Yeats’ poem Ancestral Houses, the first section of his Meditations in Time of Civil War (1923).¹⁸⁶