A Tall Order Writing the Social History of the Ancient World Essays in honor of WilliamV Harris Edited by Jean-Jacques Aubert and Zsuzsanna Virhelyi K · G · Saur Munchen · Leipzig 2005 Beitrage zur Altertumskunde Herausgegeben von Michael Erler, Dorothee Gall, Ernst Heitsch, Ludwig Koenen, Reinhold Merkelbach, Clemens Zintzen Band 216 K · G · Saur Miinchen · Leipzig Bibliografische Information der Deutschen Bibliothek Die Deutsche Bibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet tiber http://dnb.ddb.deabrufbar. © 2005 by K. G. SaurVerlag GmbH, Miinchen und Leipzig Printed in Germany Alle Rechte vorbehalcen.All Rights Strictly Reserved. Jede Art derVervielfaltigung ohne Erlaubnis des Verlags ist unzuHissig. Gedruckt auf alterungsbestandigem Papier. Druck und Bindung: Druckhaus "Thomas Miintzer" GmbH, 99947 Bad Langensalza ISBN-10: 3-598-77828-7 ISBN-13: 978-3-598-77828-5 Contents Abbreviations Vl1 Illustrations ix xi Preface William V. Harris: the fust forty years of scholarship (1965-2005) xv Religion in context 1. SETH SCHWARTZ A God of Reciprocz"ry: Torah and Social Relations in an Ancient Mediterranean Jodery 3 2. JONATHAN P. ROTH Distinguishing Jewishness in Antiquity 3. JENNIFER W. KNUST 4. GLEN L. THOMPSON 37 Jesus, an Adulteress) and the Development ofEarfy Christian Scripture Constantius II and the First Removal ofthe Altar of Victory 59 85 5. JEANjACQUESAuBERT IDu lard ou du cochon?' The Testamentum porcelli as a Jewish Anti-Christian Pamphlet 107 Imperialism and its soldiers 6. MYLES McDONNELL Anitocratic Competition) Horses) and the Spolia Opima Once Again 145 7. JAMES F.D. FRAKES An Architecture ofHuman Heads: Gallic Responses to Roman Power 161 8. RACHEL KOUSSER From Conquest to Civilization: The Rhetoric of Imperialism in the EarlY Principate 185 9. SARA ELISE PHANG Soldiers'Slaves} Virry Work' and the Social Status ofRoman Soldiers 203 V1 Law and economy in ancient society 10. DAVID B. HOLLANDER Veterans, Agriculture, and Monetization in the Late Roman Republic 11. ANNALISA MARzANO Country Villas in Roman Central ItalY: Reassessing the Evidence 241 12. SAUNDRA SCHWARTZ The Delicts ofthe Countryside in Longus' Daphnis and Chloe 263 13.JINYU LIU Roman Governments and Collegia: A New Appraisal ofthe Evidence 229 285 Historiography 14. PAUL CHRISTESEN Imagining OlYmpia: Hippias ofElis and the First OlYmpic Victor List 319 15. ZSUZSANNA VARHELYI A Sense of Change and the Historiography of the Turn from Republic to Empire 357 Afterword 377 Contributors 379 Annalisa Marzano Country Villas in Rorn.an Central Italy: Reassessing the Evidence * Introduction The Roman villa located on the hill of Poggio Gramignano in Umbria was discovered in the 1980s and has since been partially excavated. The significance of this discovery for villa studies in Central Italy was manifold: (1) it offered one of the few known examples of the usage of a particular architectural typology in domestic architecture, the Corinthian oecur, (2) it offered archaeological evidence supporting the general picture of 'villa crisis' in Central Italy starting in the Middle Empire; (3) it features an infant cemetery, established in the structure of the villa in Late Antiquity, which may be a rare attestation of an epidemic outbreak, possibly malaria. 1 The rapid and greatly detailed publication of the excavation data made these new discoveries available to the scholarly community quickly, a truly praiseworthy achievement. However, some important issues about the life of this Roman villa still remain unclear or have been only partially addressed, and are worth reconsidering, for they offer an interesting case study in methodology. In this article, using the Poggio Gramignano villa and other examples, I will demonstrate how theoretical models about the economy of Roman Italy and theoretical elaborations developed at specific villa-sites have been used to reconstruct the historical, social and economic picture of villas in Central Italy in the first centuries of the Empire without taking into account the full range of evidence available. I will challenge the theory that starting in the second century A.D. the crisis of Italian agriculture generally affected the slave-based • I would like to thank Jean-Jacques Aubert, Lisa Mignone, Andrew 1. Wilson, and Zsuzsanna Varhelyi for reading an earlier draft of this paper and giving me useful suggestions. I remain the only responsible for the remaining imperfections. 1 The hill of Poggio Gramignano is 185 m high and is situated 5 km to the south of Lugnano in Teverina, in the region of Term. The results of the excavation were published in Soren 1999. The book was reviewed by Harris 2000, who expressed some moderate skepticism about the evidence of malaria epidemic. 242 A TAIL ORDER villa system, thus causing the abandonment of many villas and opening the way to squatter occupation. I will also suggest that we should use more caution in assigning the ownership of· any well-appointed villa showing sophisticated architecture and decor exclusively to members of the senatorial elite; surely we must consider the likelihood of ownership by wealthy local elites as well. Finally, this analysis will demonstrate how new approaches to the social and economic study of villas are forcing scholars to re-consider old data and theories, and address new questions. An Example of Archaeological Interpretation In Roman times the area of Lugnano in Teverina was part of the territory of Ameria (modem Amelia). The villa of Poggio Gramignano stood in a panoramic and favorable position dominating the Tiber valley. 2 Only part of the ancient structures was excavated in several campaigns conducted by the Soprintendenza Archeologica dell'Umbria first and by the University of Arizona later. 3 The investigations revealed a portion of the pars urbana of the villa, on the southwestern slope of the hill, including a Corinthian occus with an opus scutulatum floor and another room, possibly a triclinium, with a mosaic floor. 4 Two rooms paved in opus spicatum were probably service/storage areas in relation to the occus and the triclinium. On a lower level were barrel vaulted, buttressing rooms, where, sometime in the fifth century A.D., an infant cemetery was established. Archaeological investigation of the upper part of the villa was limited to a series of soundings. This upper zone, built on the hill crest to the northwest, revealed no traces of mosaic floors, but only beaten earth and whitewashed floors. The excavators therefore postulated that this upper zone was the working area or the living quarters for slaves and/or hired workers. The site is 7 km west of modern Amelia (Ameria) and 2.5 km east of the river Tiber. D. Monacchi of the Soprintendenza Archeologica per l'Umhria conducted two campaigns, in 1982 and in 1984, following local reports of clandestine digging activity. In 1988 new excavations were undertaken under the direction of D. Soren of the University of Arizona, the results of which were published in Soren 1999. 4 This architectural typology is described in Vitr., De arch. 6.3.8-9. The room measures 7.20 x 5.10 m and had red stuccoed columns along the north, east and west sides, whereas the south side had four semi-columns. The opus scutulatum is in the central part of the room and features a black mosaic background with inserted pieces of marble and colored limestone. The marbles come from quarries in Asia Minor, Greece and Africa, while the limestone is from central Italy. In central Italy sim.ila.r paving is known for the atna of the villa of Settefinestre and of the suburban villa of Sangemini. Cf. Soren 1999,419 with bibliographical references. 2 .3 Country Villas in Roman Central ItalY 243 This seemingly puzzling placement of the pars rustica on the higher part of the complex and in the best panoramic position has been tentatively explained with the fact that the climate of the region features storms and heavy winds. The hill crest, being completely exposed to the meteorological elements, would not have been the best spot to build the pars urbana. 5 Five major chronological phases were identified. The construction of the villa occurred sometime in the second half of the first century B.C. Already in the late fIrst or early second century A.D. the ceiling of the occus (Room 4) had collapsed, perhaps because of the unstable terrain or faulty construction, and the villa seems to have fallen on hard times (Second Phase). The wall paintings and the mosaic floors of the First Phase were not updated according to new styles, while the central part of the opus spicatum floor in Room 10 was removed, perhaps to re-use the room as some kind of industrial space. 6 However, the villa remained in use until the early third century (Third Phase). ｾＭ｜ｦｴ･ｲ this date, it may have been abandoned to squatters (Fourth Phase); although pottery fmds indicate continued use, no evidence of upkeep and repairs to the structure has been detected.? Finally, in the mid-fifth century _\.D., several rooms of the villa, now in ruins, were occupied by infant burials, comprising fetuses, newborns, and one- to three-year old children. It seems that all the depositions occurred within a short period of time, and this fact, in conjunction with the results of the examination of the bone remains by forensic anthropologists, has led to the theory of an epidemic outbreak. 5 Because of the cold and humid climate, most of the villas discovered in Umbria were heated by means of hypocaust systems in the rooms of the residential part. It is not clear if the villa of Poggio Gramignano had a heating system. Among the fmds, the discovery of tegulae mammatae is reported, but not tubuli, slupensurae, or the hypocaust. For other villa-sites in Umbria, cE. Soprintendenza _\rcheologica per l'Umbria 1983 and Marzano 2004, 392-409. : Both the bricks and their embedding were removed. Three hypotheses to explain the re-use of :be space are put forward by the excavators: 1) The room was used as textile factory. 2) An olive or ｾ･ press was installed. 3) It became a granary or food storage area. Of these three hypotheses, :mly nwnber 2 can in part account for the need to remove the opus spicatum floor. Some travertine :>locks found against the walls and mortared into the floor may have been part of the support for a ?ress, but no parts of the actual press were recovered, or traces of vats or channels where to gather :he liquid resulted from the pressing. In addition to this, the supposed press would have blocked :he dootway opened in Wall S when the wall was rebuilt in the Second Phase (Soren 1999, 208). In ｾ､ to hypothesis 1, I do not see the necessity to remove the floor if looms were to be installed :n the room (as noted also in Soren 1999, 207); no loom weights or spindle whorls were recovered other. Hypothesis 3 is also not satisfactory in explaining the removal of the floor: storage of grain ｾ､･ the best isolation from terrain and hwnidity, and if a superimposed wooden-floor was used, .-by remove the opus spicatum? Only if the intention was to place sunken dolia it would have been ::.ecessary to remove floors and embedding, but the excavators did not report finding large holes or ｾｩｯ fragments. After the early third century A.D., when the villa was supposedly used by squatters, excavators think that the room may have been used as a damp and stable, since the strata appeared :ompacted, as if trampled over by heavy traffic (ibid, 113). - Soren 1999, 43-44 and 119. 244 A TAll ORDER A considerable number of roof tiles recovered in the excavations had stamps, offering precious chronological data for the building phases of the villa. All the stamps indicate provenance from figlinae urbanae and can be classified in three chronological groups: 1. Stamps dating to the late first century A.D. 2. Stamps (including one of the Asinii) dating to the mid-second century A.D. 3. Stamps dating to the second century and to the early third century A.D. 8 Other material finds were abundant, especially pottery, in particular in the strata belonging to the Fourth and Fifth Phases. These phases, one must remember, correspond to the period characterized in the excavators' view by squatters and sporadic' occupation of the villa, and the ttansfonnation of portions of the villa into an infant cemetery. In fact, many amphorae and other vessels came from this cemetery, complementing the various fragments of common ware, kitchen ware, and amphorae related to actual occupation. The majority of infant burials used amphorae for coffins, a practice also wellattested in other Late Antique cemeteries; and it is the assortment of the villa's amphorae from all phases that illuminates an interesting combination of commercial goods acquired and consumed on site and in the surrounding area. The majority of amphorae used for transportation and storage of foodstuffs date to Period Five and include amphorae for wine, olive oil, garum and other fish products. The wine amphorae are mostly Italian types, like the Spello amphora, a small-sized and flat-bottomed amphora, produced in central Italy from the I\1iddle Empire on, and particularly suited for mixed land/water transportation. 9 The amphorae for the other above-mentioned foodstuffs did not include any Italian type, but all imports from the provinces, with a predominance of products of North Africa. In addition to containers for foodstuffs, African red-slip ware was also recovered, an 'anomaly' compared to other villa sites in Umbria, where African red-slip is wholly absent, substituted by locally produced types that imitate the fonns of the African vessels. The evidence discovered at this villa has been interpreted according to the general economic and social picture established for Roman Umbria. In the first century B.C. the territory of Amelia was subject to new occupation and to the installation of villas based on slave labor, replacing a land management fonnerly characterized by smallholdings engaged in self-sufficiency agriculture This third group is the largest one, counting 22 stamped tiles out of 30. Of the 22 stamps, 16 belong to the FigJinae PubJiJianae. 9 On carts to the Tiber and then on boats to Rome. The Spello amphora had a capacity of about 15-20 liters. 8 Country Villas in Roman Central ItalY 245 and pasture. 10 D. Monacchi observes that the presence, epigraphically attested in this territory, of many freedmen of Greek and eastern origin could be explained by the need of slaves for the country villas of the region. 11 She also notes that an inscription recovered in the area of Lugnano mentions an "ager Tresianus Masonianus," related to the gens Masonia, which owned slaves,12 and that Cicero, in his oration pro Rostio Amerino, stresses the nwnber of slaves held by his client. In fact, the references to slaves in Cicero's oration are to household slaves and personal attendants. 13 In the oration there is no explicit reference to slaves used in the 13 fundi of the Rosell as agricultural manpower. Of course Cicero's silence on this point does not exclude that the fanns were based on slave labor, nor does it exclude the use of other types of labor either, such as tenants or hired workmen. Cicero was not interested in giving details about types of manpower used in Roscius' fanns; this detail was not relevant for his defense speech. What Cicero carefully stressed was the value of the properties, unjusdy expropriated from his client and sold in a mock auction for a ridiculously low sum. Monacchi is misreading Cicero when she sees his testimony as referring to slave labor employed to cultivate the land. The expectation that country villas in central Italy, in the period between the Late Republic and the Early Empire, were part of a 'villa system' based on a slave mode of production, and that this system encountered a crisis starting in the Middle Empire because of the competitions of agricultural goods from the provinces and the diminished number of slaves on the market is behind ｾＶｮ｡｣ ｨｩＧｳ and other scholars' reading of the evidence. 14 But to this crucial point and how it affected the correct reading of archaeological evidence I will return shortly. From literary or epigraphic sources we learn that proprietors in the territory of Amelia included members of the local aristocracy, like the already mentioned gens Roseto in the Late Republic, or of the provincial aristocracy in the Middle Empire, as in the case of Pliny the Younger's father-in-law. D. Monacchi in Soren 1999, 36. Ibid., 34. 12 Ibid, 37. Freedmen of the Masonii are mentioned in inscriptions; cf. elL XI 4487-4489. 13 For instance, cf. Cic., Rose. Am. 27.77: ''Vos, qui hune aecusatis, omnes eius servos habetis; unus puer victus eotidiani administer ex tanta familia." 14 For example cE. Monacehi in Soren 1999, 40: ''Verso la fme del (I sec. d.C.) si avvertono i primi sintomi di crisi nel sistema delle ville amerine.... I segni piu vistosi sono materia4zzati dallo smembramento della cella vinaria della villa d.i Pennavecehia... dalla progressiva importazione in rotte e tre Ie ville indagate... delle derrate alimentari e delle ceramiche ftni da mensa dalle province romane.... E noto infatti eome nella concorrenza eserdtata sui mereati del Mediterraneo dalle merci prodotte nelle province... e state individuato uno dei moltepliei fatton responsabili della crisi del sistema delle ville." I discuss the case of the villa of Pennavecchia infra. 10 11 246 A TALL ORDER Imperial properties are also attested by inscriptions. IS However, the fact that the local aristocracies of Ameria and Otricoli failed to enter into the senatorial rank,16 and that the villas known in the territory of Ameria show architectural typologies similar to the ones used in elite residences in the territory of Rome, has been seen as indication of the high social rank of villa-owners, thought to be members of the Roman senatorial elite. This is the main reason the ownership of the villa we are examining, featuring a Corinthian oecus, has been attributed to a high-ranking figure, very likely a senator. The desire to give a name to and define the social standing of villa-owners is the first impulse when a villa is discovered. In reality, only in few fortunate cases can we securely relate a known historical ftgure to a specific villa, at least in a given time period, since estates and villas changed owners very often. 17 Elements taken into account when trying to detennine ownership include geographic location compared to references to villas in literary sources, brickstamps, jistulae,18 and place-names. In the case of the Poggio Grarnignano 15 For Calpurnius Fabatus, eques from Comum, cf. Plio., Ep. 8.20. Inscriptions attest the presence of freedmen a rationibus of Nero, freedmen of the Aavians, and an actor of Commodus (ClL XI 4360; 4462; and 4427). 16 So far we lack any testimony of senators from Ameria. Cf. Gaggiotti and Sensi 1982, 248. The gens Roscia, who had close connection with the Roman families of the Scipiones, Caecilii Metell4 and Servilii (Cic., Rose. Am. 6.15 and 51.148), is thought to have been of equestrian rank. Cicero, ibid 2.6 and 7.20 evaluates the patrimony of Sextus Roscius to 6 million sesterces. It is current opinion that the local aristocracies failed to reach the senatorial rank because of properties owned in the area by emperors and Roman senatorial families and their commercial interests, in combination with the Augustan land allotments attested in the Liber C%niarum (1:224 Lachmann), which would have caused the fragmentation of agrarian property. Cf. Soren 1999, 40, with bibliographical references. 17 On Roman mobility in real estate ownership, cf. Finley 1976. For studies focusing on senatorial properties and wealth, cf. Shattman 1975; and Andermahr 1998. 18 A personal name in a brick-stamp can indicate: a) the name of the owner of the fig/ina were the bricks/ tiles were produced; b) the name of the owner of the fundus were the material was used for construction-the fig/ina could have been in the same praedium or somewhere else; c) the person for whom the bricks/tiles where produced, with no link to the ownership of the jig/ina or the name of the officinatores. The owner of the fig/ina could also lease it to someone else. Fundamental studies about this problem are: Manacorda 1993 and 2000. About the economic and social conclusions that can be drawn from this type of material, cf. Andreau 1996. Any assignment of ownership based on brick-stamps depends on whether one interprets the tiles/bricks as having been produced on the fundus or bought on the market. M. Gualtieri 2000 has recently presented two case-studies about villas and brick-stamps. In one case (Oppido Lucano) the stamps do not indicate the owner name, as they come from several regional sources; in the other (Ossaia) the stamps seems to indicate the various owners. Overall, brick-stamps are not a reliable indication of the ownership of the villa. Stamps onftstu/ae are traditionally considered more secure than brick-stamps in indicating the name of the owner; cE. for instance, Eck 1982, but contra, cf. Aubert 1993: 177; and Bruun 2003, esp. 494-99, for a discussion of fistulae showing a proper name in the genitive case and having uncertain provenance. In these cases he suggests the necessity to consider also the possible provenance from public balnea, manufactures, etc. and not just villas. Bruun believes also that in Country Villas in Roman Central ItalY 247 villa, the brick-stamps are all from urban ftglinae and do not offer any useful element to detennine the name of the owner. The place name Gramignano, although it shows the ending typical of Latin adjectives derived from the personal name of the owner of praedia,19 cannot be securely related to any family name. One suggestion considers the possibility of deriving the name from that of the Etruscan gens Cramna,20 but this remains a hypothesis. If we consider the quality of the fmds at this site, lacking opus sectile floors, marble revetment or statues,21 the only element that has led the excavators to attribute the villa to a senator is the typology of Room 4, the Corinthian oecus. Monacchi writes that the other few examples of Corinthian oed we know of are attested in private buildings belonging to high-ranking figures, such as Augustus, the senator L. Sextius, and the Sextilii family. 22 In her view, the oecus of Poggio Gramignano shows the owner's Hellenized tastes, along with the ostentatious desire of showing his high social standing. The Late Republican or Early Imperial date for the construction of this room, a time when the municipal organization in Ameria was still in progress, led Monacchi to suppose that the architect and specialized workmen who built this villa came from workshops in Rome. 23 But can architectural typology offer us secure indication of the social class of the owner? The degree of refinement and elegance of a mansion can prove only the wealth and taste of the owner. But whether he was a senator, an eques, a rich freedman, or a decun·o is truly hard to say. In the villa of Poggio Gramignano the only two noteworthy rooms of the pars urbana, Rooms 4 and 8, are elegant, but not extremely lavish. One could object that only a portion of the residential part was uncovered, but on the basis of the evidence currently available, this villa is not astonishingly luxurious. Members of the local elite of Ameria could have easily owned such a villa, and we don't need to imagine a level of wealth comparable to the Roscii's. some cases the genitive on the stamp indicates the manufacturer, implying the label '(ex officina) illius', rather than the owner of the water line and, therefore, of the land (= 'aqua illius,), as these stamps have always been understood. 19 As in the place name Rosciano, derived from the Rosell. 20 Soren 1999, 37. 21 Of course we ought to remember that only part of the villa was brought to light; the bath quarters, for instance, were not found. 22 Soren 1999, 422-23. This type of accus is attested in Augustus' house on the Palatine, in the villa of Settefmestre, thought to have belonged to L. Sextius (cf. Manacorda in Carandini 1985, 104-6), in the House of the Labyrinth in Pompeii, owned by the Sextilii. In all these cases, one side of the oecus opens onto either the peristyle garden or columned terrace, whereas in our case there is the wall with the semi-columns. 23 Soren 1999, 414. 248 A TALL ORDER Scholars' attitude towards elegant domestic architecture is often ambivalent. On the one hand, in most cases only the senatorial rank is taken into consideration when attempting to attribute the owner to a villa, forgetting that a variety of other people with wealth owned the same type of mansions, requested the same luxury and architectural typology, and shared the same tastes. For example, we know from literary sources that even in a vacation spot as Tusculum, long favored by the Roman nobiJitas, L. Lucullus had a knight and a freedman for neighbors, who owned villas as luxurious as his own. 24 On the other hand, when facing truly original and extravagant architectonic solutions, scholars are more prone to see the presence of emerging local elites, homines nom and newly enriched people, reputed to have been more inclined to find eccentric architectural solutions in order to claim their membership in the upper class. 25 This is what I call the 'Trimalchio, influence: the assumption that the old aristocracy was immune to bad taste and extravagance, while oversimplifying the association of certain stylistic differences and excessiveness with a precise class of people. 26 In the case of Poggio Gramignano, one could argue that the architectural 'anomaly' of the occus not opening onto the garden is an indication of the villa being constructed to be used all year round, and not mostly for summer sojoums, as would be the case if we postulate as owner a Roman senator or high-ranking figure from another geographical area. In this case, visits of the dominus would have been sporadic, and concentrated in the summer/early fall months, when people would escape from the heat of Rome and eventually be present in their country estates for the harvest or vintage. If the owner was a member of the local elite and was residing at the villa also in winter, a completely walled up oecus would have been more appropriate for the Umbrian climate. The villa could have changed owners over time. In my view, the fact that after the collapse of the ceiling of the occus we have no evidence of new wall paintings or mosaic floors may indicate a move from an owner in residence to an absentee landlord (and whether he belonged to the The information is given by Cicero in a passage of the De Legibus 3.13.30, where he discusses the need for prominent men to set a good example for the whole of society: " ... L. Lucullus ferebatur quasi commodissime respondisset, cwn esset obiecta magnificentia vilIae Tusculanae, duo se habere vicinos, superiorem equitem Romanum, inferiorem libertinum; quorum cum essent magnificae villae, concedi sibi oportere quod us qui inferioris ordinis essent liceret. Non vides, Luculle, a te id ipsum natwn ut illi cuperent? .." 25 Cf. Broise and Jolivet 1995 for discussion of an oil mill with an original octagonal plan in Asine110 locality, on the via Cassia, mid-way between Rome and Viterbo. This is so far the only known example of a structure of this kind used for utilitarian purposes. 26 Modem attitude is influenced by this topos, which appears in various literary texts, not only Petroruus' Cena Trima1chionis. For instance cf. Sen., Ep. 27.5 about Calvisius Sabinus, who had the ''wealth and spirit of a freedman" and whose good fortune was a great offence against propriety." 24 Country Villas in Rnman Central ItalY 249 same family that owned the estate in the first phase or he was a new owner it is impossible to determine), rather than the decline of the villa. The second problem in the interpretation of the discoveries at Poggio Gramignano regards the chronological tenninus chosen for the beginning of the sporadic use and squatter occupation of the villa; Most of the stamps recovered on roof tiles date to ca. A.D. 200, that is they were manufactured at this date. They indicate a considerable restoration and maintenance of the villa sometime shortly after this date, since bricks and tiles were not necessarily used in buildings immediately after the production, but may have been in storehouses for a time before being commercially distributed. If such an effort to upkeep the villa was undertaken shortly after A.D. 200, postulating the abandonment of the villa and its "limited or sporadic use,,27 by squatters at about the same time is not satisfactory. It is safer to set the abandonment after a period of about 45 years from when the repairs occurred. 28 But why was the archaeological evidence interpreted thus? The heart of the problem lies in the current theoretical models in the field of Roman social and economic history, particularly in the development of the 'villa system' as mode of exploitation of the land and the old convention that everything in the Italian economy went wrong after A.D. 200. When historians consider the 'villa system' as a mode of production, they are referring to the type of villa rustica that appeared in the Italian countryside in the time of Cato, then developed into the villa described by Varro and Columella-that is, a villa with both a pars urbana, for residential purposes, and a pars rustica, which used slave labor for the concentrated cultivation of cash crops such as vines and olive trees. It is generally accepted that the 'villa' was the manifestation of a specific socia-economic system, which, in the aftennath of the Hannibalic war, had both new capital and a considerable number of slaves at its disposal. This type of villa is said to have developed in central Italy, specifically in the regions of Latium and Etruria. The scholarly definition of the villa system, and the construction of the economic models it implies, rested heavily not only on literary sources· like the works of the Latin agronomists, but also on the results of various archaeological investigations. The work directed by Andrea Carandini in the ager Cosanus and at the site of Settefinestte has been very influential in this sense. 29 At Settefines tte, S01;'en 1999, 43. As pointed out by Harris 2000 in his review to the monograph on the Poggio Gramignano villa. 29 The disappearance of small properties to the advantage of large slave-staffed villas and the type of management and work-force used (slaves vs. seasonal hired workers or tenants) have been much debated issues in recent years. It is not possible here to analyze the problem in further details, but cf. Marzano 2004, chapter 4 (with previous bibliography). Already in 1981, when preliminary results from the survey in the ager Cosanus and the excavation at Settefinestre were available, Rathbone 1981, 20 pointed out some problems in the interpretation of the survey data as indicating 27 28 250 A TALL ORDER Carandini and his team 30 excavated a large country villa, with at least three major phases: original construction in the mid-first century B.C.; additions and changes to the floor plan in the fltst century A.D.; and decline already leading to abandonment in the late second century A.D. The reason the publication of this villa so thoroughly influenced studies on Roman agriculture and the economy of Roman Italy is that, in its second phase, the Settefinestre villa seemed to exemplify Varro's villa peifecta. The villa also had two courtyards surrounded by rooms, which the excavators detennined to be slave quarters. The site thus offered an archaeological example of the large estates fanned by slaves in Etturia as described in the literary sources. 31 The signs of decay and, ultimately, abandonment in the villa in the second century were related to the market crisis of Italian exports, especially wine, under pressure from imports from the provinces, and to the diminished number of slaves available on the market. The crisis of Italian agriculture would have caused the abandonment of many villas, whose engagement in cash-crop agriculture was no longer remunerative. This shift, combined with the tendency to concentrate the land in the hands of fewer landowners, is the main reason to consider villas-both as agricultural production units and as elite residences-in crisis in central Italy in the second and third centuries A.D. However, the existence of a crisis in Italian wine and oil production is controversial in Imperial time. As has been remarked,32 the decrease in wine and oil exports from Italy indicates only a change in the distribution patterns of these products. Even if there was such a crisis, some scholars are starting to disassociate the phenomenon from the alleged decline of the villas, denying that the end of viticulture would have also meant the end of villas.33 The assumption that the economic production of villas was in crisis by the mid-second century prompted scholars to fmd proof that villas were increasingly being abandoned. Cases of poorly maintained villas, showing the subdivision of large rooms into smaller living quarters, signs of crude repairs, the disappearance of small villas and in the proposed figure for the size of the property belonging to the Settefinestre villa. He stressed that the use of slave labor was not more economic, and that the villa-system and peasant smallholdings were complementary mode of production (15). Similar conclusions were reached by Aubert 1994, chapter 3, esp. 162-68. Cf. also Carlsen 1984, who pointed out the influence of Marxist thought on Carandini's interpretation of the data from the ag" Cosanus. The results of the field survey carried out in the Albegna valley after the excavation of Settefmestre were recently published by Carandini and Cambi 2002, reviewed by Wilson 2004. 30 Carandini 1980; 1989a; and 1989b. 31 I challenge the slave-quarters thesis and the excavators' analysis of the archaeological evidence, showing that other theses are possible, in Marzano 2004, 109-32. 32 Panella and Tchernia 1994. The first-century crisis in Italian agriculture has been rejected by Purcell 1985 and Patterson 1987. 33 Lafon 1994; Metraux 1998. Country Villas in Roman Central Ita!J 251 and the reutilization of elegant residential parts for utilitarian purposes-such as workshops and storerooms-all provided ready evidence. Their function as elite residences ceased, while squatters dwelled, in precarious living conditions, in the abandoned stnlctures. This model, also called the 'Settefinestre' model, has been widely used by scholars to interpret discoveries at various villa sites in central Italy, including at Poggio Gramignano. It is my contention that this reconstruction does not necessarily fit every site, and some details of the model are not explained in a satisfactory way. Let us consider for a moment the occupation of a given site by squatters, for instance. What are we to comprehend when we read that a villa was taken over by squatters? Are we meant to understand that people used the structures illegally and temporarily as living quarters? Or that they settled in long-tenn, cultivating land neglected and forgotten by the property owner? What kind of resources did these 'squatters' have? Were these impoverished people living from hand-to-mouth? Nowhere do we get an explanation of what we are to have in mind with the tenn 'squatters'. The vessels of imported goods recovered at villa sites for these periods indicate that the people living there were able to participate in the wide range commercial transactions necessary to acquire such goods. It seems to me that in this view the defmition of these phases of occupation as 'squatters' is unsatisfactory. That a change occurred in many instances in the type of villa occupation during the second and third centuries is clear enough-witness the change in product distribution patterns from provincial to regional markets, and the possibly considerable impact of the second-century plague on rural and urban demograph y34-but we need to reconsider the notion that the crisis of the Italian villas was total. The assumption that lower standards of living in country villas meant, by and large, that they ceased to function as units of agricultural production seems to me to be faulty. As Metraux has pointed out and as I have argued elsewhere,35 it is likely that many villas were used 'as-is'. Especially in the case of contiguous estates, proprietors would have kept as elite residence only one villa, while leaving the others, still at the center of agricultural productive fundi, to their overseers or rented to coloni. Various reasons could explain such a practice, expressly stated in literary sources. In a letter addressed to his friend Calvisius Rufus, evaluating the pros and cons of 34 Contra the idea of depopulation in Italy, at least in the flrst half of the second century, cf. Lo Cascio 2003,4-6. He theorizes a situation of overpopulation, resulting in too much pressure on the economic and natural resources, and in high availability of manpower, thus explaining the high rate of turnover in the management of fundi and the short lease contracts. 35 Metraux 1998; and Marzano 2004, 182-92. Vera 1994, 245 also pointed out that signs of abandonment in second- and third-century villas rarely mean abandonment of the land as well, explaining the phenomenon with the concentration of properties, the reorganization of the mode of production, and the growing importance of the pagus-vims structure as an administrative unit. 252 A TALL ORDER buying an estate next to his own at Tifemum Tiberinum, Pliny lists the great advantage of not having to worry about furnishing another villa, while some of the personnel could also be shared between the two estates, to considerable economic advantage. 36 Pliny was certainly not the only proprietor ready to choose this solution. The gromaticus Hyginus, also ｧｮｩｴｾｷ in the time of Trajan, refers as well to landlords who opt to maintain only a few villas as aristocratic residences, while 'abandoning' others. 37 These considerations should warn us that signs of poorer living standards in a villa are not necessarily the indication that the villa was abandoned by the owner following the crisis of agricultural production. 38 Reconstructing a New Picture In the panorama of rural villas attested by archaeology, we are now in possession of a special, even a unique case, in which we can compare archaeological data with literary accounts of a villa and its management. The case in question is the famous villa in Tuscis owned by Pliny the Younger, and mentioned in many of his letters. In recent years, the site of this villa, already identified some time ago through brick-stamps bearing the name of Pliny and an inscription of a freedwoman named Plinia Chreste, was excavated. 39 Located near S. Giustino, in Umbria, 'Pliny's villa' was actually built under Augustus by M. Granius Marcellus, a member of the senatorial elite. 40 The events of the Granii family show the economic rise, diverse economic investments, and links with political life characteristic of an average gens of negotiatores. The family was from Puteoli, but they had commercial interests stretching from the ports of the East to North Africa. Their business in Umbria arose in the period of the first triumvirate, and we can deduce that they had specific interests in Hispellum from the fact that M. Granius was 36 Plin., Ep. 3.19: " ... quod non minus utile quam voluptuosum posse utraque eadem opera, eodem viatico invisere, sub eodem procuratore ac paene iisdem actorihus habere, unam villam colere et ornare, alternam tantum tueri. Inest huic computationi sumptus supellectilis, sumptus atriensium, topiarum, fabrorum atque etiam venatori instmmenti; quae plurimum refert unum in locwn conferas an in diversa dispergas." 37 Hyg., Grom. 170, p. 124 (Behrends 2000 = Th 93): "Praeterea solent quidam complurium fund 0 rum continuorom domi.ni, ut fere fit duos aut tres agros um villae contribuere et terminus qui finiehant singulos agros relinquere: desertisque villis ceteris praeter eamcui contribute sunt vicini non contenti suis ftnibus tollunt terminus quibus possession ipsorum finitur, et eos quibus inter fundos unius domini fIDes observantur sibi defendant." 38 This topic is treated in more detail in Marzano 2004, 179-92. 39 Braconi and Uroz Saez 1999. The inscription is published in elL IX 5930. 40 Cf. Braconi and Uroz Saez 1999, 191-93, with reference to previous bibliography for details on this family. Country Villas in Roman Central ItalY 253 appointed duovir quinquennalis there. 41 The Republican villa belonging to M. Granius had a simple plan, including a pars urbana with courtyard and cubicula, a pars rustica with a large threshing-floor, and a two-storey building, probably a granary. 42 An intennediate architectural phase between Gracius' and Pliny's ownership, dated to the Julio-Claudian period, seems to have concerned mainly the residential part, with the addition of a bath complex featuring the canonical sequence of caldan·um, tepidarium, and frigidarium.43 Building activity that can be attributed to Pliny includes the addition, next to the cella vinaria, of a vat for grape-pressing, the creation of a new covered space between the granary and the gallery running along the edge of the platfonn, and the addition of two buildings to the southeast, probably fannhouses with storerooms and stables. What emerged from the excavations at this site, unlike so many others, chiefly relates to the pars rustica, rather than the pars urbana. The addition of the wine-pressing basin indicates that at least part of the fundus was planted with vineyards, and in fact the pottery finds from the site confinn a mostly wine-oriented production. Already during the Granii phase, an abundance of locally produced Dressel 2-4 amphorae demonstrates the production of wine on the estate for the market, probably both local and regional. In the late ftrst century and the early second century A.D., new types of wine amphorae appeared, also produced locally, and intended mostly for Rome's market. 44 This picture is reinforced by finds pertaining to imported goods, which, with the exception of some types of Apulian wine during the Late Republican phase, consist of processed fish products, such as garum, and olive Oil45 from Spain and North Africa. Although the references to this villa in Pliny's correspondence deal with different aspects of its economy and production-i.e., difficulties in finding elL XI 5264: he also built a temple to Venus in Hispellum. The pars fUstica had a cella vinaria with dalia, the capacity of which has been hypothesized to have been 250/300 ｨｾ and two vats for the fermentation of the must. 43 Granius' brick-stamps disappear from this site around 15 A.D. It is possible, as Uroz Saez has suggested (Braconi and Uroz Saez 1999, 191), that this fact is to be related to the accusation of de rrpetundis and maiestas leveled against Granius Marcellus by his quaestor Caepio Crispinus at the end of the former's proconsulship in Bithynia. Absolved from the accusation of maiestas, he had to respond to the charge de repetundis in front of the recuperatores and probably had to pay some pecuniary sanction. Cf. Tac., Ann. 1.74. To the subsequent building phase may he related some stamps on roof tiles reading: CAESAR, which would eventually indicate the passage of the property from Granius to the imperialftscus (ibid., 46 and 193). 404 The amphorae dated to the Plinian phase of the villa account for 50% of the whole local production recovered. • 5 Forty percent of the processed-fish amphorae come from Spain; the olive oil is from Baetica. For the second-fourth centuries, North African amphorae for both olive oil and fish products were recovered. For the finds of the excavation cf. Braconi and Uroz Saez 1999. 41 42 A 254 TALL ORDER coloni, measures to take in case of a meager hatvest due to bad weather, etc. 46-he does not really spend time discussing the utilitarian parts of the villa. The letter to Domitius Apollinaris 47 in which Pliny describes his estate contains no reference to the pars rustica. We gather only in passing that the estate was not only a place for the "exercise of mind and body" (i.e., through writing and hunting), but also for agricultural production, since Pliny mentions vineyards and "production of the land" sent to Rome via the Tiber. On the other hand, the archaeological record clearly shows a series of building projects aimed at increasing the productivity of the estate and its utilitarian zone. It is possible, for instance, that the acquisition of contiguous estates 48 resulted in the need for more covered working areas and/or more storage areas (see the addition of the above-mentioned covered space measuring at least 40 by 3.4 m).49 There are other known villa-sites in the modem region of Umbria, but none has been systematically excavated, with the result that data pertaining to them are incomplete and rather fragmentary. Nonetheless, comparison with the finds from Pliny's villa can be illuminating. As I have stated above, in some cases the interpretation of the archaeological data reveals preconceptions about villas and Italian agriculture in the second century, which the results of the excavations at S. Giustino help to correct. For instance, a villa in the territory of Ameria, at Pennavecchia,50 built sometime in the flrst century B.C., is thought to shows signs, towards the end of the flrst century A.D., of "crisi nel sistema di queste ville a conduzione schiavistica."sl The signs of crisis in this specific case would be: 46 Plin., Ep. 3.19. Or, according to another interpretation, penuria c%norum alluded to the economic difficulties of the c%ni; cf. E. La Cascio, "The Economy of Roman Italy according to Pliny the Younger," paper read at the Symposium in Honor of J.H. D'Arms, 26 October 2002, Cohunbia University, New York. 47 Plin. Ep. 5.6. 48 Cf. the letter to Calvisius Rufus quoted supra n. 36. 49 Bracom and Uroz Saez 1999, 36 relate this reorganization of the villa complex to Pliny's use of metayage in his properties (the change from a five-year contract, with fixed rent paid in currency, to rent paid in kind with part of the harvest), requesting more storage space for the part of the fructus that was owed to him (probably a portion similar to the ftgures given in the lex Manciana, CILVIII 25902: one-third each of grain, oil, and wine, a bit less for other goods). The system also dictated the need, on Pliny's part, for a procurator and various attores to oversee the works; Plin., Ep. 9.37: "medendi una ratio, si non nummo sed partibus locem, ac deinde ex meis aliquos operas exactores, custodies fructibus ponaro." 50 As we have seen in the case of Pliny's villa and of the Poggio Gramignano villa, the establishment of most villas in Umbria dates to the first century B.C. The major villas were distributed along the Tiber and its tributaries. 51 Monacchi 1991, 183. 3 Country Vii/as in Roman Central ItalY 255 a) The removal of the dolia from the cella vinaria, which is filled in to allow the construction of a colonnade, thus indicating the cessation of vine production and the substitution of imported wine. b) The increased importation of foodstuffs from the provinces, as attested by amphora finds, a sign of the competition on the market of products from the provinces and of the progressive disappearance of Italian products. But if we combine this report with the more complete one coming from S. Giustino, different conclusions are possible. First of all, the removal of the earlier cella vinaria need not imply the total cessation of wine production on that estate or in Umbria in general. Only part of the vilJa has been excavated, so it is possible that the cella vinaria was simply moved someplace else. ｬ｡ｲｵｴ｣･ ｩｨ ｜ｾ restorations and changes in the plan, sometimes even drastic ones, were quite common-as we shall see below for the villa at Ossaia-and it is difficult to understand them fully when only part of the complex is known. At Pliny's villa, for instance, during Pliny's phase the hypocaust of the baths is abolished and filled in with debris from other parts of the complex, while the floor of the frigidarium is ruined by the installation of a drain. 52 From Pliny's epistolary we know that he visited and resided often in his property, with a high degree of comfort and elegance, so that the changes in the baths cannot be indication of the mansion's decline. We also know that the production of wine on the estate not only continued, but probably increased in size, if the thesis is correct that the new treading vat was double the size of the previous one. 53 The amphorae recovered in the villa at Pennavecchia show the same pattern for the movement of goods as that observed at S. Giustino and Poggio G ramignan 0: what is imported from the provinces is olive oil, olives, and garum. 54 No provincial amphorae for wine were recovered, and this fact makes me doubt the conclusion that wine was no longer produced at this site. Pliny's villa has shown that a new type of locally produced wine amphora appeared in the second century, the flatbottomed and small Spello amphora, the same type recovered at Poggio Gramignano. Since we know that wine was a fundamental part of ancient dietary and social habits, the lack of wine amphorae at Pennavecchia might be explained in various ways. The above-mentioned amphorae could have also been in use at Pennavecchia, containing wine either produced on the same property or bought from local estates, but they eluded identification. Or the estate could have used wooden barrels, which by their very nature would not 52 The relative chronology is secure because the bottom of the channel is paved with tiles stamped with Pliny's name (CPCS). S3 Braconi and Uroz Saez 1999,35. The vat, faced in cocciopesto measured 5 x 4 m. 54 The amphora-types discovered at this site are Africana I, Africana II-A, and spatheia. 256 A TALL ORDER have survived this long. In any case, it seems to me too much of a stretch to infer from importation of oil and gart/m that the villas were in crisis. After all, these products were never typical of the region. In the literary sources, for instance, the territory of Ameria is praised for the production, in the mid-first century A.D. or before, of apples, pears, and willOWS,55 not oil or garum. The presence of garum and olive oil from Spain and Africa fits in with the general movement of goods toward Rome. For an area that used the Tiber as a major transportation route, it is logical that whatever merchandise reached Rome could easily be redistributed along the Tiber valley. It would then be easier to buy olive oil from Baetica or Africa than, say, from Apulia. The boats would bring foodstuffs up the Tiber and return with wine for Rome. This type of exchange, emphasizing the acquisition of provincial foodstuffs, would also explain why in Umbria, with the exception of Poggio Gramignano, we do not find abundant African red slip ware, which was used to fill up the cargo ships transporting oil and garum from Africa. Imported table wares and other sorts of pottery were replaced by local products: in fact, a great variety of red slip ware, manufactured in imitation of African shapes, is attested all over central Italy. 56 The same types of observations are valid for the site of Poggio Gramignano, examined in the first part of this article. The results of the pottery analysis at this site show that most of its wine came from Italy, with no significant importation from abroad. For oil, garum, and other fish products, on the other hand, the pottery analysis indicates the opposite. The large number of finds, as pointed out by Martin, also indicates that the commtmity on that site was still able, in the fourth and fifth centuries, to participate in a wide-ranging commercial network, which points to the necessity of re-defining what is meant when tallcing of the 'decline of the villa' in the Late Empire. Not far from Tifemum Tiberinum, another recently excavated villa site has yielded interesting data-the villa at Ossaia, near Cortona. 57 In this case, the flexibility with which owners regarded the use and distribution of space at their villas is remarkable. At Ossaia, sometime between the first and third centuries A.D., the villa underwent major architectural changes. Previously residential rooms were turned into some sort of workshop; another part of the villa fell into complete disuse, while some rooms indicate a redistribution of space by means of brick-works. Columella, Rust. 4.30.4; and 5.10.19; Plin., NH 15.50.55, 58, and 59; and 16.177. Willows branches were used in the manufacture of baskets and to tie vines to stakes. 56 Cf. also the same scenario in Tuscany (Chianciano Terme, Mezzomiglio); cf. Soren forthcoming. Martin in Soren 1999, 361 notes that the attestations of African red slip ware at Lugnano, especially up to the third century, partially attenuate this argument. 57 Fracchia and Gualtieri 1996. 55 Country Villas in Roman Central Itafy 257 If this part of the complex alone had been excavated, it would have fallen into the general trend of decline assumed for country establishments in Italy. But in this case this is not a sign of the villa's 'decline' or its ceasing to be used as an elegant residence, for elegant mosaics chronologically contemporary with the workshop phase were discovered in other parts, and a beautiful opus sectile floor was added in a subsequent third-century phase. Moreover, the amount of pottery recovered for the third and fourth centuries was greater than for other chronological phases, including both imported and locally produced types, such as African red slipware and Middle Adriatic teTTa sigillata. These data not only show that the villa was still economically 'active', but also that it was used, as before, as a residence by its dominus. Indeed, the evidence offered by the elegant and expensive marble floor roles out the supposition of 'squatters' to account for the workshop or the downsizing in other parts of the villa. These data demonstrate how changes in ownership 58 or in the demands of the market could lead to changes in the use of space in a villa, eventually affecting also a conversion of the residential quarters but not a substantial change in the nature of the property. It remained a unit fundamentally devoted to economic enterprises, but at the same time providing the owner with a pleasant retreat. The Ossaia case shows that the use of space in a villa could be very flexible and that when only part of a complex is excavated it is very difficult to assess the true history of the complex. We do not know what kind of evidence we would have if the remaining portion of the villas at Poggio Gramignano or Pennavecchia were excavated. Conclusions WhenJ.H. D'Anns published his seminal study on Roman villas in the Bay of Naples the main questions he was asking concerned the assessment of the social standing of villa owners in the area. On the basis of a meticulous analysis of literary texts, he detennined who were the various owners of the 'pleasure homes' that dotted the Bay of Naples and the social network they partook. Published ten years later, Carandini's study on Settefinestre marked a clear change. Moving the focus from maritime villas to country villas, the main attention in this study was given to the economy and production of the villa 58 For this villa, on the basis of the brick-stamps recovered, ownership by the Vibii Pansae, a family native to Perugia, has been suggested, possibly followed by members of the Augustan family, should the brick-stamp CAESARVM refer to Gaius and Lucius Caesar. The excavators note that it is very tempting to relate the installation of the workshop in Phase 2 with a stamp naming a freedman of Aulus Gellius. 258 A TALL ORDER and the type of management it had. The interest in villas as place of production and not merely vacation retreats also changed the way we looked at coastal villas. D'Anns was the first to recognize that even the villas in the bay of Naples, always considered retreats for the practice of otium, were not completely separated from economic aspects, but had a series of satellite fanns engaged in agricultural production for the main villa and for the market as well. 59 In recent years, the type of questions we ask when studying Roman villas has changed. The considerable amount of new archaeological data that became available on villas in Italy in the past years is forcing scholars to intetpret and systematize them. For instance, the discovery in coastal villas of production quarters has indicated the important role that these units had as economic investments. What we are now asking in studying Roman villas in Italy is not only what kind of luxurious architecture and decor surrounded the owner, but what was produced on a given estate; if the production was for internal consumption only or also for commercial distribution; how long was the villa used and how the usage changed over time; the role of villa establishments in the territorial organization that lead to the Late Imperial and then medieval organization of the territory. If we look at villas excavated in the past, very often we discover that much useful infonnation was either neglected and not recorded, because it was deemed unimportant at the time, or, if recorded, it was not taken into consideration when publishing the excavation reports. For example, in the excavation of the villa near Volterra60 under the direction of E. Paribeni in the 1930s, following fascist enthusiasm for the 'golden age' of the Roman Empire, no record was made of the many amphora fragments dating to the :Middle and Late Empire recovered during the excavation,61 since the focus and interest at the time was exclusively in the Augustan phase and in the residential part of the villa. For this reason, much material studied long ago is in need of a re-evaluation; it is important to insert it into a larger context and to compare it with the new data available, trying to better defme what kind of evidence we are facing. The case of the Poggio Gramignano villa and the proposed date of abandonment we have discussed in the first part of this article is one of many possible examples of archaeological evidence being interpreted to fit existing theories. In this case, the proposed breakdown around A.D. 200 seems to owe more to the idea that the Italian economy at that date was in deep crisis than to the actual archaeological evidence. Similarly, the economic picture D' Arms 1977; and 1981. At Pieve Vecchia d.i Casale Marittimo. 61 Shepherd 1998 and Marzano 2004, 391. Shepherd points out that numerous fragments are visible in the photographs of the excavation. 59 60 Country Vtilas in Roman Central ItalY 259 reconstructed for the partially excavated villa at Penna Vecchia, supposing a cessation of viticulture and therefore a crisis of the economy of the villa, rests on the idea that the cessation of Italian wine exports indicated the cessation of wine production. 'The different reading given to the data on the economic picture of the villa site of Pennavecchia in the light of the new discoveries at Pliny's villa strongly reminds us of the possibility that new data will add substantial infonnation to reshape and better define a given pictur.e. The analysis of the various case-studies presented in this article aimed to show the difficulty in interpreting archaeological data without the influence and suggestion of pre-existing theoretical models. Such models are a valuable, necessary tool, but at the same time we should be ready to adjust and change them according to the specific evidence we are facing, and not to try to interpret the evidence according to the model at any cost. As observed by I. Hodder, archaeologists use inductive methods to construct an understanding of historical meaning, but the extent to which their interpretations succeedas well as the universal validity of the results-ultimately depends on the richness of the data and on what the excavator is 'looking for'.62 In this article I tried to explain why we have been interpreting a certain class of archaeological data in a certain way by using as case studies some well-known villa sites. It is not the specific results of the excavations, conducted with great care and diligence, that I meant to challenge, but the general methodology behind the interpretation of archaeological evidence pertaining to villa sites that we have been using when dealing with sites in central Italy. Works cited Andennahr, A.M. 1998. Totus in praediis. Senatorischer Grundbesitz in Italien in der Fmhen und Hohen Kaiserzeit. Bonn. Andreau, J. 1996. "Les inscriptions et l'economie romaine," ]RA 9:409-14. Aubert, J.-J. 1993. ''Workshop managers," in The Inscribed Economy: Production and Distribution in the Roman Empire in the ught of instrumentum domesticum, edited by W.V. Harris, 171-81. JRA Supplement 6. Ann Arbor. 1994. Business Managers In Ancient Rome: A Social and Economic Aubert, ｊＮｾ Stut!J oflnstitores, 200 B.C.-A.D. 250. Columbia Studies in the Classical Tradition 21. Leiden. Behrends, O. et a!., trans. 2000. Hygin, j'ceuvre gromatique: Corpus agrimensorum Roman0 rum V Hyginus. Cost Action G2. Paysages antiques et structures rurales. Luxembourg. 62 Hodder 1991, 191. A TAlL ORDER 260 Braconi, P., and J. Uroz Saez, eds. 1999. La villa di Plinio il Giovane a San Giustino. Ponte S. Giovanni (perugia). Braise, H. and V. Jolivet. 1986. "Interventi di scavo in localici Asinello e nell'abitato eli Musama," in Archeologia nella Tuscia II: atti degli incontn' di studio organizzati a Viterbo (1984), edited by A. Emiliozzi Morandi and A. M. Sgubini Moretti, 113-15. Rome. Bruun, C. 2003. "Le fistule aquarie e i proprietari terrieri nel suburbium di Roma," in Suburbium. II suburbio di Roma dalla crisi del sistema delle ville a Gregorio Magno, edited by P. Pergola et aI., 485-501. CEFR 311. Rome. Carandini, A. 1980. "II vigneto e la villa del fonda di Settefinestte nel Casano. Un caso di produzione agricola per il mercato transmarino," MAAR 36:1-10. Carandini, A., ed. 1985. Sette.ftnestre: una villa schiavistica nell'Etrutia romana. Modena. Carandini, A. 1989a. "Italian Wine and African Oil: Commerce in A World Empire," in The Birth of Europe: Archaeology and Social Development in the First Millennium A.D., edited by K. Randsborg, 16-24. Rome. Carandini, A. 1989b. "La villa romana e 1a piantaggiane schiavistica," in Storia di Roma. Caratten e moifologie, vol. IV, edited by A. Schiavone, 101200. Torino. Carandini, A. 1994. "I paesaggi agrari dell'ltalia romana visti a partire dall'Etruria," in L'Italie d'Auguste a Diocletien. Actes du Colloque International (&me, 25-28 mars 1992), 167-74. CEFR 198. Rome. Carandini, A. and F. Cambi. 2002. Paesaggi d'Etrutia. Valle dell'Albegna, Valle d'Oro, Valle del Chiarore, Valle del TaJone. Progetto di ricerca italo-britannico seguito allo scavo di Sette.ftnestre. Rome. Carandini, A. and S. Settis. 1979. Schiavi epadroni nell'Etrutia romana. La villa di Sette.ftnestre dalln scavo alia mostra. Bari. Carlsen, J. 1984. "Considerations on Cosa and Ager Cosanus," ARID 13:49- 58. D'Anns, J.H. 1977. "Ville rusriche e ville di 'orium'," in Pompei 79. Racco/ta di studiper il decimonono centenario del'･ｲｵｾｯｮ vesuviana, edited by F. Zevi, 65-86. Napoli. D'Anns, J.H. 1981. Commerce and Social Standing in Ancient Rome. Cambridge, MA. Eck, W. 1982. "Die fistulae aquanae der Stadt Rom. Zum Einfluss des sozialen Status auf administratives Handeln," in Atti del Col/oquin ･ｬ｡ｮｯｾ ｲ･ｴｮｩ AIEGL su epigrafia ed ordine senatorio, 197-225. Tituli 4. Rome. Finley, M.I., ed. 1976. Studies in Roman Properry. Cambridge/New York. Fracchia, Hand M. Gualtieri. 1996. "The Imperial Villa at Ossia," EchCI 15:157-200. Country Villas in Roman Central ItalY 261 Gaggiotti, M. and L. Sensi. 1982. "Ascesa al senato e rapporto con i territori di origine. Italia: Regio VI (Umbria)," in Atti del Colloquio ･ｬ｡ｮｯｾ ｲ･ｴｮｩ AlEGL su epigrajia ed ordine senatorio, 245-74. Tituli 5. Rome. Gualtieri, M. 2000. "Figlinae, domi nobiles ed approvvigionamento di laterizi nell'Italia centro-meridionale: due casi di studio," in La brique antique et medievale. Production et commercialisation d'un materiau. Actes du colloque international (Saint- Cloud, 16-18 novembre 1995, edited by P. Boucheron, H. Broise, and Y. Thebert, 329-40. CEFR 272. Rome. Harris, W.V. 2000. Review of A Roman Villa and a Late-Roman Infant Cemetery: Excavation at Poggio Gramignano, lJignano in Teverina, by D. and N. Soren, eds. BMCR (2000). http://ccat.sas.upenn.edu/bmcr/2000/2000-10-27.html (27 Oct. 2000). Hodder, I. 1991. &ading the past. CUTTent approaches to interpretation in archaeology. Cambridge/New York. Lafon, X. 1994. "Les villas de rItalie imperiale," in L1talie d'Auguste a Diocletien. Actes du Colloque International organise par L 'Ecole francaise de Rome (Rome, 25-28 mars 1992),219-26. CEFR 198. Rome. Manacorda, D. 1993. "Appunti. sulla bollatura in eta romana," in The Inscribed Economy: Production and Distribution in the Roman Empire in the Light of Instrumentum Domesticum, edited by W.V. Harris, 44-54. Ann Arbor. :Manacorda, D. 2000. "I diversi significati dei bolli laterizi. Appunti e riflessioni," in La brique antique et medievale. Production et commercialisation d'un matinau. Actes du colloque international (Saint- Cloud, 16-18 novembre 1995), edited by P. Boucheron, H. Broise, and Y. Thebert, 127-60. CEFR 272. Rome. Marzano, A. 2004. Villas and Roman Society in Centralltafy:from the Late Republic to the Middle Empire. Ph.D. diss., Columbia. Metraux, G.P.R. 1998. "Villa rustica alimentaria et annonaria," in The Roman Villa. Villa Urbana. Symposium on Classical Architecture, edited by A. Frazer, 1-19. Philadelphia. ｾｯｮ｡｣ ｨｩＬ D. 1991. "La cultura materiale delle ville romane del territorio amerino in eta tardo anrica," in L 'Umbria meridionale fra tardo-antico ed altomedioevo. Atti del Convegno di Studio (Acquasparta 6-7 maggio 1989), edited by G. Binazzi, 181-95. Assisi. Panella, C., and A. Tchemia. 1994. "Produits agricoles transportes en amphores," in L 1talie d'Auguste a Dioclitien. Actes du CoJloque International organisepar L'Ecolefrancaise de Rome (Rome, 25-28 mars 1992),145-65. CEFR 198. Rome. Patterson, J.R. 1987. "Crisis: What Crisis? Rural Change and Urban Development in Imperial Apennine Italy," PBSR 42:115-46. Purcell, N. 1985. "Wine and Wealth in Ancient Italy," IRS 75: 1-25. 262 A TALL ORDER Rathbone, D.W. 1981. "The Development of Agriculture in the 'Ager Cosanus' During the Roman Republic: Problems of Evidence and Interpretation," IRS 71:10-23. Shatzman, I. 1975. Senatorial Wealth and Roman Politics. CollLatomus 142. Bruxelles. Shepherd, E.]. 1998. "Portrait of the Archaeologist as a Young Man. Enrico Paribeni e 10 scavo eli Casalmarittimo," in In memoria di En17:co Paribeni, edited by G. Capecchi, 427-50. Rome. Soprintendenza Archeologica per l'Umbria. 1983. Ville e insediamenti mstici di eta romana in Umbria. Perugia. Soren, D., and N. Soren, eds. 1999. A Roman Villa and a Late-Roman Infant Cemetery: Excavation at Poggio Gramignano, Lugnano in Teverina. Rome. Soren, D., ed. Forthcoming. The site of Mezzomiglio. British Archaeological Reports Series. Vera, D. 1995. "Dalla Villa perfecta alia villa eli Palladia: sulle ttasfonnaziorii del sistema agrario in Italia Era principato e dominato," Athenaeum 83:189211 and 331-56. Wilson, A. 2004. "Tuscan Landscapes: Surveying the Albegna Valley," IRA 17.2:569-76.