Results of the New Excavations, 1997-2001

Bernard Frischer

Monica De Simone

New Excavations at Horace's Villa, 1997-2001

Contents

Background

Bath Complex

Quadriporticus and Garden

Residence

Conclusion

This site is
sponsored by

The Steinmetz
Family of
Los Angeles

The Samuel H.
Kress Founda-
tion

The Vincenzo
Romagnoli
Group

The Creative Kids
Education Foundation

Transportation
provided by
Alitalia

This site is
hosted by
UCLA's
Humanities
Computing
Facility

Last revised:
Jan. 1, 2002

Background

The new excavations (fig. 1 [high resolution (196k); low resolution (96k)]), initiated in the summer of 1997, have brought to light much new data about the archaeological site, clarifying some old problems and at the same time posing new questions and providing new directions for research. It is no exaggeration to begin by stating that the project has significantly altered our understanding of the site. For the student of Horace, the most important result is doubtless the finding that almost everything hitherto considered to date to the Augustan age really dates to the period 75/125 A.D. In compensation for this loss, the project found new features dating to the first century B.C. and thus belonging to a phase in the life of the villa that the poet could have experienced. The project found no new evidence either supporting or contradicting Horatian ownership of the villa, which remains as problematic as ever.

Most of the initial objectives were met. These were to:

1. Provide a close reading of the Licenza Valley through the early medieval period based on an archaeological survey and on archival research

2. Establish the natural property lines of the Villa of Horace as well as the ancient access road from the Via Licinese

3. Look for evidence of the pars rustica of the villa

4. Verify the state plan of existing structures and reexamine the different construction phases of the villa

5. Explore unexcavated areas of the complex, including the garden area in the peristyle and the western hillside

6. Publish a final excavation report integrating the earlier studies since the 18th century

7. Conserve the small finds collected during the project

8. Create a virtual reality model to be posted on a Web site

Fieldwork techniques and operations were to include:

9. Field survey and collection of surface finds

10. Magnetic and electric prospections

11. Borings, soundings, excavations

12. Palaeobotanical studies

Objectives that were not met include: survey of the Licenza Valley (1); determination of the natural property lines of the villa as well as the villa's access road from the Via Licinese (2); finding the pars rustica of the villa (3); and creation of a 3D computer model of the villa (8). The main obstacle to achieving these objectives was lack of funding, which prevented the participation of the co-principal investigators invited to undertake these tasks.

On the other hand, several new objectives were added to the initial list and were successfully realized, including creation of a World Wide Web Site; creation of a video documentary about the project; installation of a "green" retaining wall to protect the western slope of the site in an environmentally friendly and aesthetically pleasing way; and installation of new signs and benches on the site. More importantly, the originally planned workload of 1080 man days of fieldwork (1997: 216; 1998: 432; 1999: 432) could be increased to 2450 man days (1997: 216; 1998: 432; 1999: 1440; 2000: 170; 2001: 180). The additional manpower enabled us to keep pace with the work necessitated by objectives (4) and (5), the scope of which turned out to be far greater than anticipated. This happened because, as noted above, it was not possible (as we originally thought) merely to fill in gaps in our knowledge left by the earlier excavations, but we had to re-excavate and reconsider the earlier work once it became clear that our predecessors had misinterpreted their finds.

In earlier versions of this Web site we have provided year-by-year accounts of our work (click here for the 1997 and 1998 reports). These are now updated and integrated with the results from the 1999, 2000, and 2001 campaigns in the following synthesis, which is organized topically.

Bath Complex (rooms 35,37,38,39,40,50,51)

The bath complex is located along the northwest limit of the archaeological zone. It was largely excavated by Pasqui from 1911 to 1914. As early as the time of Pasqui two different building phases were distinguished. The older of the two, located to the north, comprises the rooms 32-34 and is constructed in opus reticulatum and in brick. This area was completely excavated by Pasqui. In a later period were added other rooms (35-53), which like the earlier thermal rooms contained plumbing and heating. We determined that these more recent rooms were in part built on top of earlier structures with a different function. This second phase differs from the first by a more extensive use of brick and by a greater complexity of the architectural design. One room (53) is of great architectural interest. Long interpreted as a vivarium, or pool for raising fish, we determined that it is actually a laconicum, or hot room.

Pasqui did not completely excavate this area, and so we concentrated on a series of adjacent rooms that he either did not study at all (35,37,50,51) or only in part (38,39,40). Moreover (as our comment on room 53 suggests) we also did a thorough wall census of all the existing walls in the Bath Complex with three goals in mind: to distinguish ancient remains from modern restorations; to establish a relative chronology of the walls; and to determine the functions of the various rooms.

The excavation of rooms 38, 39, and 40--which were previously interpreted as late-antique additions to the thermal complex--has permitted us to recognize several earlier phases. In the first phase--loosely dated to the second half of the first century B.C. on the basis of the scarce pottery finds--there was just one large room covering this area. It was rectangular in shape and had an impluvium in the middle. In the second phase, the pavement level of this room was raised and covered with a black and white mosaic, and the room itself was enlarged with the addition of a large apsidal basin (37). The basin had benches on the side walls, a central stepped fountain, and was paved with a mosaic, of which only white tesserae survive. We interpret the structure in its second phase as a frigidarium (a room used for bathing in cold water). In this phase, therefore, the area was radically transformed. The earlier atrium was demolished and replaced with a frigidarium, to the south were built the service room 50 and the biapsidal room 51. To the north a portico was constructed with seven pillars aligned north-south. It presumably faced a garden to the west. The water supply was provided by a series of lead pipes bringing fresh water from the west. The stratigraphic data and the stamps on the pipes (with the name of the manufacturer--C. Iulius Priscus--and of the owner of the villa--P. <H>ostilius Firminus) permit us to date this phase to ca. 75/125 A.D. Publius Hostilius Firminus was either the Hostilius Firminus mentioned in two letters of Pliny the Younger as a corrupt legate in Africa and senator (cf. epist. 2.11.23f and 2.12.1-5), or else he was a close relative of the senator, since the name is quite rare. Pliny's letters date to 100 A.D. , which accords well with the post-quem date provided by the Flavian pottery associated with the pipes. In late antiquity (IV-V centuries A.D.) the bath complex fell out of use and underwent another major transformation in a third phase with the creation of walls subdividing the space in room 50 and with the insertion in the area of room 40 of a small cemetery. Two burials were identified. Of special interest is the tomb "a cappuccina" of a male (cf. fig. 3). We commissioned radiocarbon datings of the remains from the University of Arizona. The burials fall into the period 275/350 A.D., a period that accords well with analogous developments elsewhere at villa sites in the Campagna Romana. In the fourth phase, dating to the VI-IX centuries, there are signs of new life on the site as the large room that originally was an atrium and then part of the frigidarium now is subdivided by walls and broken up into the smaller rooms 38, 39, and 40. The apsidal basin (37) is now definitively put out of service and is partially covered by a wall. Room 40 is directly supplied with running water. The Bath Complex in general is remanaged in this phase, with some rooms restored for new purposes and other rooms pillaged for building materials and abandoned. The fourth phase is perhaps to be associated with a monastic settlement (in large part destroyed in Pasqui's excavations), to which can be referred a series of early medieval ditch graves found in area 35 and in some medieval walls still to be seen the the western slope above rooms 50 and 51.

Among the notable finds from these rooms were three statuettes found in the apsidal basin (37). They are all in white marble and of comparable dimensions. Fragmentarily preserved are: an Eros of the "Eros Thanatos" type (fig. 4); a Venus of the Capitoline or Medici type (fig. 5); and a head of Hercules, possibly of the Lenbach type (fig. 6).

Plan showing the garden and quadriporticus

The Quadriporticus and Garden

Several areas were excavated around the large quadriporticus (fig. 1: 23,54,55 [high resolution (196k); low resolution (96k)]). The remains of the quadriporticus visible today are built in opus reticulatum, but our wall census has ascertained that these are almost entirely modern restorations. In the southern parts of the east and west corridors and the entire length of the southern corridor the walls are preserved only at the foundations, sealed by a restoration cap. The excavations in the western corridor (23a,b) brought to light the first phase of building in this area: an older stone structure over which the western back wall of the quadriporticus was built. The earlier structure was carefully constructed in opus incertum; in one spot (23a) some red painted plaster is preserved. Even if its function is unknown, this wall in opus incertum gives clear evidence of a phase preceding the construction of the quadriporticus--a phase not attested before our excavations. The wall was used as the foundation for this part of the quadriporticus in the second building phase, during which the floor level was significantly raised. Also datable to the second phase is the drain exposed at the foot of the northwest staircase leading from the corridor to the residence and rising higher than the floor of phase one (23a). The raising of the surface level in this part of the villa is encountered elsewhere on the site at the same period. Giving an absolute chronology to this phase is difficult in view of the scarce ceramic finds and the presence of some residual material.

The investigation conducted in the eastern corridor excluded a phase analogous to the first phase in the western corridor. The eastern part, which drifts slightly to the southwest in orientation, presents a series of interpretative problems partially connected to various restorations in the twentieth century and complicated by the presence here and there of intrusive constructions. These include a brick structure (55b) inserted into the corridor on axis with the great pool (25) in the middle of the garden. This structure can be identified as either a fountain or a theatral screen. The latter is perhaps more likely since no evidence of a water supply or of a drain was found. In either case, its structural function was to retain the steep hillside to the east. The same is likely true of the other structures inserted into the eastern corridor (55a, etc.). Dating of this phase to a precise period is not possible given the total lack of stratigraphy. Based on the similar construction technique found here and, e.g., in the apsidal basin (37) we propose a date somewhere in the period 75/125 A.D.

Area 54 on the south side of the quadriporticus was excavated in order to verify the hypothesis of various scholars that the main entrance to the villa was to be found here. The investigations led to the identification of the foundation of the wall, to which were bonded buttresses similar to those present on the exterior wall of the western corridor, except that they are set closer together (see illustration above). In fact the buttresses are so close together (1.20 meters apart) that it is difficult to imagine a monumental entrance here. On the other hand, we cannot exclude the possibility that there was an opening for at least a secondary doorway giving access from the garden and quadriporticus to the adjacent land to the south.

The vast area surounded by the quadriporticus was the object of various interventions from 1997-1999 that were mainly intended to verify traces of the garden, which we assumed filled much of the space within the limits of the quadriporticus. We also were interested in determining the chronological relationship of the piscina in the middle of this area to the rest of the villa. Excavations and borings established various phases for the garden. Special attention was paid to area 24 at the foot of the central staircase (13) leading from the garden to the residence. Here Kathryn Gleason and her team from Cornell University dug a trial trench to see if remains of the garden could be found. In area 24, two levels of planting were found. The earliest can be dated by pottery finds to the last half of the first century B.C. A new and more complex organization of the green area was detectable in the late first century A.D. Among the various finds useful for reconstructing the design of the garden, the most important was the discovery of a flowerpot still in situ. It dates to the Flavian period (see illustration above). Two other containers for plant materials were found on a line oriented north-south and located at the foot of the central staircase leading from the residence to the garden. It is known that the Romans often planted flowers and other plants in such containers, aligning them at regular intervals on a grid system in order to create ornamental areas along pathways and canals in their gardens.

In contrast to the garden, the piscina yielded only very little material useful for dating. This structure, built in opus signinum, had both a decorative function in the middle of the garden and also served as an open-air cistern for storing water. It may also have been used for fish cultivation. In the late phase of the villa, when the site was no longer inhabited, the piscina was stripped of its decorataive elements (which might have included statues and marble revetment) and used as a waste dump, as is attested by the many fragments of marble and building material constituting its fill.

Plan showing the residence

The Residence

One of the main objectives of the project was to identify the complete extent of the complex. Various considerations led us to doubt the validity of the plan of Pasqui's excavations published in 1926 by Lugli and frequently reproduced thereafter. Particularly suspicious was the north wall that seemingly closed off the residence. Our wall census showed that the wall was largely a modern invention; our archival research brought to light documents confirming this observation; and our soundings in 1999 proved that three of the walls did in fact continue running northwards (cf. the soundings beyond areas 8,17, and 26 on the plan above). It is not possible on the basis of our current knowledge to hazard a guess as to the date or function of these walls, but their discovery clearly necessitates a rethinking of the plan and design of the villa.

A drawing found in the archives, which dates from ca. 1912 (see illustration above), occasioned a small excavation inside the residence in 2001. At the beginning of the project, we had decided not to dig in the residence because we knew that it had been thoroughly excavated by Pasqui from 1911 to 1914 and so we did not expect to find any undisturbed ancient stratigraphy here. But the document we found surprised us: it showed structures below the surface of the atrium (room 12) that were hitherto unmentioned in the scholarly literature (cf. the upper left-hand part of the drawing for the these structures).

While we knew that Pasqui had destroyed ancient stratigraphy and had heavily restored the site--in some cases, as we have just seen, even inventing new walls--we did not know that he had reburied some of the authentic ancient features he did find. But a glance at the atrium as it appeared in 2001 before our excavation (see illustration above; and note that the photograph and drawing have the same orientation and viewpoint) will confirm that the features in the 1912 drawing were no longer visible on the surface. So we were naturally curious to know whether they could still be found, if the area was excavated. And, even more importantly, we wished to know if the purpose and date of the structures could still be determined, despite Pasqui's destructive methods of digging.

As the above photograph shows, the excavation of July 2001 confirmed that the features drawn in 1912 still exist and predate the walls visible on the surface today. In particular, the square structure in the top corner of the drawing turned out to be part of a water basin. It was presumably rectangular in shape, but this could not be confirmed since, for a variety of reasons, it was not possible to trace its walls beyond room 12. Other structures were also found whose existence had never been suspected. The basin is interesting for several reasons. Although no dating materials were found in association with it, we can put it securely into a relative chronology that is earlier than the building phase visible on the surface today and that probably corresponds to the first garden layer and the atrium phase of the imperial bath complex. This adds weight to our redating of the remains currently visible in the residence, on other grounds, to the period 75/125 A.D.--a significant change in our understanding of the complex, since Pasqui and all later scholars have considered the residence to be Augustan in date. Moreover, the function of the basin may have been to ornament the garden in the first (late-Republican) phase; alternatively, it may have had a more practical use. In the latter case, it would be our first evidence of the long-sought pars rustica (i.e., working farm area) of the villa. At the moment, opinion is divided about how to interpret the function of this water feature.

Be that as it may, the excavation of 2001 was also important because it confirmed the reliability of the 1912 drawing and also the fact that Pasqui had not destroyed the features he reburied. Since the 1912 document was found in a cache of several such drawings, some of which show other features no longer visible in the residence, this is an encouragement to future excavators to undertake other interventions in the residence.

Conclusion

The Horace's Villa Project began in 1997 with the notion that after two major excavations earlier in the twentieth century, all that was left to do on the site was to fill in some gaps in our knowledge and to clarify some remaining questions about details of the design and phasing of the villa. As the project developed, it became clear that our mission had shifted significantly, as evidence kept coming to light all around the site that the earlier excavators had misdated many features to the Augustan age--doubtless in an attempt to make Horatian ownership more likely. Finding structures in opus reticulatum, they leapt to the conclusion that these were datable to the late first century B.C. and stopped digging before reaching virgin soil. Application of the technique of stratigraphic excavation and reconsideration of the material found by our predecessors (especially the mosaics and the fresco fragments) suggested that what the earlier excavators considered Augustan should really be assigned to the period 75/125 A.D., and study of the lead waterpipe inscriptions even suggested one owner who nicely fits into this period--Publius Hostilius Firminus. Meanwhile, the case for Horatian ownership at an earlier phase was, if anything, strengthened, since we also brought to light some hitherto unknown features all around the site that did date to the first century B.C. Hence, much to our surprise (not to say consternation!) what began as a small project to put the final touches onto the large canvas painted by our predecessors turned out to be a feasibility study for a whole new major study of the site.

Taken as a whole, our new finds--important as they are--are only a drop in the proverbial bucket. Limitations of time and money (for which we had only ourselves to blame, given our intial limited understanding of our mission, which determined our budget) made it impossible to do more than undertake some test excavations and soundings. If these bore fruit (and they did), then they constitute no more than a strong incentive to return to the site in the future with much more time, manpower, and funding in order to explore the 90-95% of the site where we estimate that no one has yet dug down to virgin soil. This includes the garden, where Prof. Gleason's painstaking work in one small area has proven that data survives on two levels that could permit a full reconstruction of the design of the quadriporticus park. But it also encompasses the baths and residence, where we have shown that late-Republican structures still are to be found beneath the imperial walls. These three areas would require substantial financial support since the garden is large (ca. 40 meters x 80 meters, with an average overburden of 1.5 to 2.0 meters); and in the residence and baths excavation would often mean cutting through substantial imperial remains, which would in itself be expensive and would also necessitate a large investment in conservation.

That someone will someday undertake this major new effort is inevitable, given the importance of the site and the fact that what makes it important--the villa as Horace knew it--is just starting to emerge from the soil. Whether that day will be in the immediate or distant future, only time will tell.