Navigation – Plan du site

AccueilNuméros135-2VariaEarthen Architecture and Seismic ...

Varia

Earthen Architecture and Seismic Impact at the Roman Villa of Rufio (Giano dell’Umbria, Italy)

Jaime Molina Vidal, María Pastor Quiles (corresponding author) et Daniel Mateo Corredor
p. 521-543

Résumés

La villa romaine de Rufio, construite à l’époque augustéenne, a été occupée jusqu’à la fin du Ier siècle-début du IIe siècle de notre ère. Cet article aborde l’une des parties les plus fondamentales, mais souvent presque invisibles, des formes architecturales de ce complexe, à savoir ses structures érigées en terre. Leur étude laisse supposer qu’il s’agissait de murs en terre battue. Cette technique de construction peut être difficile à identifier dans les contextes archéologiques, et ne figure pas parmi les plus mises en valeur dans l’architecture romaine. Pourtant, on sait qu’elle a été utilisée dans tout l’Empire et dans différents sites de la péninsule italienne. Les caractéristiques montrées par les vestiges architecturaux de cette villa lors de sa fouille peuvent être liées à l’activité sismique, étant donné qu’elle se trouve dans un territoire à haut risque sismique où l’architecture en terre, ainsi que d’autres éléments de construction, pourraient avoir joué un rôle dans ce sens au cours de l’époque romaine. L’utilisation de murs de terre massifs, l’adoption d’autres techniques et éléments de construction et une série de dommages, de déplacements et de déformations peuvent être interprétées comme une séquence de phases de construction, peut-être motivées par différents événements sismiques.

Haut de page

Notes de l’auteur

This work has been supported by diverse institutions and developed inside the projects PGC2018–099843–B–I00 (MCIU/AEI/FEDER, UE) and PID2019–107264GB–I00. MPQ’s research was funded by the Spanish Ministry of Science and Innovation (FJC2019–039469–I) and by Next Generation-EU (MARSALAS21-19).

Texte intégral

I. Introduction

  • 1 Llidó – Molina 2012.
  • 2 Suet., Iul., 76.

1The Roman villa of Rufio owes its name to the inscription found in the villa that refers to its owner, C. Iulius Rufio,1 a character that can be identified with the son of a freedman very close to Caesar, mentioned by Suetonius in De vita Caesarum.2 This architectural complex was built on a hillside, with a total height difference of about 8 m. It is located in the municipality of Giano dell’Umbria, in the Umbria region, in a rural environment well connected to the surrounding territory. In fact, the villa was directly installed on the ancient Via Flaminia, on the Regio VI, between mansio Ad Martis (Massa Martana) and Mevania (Bevagna) (fig. 1).

Fig. 1. Location map of the Villa of Rufio (figure by the authors).

Fig. 1. Location map of the Villa of Rufio (figure by the authors).
  • 3 Molina et al. 2017.

2The first archaeological interventions at the site began in the late 1990s and early 2000s by the Italian archaeological company Kronos, overseen by the Soprintendenza, and excavations began in 2002. From 2007 to 2013, the University of Alicante joined the project with a team led by J. Molina Vidal. In this site, the existence during the first phase of the complex of an ergastulum in area 1, or pars rustica,3 where it has been possible to excavate in an open area, stands out. In the pars urbana and, to a lesser extent, in the pars fructuaria, targeted interventions have been carried out (fig. 2).

3It was a slave-owning villa of about 9000 m2 built at the end of the 1st century BC (phase 1). In the second half of the 1st century AD, the villa underwent major restructuring with a total remodelling of the pars rustica (phase 2). Firstly, rooms A3, 4, 6, 7, 8 and 9 were rebuilt with opus mixtum, with double courses of bricks between the stone blocks, at the same time that the ergastulum was abandoned and new baths were built in A6 and A7 (phase 2a). Then, the praefurnium of these baths in A4 was no longer in use and the walls of diverse rooms in this area, such as A3, were rebuilt with a different type of soil (phase 2b). The complex was abandoned with evident signs of destruction at the end of the first or beginning of the 2nd century AD (phase 3).

Fig. 2. Ground plan of the Villa of Rufio, result of the different building interventions (figure by the authors).

Fig. 2. Ground plan of the Villa of Rufio, result of the different building interventions (figure by the authors).
  • 4 Grau – Molina 2008, 2009, 2010, 2011.
  • 5 Llidó – Molina 2012; Molina et al. 2017; Molina – López 2021.

4The set of data obtained on the Villa of Rufio after a long period of archaeological interventions4 and research5 have made it possible to document the construction of earthen walls in different parts of the complex and to propose their existence in other spaces where these remains have not been preserved.

  • 6 Molina – López 2021.
  • 7 Stiros 1996; Rodríguez-Pascua et al. 2011; Arrighetti 2015; among others.

5This earthen architecture, completely integrated into the complex itself, must be understood as part of its construction sequence, its phases, and transformations,6 but it must also be related to other factors of a geographical nature. As is well known, the region of Umbria, in the centre of the Italian Peninsula, is an area of outstanding seismic activity, which is reflected in certain characteristics of the architecture of this territory. The seismic phenomenon was also present in ancient times and left various sorts of traces in the archaeological record.7 A varied set of elements identified in the structures of the Villa of Rufio and aspects of its architectural characteristics allow us to consider the influence and impact of the seismic phenomenon on them.

II. The earthen structures of the Villa of Rufio

1. The archaeological evidence

  • 8 Unfortunately, due to excavation constraints geoarchaeological analyses to provide a deeper insigh (...)

6Remains of earthen walls have been found in both the pars rustica and the pars urbana. They can be associated with different construction phases, a large part of them corresponding with the first ground plan from the Augustan period. The structures identified as earthen walls are around 60 cm thick and would have a variable composition,8 with an earthen material of different colours, with stones and reused ceramic remains and were built with or without stone plinths.

7The remains of earthen walls identified in different rooms (“ambientes”, A) (fig. 2) in area 2 or pars urbana (A24, 25 and 26, as well as A58) undoubtedly correspond to the first construction period, when most of the distribution and design of the spaces throughout the villa took shape.

  • 9 Molina – López 2021.

8The preserved remains of A24, 25 and 26 (fig. 3A), which would lead to a porticoed corridor next to the hortus, show the existence of earthen walls of which only the lower part is preserved (Stratigraphic Units, SU 695, 553, 556), as well as the associated coatings preserved in situ, stuccoes with evidence of having been painted, with remains of red colour. These coatings directly delimit the space that would have been occupied by the walls, built with earth, stones in their composition and without a stone plinth. These rooms of the pars urbana were paved with mosaics with geometric motifs dated in Augustan times.9

9There is also evidence of earthen elevations in A58, a space that was also paved with a mosaic of geometric motifs and which is interpreted as the triclinium. In the excavated section of the back wall of the room, a stone plinth has been preserved, the base of which shows remains of what would have been its covering. Leaning against the plinth and on top of the mosaic pavement, a thick layer of collapsed building material was found, with abundant fragments of stucco and some ceramic architectural remains (tiles/bricks) and, on top of these, a thick layer of homogeneous soil (SU 1064) (fig. 3B) with orange and white aggregates. Based on the set of characteristics shown in the room, it can be considered that this layer corresponds to the collapse of the earthen elevation. The excavated remains of the two lateral walls show that these elevations would also have been made of earth, with some stones inside them. The side wall to the west would appear to have had stones at its base, which would have contributed to the insulation of the upper earth construction, with some squared blocks observed on its outermost face. The other side wall (SU 1059) has little stone remains (fig. 3C) and would probably not have had a plinth, as well as the earthen walls of rooms A24, A25 and A26 located a few metres from it.

Fig. 3. Remains of earthen walls from the pars urbana corresponding to the first construction phase of the villa (Augustan period). A. Rooms A25 and 26, viewed from the northwest. B-C. Room A58 (figure by the authors).

Fig. 3. Remains of earthen walls from the pars urbana corresponding to the first construction phase of the villa (Augustan period). A. Rooms A25 and 26, viewed from the northwest. B-C. Room A58 (figure by the authors).
  • 10 Molina et al. 2017.
  • 11 Adam 1984.

10The rest of the available archaeological documentation of earthen walls in this enclave is concentrated in the ergastulum (area 1 or pars rustica) of the first phase, in the Augustan period.10 Here, except for the large retaining wall of the terrace built with stone blocks up to a height of 2.5 m, which has been preserved in a very remarkable way, we can consider that all the elevations were built with earth, in most cases with lower parts or plinths of masonry, in “petit appareil”,11 and in others of opus mixtum.

11In almost all the walls in area 1, only a base of small stone blocks was found, as a plinth, with a straight upper surface. Other sections of walls between thresholds had no plinths, but a low base of small stones. However, certain elements built here using stone and opus mixtum, such as two of the side walls of the peristyle (A9), or an opus mixtum pillar in A7, show a considerably greater height (fig. 4). These indications suggest that the upper part of the elevations of these rooms in the pars rustica, as could also have been the case in the pars urbana, were built with earth from the first phase. The flat surface where the stone base ends would have served to continue the elevation with earth and insulate it from the rising humidity of the ground.

Fig. 4. Different heights preserved in the stone and opus mixtum structures in area 1 or pars rustica. In the foreground, rooms 6, 7 and 10, viewed from the northeast (figure by the authors).

Fig. 4. Different heights preserved in the stone and opus mixtum structures in area 1 or pars rustica. In the foreground, rooms 6, 7 and 10, viewed from the northeast (figure by the authors).

12In certain rooms in this area (A3, 4, 6, 7, 8 and 9), part of the walls are made of opus mixtum, built during a second construction phase (second half of the 1st century AD), related to the abandonment of the ergastulum and the construction of new baths in rooms 6 and 7, with the praefurnium being installed in A4 (phase 2a). However, the amortisation of this praefurnium and the reconstruction of the walls of A3, built up with earth (SU 339, 340, 341) on lower parts of opus mixtum and masonry, suggest a second stage in this remodelling (phase 2b). These earthen walls that enclose A3 are partly composed by a dark brown earth (fig. 8A and D), as well as those that separated A6 and A7 (SU 47), and A7 from A8 (SU 48) (fig. 5A) and the ones detected in A8b (fig. 5D).

13On the other hand, in the aforementioned opus mixtum pillar (fig. 4) located in A7, one of its sides, unfinished, shows that this fabric actually lines a core of cobbles, larger stones and reused ceramic construction remains, arranged in a mortar matrix. This side of the pillar would possibly correspond to a continuation of the earthen elevation, and the pillar would close and protect this earthen wall from its side. Similarly, the stone structure surrounding the peristyle (A9), with parts of opus mixtum associated with the second phase remodelling at the top, is in fact the outer lining of an earthen body, which could be all or part of the pre-existing one erected in the first phase.

14In addition, on the north side of A8, the peristyle corridor A8b, different sections of earthen walls have been identified, including those of A51-71 and 53-70, between their respective thresholds, visible through large slabs. These walls have both light-coloured earth, stones, and white aggregates, as well as dark brown sediment and ceramic building material inside (fig. 5B). Among these sections of earthen walls is SU 486 (fig. 5C), in which a series of horizontal lines appear to be distinguished in the body of light brown earth, which could be due to the alignment of the soil particles as a result of the building technique employed. As observed during the excavations in this area, only one of these earthen wall sections has a solid base of squared stone blocks, while the one next to the latter has only a few small stones arranged in its lower part.

Fig. 5. Remains of earth elevations in area 1 or pars rustica. A. Separating rooms A7 and A8. B-D. Along the corridor 8b (figure by the authors).

Fig. 5. Remains of earth elevations in area 1 or pars rustica. A. Separating rooms A7 and A8. B-D. Along the corridor 8b (figure by the authors).

2. Interpreting the construction technique employed

  • 12 Algorri – Vázquez 1996, p. 22.
  • 13 Guerrero 2011, p. 9.
  • 14 Guerrero – Boto de Matos 2013.
  • 15 Roux – Cammas 2007; Aurenche et al. 2011; Knoll et al. 2019, p. 19.
  • 16 Minke 2001, p. 61-62; Chazelles 1999, p. 232; 2016, p. 9.

15The walls presented are massive earthen elevations, with other materials inside, both natural and anthropic. The earth would apparently have been arranged with a low degree of humidity (macroscopically observing no layers, lumps or differentiated mud units), without the addition of plant stabilisers and an approximate thickness of 60 cm. Taking these characteristics into account, it is possible that they were built using a formwork. Rammed earth (also known, commonly in Italy, as “pisé”, from the French term “pisé de terre”), is the best-known earth construction technique that uses these large moulds (usually of wood) to build walls by compacting the earth inside them with the use of a tool (a rammer). For rammed earth walls, soil with a low degree of humidity is used. It has been established that rammed earth walls can be distinguished based on features such as the presence of horizontal and/or vertical formwork imprints (not to be confused with cracks or interfaces between successive layers or heights built at the same time in walls of kneaded or stacked earth), as well as putlogs (the holes of the needles that pass through the formwork). It has also been proposed that it would be the evidence of successive ramming units what would certainly imply the construction with rammed earth.12 However, contrary to what is usually considered, rammed earth is not the only earthen construction technique in which a formwork can be used. Kneaded mud can be placed inside formworks to build the walls by means of the so-called in Spanish “barro colado”,13 “barro vertido”,14 and “bauge coffrée” in French.15 Also, the needle holes in the formwork are not always an indicator, as it is possible to build with rammed earth using formworks that do not use them.16

16It would be the assessment of the above mentioned features that would allow the identification of rammed earth construction, but as is well known, the identification in archaeological contexts of this form of earthen building is difficult. In these cases, recognising the marks of formwork units or the presence of needle holes in the remains of earthen elevations is often not possible and certainly not simple, as there is not enough height of the elevation preserved or simply no trace of the standing structures, being these collapsed and in the form of strata.

  • 17 Mileto – Vegas López-Manzanares – Cristini 2012.
  • 18 Font – Hidalgo 1990, p. 33.
  • 19 Castellarnau – Rivas 2013, p. 104.
  • 20 Chazelles 1990; 2003, fig. 5.

17In relation to these indicative features of rammed earth, in the case of the Villa of Rufio we have already mentioned the homogeneity of the earth and the fact that it appears to have been employed with a low degree of humidity. On the surface of one of the excavated earthen elevations of A8, a series of horizontal lines can be distinguished which could correspond to the alignment of the soil particles due to ramming (fig. 5C), although these extents need to be further investigated, mainly by means of micromorphological analysis. On the other hand, the presence of solid elements, such as stones, could be seen as inappropriate in the construction of rammed earth walls. However, variants of the rammed earth technique such as the “Valencian tapia” use pebbles as stabilising elements,17 while other variants are made by directly incorporating courses of blocks.18 Ethnographic works record the importance of the earth used to build with this technique containing stones and pebbles.19 In fact, part of the rammed earth elevations of Empúries (Girona, Spain), the best-known and maybe the first rammed earth walls in the Iberian Peninsula, on stone plinths and with a thickness of 50 cm, have stones of considerable size inside.20

  • 21 Rodríguez et al. 2021, p. 163-166.

18On the other hand, as opposed to the most widespread image of rammed earth walls as being formed by regular and horizontal compacted layers, it must be considered that these elevations can also be built from successive contributions whose termination and separation is irregular and even curved or inclined, as it can be seen in some of the earthen walls documented in the Villa of Rufio. These earthen wall remains are in A8b, as well as at its junction with A3 (fig. 5B and 5D), suggesting that they were reconstructed during the second phase, mainly with the darker earth (phase 2b). The inclination between the different earthen parts probably corresponds to that reconstruction. However, it has been pointed out that the construction of rammed earth walls with an inclined upper surface, like trapezoidal blocks, would facilitate the union between the different parts of the wall, reinforcing its resistance21.

3. The use of rammed earth in the Roman world

  • 22 Pesando 2005, p. 87; 2011, p. 91; 2013, p. 123; Pesando – Guidobaldi 2006; Battaglini – Diosono 20 (...)
  • 23 Algorri – Vázquez 1996, p. 22; Perello 2015; Pastor 2017.

19The technique of rammed earth, for which the term opus formaceum22 has recently been coined based on the ancient descriptions, is very little present in the numerous works dealing with Roman architecture. As is the case with many other historical periods, in the bibliography on ancient Rome the term “rammed earth” is often mentioned in a vague way and often together with another term, “adobe”, referring to the existence of earthen constructions or their remains, but without considering the existing differences between these different ways of building. In more than a few cases, the presence of walls of this material is cited, but determining and specifying the construction technique is more challenging, not only because of the misuse of terminology that occurs,23 but also because of the difficulties already mentioned in identifying the rammed earth technique in archaeological contexts.

  • 24 Varro, Rust., 1, 14, 4.
  • 25 Plin., Nat., 35, 48.
  • 26 Pallad., 1, 34, 4.
  • 27 Desbat 1993, p. 147; Chazelles 2003, p. 8; 2016, p. 13.
  • 28 Isid., Orig., 15, 9, 5.

20Rammed earth construction in the Roman world is mentioned by Varro,24 Pliny the Elder25 and Palladius,26 who indicate its use in North Africa and the Iberian Peninsula, as well as in the south of the Italian Peninsula during the 1st century BC, although, as has been pointed out,27 they seem to mean that it would be a technique more typical of other territories than of Italy. Rammed earth construction in the Roman world is also mentioned by Isidore of Seville.28

  • 29 Desbat 1993, p. 153; Chazelles 1999, p. 243; 2003, p. 8; 2016, p. 14-16; Russell – Fentress 2016, (...)
  • 30 Chazelles 2016, p. 18, 22, fig. 11.
  • 31 Chazelles 1990, 2016.

21Regarding its archaeological documentation in the territories mentioned, in present day Tunisia constructions are known that were probably built with rammed earth at various sites during the last centuries of the 1st millennium BC, such as Kerkouane, Carthage, Utica or Thysdrus.29 A particularly clear case of the use of rammed earth would have been identified at Rirha (Morocco).30 In the Iberian Peninsula, the most relevant example is that of the Augustan dwellings at Empúries (Girona).31 Rammed earth construction is noted in many other buildings in Roman Hispania and from the Republican period onwards, pointed on many occasions from walls of which only stone plinths with regular upper parts have been identified.

  • 32 Battaglini – Diosono 2010, p. 229; Russell – Fentress 2016, p. 138; Ceccarelli 2018, p. 33; 2019, (...)
  • 33 Regoli 1985, p. 64-65.
  • 34 Fentress et al. 2003, p. 21.
  • 35 Bataglini – Diosono 2010, p. 227; Pesando 2011, p. 91, fig. 13; Mogetta 2021, p. 192.
  • 36 Carfora – Ferrante – Quilici Gigli 2013.
  • 37 Ferrante 2014; Quilici Gigli 2014.
  • 38 Pesando 2005, p. 87; 2011, p. 92, fig. 20; 2013, p. 123, fig. 9.
  • 39 Staffa 1996, p. 111, 113, fig. 60.

22As for the Italian Peninsula, the use of rammed earth in the Roman period has been noted in various settlements32. We can highlight the case of Settefinestre (Tuscany), where building typologies were systematically classified, noting that most of the internal walls and annexes would have had a stone plinth and the rest of the elevation would have been made of “terra pressata”.33 Rammed earth walls would have been documented at Cosa (Tuscany), from the Republican period, from the first half of the 2nd century BC to the 1st century BC, with or without stone plinths and plastered;34 at Fregellae (Lazio), in the 3rd century BC, on a stone foundation and tiled plinths and then plastered;35 at Norba (Lazio), between the 3rd and 1st centuries BC, on internal walls and up to 40 cm thick,36 particularly well preserved in Domus VI,37 or in Pompeii (Campania), in the Protocase of the Granduca Michele and the Centauro, from the mid-3rd century BC.38 It would have been also employed in the settlement of Penne-viale Ringa (Pescara, Abruzzo), late 2nd-1st century BC.39

III. Architecture and seismic evidence in the Villa of Rufio

1. The territory of the Villa of Rufio: a seismic zone today and in Antiquity

  • 40 Guidoboni – Comastri – Traina 1994.
  • 41 E.g., Stiros – Jones 1996.
  • 42 Rovida et al. 2021.

23It is well known that the Italian Peninsula is an area of high seismic activity. Logically, these episodes also occurred frequently in Antiquity,40 as shown by archaeological evidence in different parts of the Mediterranean41. The region of Umbria is among the most affected by earthquakes, with several episodes of great magnitude occurring in short intervals of time and even in the same year, as evidenced by the historical record.42

  • 43 Where 1 is the value for the highest risk and 4 is the lowest value (https://rischi.protezionecivi (...)
  • 44 Rovida et al. 2021.

24In particular, the Villa of Rufio is in Giano dell’Umbria, an area of moderate risk according to the latest Italian seismic classification43 and close to active and inactive seismogenic faults, such as the Martana fault (fig. 6). If we focus on the last 130 years, the best-known period, 43 earthquakes of at least intensity 3 have been recorded in this municipality during this time, 12 of them with a magnitude greater than 5 (an average of 1 every 11 years).44

  • 45 Camassi 2004; Galli – Molin 2013; Conte 2019.
  • 46 Obseq., 46 and 61.
  • 47 Gell., 4, 6, 2.

25The information transmitted by classical authors on seismic movements in Antiquity is very scarce and mainly focused on earthquakes experienced in the city of Rome.45 In this sense, in the Umbria region very few of these events are cited in ancient sources. Particularly noteworthy is the earthquake of 99 BC in Nursia (Norcia), which caused the destruction of a sacred building, and the one that occurred in 63 BC in Spoletium (Spoleto). Both episodes are mentioned in the work of Obsequens,46 which records the “prodigies” between 249 and 11 BC and can be considered the first compilation of earth movements in the Italian Peninsula. Through Aulus Gellius we know that the earthquake of 99 BC was noted in Rome: “the spears of Mars had moved in the sanctuary in the Regia”.47

  • 48 Cic., Catil. 3, 8.
  • 49 Obseq., 59.

26Similarly, Cicero’s vague mention of the existence of earthquakes in 63 BC, the year of his consulship, can be related to the Spoletium earthquake: “to pass over the falling of thunderbolts and the earthquakes, to say nothing of all the other portents which have taken place in such number during my consulship”.48 Obsequens49 work also mentions another episode in 76 BC in Reate (Rieti), situated in the north of Latium and relatively close to the Villa of Rufio.

Fig. 6. Supra: map of central Italy overlaying the main seismogenic zones (seismic data extracted from: http://diss.rm.ingv.it/​diss/​ (figure by the authors). Infra: map of geological and seismic activity in the southern part of the Tiber basin (Bonini et al. 2003, fig. 1).

Fig. 6. Supra: map of central Italy overlaying the main seismogenic zones (seismic data extracted from: http://diss.rm.ingv.it/​diss/​ (figure by the authors). Infra: map of geological and seismic activity in the southern part of the Tiber basin (Bonini et al. 2003, fig. 1).
  • 50 Ciotti 1978; Petraccia Lucernoni 1996; AE 1996, 601.

27Sometimes epigraphy also serves as a source of information on the existence of seismic movements. This is the case of an epigraph found in the castle of Montoro (Terni), which in Antiquity belonged to the territorium of Carsulae, a city located on the Via Flaminia, less than 30 km from the Villa of Rufio. This inscription, dated to the first half of the 3rd century AD, refers to the restitution of a mithraeum50Spe[l]aeum vi motu terrae diruptum ex suo omni i[m]pen / sa re[f]ecit”.

  • 51 Morigi 1997; Bonini et al. 2003.
  • 52 Bottari – Sepe 2013, Bottari et al. 2017.

28Archaeological evidence has also been found that points to the presence of earthquakes in this area. However, as in the case of the Villa of Rufio, the linking of certain evidence from the archaeological record with the presence of seismic events is not without its difficulties. In the city of Carsulae, episodes of great destruction have been recorded which could be related to different earthquakes, including the one referred to in the Montoro epigraph. The existence of a seismic event at the beginning of the 1st century AD linked to the collapse and subsequent reconstruction of a section of the Via Flaminia has also been suggested, and a strong earthquake has been argued as the probable cause of the abandonment of this settlement in the 4th/5th century AD.51 However, the direct relationship between the major alterations observed in Carsulae and earthquakes has been addressed by other authors, who question the direct relationship between an earthquake and the abandonment of this city.52

  • 53 The Historia Augusta (Gord., 26, 1-2) mentions a major earthquake around that date but does not sp (...)
  • 54 Ponzi 1983; Marzano 2007.

29Similarly, in Orticello, a suburban villa of the ancient Hispellum (Spello), evidence has also been found of a strong destructive episode that led to its abandonment and which has been hypothetically linked to the earthquake of AD 24253 and the aforementioned event around 346 AD. The Roman villa of Peruggia Vecchia (Deruta) also has remains that point to its destruction by a seismic event in the 1st century AD.54

  • 55 Diosono 2009.

30A well-documented example is the villa of San Silvestro (Cascia) in the ager nursinus, where there is evidence of an earthquake that has been linked to the Nursia earthquake of 99 BC, as well as further destruction by another seismic movement somewhere between AD 100 and 300.55

2. Features in the architectural remains of the villa with possible origins in seismic action

31The specialised bibliography on the archaeological detection of earthquake effects in Antiquity points out some indicators manifested in a pattern or a whole series of affected and/or deformed elements, beyond isolated cracks or displaced features. In the Villa of Rufio, a series of elements have been identified, ranging from damage in structures to several falls and displacements of architectural parts (fig. 7), which may have been caused by several seismic episodes separated by a short time interval. These elements have been detected along the whole area 1 or pars rustica where a better stratigraphic depth was preserved, in A44, 43, 6, 7, 4, 3, 59, 8 and 9, together with other possible indicators detected in different parts of area 2 or pars urbana, as well as in the pars fructuaria.

Fig. 7. Plan of the villa with indication of the main signs of destruction that could be related to seismic activity (figure by the authors).

Fig. 7. Plan of the villa with indication of the main signs of destruction that could be related to seismic activity (figure by the authors).
  • 56 Stiros 1996, p. 151, fig. 13.

32On the one hand, in area 1, a crack that runs through several walls and rooms is noteworthy (fig. 8A): firstly, observed in the back of the brick arch that separates A44 and A43 (fig. 7, no. 1; fig. 8B). In arches built with stone or brick joined by mortar, it is common for a seismic event to result in a crack.56 This crack is again visible in the retaining wall at A6, as well as in the opus mixtum wall of this same room (fig. 8C), and in the lower part of masonry reused and reintegrated in the wall built between A4 and A3 during the remodelling of this area and the construction of the new baths. In this same separation between A4 and A3, next to where this crack can be observed, a series of dark earth layers can be distinguished in an inclined position (fig. 7, no. 9; fig. 8D), which could also be related to displacements associated with these phenomena and seem to have been integrated into this wall during its remodelling in phase 2b, following a destruction episode materialised in A6.

Fig. 8. A. View of the rooms in area 1 showing a continuous crack. B. Crack in the arch. C. A6, with cracks in the walls, breakage, and undulation of the interior surface of the hypocaust. D. Crack running through the blocks and inclined strata in a rebuilt wall in A3 (figure by the authors).

Fig. 8. A. View of the rooms in area 1 showing a continuous crack. B. Crack in the arch. C. A6, with cracks in the walls, breakage, and undulation of the interior surface of the hypocaust. D. Crack running through the blocks and inclined strata in a rebuilt wall in A3 (figure by the authors).
  • 57 E.g., Stiros 1996, p. 150, fig. 5; Spence 1998; Pecchioli – Cangi – Marra 2018.
  • 58 Rodríguez-Pascua et al. 2011, p. 22.

33The first and third sections of the crack have an inclined orientation from right to left, with the second and fourth sections slightly inclined in the opposite direction (see fig. 8A). In this sense, in walls composed of blocks that are affected by an earthquake, such as masonry or bricks, like many of those built during Roman Antiquity, the cracks formed are diagonal.57 Moreover, cracks caused by earthquakes do not only run along the block joints, but also through part of the blocks themselves,58 which is observed in different points in the villa (fig. 8D).

  • 59 Stiros 1996, p. 135, fig. 6a.

34Following the axis of the crack, in A6 we can point out that during its excavation a large fallen and broken construction remain was found (fig. 7, no. 3; fig. 9A), corresponding to the displaced stone threshold of A43. It is also worth noting the presence of an inclined stone base (fig. 7, no. 10; fig. 9B) a few meters away, in A4 and behind the praefurnium, a form that vertical elements can adopt when affected by earthquakes, oscillating with respect to the vertical axis.59 On the other hand, on the inner surface of the hypocaust of the baths in A6, a large breakage was detected (fig. 7, no. 4; fig. 9C) which can be associated with the fall of an element from the upper part of the terrace, as documented as well by the fallen threshold of A43 already mentioned, in the same area of the room. This brick inner surface was also found showing a considerable undulation (see fig. 8C), as well as other paved spaces in the villa (fig. 7; nos. 19-21). Likewise, the threshold that separates A8 from A54 is fractured in its central part (fig. 7, no. 12; fig. 9D), as is the thick slab just next to it, on the other side of the earthen wall SU 339 (see fig. 5B). These breaks may have been due to movements of the ground as there is no evidence that, in this case, they could have been caused by an element fallen on them.

Fig. 9. A. Large threshold for the entrance to the ergastulum found fragmented and displaced several metres from its original position (A6). B. Base of stone blocks in an inclined position in A4. C. Breakage of the inner surface of the hypocaust (A6). D. Fractured threshold, between A8 and A54 (figure by the authors).

Fig. 9. A. Large threshold for the entrance to the ergastulum found fragmented and displaced several metres from its original position (A6). B. Base of stone blocks in an inclined position in A4. C. Breakage of the inner surface of the hypocaust (A6). D. Fractured threshold, between A8 and A54 (figure by the authors).

35In addition, inside A59 a completely collapsed masonry structure was documented (fig. 7, no. 11; fig. 10), that would have been built to wall off an opening between two sections of earthen walls aligned with the earthen walls SU 339 or 486 in A8b or the peristyle corridor.

Fig. 10. Remains found during the excavation of A59, showing a collapsed masonry elevation. Top view and detail of the fracture surface (figure by the authors).

Fig. 10. Remains found during the excavation of A59, showing a collapsed masonry elevation. Top view and detail of the fracture surface (figure by the authors).
  • 60 Stiros 1996, p. 149, 152; Rodríguez-Pascua et al. 2011, p. 22; Galli – Molin 2013; Arrighetti 2015 (...)
  • 61 Giuliani 2011, p. 48-49.

36In A9, the peristyle, different brick columns were found fallen inside, towards the east, and with two different orientations (fig. 7, no. 16). It is also worth noting the rotated position in which a quadrangular base of a fallen column was found also inside the peristyle (fig. 7, no. 14; fig. 11B). The characteristics of this find seem to respond to a sudden fall of the element, turned and displaced inside the peristyle on its base. During the excavation of A6 already discussed, a rectangular slab was also found in this rotated position (fig. 7, no. 7). The existence of several columns fallen in parallel, and the rotation displacements of the columns are related to the effects of earthquakes.60 During the excavations in the peristyle, it was also documented that two of the slabs, on both sides of the peristyle, were displaced and fallen, one of them inside this courtyard (fig. 7, no. 17) and the other in corridor A8b (fig. 7, no. 18; fig. 11D), displacements that also seem to respond to sudden movements. Additionally, in the stone base of the wall that closes the peristyle on the south-southwest side, there is a deformation in the arrangement of the blocks, and cracks that run through them (fig. 7, no. 13; fig. 11C) that could have been caused by a fold (anticline). The fact that the peristyle is not surrounded by free-standing columns but enclosed by a solid wall at its base with the columns above it, could be considered an anti-seismic architectural element, as is the documented closure of free-standing column peristyles by joining the lower part of the columns by a continuous wall after an earthquake.61

Fig. 11. A. Some of the fallen columns inside the peristyle (A9) at the time of its excavation in 2004. B. Column remains fallen in its base with the quadrangular slab in a rotated position, in A9. C. Deformation observed in one of the walls enclosing the peristyle. D. Peristyle slab found displaced towards A8b (figure by the authors).

Fig. 11. A. Some of the fallen columns inside the peristyle (A9) at the time of its excavation in 2004. B. Column remains fallen in its base with the quadrangular slab in a rotated position, in A9. C. Deformation observed in one of the walls enclosing the peristyle. D. Peristyle slab found displaced towards A8b (figure by the authors).

37On the other hand, it is worth noting the irregularity that can be seen in the profile of several of the dividing walls that start from the terracing wall towards the peristyle or courtyard of the ergastulum, on both sides of A10, or even in the channel itself inside the peristyle (fig. 12A). This can also be observed, especially in the walls of saggio XVI-XVII (fig. 7, no. 23; fig. 12B), or in area 2 or pars urbana, in A27-35 (fig. 7, no. 22; fig. 12C), where the zigzagging of the trajectory of the structures is evident. In this area 2, it is also worth noting the appearance of isolated signs of fire in the villa, in A81, where a pavement covered with a regular layer of ash and charcoal was found.

Fig. 12. Undulations in the profile of different structures. A. Channel inside A9. B. Saggio XVI-XVII. C. A27-35 (figure by the authors).

Fig. 12. Undulations in the profile of different structures. A. Channel inside A9. B. Saggio XVI-XVII. C. A27-35 (figure by the authors).
  • 62 Rodríguez-Pascua et al. 2011; Lazar et al. 2020.

38We have already noted the presence of undulations in different areas of the villa, paved with brick, such as the one discussed for A6, to which we should add the spicatum pavements of A5 and A78-79 (fig. 7, nos. 20-21), where rather than undulations in these architectural elements, we can observe a real folding, forming an obtuse angle in the pavement. In this sense, the deformations seen in canalisations such as that of A39, in area 1 or pars rustica (fig. 7, no. 24; fig. 13A), or in saggio XX (fig. 7, no. 19; fig. 13B-C), in the pars fructuaria, which can be associated with movements and folding of the ground, are also notable. Folding in pavements in archaeological contexts is also considered an effect of seismic events, as are undulations in walls and other structures.62

Fig. 13. Undulations in canalisations. A. A39. B-C. Saggio XX (figure by the authors).

Fig. 13. Undulations in canalisations. A. A39. B-C. Saggio XX (figure by the authors).

3. Earth walls and other architectural solutions in relation to seismic activity

  • 63 Galdieri 1986; Correia – Dipasquale – Mecca 2011.
  • 64 Minke 2001; Guerrero – Boto de Matos 2013, p. 8.

39The seismic factor is a reality with a long history in the centre of the Italian Peninsula. Its incidence is reflected not only in the destruction of buildings, but also in preventive architectural strategies in relation to seismic activity and the dangers it poses to the population. Thus, the Umbria region is one of the areas in Italy where earth has historically been used for building, with rammed earth (“pisé”, “terra pressata”) standing out,63 a technique valued for the seismic-resistant characteristics that can have,64 since massive earth walls would be more stable and flexible than those made up of individual pieces, such as adobes, fired bricks or stone blocks.

  • 65 Stiros 1996, p. 150.
  • 66 Spence 1998, p. 31, fig. 2.5; Dipasquale – Mecca 2015, p. 70, fig. 5.
  • 67 Jorquera 2014, p. 27.
  • 68 DeLaine 1992, p. 192; Stiros 1995, p. 731; 1996, fig. 10; Amici 2011; Giuliani 2011.

40To reinforce masonry walls against seismic events, wooden elements, rows of bricks or metal reinforcements are usually introduced.65 Iron elements with different morphologies are commonly added in the stone facades of Umbrian architecture to increase the bond between elements and reinforce them against earthquakes,66 antiseismic elements that are also present in traditional architecture in other regions of the world.67 The use of metal as architectural reinforcement was already employed in classical Antiquity and the Roman world.68 In the Villa of Rufio, the addition of nails (fig. 14) in the second phase of the peristyle wall (A9) is notorious, added to fix the mortar layer that served as a base for the exterior render to the stone blocks and bricks of the opus mixtum.

Fig. 14. Construction reinforcements in A9, with the introduction of rows of bricks between the stone blocks, as well as metal nails (figure by the authors).

Fig. 14. Construction reinforcements in A9, with the introduction of rows of bricks between the stone blocks, as well as metal nails (figure by the authors).
  • 69 Dipasquale – Mecca 2015, p. 69, fig. 2.

41The antiseismic recommendation of introducing brick courses in walls made up of blocks, fully coincides with the opus mixtum technique used in this area of the villa from the second construction phase onwards, with horizontal double courses of bricks interspersed between the stone blocks. These brick courses, commonly used in Roman times and also present in traditional Italian architecture, not only reinforce the walls composed of single blocks, but also prevent the extension of possible cracks along them.69

IV. Conclusions

42The difficulties that can arise in the identification of earthen walls in archaeological contexts are known, as well as the little attention they often receive from the discipline, especially regarding the Roman world. The excavation and research at the Villa of Rufio has made it possible to detect the construction of earthen walls in different parts of its ground plan and to suggest their presence in others, with the result that soil would have been used in most of the elevations built in this complex. These solid earth walls have characteristics that allow us to interpret the use of the rammed earth technique, a form of construction in which sediment is generally used with little water and without the addition of vegetable matter, rammed inside formworks. This construction system is not exceptional in the Roman world or in the Italian Peninsula, despite its scarce representation in traditional archaeological research and up to the present day, in which these architectural forms have hardly been taken into account.

  • 70 Stiros 1996, p. 145.

43On the other hand, the destruction signs and alterations identified in the structures of the villa can be related to seismic phenomena. Given evidence of earthquakes in a settlement, not all of them should necessarily be assigned to a single event but may more likely be the result of several seismic factors together with those of other types.70 In these pages we have pointed out the incorporation of construction techniques and practices in the Villa of Rufio that are not only adequate, but possibly also adapted to the seismic characteristics of the territory in which it was built.

44Solid earth walls, such as those made of rammed earth, can be an essential element of these seismic-resistant construction solutions, especially in comparison with other techniques and when they are reinforced, and these are present from the first phase of construction. In addition, in the first remodelling in the villa (phase 2), new architectural resources were adopted in this sense, such as the reinforcement of the masonry walls, especially employed in the face of seismic movements. This is the case of the opus mixtum structures, with a double row of bricks between the stone blocks, as well as the metal reinforcements in this type of masonry documented in the peristyle and probably corresponding to this second phase. It cannot be ruled out that phase 2, during which these antiseismic elements were adopted, was a response to a seismic event, with a hypothetical first destruction, to which the arch crack could be associated. However, most of the architectural damage addressed in this paper must have occurred after this second construction phase or remodelling of opus mixtum, which took place at some point in the second half of the 1st century AD, considering, among other features, the crack that runs through the structures of different rooms built at that time and with that technique.

  • 71 Tomassini 2021.
  • 72 Dessales 2019; Dessales et al. 2022. For more information, see: http://recap.huma-num.fr/webpublic (...)

45Shortly afterwards, the rooms in which the second phase of opus mixtum construction took place underwent remodelling and repair (phase 2b). It is during this remodelling when the oven was amortised and probably when a whole series of walls in area 1 were rebuilt using dark earth, not only those of A3 (fig. 8A, C, D). The reason for such a short period between renovations could be due to destructions caused by seismic movements that led to the repair of the walls that would have been previously built with opus mixtum in A6. This hypothesis is supported by the characteristics of destructive signs that would have motivated this phase of reconstruction (2b), such as the breakage of the interior surface of the hypocaust of A6 due to the fall of an element and the characteristics observed in the wall of A3, with the apparent integration of inclined strata and the crack that runs through the stone blocks at its base. It is in this phase of destruction in this area of the complex that the accumulation of construction material that would have been removed, stored, and probably prepared for reuse, such as the marble slabs found in this way in A14, would have taken place. In areas of traditional seismic risk, repairs and renovations detected in Roman buildings are often understood as the result of possible seismic phenomena, as it is the case of Ostia71 and especially Pompeii, with the project “Reconstruire après un séisme. Expériences antiques et innovations à Pompéi” (RECAP).72

46In any case, most of the architectural damage and fragmented, displaced or fallen elements that have been detected in the villa during its excavation can be attributed to a seismic event subsequent to this phase 2b. This event would mean the abandonment of this area of the complex, as these architectural parts and elements were not restored or repaired again, such as the displaced slabs and fallen columns of the peristyle, the fallen threshold on A6, as well as probably most of the folding and deformations in structures such as the pavements and canalisations, which could hardly continue to be used after them (table 1).

Table 1. Summary of the building phases of the Villa of Rufio and the possible seismic episodes (figure by the authors).

Phases Chronology Main building activity Possible seismic activity at the end of phase Earthquake indicators
1 From the end of the 1st century BC Construction of the villa: pars urbana, pars fructuaria, ergastulum in pars rustica Earthquake? (middle of the 1st century AD) – Arch crack
– Implementation of new techniques and elements: opus mixtum, metal reinforcements
2a 2nd half of the 1st century AD Remodelling in pars rustica with new baths (A4, 6, 7) and opus mixtum Earthquake (2nd half of the 1st century AD) – Continuous crack in opus mixtum structures
– Damage in hypocaust (from threshold fallen from above)
2b 2nd half of the 1st century AD – Repair of opus mixtum walls (dark earth)
– Collection and accumulation of used building materials in certain spaces (such as marbles in A14)
Earthquake (end of the 1st-beginning of 2nd century AD) – Breakage and fall of elements (columns, slabs, threshold) – Foldings and deformations
3 End of the 1st-beginning of 2nd century AD – Abandonment of pars rustica and filling of rooms with architectural rubble from pars urbana – Ramps?

47Finally, the dumping and filling of rooms in this area 1 with construction material and levelling of the terrain would have taken place at a third time of occupation or use of the villa, which would not have gone beyond the end of the 1st century AD or the beginning of the 2nd century AD, given the scarcity of material remains from after the Flavian period found in the complex. Even what looks like ramps can be seen in the middle of collapsed constructions (A9), which would indicate a phase of abandonment.

48Thus, we can consider at least two episodes of destruction in the villa associated with seismic phenomena in the second half of the 1st century AD and separated by a short period of time: the first of these is reflected in the repair of the previously built opus mixtum structures (phase 2b), and the second is associated with most of the architectural damage and destruction that we have dealt with and would lead to the definitive abandonment of the villa. Nevertheless, the antiseismic nature of the architectural solutions adopted in the first transformation, from phase 1 to phase 2, can be understood in the knowledge of the seismic risk of this territory – we should remember the earthquakes in Nursia in 99 BC or Spoletum in 63 BC –, without ruling out the possibility that some episode could have affected the villa prior to the two seismic events mentioned above and caused the adoption of the opus mixtum technique and the metal reinforcements. All this allows us to consider the central role in the construction of this complex of an accumulated and transmitted building experience that reflects the knowledge of these events and their effects, with the aim of reducing the damage that can be caused by this factor and responds to it through architecture.

Haut de page

Bibliographie

Adam 1984 = J.-P. Adam, La construction romaine. Matériaux et techniques, Paris, 1984.

Algorri García – Vázquez Espí 1996 = E. Algorri García, M. Vázquez Espí, Enmienda a dos de los errores más comunes sobre el tapial, in A. de las Casas Gómez (ed.), Actas del primer congreso nacional de historia de la construcción (Madrid, 19-21 octubre 1996), Madrid, 1996, p. 19-23.

Amici 2011 = C. M. Amici, L’utilizzazione delle catene metalliche nei sistemi voltati di età romana, in C. Giardino (ed.), Archeometallurgia. Dalla conoscenza alla fruizione, Bari, 2011, p. 222-238.

Arrighetti 2015 = A. Arrighetti, L’archeosismologia in architettura. Per un manuale, Florence, 2015.

Aurenche et al. 2011 = O. Aurenche, A. Klein, C.-A. de Chazelles, H. Guillard, Essai de classification des modalités de mise en œuvre de la terre crue en parois verticales et de leur nomenclature, in C.-A. de Chazelles, A. Klein, N. Pousthomis (ed.), Échanges transdisciplinaires sur constructions en terre crue, III, Les cultures constructives de la terre crue. Actes de la table ronde de Toulouse (16-17 mai 2008), Montpellier, 2011, p. 13-34.

Battaglini – Diosono 2010 = G. Battaglini, F. Diosono, Le domus di Fregellae: case aristocratiche di ambito coloniale, in M. Bentz, C. Reusser (ed.), Etruskisch–italische und römisch–republikanische Häuser, 2010, Wiesbaden, p. 217-231.

Bonini et al. 2003 = M. Bonini, C. Tanini, G. Moratti, L. Piccardi, F. Sani, Geological and archaeological evidence of active faulting on the Martana Fault (Umbria-Marche Apennines, Italy) and its geodynamic implication, in Journal of quaternary science, 18, 2003, p. 693-708.

Boschi et al. 2007 = E. Boschi, E. Guidoboni, G. Ferrari, G. Valensise, I terremoti dell’Appennino Umbro-Marchigiano, Bologna, 2007.

Bottari – Sepe 2013 = C. Bottari, V. Sepe, The role of earthquakes, landslides and climate changes in the abandonment of the Roman Carsulae site (Tevere basin – Central Italy), in Quaternary international, 308-309, 2013, p. 105-111.

Bottari 2017 = C. Bottari, D. Aringoli, R. Carluccio, C. Castellano, F. D’Ajello Caracciolo, M. Gasperini, M. Materazzi, I. Nicolosi, G. Pambianchi, P. Pieruccini, V. Sepe, S. Urbini, F. Varazi, Geomorphological and geophysical investigations for the characterization of the Roman Carsulae site (Tiber basin, Central Italy), in Journal of applied geophysics, 143, 2017, p. 74-85.

Camassi 2004 = R. Camassi, Catalogues of historical earthquakes in Italy, in Annals of geophysics, 47/2-3, 2004, p. 645-657.

Carfora – Ferrante – Quilici Gigli 2013 = P. Carfora, S. Ferrante, S. Quilici Gigli, Tecniche costruttive in epoca medio-tardo repubblicana: il caso di Norba, in F.M. Cifarelli (ed.), Tecniche costruttive del tardo ellenismo nel Lazio e in Campania. Atti del Convegno Segni (3 dicembre 2011), Roma, 2013, p. 93-102.

Castellarnau Visús – Rivas 2013 = A. Castellarnau Visús, F.A. Rivas, Las dos caras de una misma técnica constructiva. Registro y trasmisión de la construcción en tapia, in En construcción con tierra. Patrimonio y vivienda. X CIATTI. Congreso de arquitectura de tierra en Cuenca de Campos, Valladolid, 2013, p. 97-108.

Ceccarelli 2018 = L. Ceccarelli, La terra cruda in Italia come materiale da costruzione in ambito architettonico e produttivo di epoca antica, in Terra magazine, Milan, 2018, p. 30-42.

Ceccarelli 2019 = L. Ceccarelli, Tecniche di costruzione in terra cruda di epoca etrusca e romana, in S. Sabbadini (ed.), Terra book, Milan, 2019, p. 34-41.

Chazelles 1990 = C.-A. de Chazelles, Les constructions en terre crue d’Empúries à l’époque romaine, in Cypsela, 8, 1990, p. 101-118.

Chazelles 1999 = C.-A. de Chazelles, À propos des murs en bauge de Lattes : problématique des murs en terre massive dans l’Antiquité, in M. Py (ed.), Recherches sur le quatrième siècle avant notre ère à Lattes, Lattes, 1999, p. 229-254.

Chazelles 2003 = C.-A. de Chazelles, Témoignages croisés sur les constructions antiques en terre crue : textes latins et données archéologiques, in Techniques et culture, 41, 2003, p. 1-27.

Chazelles 2016 = C.-A. de Chazelles, Recherches sur les origines de la construction en pisé en Occident, in Architecture en terre crue. Actes du colloque (Tunis, 6 février 2015), Tunis, 2016, p. 9-44.

Ciotti 1978 = U. Ciotti, Due iscrizioni mitriache inedite, in M. de Boer, A.T. Edridge (ed.), Hommage a Maarten J. Vermaseren I, Leiden, 1978, p. 233-246.

Conte 2019 = A.R. Conte, Per una cronologia degli eventi naturali a Roma dal VII secolo a.C. al V secolo d.C., in Traces in time, 8, 2019, p. 1-23.

Correia – Dipasquale – Mecca 2011 = M. Correia, L. Dipasquale, S. Mecca (ed.), Terra Europae. Earthen architecture in the European Union, Pisa-Brussels, 2011.

DeLaine 1992 = J. DeLaine, Design and construction in Roman imperial architecture. The baths of Caracalla in Rome, PhD, University of Adelaide, 1992.

Desbat 1993 = A. Desbat, La construction en terre à l’époque romaine, in A. Greffier-Richard (ed.), Matières à faire. Actes des séminaires publics d’archéologie, année 1991, Besançon, 1993, p. 147-160.

Dessales 2019 = H. Dessales, RECAP. Reconstruire après un tremblement de terre, in M. Devès, P. Bougeault, Risques et catastrophes naturels, in Cahiers de l’ANR, 10, 2014, p. 35.

Dessales et al. 2022 = H. Dessales, J. Cavero, A. Tricoche, Post-earthquake reconstruction: mapping and recording repairs in ancient Pompeii, in S. D’Amico, V. Venuti (ed.), Handbook of cultural heritage analysis, Cham, 2022, p. 1805-1821.

Diosono 2009 = F. Diosono, Cascia: i templi ed il forum di Villa San Silvestro. La Sabina dalla conquista romana a Vespasiano, in Forma Urbis, 14/7-8, 2009, p. 8-18.

Dipasquale – Mecca 2015 = L. Dipasquale, S. Mecca, Local seismic culture in the Mediterranean region, in M. Correia, P.B. Lourenço, H. Varum (ed.) Seismic retrofitting. Learning from vernacular architecture, London, 2015, p. 67-76.

Fentress et al. 2003 = E. Fentress, J. Bodel, A. Rabinowitz, R. Taylor, Cosa in the Republic and Early Empire, in E. Fentress (ed.) Cosa V. An intermittent town. Excavations 1991-1997, Ann Arbor, 2003, p. 13-62.

Ferrante 2014 = S. Ferrante, Norba: la domus VI nella sequenza stratigrafica e nelle sue parti, in Norba. Domus e materiali, Rome, 2014, p. 37-58.

Font i Mezquita – Hidalgo i Chulio 1990 = F. Font i Mezquita, P. Hidalgo i Chulio, El tapial. Una tècnica constructiva mil·lenària, Castellón, 1990.

Galadini – Galli 2004 = F. Galadini, P. Galli, The 346 A.D. earthquake (Central-Southern Italy): an archaeoseismological approach, in Annals geophysics, 47/2-3, 2004, p. 885-905.

Galdieri 1986 = E. Galdieri, Arquitectura de tierra – histórica y moderna – en Italia, in Informes de la construcción, 377, 1986, p. 51-53.

Galli – Molin 2013 = P.A.C. Galli, D. Molin, Beyond the damage threshold: the historic earthquakes of Rome, in Bulletin of earthquake engineering, 12/3, 2013, p. 1277-1306.

Giuliani 2011 = C.F. Giuliani, Provvedimenti antisisimici nell’antichità, in Rivista di topografia antica, 21, 2011, p. 25-52.

Grau Mira – Molina Vidal 2008 = I. Grau Mira, J. Molina Vidal, La villa de Rufión (Giano dell’Umbria, Italia): producción y territorio en la vía Flaminia (campaña 2007), in Excavaciones en el exterior 2007, in Informes y trabajos, 1, 2008, p. 77-81.

Grau Mira – Molina Vidal 2009 = I. Grau Mira, J. Molina Vidal, La villa de Rufio (Giano dell’Umbria, Italia) y su inserción territorial: hipótesis para una investigación combinada (campaña 2008), in Excavaciones en el exterior 2008, in Informes y trabajos, 3, 2009, p. 111-116.

Grau Mira – Molina Vidal 2010 = I. Grau Mira, J. Molina Vidal, La villa de Rufio (Giano Dell’Umbria, Italia): la delimitación del área residencial (campaña 2010), in Excavaciones en el exterior 2009, in Informes y trabajos, 5, 2010, p. 178-187.

Grau Mira – Molina Vidal 2011 = I. Grau Mira, J. Molina Vidal, La villa de Rufio (Giano dell’Umbria, Italia): el pabellón de servicio y las áreas periféricas, in Excavaciones en el exterior 2010, in Informes y trabajos, 7, 2011, p. 159-165.

Guerrero Baca 2011 = L.F. Guerrero Baca, Pasado y porvenir de la arquitectura de tapia, in Bitácora arquitectura, 22, 2011, p. 6-13.

Guerrero Baca – Boto de Matos 2013 = L.F. Guerrero Baca, J.G. Boto de Matos, El legado de la construcción tradicional con tapial, in Gaceta del Instituto del patrimonio cultural del Estado de Oaxaca, 24, 2013, p. 4-11.

Guidoboni 1989 = E. Guidoboni, I terremoti prima del Mille in Italia e nell’area mediterranea. Storia, archeologia, sismologia, 1989, Bologna.

Guidoboni et al. 1994 = E. Guidoboni, A. Comastri, G. Traina, Catalogue of earthquakes in the Mediterranean region up to the 10th century, Rome-Bologna, 1994.

Jorquera 2014 = N. Jorquera, Culturas sísmicas: estrategias vernaculares de sismorresistencia del patrimonio arquitectónico chileno, in Arquitecturas del Sur, 46, 2014, p. 18-29.

Knoll et al. 2019 = F. Knoll, M. Pastor Quiles, C.A. de Chazelles, L. Cooke, On cob balls, adobe and daubed straw plaits. A glossary on traditional earth building techniques (walls) in four languages, Halle, 2019 (Tagungen des Landesmuseums für Vorgeschichte Halle, 18).

Lazar et al. 2020 = M. Lazar, E.H. Cline, R. Nickelsberg, R. Shahack-Gross, A. Yasur-Landau, Earthquake damage as a catalyst to abandonment of a Middle Bronze Age settlement: Tel Kabri, Israel, in PLOS ONE, 15/9, 2020, online: https://doi.org/10.1371/journal.pone.0239079.

Llidó López – Molina Vidal 2012 = F. Llidó López, J. Molina Vidal, Gaius Iulius Rufio, propietario en la vía Flaminia, entre Suetonio y la epigrafía, in Epigraphica, 74, 2012, p. 75-82.

Marzano 2007 = A. Marzano, Roman villas in central Italy. A social and economic history, Leiden-Boston, 2007 (Columbia studies in the classical tradition, 30).

Mileto – Vegas López-Manzanares – Cristini 2012 = C. Mileto, F. Vegas López-Manzanares, V. Cristini, Refuerzos y mejoras: variantes constructivas de la tapia en España, in Actas del Congreso iberoamericano de arquitectura y construcción en tierra, SIACOT 2012 (Lima, 23-27 abril 2012), Lima, 2012, p. 10-20.

Minke 2001 = G. Minke, Manual de construcción para viviendas antisísmicas de tierra, Kassel, 2001.

Mogetta 2021 = M. Mogetta, The origins of concrete construction in Roman architecture. Technology and society in Republican Italy, Cambridge, 2021.

Molina Vidal et al. 2017 = J. Molina Vidal, I. Grau Mira, F. Llidó López, J.F. Álvarez Tortosa, Housing slaves on Roman estates: a proposed ergastulum at the villa of Rufio (Giano dell’Umbria), in Journal of Roman archaeology, 30, 2017, p. 387-406.

Molina Vidal 2021 = J. Molina Vidal, M.J. López Medina, La villa de Rufio (Giano dell’Umbria, PG-Italia): fases constructivas y desarrollo de un modelo productivo esclavista, in Archivo español de arqueología, 94, 2021.

Morigi 1997 = A. Morigi, Carsulae. Topografia e Monumenti, Rome,1997.

Pagana 2011 = P. Pagana, Sismicità storica in Umbria. Ricostruzione e studio dei principali terremoti verificatisi a partire dal III secolo a.C., Perugia, 2011.

Pastor Quiles 2017 = M. Pastor Quiles, La construcción con tierra en arqueología. Teoría, método, técnicas y aplicación, Alicante, 2017.

Pecchioli – Cangi – Marra 2018 = L. Pecchioli, G. Cangi, F. Marra, Evidence of seismic damages on ancient Roman buildings at Ostia: an arch mechanics approach, in Journal of archaeological science: reports, 21, 2018, p. 117-127.

Perello 2015 = B. Perello, Pisé or not pisé ? Problème de définition des techniques traditionnelles de la construction en terre sur les sites archéologiques, in ArchéOrient, 2015, online: https://archeorient.hypotheses.org/4562.

Pesando 2005 = F. Pesando, Il progetto Regio VI: le champagne di scavo 2001–2002 nelle insulae 9 e 10, in P.G. Guzzo, M.P. Guidobaldi (ed.), Nuove ricerche archaeologiche a Pompei ed Ercolano, Naples, 2005, p. 73-96.

Pesando 2011 = F. Pesando, L’ars struendi nella precettistica catoniana (agr.14), in A. Roselli, R. Velardi (ed.) L’insegnamento delle technai nelle culture antiche (Ercolano 23-24, marzo 2009), Pisa-Rome, 2011, p. 85-94.

Pesando 2013 = F. Pesando, Pompei in età sannitica. Tipologia, uso e cronologia delle tecniche edilizie, in F.M. Cifarelli (ed.), Tecniche costruttive del tardo ellenismo nel Lazio e in Campania. Atti del Convegno Segni (3 dicembre 2011), Rome, 2013, p. 117-123.

Pesando – Guidobaldi 2006 = F. Pesando, M.P. Guidobaldi, Pompei, Oplontis, Ercolano, Stabiae, Rome, 2006.

Petraccia Lucernoni 1996 = M.F. Petraccia Lucernoni, Iscrizione mitraica di Montoro, in Epigraphica, 58, 1996, p. 51-59.

Ponzi 1983 = C. Ponzi, Ville e insediamenti rustici di età romana in Umbria, Perugia, 1983.

Quilici Gigli 2014 = S. Quilici Gigli, Ricerche a Norba sull’edilizia privata: la domus VI, in S. Quilici Gigli, L. Quilici (ed.), Norba. Domus e materiali, Rome, 2014, p. 7-36.

Regoli 1985 = E. Regoli, Tecnica e tipologia delle construzioni, in A. Carandini, Settefinestre. Una villa schiavistica nell’Etruria romana, I, La villa nel suo insieme, Modena, 1985, p. 61-73.

Rodríguez Cano et al. 2021 = L. Rodríguez Cano, L.F. Guerrero Baca, R. Rosas Salinas, A. Ramírez Rodríguez, Después del sismo. Saberes tradicionales de la Mixteca poblana. Estudios y experiencias, Mexico City, 2021.

Rodríguez-Pascua et al. 2011 = M.A. Rodríguez-Pascua, R. Pérez-López, J. Giner-Robles, P.G. Silva, V.H. Garduño-Monroy, K. Reicherter, A comprehensive classification of earthquake archaeological effects (EAE) in archaeoseismology: application to ancient remains of Roman and Mesoamerican cultures, in Quaternary international, 242, 2011, p. 20-30.

Roux – Cammas 2007 = J.C. Roux, C. Cammas, La bauge coffrée : appréhension d’un mode de construction inédit dans la ville protohistorique de Lattes, Hérault (deuxième quart du IV s. av. notre ère), in H. Guillaud, C.-A. de Chazelles, A. Klein (ed.), Les constructions en terre massive, pisé et bauge, Montpellier, 2007 (Échanges transdisciplinaires sur constructions en terre crue, 2), p. 87-98.

Rovida 2021 = A. Rovida, R. Camassi, P. Gasperini, M. Stucchi (ed.), Catalogo parametrico dei terremoti italiani (CPTI15), versione 3.0, in Istituto nazionale di geofisica e vulcanologia (INGV), 2021, online: https://doi.org/10.13127/CPTI/CPTI15.3.

Russell – Fentress 2016 = B. Russell, E. Fentress, Mud brick and pisé de terre between Punic and Roman, in J. DeLaine, S. Camporeale, A. Pizzo (ed.), Arqueología de la construcción, V, Man-made materials, engineering and infrastructure. proceedings of the 5th International workshop on the archaeology of Roman construction, Oxford, April 11-12, 2015, Madrid, 2016, p. 131-43.

Soricelli 2009 = G. Soricelli, La provincia del Samnium e il terremoto del 346 d.C., in A. Storchi Marino, G.D. Merola (ed.), lnterventi imperiali in campo economice e sociale. Da Augusto al Tardoantico, Bari, 2009 (Pragmateiai, 18), p. 245-262.

Spence 1998 = R. Spence (ed.), The Umbria Marche earthquakes of 26 September 1997. A field report by EEFIT, London, 1998.

Staffa 1996 = A.R. Staffa, Una tecnica costruttiva di antichissima origine: le case di terra, in Le valli della Vibrata e del Salinello, Pescara, 1996 (Documenti dell’Abruzzo Teramano, 1), p. 111-119.

Stiros 1995 = S.C. Stiros, Archaeological evidence of antiseismic constructions in Antiquity, in Annali de geofisica, 38/5-6, 1995, p. 725-736.

Stiros 1996 = S.C. Stiros, Identification of earthquakes from archaeological data: methodology, criteria and limitations, in Stiros – Jones 1996, p. 129-152.

Stiros – Jones 1996 = S.C. Stiros, E. Jones (ed.) Archaeoseismology, Athens, 1996 (Fitch laboratory occasional paper, 7).

Tomassini 2021 = P. Tomassini, Construire, réparer, consolider : stratégies de résilience de l’architecture ostienne entre le IIe et le IIIe siècle ap. J.-C., in MEFRA, 133/2, 2021, p. 427-455.

Haut de page

Notes

1 Llidó – Molina 2012.

2 Suet., Iul., 76.

3 Molina et al. 2017.

4 Grau – Molina 2008, 2009, 2010, 2011.

5 Llidó – Molina 2012; Molina et al. 2017; Molina – López 2021.

6 Molina – López 2021.

7 Stiros 1996; Rodríguez-Pascua et al. 2011; Arrighetti 2015; among others.

8 Unfortunately, due to excavation constraints geoarchaeological analyses to provide a deeper insight into soil composition and material sources are not available for this research.

9 Molina – López 2021.

10 Molina et al. 2017.

11 Adam 1984.

12 Algorri – Vázquez 1996, p. 22.

13 Guerrero 2011, p. 9.

14 Guerrero – Boto de Matos 2013.

15 Roux – Cammas 2007; Aurenche et al. 2011; Knoll et al. 2019, p. 19.

16 Minke 2001, p. 61-62; Chazelles 1999, p. 232; 2016, p. 9.

17 Mileto – Vegas López-Manzanares – Cristini 2012.

18 Font – Hidalgo 1990, p. 33.

19 Castellarnau – Rivas 2013, p. 104.

20 Chazelles 1990; 2003, fig. 5.

21 Rodríguez et al. 2021, p. 163-166.

22 Pesando 2005, p. 87; 2011, p. 91; 2013, p. 123; Pesando – Guidobaldi 2006; Battaglini – Diosono 2010, p. 229; Ceccarelli 2018, p. 33; 2019, p. 36; Mogetta 2021, p. 256.

23 Algorri – Vázquez 1996, p. 22; Perello 2015; Pastor 2017.

24 Varro, Rust., 1, 14, 4.

25 Plin., Nat., 35, 48.

26 Pallad., 1, 34, 4.

27 Desbat 1993, p. 147; Chazelles 2003, p. 8; 2016, p. 13.

28 Isid., Orig., 15, 9, 5.

29 Desbat 1993, p. 153; Chazelles 1999, p. 243; 2003, p. 8; 2016, p. 14-16; Russell – Fentress 2016, p. 135-136.

30 Chazelles 2016, p. 18, 22, fig. 11.

31 Chazelles 1990, 2016.

32 Battaglini – Diosono 2010, p. 229; Russell – Fentress 2016, p. 138; Ceccarelli 2018, p. 33; 2019, p. 39-40.

33 Regoli 1985, p. 64-65.

34 Fentress et al. 2003, p. 21.

35 Bataglini – Diosono 2010, p. 227; Pesando 2011, p. 91, fig. 13; Mogetta 2021, p. 192.

36 Carfora – Ferrante – Quilici Gigli 2013.

37 Ferrante 2014; Quilici Gigli 2014.

38 Pesando 2005, p. 87; 2011, p. 92, fig. 20; 2013, p. 123, fig. 9.

39 Staffa 1996, p. 111, 113, fig. 60.

40 Guidoboni – Comastri – Traina 1994.

41 E.g., Stiros – Jones 1996.

42 Rovida et al. 2021.

43 Where 1 is the value for the highest risk and 4 is the lowest value (https://rischi.protezionecivile.gov.it/it/sismico/attivita/classificazione-sismica).

44 Rovida et al. 2021.

45 Camassi 2004; Galli – Molin 2013; Conte 2019.

46 Obseq., 46 and 61.

47 Gell., 4, 6, 2.

48 Cic., Catil. 3, 8.

49 Obseq., 59.

50 Ciotti 1978; Petraccia Lucernoni 1996; AE 1996, 601.

51 Morigi 1997; Bonini et al. 2003.

52 Bottari – Sepe 2013, Bottari et al. 2017.

53 The Historia Augusta (Gord., 26, 1-2) mentions a major earthquake around that date but does not specify its location: “There was a severe earthquake in Gordian’s reign – o severe that whole cities with all their inhabitants disappeared in the opening of the ground. Vast sacrifices were offered through the entire city and the entire world because of this. And Cordus says that the Sibylline Books were consulted, and everything that seemed ordered therein done; whereupon this world-wide evil was stayed” (LCL trans. by D. Magie 1924).

54 Ponzi 1983; Marzano 2007.

55 Diosono 2009.

56 Stiros 1996, p. 151, fig. 13.

57 E.g., Stiros 1996, p. 150, fig. 5; Spence 1998; Pecchioli – Cangi – Marra 2018.

58 Rodríguez-Pascua et al. 2011, p. 22.

59 Stiros 1996, p. 135, fig. 6a.

60 Stiros 1996, p. 149, 152; Rodríguez-Pascua et al. 2011, p. 22; Galli – Molin 2013; Arrighetti 2015; Pecchioli – Cangi – Marra 2018.

61 Giuliani 2011, p. 48-49.

62 Rodríguez-Pascua et al. 2011; Lazar et al. 2020.

63 Galdieri 1986; Correia – Dipasquale – Mecca 2011.

64 Minke 2001; Guerrero – Boto de Matos 2013, p. 8.

65 Stiros 1996, p. 150.

66 Spence 1998, p. 31, fig. 2.5; Dipasquale – Mecca 2015, p. 70, fig. 5.

67 Jorquera 2014, p. 27.

68 DeLaine 1992, p. 192; Stiros 1995, p. 731; 1996, fig. 10; Amici 2011; Giuliani 2011.

69 Dipasquale – Mecca 2015, p. 69, fig. 2.

70 Stiros 1996, p. 145.

71 Tomassini 2021.

72 Dessales 2019; Dessales et al. 2022. For more information, see: http://recap.huma-num.fr/webpublic/?lang=fr.

Haut de page

Table des illustrations

Titre Fig. 1. Location map of the Villa of Rufio (figure by the authors).
URL http://journals.openedition.org/mefra/docannexe/image/16239/img-1.jpg
Fichier image/jpeg, 238k
Titre Fig. 2. Ground plan of the Villa of Rufio, result of the different building interventions (figure by the authors).
URL http://journals.openedition.org/mefra/docannexe/image/16239/img-2.jpg
Fichier image/jpeg, 171k
Titre Fig. 3. Remains of earthen walls from the pars urbana corresponding to the first construction phase of the villa (Augustan period). A. Rooms A25 and 26, viewed from the northwest. B-C. Room A58 (figure by the authors).
URL http://journals.openedition.org/mefra/docannexe/image/16239/img-3.jpg
Fichier image/jpeg, 459k
Titre Fig. 4. Different heights preserved in the stone and opus mixtum structures in area 1 or pars rustica. In the foreground, rooms 6, 7 and 10, viewed from the northeast (figure by the authors).
URL http://journals.openedition.org/mefra/docannexe/image/16239/img-4.jpg
Fichier image/jpeg, 167k
Titre Fig. 5. Remains of earth elevations in area 1 or pars rustica. A. Separating rooms A7 and A8. B-D. Along the corridor 8b (figure by the authors).
URL http://journals.openedition.org/mefra/docannexe/image/16239/img-5.jpg
Fichier image/jpeg, 560k
Titre Fig. 6. Supra: map of central Italy overlaying the main seismogenic zones (seismic data extracted from: http://diss.rm.ingv.it/​diss/​ (figure by the authors). Infra: map of geological and seismic activity in the southern part of the Tiber basin (Bonini et al. 2003, fig. 1).
URL http://journals.openedition.org/mefra/docannexe/image/16239/img-6.jpg
Fichier image/jpeg, 960k
Titre Fig. 7. Plan of the villa with indication of the main signs of destruction that could be related to seismic activity (figure by the authors).
URL http://journals.openedition.org/mefra/docannexe/image/16239/img-7.jpg
Fichier image/jpeg, 301k
Titre Fig. 8. A. View of the rooms in area 1 showing a continuous crack. B. Crack in the arch. C. A6, with cracks in the walls, breakage, and undulation of the interior surface of the hypocaust. D. Crack running through the blocks and inclined strata in a rebuilt wall in A3 (figure by the authors).
URL http://journals.openedition.org/mefra/docannexe/image/16239/img-8.jpg
Fichier image/jpeg, 402k
Titre Fig. 9. A. Large threshold for the entrance to the ergastulum found fragmented and displaced several metres from its original position (A6). B. Base of stone blocks in an inclined position in A4. C. Breakage of the inner surface of the hypocaust (A6). D. Fractured threshold, between A8 and A54 (figure by the authors).
URL http://journals.openedition.org/mefra/docannexe/image/16239/img-9.jpg
Fichier image/jpeg, 434k
Titre Fig. 10. Remains found during the excavation of A59, showing a collapsed masonry elevation. Top view and detail of the fracture surface (figure by the authors).
URL http://journals.openedition.org/mefra/docannexe/image/16239/img-10.jpg
Fichier image/jpeg, 169k
Titre Fig. 11. A. Some of the fallen columns inside the peristyle (A9) at the time of its excavation in 2004. B. Column remains fallen in its base with the quadrangular slab in a rotated position, in A9. C. Deformation observed in one of the walls enclosing the peristyle. D. Peristyle slab found displaced towards A8b (figure by the authors).
URL http://journals.openedition.org/mefra/docannexe/image/16239/img-11.jpg
Fichier image/jpeg, 352k
Titre Fig. 12. Undulations in the profile of different structures. A. Channel inside A9. B. Saggio XVI-XVII. C. A27-35 (figure by the authors).
URL http://journals.openedition.org/mefra/docannexe/image/16239/img-12.jpg
Fichier image/jpeg, 242k
Titre Fig. 13. Undulations in canalisations. A. A39. B-C. Saggio XX (figure by the authors).
URL http://journals.openedition.org/mefra/docannexe/image/16239/img-13.jpg
Fichier image/jpeg, 279k
Titre Fig. 14. Construction reinforcements in A9, with the introduction of rows of bricks between the stone blocks, as well as metal nails (figure by the authors).
URL http://journals.openedition.org/mefra/docannexe/image/16239/img-14.jpg
Fichier image/jpeg, 181k
Haut de page

Pour citer cet article

Référence papier

Jaime Molina Vidal, María Pastor Quiles (corresponding author) et Daniel Mateo Corredor, « Earthen Architecture and Seismic Impact at the Roman Villa of Rufio (Giano dell’Umbria, Italy) »Mélanges de l'École française de Rome - Antiquité, 135-2 | 2023, 521-543.

Référence électronique

Jaime Molina Vidal, María Pastor Quiles (corresponding author) et Daniel Mateo Corredor, « Earthen Architecture and Seismic Impact at the Roman Villa of Rufio (Giano dell’Umbria, Italy) »Mélanges de l'École française de Rome - Antiquité [En ligne], 135-2 | 2023, mis en ligne le 10 avril 2024, consulté le 14 juin 2024. URL : http://journals.openedition.org/mefra/16239 ; DOI : https://doi.org/10.4000/mefra.16239

Haut de page

Auteurs

Jaime Molina Vidal

University of Alicante – jaime.molina@ua.es

María Pastor Quiles (corresponding author)

University of Alicante – Catalan Institute for Classical Archaeology (ICAC) – m.pastor@ua.es

Daniel Mateo Corredor

University of Alicante – daniel.mateo@ua.es

Haut de page

Droits d’auteur

CC-BY-NC-ND-4.0

Le texte seul est utilisable sous licence CC BY-NC-ND 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search