This paper presents the contextualized results of the latest excavations at the site of Casa Herrera (Mérida, Spain). Casa Herrera is one of the best examples of a late antique site in the Iberian Peninsula, not only because of the degree of preservation of the remains but also because of its long chronological sequence, which runs from the first to the ninth century. The excavations of the surroundings of the funerary basilica and the Roman aqueduct have unearthed the remains of a handful of buildings that could be linked to a rural monastic community from the late Roman and Visigothic periods. The site has an Umayyad phase where settlement clusters around the basilica before being finally abandoned during the ninth century.

The city of Mérida (ancient Augusta Emerita) is one of the main points of reference in late antique and Visigothic studies, not only for its urban monuments and its well-studied archaeological sequence but also for the written accounts collected in the seventh century and the numerous inscriptions that are still being discovered to this day.1 Its territory has received similar attention, but among the known rural sites, one stands out. The site of Casa Herrera, famous for its funerary basilica, has been for many years one of the best examples of early Christian, rural archaeology in Spain. While the basilica has been thoroughly studied and excavated, very little was known about its surrounding context, and most discussions were focused on its architecture and decorations. Small but selective excavations carried out between 2007 and 2013 under the auspices of the Consorcio de Mérida (the public entity that manages the archaeological heritage of the city and its territory) and the Oxford Classics Faculty, however, focused on the buildings that surround the basilica. These digs have revealed a number of structures coeval with the basilica and the remains of one of the urban aqueducts, casting new light on the peri-urban landscape of Mérida and on rural Christianity in the Visigothic period.

In this paper, we will present and analyze the results of the Oxford-Consorcio 2012–2013 campaigns. This will include discussion of the structures, the chronological sequence, and the finds. It will be possible to give an overall and preliminary interpretation of Casa Herrera not only as a suburban rural settlement linked to a monastic community but also as a multiperiod site during the late antique longue durée that includes a Roman, a Visigothic, and an early Islamic phase.

Located on a gentle slope in one of the hilly valleys that surround Mérida, the site of Casa Herrera is 6 km NE from the Roman city, along the unpaved road to Mirandilla (Figure 1). The site was discovered and first excavated in 1943 by José Serra i Rafols, the then commissar for excavations at Mérida. His excavations unearthed not a Roman villa (as he expected) but a double-apsed church linked to a small necropolis, and while he never published his results, in his notes he suggested, as we do here, that the site could have been a monastery.2

Figure 1.

Map of Mérida and surroundings, indicating the location of the site of Casa Herrera with the city’s network of roads and aqueducts.

Figure 1.

Map of Mérida and surroundings, indicating the location of the site of Casa Herrera with the city’s network of roads and aqueducts.

Close modal

The basilica itself is a double-apsed building, a type usually highlighted as North African, but with a strong and older tradition in the Iberian southwest, including the basilicas of Torre de Palma and Mértola (both in Portugal) and El Germo and San Pedro de Alcántara (both in Andalusia).3 In its first phase, the building was 25 m long and 15 m wide with a central nave and two side naves separated by columns (Figure 2). The perimetral walls are 0.8 m wide and made of mortared rubble. The outside was limewashed, and the inside was decorated with carved marbles, both reused directly from Roman buildings and reworked Roman pieces. The roof was tiled, and the floor was paved in opus signinum (waterproof mortar).

Figure 2.

(a) Plan of Casa Herrera (based of Caballero and Ulbert’s plans).

(b) Photo of the basilica, looking east.

Figure 2.

(a) Plan of Casa Herrera (based of Caballero and Ulbert’s plans).

(b) Photo of the basilica, looking east.

Close modal

The main (eastern) apse held a liturgic altar in its center, identified by the imprints of a central ara and four corner stipites on the floor, although the broken remains of one of the stipites has been found in situ, still bound to the opus signinum, probably broken off when the basilica was dismantled in the Umayyad period. As with the rest of the sculpture of Casa Herrera, this altar was made entirely in white (recycled) marble in the local Mérida workshop (the most important carving center in the late antique peninsula).4 Besides the in situ remains, an altar slab excavated in the 1940s has been found recently in the magazines of the National Museum of Roman Art in Mérida.5 Its shape (rectangular with a triple frame) follows the Roman tradition of the cyma inversa, and its iconography (with a leaf in each corner of the slab) belongs to one of the most popular typologies of early Christian Iberian slabs. This apse formed, as in most late antique Hispanic basilicas, the sanctuary, into which only ordained members of the congregation were allowed. It was separated by marble chancel screens, following the rules set down in late antique Hispanic church councils.

The counter-apse did not hold a liturgical altar but rather a memorial with a sigmatic mensa (Figure 3). The sigmatic slab is another example of local production in local Borba marble.6 While the shape imitates East-Roman patterns, the carved decoration of an arch on columns is typical from the Mérida workshop. Serra i Rafols left in his unpublished notes a sketch of a stepped, trapezoidal base found next to this slab. We have been able to identify this piece, made in the same marble, hidden in the magazines in the museum,7 and it is clear now that it probably supported the sigmatic mensa. This would make the altar of the counter-apse the first example of a type widely known elsewhere in the West but until now unknown in the Iberian Peninsula.8

Figure 3.

Sigmatic mensa from Casa Herrera. Photo by Luis Caballero, with permission.

Figure 3.

Sigmatic mensa from Casa Herrera. Photo by Luis Caballero, with permission.

Close modal

The apses were located at either end of the central nave, which had a double colonnade, and which was filled with burials, clustering around the memorial mensa.

The chronology given to this first phase is circa 500, as suggested by the presence of fifth-century African Red Slip ware (ARS) pottery and the post quem dates given by the inscriptions, one of which, that of one Asella, gives a date of VIII id(us) Septem/bres (a)era DLX/a = 6 September 523 CE.9

In the sixth century, the basilica went through a phase of renovation and expansion, in a process that has traditionally been linked to the reinforcement of episcopal power during the period of Visigothic state formation.10 Two more lateral naves were added (increasing the total to five, even if the outmost ones seem to have been used as separate rooms) to a full width of 22 m. In the NE corner, by the main apse, another, smaller altar and a stepped, immersion baptistery were built. The baptistery was framed by six marble pilasters that held a small cupola. One of these pilasters is a failed attempt at creating a rounded column out of a recycled, prismatic, Roman block.11 The central nave was, furthermore, modified so that the intercolumniations were blocked by low walls, creating a third liturgical area next to the sanctuary that formed a corridor connection both apses, perhaps linked to processions.

Serra i Rafols dated the abandonment of the basilica to the Visigothic period, but his excavations had only focused on the inside of the building. In the 1970s the site was excavated again, in this case by Luis Caballero and Thilo Ulbert, who focused on the burials inside the basilica, paying more attention to the later phases.12 These excavations discovered evidence for an Islamic, Umayyad phase, during which the basilica underwent further transformations. After being partially dismantled, and having its doorways blocked, the nave seems to have been turned into a secluded and self-enclosed space. The presence of Kufic Arabic graffiti datable to the ninth/tenth century (Figure 4), inscribed on the shafts of the columns of the central nave, shows at least six different hands writing complaints, including a hypothetical “there is no salvation/hope” (3.7: lā khalāṣ) and a more certain “may God have mercy on Amin, son of Fityan ib[n …]” (4.10: t-r-ḥ-m Allāh Amīn / Ibn Fityān I- / -bn), which has led researchers to suggest that it was turned into a prison.13

Figure 4.

Column with Arabic inscription.

Figure 4.

Column with Arabic inscription.

Close modal

The German team carried out further narrow sondages around the basilica in the early 1980s, which confirmed that the large stones that could be seen littered around the site belonged to other buildings,14 although it was impossible with that method to go into any elaborate interpretations. More recently, larger sondages were carried out again in 2007 and 2008 by Tomás Cordero and Isaac Sastre, who identified traces of walls linked to three structures (B3–5) similar to those briefly discussed by Ulbert.15 The identified structures (Figure 5) were small, rectangular buildings, roughly 6 × 5 m, of which only the foundations (set directly on the natural clay) had been preserved. The earliest structure (B5) can be dated to the late fifth century, and the two others (B3 and B4) to the sixth/seventh century. There was evidence for an early Umayyad (eighth-century) phase, but this was inferred from later layers without any associated structures. It was not possible to define what the purpose of these rectangular structures was, as no archaeological deposits associated with usage were identified. These excavations also unearthed a section of a Roman aqueduct.16

Figure 5.

Structures identified in Area 2.

Figure 5.

Structures identified in Area 2.

Close modal

With the information from previous excavations, in 2011 we put forward the main lines for a research project, which served as the basis for a small set of excavations over the summers of 2012 and 2013 in three areas that we recognized as having the best archaeological potential to diachronically explain the site and what role the basilica played in it.

Area 1

The first area (Area 1) included two small sondages, one behind the eastern apse (where there was no archaeology other than a pile of tegula fragments) and another behind the counter-apse (both 5 × 5 m), with hopes of identifying burial remains.17

The stratigraphic sequence of the western (counter-apse) trench was simple, due to its shallow depth. Under the thin topsoil there was a homogeneously dark, plowed soil that sealed the trench, in which there were finds ranging from the first to the nineteenth century. (This plow soil sealed also all the other excavated trenches.)

Under this layer, a 5 m long and 0.8 m wide, NW-SE wall was found, almost parallel to the short side of the basilica but truncated by an E-W robbing trench (Figure 6). This wall was made of two faces of large, unworked stones bound with lime and clay mortar (Figure 7). It continued the line of some remains identified in one of Ulbert’s test pits, although the overall shape of the structure cannot be guessed. Collapsed remains of this wall were found toward the SW, following the slope—this might also indicate where the inside of the building was, suggesting that this was a wall enclosing a building separated from the basilica rather than a construction butting against it. The collapse and the robbing event must postdate the ninth century because the backfill of the robbing trench contained mixed pottery from both the Visigothic and Umayyad phases.

Figure 6.

Plan of the trenches excavated around the basilica.

Figure 6.

Plan of the trenches excavated around the basilica.

Close modal
Figure 7.

Main wall from Area 1.

Figure 7.

Main wall from Area 1.

Close modal

Under the debris of the collapse there were two main layers. The earliest one was found on top of the natural soil and butting against the wall. This layer was a dark brown silty clay, and its finds suggest a sixth- to seventh-century date. On top of this and under the collapse there was another layer in a sandier and lighter-colored matrix that can be dated with certainty to the ninth century (red-gloss ware). The nature of the finds of these two layers (overabundance of tableware, cookware, and animal bone) suggest that this was an area of food preparation and consumption, and most likely part of a dwelling structure. Chronologically, the finds suggest that the building had an extended period of use, that it was most likely in use during the later Visigothic period (coinciding with the second phase of the basilica), and that the dwelling function was preserved into the Umayyad phase.

This chronology is further confirmed by the presence of two carved marble items in the wall and the collapse debris. The first one is a small, rectangular, hollowed box measuring 21.5 × 16 × 9 cm, with a 4 × 10 cm hole (Figure 8), and the chisel lines and irregular surfaces on all of its sides show that it was never finished and was then turned into construction material. This suggests that at least the last stages of marble carving for the decorative sculpture of Casa Herrera were done on-site. Because of its dimensions and material, this piece seems to have been a loculus for an altar, the “relic box” in which the sacred remains were deposited as part of the consecration rite.18 The box was most likely meant to be hidden under the base of one of the altars. This kind of reliquary has been found in late antique churches across the peninsula, like those of El Berrueco (Madrid), Las Tapias (La Rioja), or Monte da Cegonha (Portugal).19 A marble slab and a stepped trapezoidal base, both damaged, broken, or unfinished, were also reused as building material. The presence of these recycled fragments would suggest that this structure was built at the same time as the basilica was extended.

Figure 8.

Possible marble reliquary.

Figure 8.

Possible marble reliquary.

Close modal

Lastly, it is worth mentioning that this main wall was built atop the razed foundations of a previous structure. The previous wall was only preserved as broken tiles and smaller stones imbedded into the natural clays, forming a curved wall, barely 0.3 m wide, that can only be dated as preceding the second phase of the basilica and only perhaps coeval with the first one.

Area 2

Area 2 is located 60 m S of the basilica and 10 m to the NE of the 2007 excavation, where one large monolithic granite threshold was visible on the surface (Figure 5). We identified the remains of two different structures with three overall phases, mirroring the sequence identified in the 2007 intervention.

The chronological sequence of Area 2 is best understood horizontally rather than vertically because plowing has removed almost all the stratigraphic sequence. The identified structures are preserved only at their foundations, and their sequence is clear because they appear butting and truncating each other. Similarly, all the sealed contexts under the plowed soil are layers trampled into the natural clays that can be best interpreted as floor preparations. As such, the finds only offer a date for the construction of each associated building. The lack of use and abandonment layers limits the extent to which we can interpret the functioning of the site beyond their construction.

The building with the granite threshold (B1a) was a small, rectangular structure, measuring 10 × 6 m, with less than 0.1 m of its height preserved. The perimetral walls of this building were 0.6 m wide and made of mortared rubble and occasional tile, although it is likely that the rest of the wall was built in pisé. Considering the large numbers of tegulae and imbrices, both reused Roman (16 fragments) and new, late antique types (460 fragments), B1a most probably had a tiled roof. The 1.8 meter-wide granite threshold (Figure 9) was located at the middle of the NW wall, and it gave access to a central room as wide as the stepping-stone that ran for the full depth of the building. The threshold has two square slots at either end, and the left one is truncated by a straight, diagonal groove. This room served as a distribution hall onto which two sets of side rooms (roughly 2.5 × 2 m) opened on either side. The rooms were separated by walls or rubble bound by clay (not mortar). The presence of a posthole in the north room butting the partition wall would suggest that the building had a pitched roof following the central longitudinal axis. The finds retrieved from the trampled layers that can be typologically dated are overwhelmingly fifth-century in date (Casa Herrera types [CH] 7, 29A, 35, 48) with a couple residual, earlier Roman forms (CH 5, 37, 73). It should be noted, however, that a dozen sherds belonging to coarser ware found in this building may be (because of the clays and the inclusions) roughly dated to the fifth/sixth century. This may, overall, suggest a date circa 500 for the construction of building B1a, and the likely implication is that this structure was built at the same time as the basilica, especially since the construction technique (use of high-quality lime mortar to bind two-faced, hewn rubble foundations) is the same—barring the dimensions, and B1 does not seem to reuse any of the larger voussoirs of the aqueduct’s vault. This, in turn, would also correspond to the earliest structure of the 2007 excavation (B5), with which it is aligned.

Figure 9.

Granite threshold, B1a.

Figure 9.

Granite threshold, B1a.

Close modal

B1a was extended at a later stage (B1b) when an extra room was added, abutting the SW wall whilst B1a was still standing (and, presumably, in use). The finds from the layer contained by its walls are still mostly fifth century in date, but there is an increasing number of coarser ware, which may suggest a later chronology, perhaps in the sixth century. As B1a, the foundations of this extension are made of mortared rubble. This extension was later truncated by a new building (B2), which is built following a slightly different alignment and does not use lime mortar as the binding element. Sequentially, this building might be either later Visigothic or Umayyad in date (mirroring the more complete sequence of B3-4-5), but there was no material culture associated with it to confirm this suggested date.

Without clear layers of occupation, it is difficult to guess what the purpose of these structures was. There are two pieces of evidence, however, that may indicate that these were storage or working spaces rather than dwellings. First is the concentration of amphora remains in B1a. These fragments are very eroded and broken and clearly are part of a secondary deposition (in the floor-leveling layer), but may still reflect reused waste from the settlement. Second is the significant presence of slag on an ashy bed in the central hall of B1 (context 2013). It should be emphasized that the layout is comparable to storage buildings found in other late Roman rural sites from the Iberian Peninsula—found for example at El Pelicano and Loranca—a type that continues into the Visigothic period in sites like Gózquez.20

Area 3

Area 3 was opened in 2012 and expanded in 2013 (Area 31). It consisted of a 5 × 5 m trench, dug next to the 2008 trench that had unearthed a section of the aqueduct and within sight of the remains visible in the roadside ditch beyond the site’s fence. Area 31 was excavated 10 m farther up the hill toward the NE, and it occupied an area of 4 × 4 m (Figure 10).

Figure 10.

Plan of the trenches excavated in Area 3.

Figure 10.

Plan of the trenches excavated in Area 3.

Close modal

The aqueduct itself is built by cutting a flat-based, leveled trench into the bedrock (Figure 11). The construction ditch is 1.75 m wide at the top, narrowing down to just 0.75 m, which is the width of the aqueduct built inside a boxed-cut trench, into which the specus was built. This is made of two different walls of faced rubble (0.3 m wide each), on top of which the top-quality opus signinum lining was added once the right gradient had been calculated and established. The specus itself is rectangular in profile (0.24 m wide and 0.28 m deep), and its bottom corners are reinforced. On top of the side walls, but apparently not on top of the opus signinum lining, two other walls of mortared rubble were built, on which sat the vault. These walls have only been identified in Area 31, as the vault does not fit perfectly on them. The vault itself was made with opus caementicium and large, unworked stones (the likes of which appear reused in the walls of the basilica) fitted as rough voussoirs 0.3 m thick. The vault itself was later covered with a mixed backfill of rocks and clay, perhaps to protect the vault from damage. The aqueduct was built together with a parallel road of beaten earth and metaled with gravel. It is impossible to calculate the full depth of the vaulted conduit because we do not have a full cross-section; however, the calculated height difference between the extrados of the vault and the base of the specus is 0.8 m.

Figure 11.

Aqueduct remains in Area 3.

Figure 11.

Aqueduct remains in Area 3.

Close modal

Despite its apparent small size, it is most likely that the aqueduct was part of the public urban supply network of Mérida. Our current understanding is that the Casa Herrera conduit was one of the four main branches that tapped into the northern aquifers and that converged into the one conduit now known as the San Lázaro aqueduct. The caput, furthermore, has been identified as a perennial spring 1.3 km toward the NE.21

At one point during the Roman period, a longitudinal crack, 3 m long (probably due to a bad curing of the mortar), appeared on the vault. This was later repaired (at an unknown date) by the smothering of a dark orange, very clayey sandy-lime mortar (Figure 12). The lack of sinter concretions on the specus walls in an area of very hard water suggests that the aqueduct was undergoing cleaning and maintenance until the conduit went out of use.22

Figure 12.

Mortar repairs on the vault in Area 31.

Figure 12.

Mortar repairs on the vault in Area 31.

Close modal

One of the most interesting findings was that the robbing and dismantling of the vault started somewhere in the fifteen meters that separate Areas 3 and 31. The backfill inside the robbing trench is a mix of late Roman and Visigothic period material, but overall of sixth/seventh century date (types CH 12, 17, 32). The context in which these are found (3014) is an orangish-brown, malleable, sticky, silty clay, which suggests that it was formed by flowing water carrying sediment rather than by the result of deliberate dumping of rubbish. This would point toward a last moment of unregulated water flow through the aqueduct into the sixth century, perhaps coinciding with the construction of the basilica.

After this spoliation and siltation of the conduit, a thick (0.4 m), dark abandonment layer sealed the remains of the conduit, with materials that suggest a ninth-century date, consistent with the abandonment of the other areas of the site.

Ceramics: Classification and Chronological Overview

The ceramic finds were cleaned on-site and processed after the campaigns were over. The sherds of each context were first separated, after a preliminary visual assessment, between common ware, Samian ware, amphorae, and building material. Considering the limited stratigraphy and the lack of absolute chronologies, we opted to follow the well-developed typological sequence of the late antique forms of the urban sites of Mérida.23 This chronological classification was based on a visual and tactile assessment of a number of characteristics:

  • Early Roman (first–fourth century): Includes finely decanted clays or clays with small, measured inclusions worked with a fast wheel and fired to a high temperature, ranging in color from orange to brownish gray and creamy brown. Includes all mold-made, red-gloss (Samian) productions. The fragments present regular break lines and have a dull metallic sound. Walls tend to be slipped, although painting also appears.

  • Late Roman (fifth–early sixth century): Lines and angles on lips, feet, and bases are still sharp, and walls tend to be straight—both as a result of being thrown on a fast (foot-powered) wheel. However, the mineral inclusions in the clay tend to be bigger and more noticeable. The firing is still homogeneous, but at lower temperatures, which results in browner and darker fabrics and irregular break lines. Slips are less frequent.

  • Post-Roman (late fifth / early sixth–mid eighth century): Includes Visigothic and Visigothic/Umayyad transitional ware. Walls become thicker, and clays are unrefined, with many mineral inclusions (quartz and especially mica). Shapes are curvier, tending to globular. Rounded bases appear. Irregular and mixed firing (including mixed oxidation and reduction) become the norm, and fabrics range from very dark orange to gray, but mostly brown. Pots are either thrown on a wheel or shaped by hand. Walls are either untreated or polished. Combed and spatulated decorations. Handles are thick and curve downward from the rim.

  • Umayyad (mid eighth–ninth century): Reintroduction of professionalized techniques, including fast wheel, refined clays, and high-temperature kilns. Walls become thinner and straighter, resulting in slenderer forms with clearly marked angles. Tableware are decorated with dark-red slips. Handles are thinner and curve above the edge of the rim.

After this preliminary classification, the potsherds were separated into nondiagnostic walls and significant fragments (including bases, handles, rims, and decorated walls), out of which diagnostic sherds were later chosen. Nondiagnostic fragments were counted and grouped by rough chronology; significant sherds were photographed, numbered, and indexed, while diagnostic sherds were also drawn and classified (see Table 1).

Table 1.

Casa Herrera typologies

CH typeCodeShapePeriodDate rangeCf
8265/1011/9, 12, 21 Closed bowl Umayyad 8th–9th Alba & Gutierrez 2008, fig. 6 
2A 8265/2012/4 Closed bowl Post-Roman 5th–6th  
2B 8284/3010/3 Closed bowl Early Roman 4th Bustamante, type 11.6 
8265/2015/5 Closed bowl Late Roman 5th  
8285/2106/3 Closed bowl Late Roman 5th  
8265/2018/1 Closed bowl Early Roman 1st–4th  
8285/2108/6 Open bowl Post-Roman 5th–6th  
8265/2008/7 Open bowl Late Roman 5th  
8265/2001/4 Open bowl Late Roman 5th  
8285/2105/3 Open bowl Early Roman 1st–2nd  
10 8285/2116/4 Open bowl Late Roman 5th  
11 8265/1011/27 Cup Umayyad 8th–9th Alba & Feijoo 2003, type 11-vaso 1 
12 8265/1011/18 Cup Umayyad 9th Alba & Feijoo 2003, type 11-vaso 2 
13A 8265/1011/7 Cup Post-Roman early 8th Alba & Feijoo 2003, type 8-vaso 
13B 8265/1006ii/15, 8265/3002/3 Cup Post-Roman 6th–7th Alba & Feijoo 2003, type 8-vaso 
13C 8285/2110/3 Cup Post-Roman 6th Alba & Feijoo 2003, type 8-vaso 
13D 8265/1002/3, 8265/1004/2 Cup Late Roman 5th  
14 8265/3002/13 Cup Late Roman 5th  
15 8284/3014/1 Cup Post-Roman 6th–7th Alba & Feijoo 2003, type 8-vaso 
16 8265/3000/29, 31 Bowl base Early Roman 1st–4th  
17 8265/2001/1 Bowl base Early Roman 1st–4th  
18 8265/3002/12 Plate Early Roman 1st–4th Bustamante 8.5? 
19 8284/3010/2 Plate Early Roman 4th  
20 8285/2125/2 Plate Late Roman 4th–5th  
21 8285/2106/6 Tray Early Roman 1st–4th  
22 8285/2110/4 Tray Post-Roman 6th–7th Alba & Feijoo 2003 9. baños 
23 8265/1011/7 Tray Umayyad late 8th  
24 8265/1005/1 Tray Post-Roman 6th–7th  
25 8265/1009/1 Tray Late Roman 5th  
26A 8265/1011/10 Tray Umayyad 8th–9th Alba & Feijoo 2003 12.B 
26B 8265/1011/8 Tray Umayyad 8th–9th Alba & Feijoo 2003 12.B 
27A 8265/1011/1 Long-necked pot Umayyad 8th–9th Alba & Gutierrez 2008, fig. 6 
27B 8265/1066ii/100 Long-necked pot Post-Roman 6th–7th Alba & Feijoo 2003, type 8, A6 
27C 8265/1000/4 Long-necked pot Post-Roman 6th  
28 8265/1002/1 Long-necked pot Late Roman 5th  
29A 8265/2017/1, 8285/2116/5 Short-necked pot Late Roman 5th  
29B 8265/3000/2 Short-necked pot Early Roman 1st–4th Bustamante, type 4.25 
30A 8285/2102/8 Short-necked pot Post-Roman 6th Alba & Feijoo 2003, type 8, A1 
30B 8284/3005/1 Short-necked pot Late Roman 5th  
30C 8265/3000/1 Short-necked pot Early Roman 1st–4th  
31 8265/3002/1 Short-necked pot Late Roman 5th  
32 8265/1011/5, 11 Triangular rim pot Post-Roman early 8th Alba & Gutierrez 2008, fig. 5 
33 8265/1011/2 Triangular rim pot Post-Roman early 8th Alba & Gutierrez 2008, fig. 5 
34 8284/3014/2 Triangular rim pot Late Roman 5th  
35 8265/2008/4 Triangular rim pot Late Roman 5th  
36 8265/1009/2 Other pot Post-Roman 6th–7th  
37 8265/2007/1 Other pot Early Roman 1st–2nd Bustamante, type 4.21 
38 8265/1011/3, 23 Other pot Umayyad 9th Alba & Feijoo 2003, type 10, A2 
39 8265/3001/24 Other pot Late Roman 5th  
40 8265/2001/3 Other pot Early Roman 1st–4th  
41A 8265/3001/8 Convex-based pot Post-Roman 7th–8th Alba & Feijoo 2003, type 8 
41B 8285/2117/1 Convex-based pot Late Roman 5th  
42 8265/1006ii/19, 21 Convex-based pot Post-Roman 6th–7th Alba & Feijoo 2003, type 8 
43A 8265/1011/39 Flat-based pot Umayyad 8th–9th Alba & Gutierrez 2008, fig. 6 
43B 8284/3011/1 Flat-based pot Late Roman 5th  
44 8265/1007/4, 6, 8265/3000/27 Flat-based pot Post-Roman 5th–7th Alba & Feijoo 2003, type 8 
45 8265/1011/42 Flat-based pot Umayyad late 8th  
46A 8265/1011/34, 45 Flat-based pot Umayyad 8th–9th Alba & Feijoo 2003, type 10, A1/2 
46B 8265/1006ii/17 Flat-based pot Post-Roman 6th–7th  
47 8265/1011/44 Flat-based pot Umayyad late 8th  
48 8265/2007/5 Flat-based pot Late Roman 5th  
49 8265/2001/5 Flat-based pot Late Roman 5th  
50 8265/3000/10 Flat-based pot Early Roman 1st–4th  
51 8265/3000/26 Other pot Post-Roman 6th–7th  
52 8265/3000/27 Other pot Late Roman 5th  
53 8265/1006ii/23 Jar Post-Roman 7th–8th  
54 8265/3002/10 Jar Late Roman 5th  
55 8265/3002/8 Jar Late Roman 5th  
56 8265/3001/3-4-5-9, 16 Jar Post-Roman 6th Alba & Feijoo 2003, type 8-cántaro 
57A 8265/1011/7 Juglet Umayyad 9th Alba & Feijoo 2003, type 11-jarro 4 
57B 8265/1011/76 Juglet Umayyad 9th Alba & Feijoo 2003, type 11-jarro 4 
58A 8265/1011/75 Slipped jar Umayyad 8th–9th  
58B 8285/2100/2 Slipped jar Umayyad 8th–9th Alba & Gutierrez 2008, fig. 6 
59 8265/1011/4 Other jar Umayyad 8th–9th Alba & Feijoo 2003, type 10-cántaro 6 
60 8265/1002/2 Other jar Late Roman 5th  
61 8284/3010/5 Other jar Umayyad 8th–9th  
62 8265/1007/1, 8265/1011/73 Slipped jar Umayyad 9th Alba & Gutierrez 2008, fig 7 
63A 8265/1007/5 Globular jar Post-Roman 6th–7th Alba & Feijoo 2003, type 9-jarro 
63B 8265/1002/4 Globular jar Post-Roman 6th–7th Alba & Feijoo 2003, type 9-jarro 
64 8285/2110/1 Globular jar Early Roman 4th  
65 8265/3001/7 Footed jar Early Roman 4th  
66 8285/2116/2 Bottle Late Roman 5th Bustamante, p. 431, type 25 
67 8285/2102/9-10-16 Bottle Late Roman 5th Bustamante, p. 431, type 6.15 
68 8265/1011/6 Storage Umayyad 8th–9th  
69 8265/1066ii/16 Storage Post-Roman 6th  
70 8265/1066ii/13 Storage Post-Roman 5th–6th  
71 8285/2105/2 Amphora Late Roman 5th–6th LRA 4 = Almagro 54 
72 8285/2106/1 Glass unguentarium Late Roman 5th  
73 8265/2016/1 Sigillata Early Roman 1st–2nd  
74 8285/2110/2 Sigillata Late Roman 470s Paz 83, immit. Hayes 87A? or 61A? 
75 8285/2102/2 Sigillata Early Roman 1st–2nd Poss. Drag. 37 
76 8265/3001/1 Sigillata Early Roman 1st–4th  
77 8265/3001/2 Sigillata Early Roman 1st–4th  
CH typeCodeShapePeriodDate rangeCf
8265/1011/9, 12, 21 Closed bowl Umayyad 8th–9th Alba & Gutierrez 2008, fig. 6 
2A 8265/2012/4 Closed bowl Post-Roman 5th–6th  
2B 8284/3010/3 Closed bowl Early Roman 4th Bustamante, type 11.6 
8265/2015/5 Closed bowl Late Roman 5th  
8285/2106/3 Closed bowl Late Roman 5th  
8265/2018/1 Closed bowl Early Roman 1st–4th  
8285/2108/6 Open bowl Post-Roman 5th–6th  
8265/2008/7 Open bowl Late Roman 5th  
8265/2001/4 Open bowl Late Roman 5th  
8285/2105/3 Open bowl Early Roman 1st–2nd  
10 8285/2116/4 Open bowl Late Roman 5th  
11 8265/1011/27 Cup Umayyad 8th–9th Alba & Feijoo 2003, type 11-vaso 1 
12 8265/1011/18 Cup Umayyad 9th Alba & Feijoo 2003, type 11-vaso 2 
13A 8265/1011/7 Cup Post-Roman early 8th Alba & Feijoo 2003, type 8-vaso 
13B 8265/1006ii/15, 8265/3002/3 Cup Post-Roman 6th–7th Alba & Feijoo 2003, type 8-vaso 
13C 8285/2110/3 Cup Post-Roman 6th Alba & Feijoo 2003, type 8-vaso 
13D 8265/1002/3, 8265/1004/2 Cup Late Roman 5th  
14 8265/3002/13 Cup Late Roman 5th  
15 8284/3014/1 Cup Post-Roman 6th–7th Alba & Feijoo 2003, type 8-vaso 
16 8265/3000/29, 31 Bowl base Early Roman 1st–4th  
17 8265/2001/1 Bowl base Early Roman 1st–4th  
18 8265/3002/12 Plate Early Roman 1st–4th Bustamante 8.5? 
19 8284/3010/2 Plate Early Roman 4th  
20 8285/2125/2 Plate Late Roman 4th–5th  
21 8285/2106/6 Tray Early Roman 1st–4th  
22 8285/2110/4 Tray Post-Roman 6th–7th Alba & Feijoo 2003 9. baños 
23 8265/1011/7 Tray Umayyad late 8th  
24 8265/1005/1 Tray Post-Roman 6th–7th  
25 8265/1009/1 Tray Late Roman 5th  
26A 8265/1011/10 Tray Umayyad 8th–9th Alba & Feijoo 2003 12.B 
26B 8265/1011/8 Tray Umayyad 8th–9th Alba & Feijoo 2003 12.B 
27A 8265/1011/1 Long-necked pot Umayyad 8th–9th Alba & Gutierrez 2008, fig. 6 
27B 8265/1066ii/100 Long-necked pot Post-Roman 6th–7th Alba & Feijoo 2003, type 8, A6 
27C 8265/1000/4 Long-necked pot Post-Roman 6th  
28 8265/1002/1 Long-necked pot Late Roman 5th  
29A 8265/2017/1, 8285/2116/5 Short-necked pot Late Roman 5th  
29B 8265/3000/2 Short-necked pot Early Roman 1st–4th Bustamante, type 4.25 
30A 8285/2102/8 Short-necked pot Post-Roman 6th Alba & Feijoo 2003, type 8, A1 
30B 8284/3005/1 Short-necked pot Late Roman 5th  
30C 8265/3000/1 Short-necked pot Early Roman 1st–4th  
31 8265/3002/1 Short-necked pot Late Roman 5th  
32 8265/1011/5, 11 Triangular rim pot Post-Roman early 8th Alba & Gutierrez 2008, fig. 5 
33 8265/1011/2 Triangular rim pot Post-Roman early 8th Alba & Gutierrez 2008, fig. 5 
34 8284/3014/2 Triangular rim pot Late Roman 5th  
35 8265/2008/4 Triangular rim pot Late Roman 5th  
36 8265/1009/2 Other pot Post-Roman 6th–7th  
37 8265/2007/1 Other pot Early Roman 1st–2nd Bustamante, type 4.21 
38 8265/1011/3, 23 Other pot Umayyad 9th Alba & Feijoo 2003, type 10, A2 
39 8265/3001/24 Other pot Late Roman 5th  
40 8265/2001/3 Other pot Early Roman 1st–4th  
41A 8265/3001/8 Convex-based pot Post-Roman 7th–8th Alba & Feijoo 2003, type 8 
41B 8285/2117/1 Convex-based pot Late Roman 5th  
42 8265/1006ii/19, 21 Convex-based pot Post-Roman 6th–7th Alba & Feijoo 2003, type 8 
43A 8265/1011/39 Flat-based pot Umayyad 8th–9th Alba & Gutierrez 2008, fig. 6 
43B 8284/3011/1 Flat-based pot Late Roman 5th  
44 8265/1007/4, 6, 8265/3000/27 Flat-based pot Post-Roman 5th–7th Alba & Feijoo 2003, type 8 
45 8265/1011/42 Flat-based pot Umayyad late 8th  
46A 8265/1011/34, 45 Flat-based pot Umayyad 8th–9th Alba & Feijoo 2003, type 10, A1/2 
46B 8265/1006ii/17 Flat-based pot Post-Roman 6th–7th  
47 8265/1011/44 Flat-based pot Umayyad late 8th  
48 8265/2007/5 Flat-based pot Late Roman 5th  
49 8265/2001/5 Flat-based pot Late Roman 5th  
50 8265/3000/10 Flat-based pot Early Roman 1st–4th  
51 8265/3000/26 Other pot Post-Roman 6th–7th  
52 8265/3000/27 Other pot Late Roman 5th  
53 8265/1006ii/23 Jar Post-Roman 7th–8th  
54 8265/3002/10 Jar Late Roman 5th  
55 8265/3002/8 Jar Late Roman 5th  
56 8265/3001/3-4-5-9, 16 Jar Post-Roman 6th Alba & Feijoo 2003, type 8-cántaro 
57A 8265/1011/7 Juglet Umayyad 9th Alba & Feijoo 2003, type 11-jarro 4 
57B 8265/1011/76 Juglet Umayyad 9th Alba & Feijoo 2003, type 11-jarro 4 
58A 8265/1011/75 Slipped jar Umayyad 8th–9th  
58B 8285/2100/2 Slipped jar Umayyad 8th–9th Alba & Gutierrez 2008, fig. 6 
59 8265/1011/4 Other jar Umayyad 8th–9th Alba & Feijoo 2003, type 10-cántaro 6 
60 8265/1002/2 Other jar Late Roman 5th  
61 8284/3010/5 Other jar Umayyad 8th–9th  
62 8265/1007/1, 8265/1011/73 Slipped jar Umayyad 9th Alba & Gutierrez 2008, fig 7 
63A 8265/1007/5 Globular jar Post-Roman 6th–7th Alba & Feijoo 2003, type 9-jarro 
63B 8265/1002/4 Globular jar Post-Roman 6th–7th Alba & Feijoo 2003, type 9-jarro 
64 8285/2110/1 Globular jar Early Roman 4th  
65 8265/3001/7 Footed jar Early Roman 4th  
66 8285/2116/2 Bottle Late Roman 5th Bustamante, p. 431, type 25 
67 8285/2102/9-10-16 Bottle Late Roman 5th Bustamante, p. 431, type 6.15 
68 8265/1011/6 Storage Umayyad 8th–9th  
69 8265/1066ii/16 Storage Post-Roman 6th  
70 8265/1066ii/13 Storage Post-Roman 5th–6th  
71 8285/2105/2 Amphora Late Roman 5th–6th LRA 4 = Almagro 54 
72 8285/2106/1 Glass unguentarium Late Roman 5th  
73 8265/2016/1 Sigillata Early Roman 1st–2nd  
74 8285/2110/2 Sigillata Late Roman 470s Paz 83, immit. Hayes 87A? or 61A? 
75 8285/2102/2 Sigillata Early Roman 1st–2nd Poss. Drag. 37 
76 8265/3001/1 Sigillata Early Roman 1st–4th  
77 8265/3001/2 Sigillata Early Roman 1st–4th  

Early Roman remnants

Early Roman ceramics are surprisingly numerous, but when they appear in sealed contexts, their presence is very clearly residual and eroded; the vast majority of these appear in the plow soil (59% of the 270 datable sherds come from the abandonment layer of Area 3, mixed with later materials). In total, 24.3% of the common ware and 34% of the ceramic building materials can be dated to this phase.24 Fifteen of these sherds were deemed diagnostic. Half of these seems to be tableware (bowls, plates), while the other half seems to correspond to cookware. Regarding early Roman Samian ware, eight very worn and eroded fragments were retrieved, mostly south Gaulish or Hispanic (as the colors of the clays suggest).

Late Roman remnants

The material for the late Roman phase is the second most numerous, and it consists overwhelmingly of common ware. With a total of 290 fragments (26.1% of the total), 76 of these were significant and 28 were deemed diagnostic. These sherds tend to appear in better condition, and they show a wider variety of types. While bowls and cooking pots are the most numerous, we also find jugs and bottles of this chronology. Over half of this material (55%) came from Area 2, mostly from the floors of B1a. One fragment of late Hispanic red gloss (terra sigillata hispánica tardía) and one fragment of imitation red gloss (cerámica de imitación de sigillata), both dated to the fifth century, represent the only two examples of late Roman fine ware.

It should be noted that 61 fragments of amphorae were found in the excavations. They have pinkish-orange clays with mineral inclusions (including white quartz, reddish specks, and mica) and a cream-colored exterior. Of these, only one sherd (CH 71) was diagnostic, a rim that suggests that it might have been of late antique date. The other fragments were also small and probably belonged to late-antique spatheia. The presence of these imported amphorae clearly points toward the connection between the site and the city; probably the latter supplying the former rather than the inhabitants of Casa Herrera directly participating in long-distance trade.

Post-Roman remnants

The ceramics of the Visigothic period, including those of the early eighth century that follow up with pre-Umayyad traditions, represent 22% of the total, with 242 fragments, 55 significant and 30 diagnostic. The types of this phase mirror those of the fifth century: tableware (bowls, plates, jugs, trays) and cooking pots, but also larger storage vessels. The fragments dated to this phase are evenly distributed between Areas 1 and 2.

Umayyad remnants

Lastly, Umayyad sherds are the most abundant (305 fragments, 27.5% of the total), all but two of which come from Area 1. There were 77 significant fragments and 30 diagnostic. The standardization and professionalization of production in this period is evident in the narrowed range of repeated typologies of this phase. Bowls are limited to closed, angular types; drinking cups and mugs are straighter; and cooking pots have flat bases and long necks.

The Casa Herrera Typologies

Typologies for common ware in the late antique and early medieval periods are not always useful for cross-chronological comparative purposes because of the extremely local (and, often, amateur or domestic) nature of production.25 The classification of the excavation’s ceramics, however, is still relevant to understanding the evolution of Casa Herrera throughout its main phases.

Bowls and cups

We used bowl and cup to designate a large variety of forms linked to the individual consumption of liquids (which could include drinking but also stews and soups). Closed forms (CH 1–5), for instance, are abundant in the late Roman period and probably into the sixth century, only to become dominant in the Umayyad period, particularly CH 1, the red-slipped truncated-biconical bowl (Figure 13). Open forms (CH 6–10) are absent from the last phase, but they are as common in the late Roman as they are throughout the Visigothic period (Figure 14). A number of smaller examples, better described as cups, have been found (CH 11–17), and while not a full profile was found, Visigothic types CH 13 and 15 probably could belong to the same form: a lipped cup with a narrow base and a carinated shoulder. Roman and late Roman cups would have had a footed base (Figures 15 and 16).

Figure 13.

Cups and bowls, types CH 1–5.

Figure 13.

Cups and bowls, types CH 1–5.

Close modal
Figure 14.

Cups and bowls, types CH 6–10.

Figure 14.

Cups and bowls, types CH 6–10.

Close modal
Figure 15.

Chronotypological diagram showing Casa Herrera bowls and cups.

Figure 15.

Chronotypological diagram showing Casa Herrera bowls and cups.

Close modal
Figure 16.

Cups and bowls, types CH 11–17.

Figure 16.

Cups and bowls, types CH 11–17.

Close modal

Plates and basins

Plates and trays are far less common than bowls (CH 18–20), and they are all residual, Roman or late-Roman examples (Figure 17). When compared to the overabundance of bowls, this might indicate a transition in eating patterns, moving on from sharing food from platters to having individual portions.26

Figure 17.

Plates and trays, types CH 18–20.

Figure 17.

Plates and trays, types CH 18–20.

Close modal

Similarly, each phase has produced at least one type of wide, open, thick-walled vessels (CH 21–26). Their shape would indicate that these are not for storage, and it is likely they were basins of various description. Some might have been linked to washing or bathing (CH 22), others perhaps were trays or mixing bowls (Figure 18).

Figure 18.

Basins, types CH 21–26.

Figure 18.

Basins, types CH 21–26.

Close modal

Cooking pots

The widest range of forms belongs to cooking pots (Figure 19). One of the main types of cooking pot are the S-shaped or long-necked pots (CH 27–28), which begin in the late Roman period and continue to appear in the ninth century (Figure 20). Umayyad examples can develop quite elongated necks and narrow mouths. Short-necked pots (CH 29–31), on the contrary, are mostly late Roman in date, linking back to earlier, professionally made forms (Figure 21). Pots with triangular or almond-shaped rims appear both in the fifth (CH 34–35) and the eighth centuries (CH 32–33), the former still tapping from Roman short-necked types, but the latter seeming to be a modification of the S-shaped, long-necked cooking pot (Figure 22).

Figure 19.

Chronotypological diagram showing Casa Herrera cooking-pot rims.

Figure 19.

Chronotypological diagram showing Casa Herrera cooking-pot rims.

Close modal
Figure 20.

Long-necked cooking pots, types CH 27–28.

Figure 20.

Long-necked cooking pots, types CH 27–28.

Close modal
Figure 21.

Short-necked cooking pots, types CH 29–31.

Figure 21.

Short-necked cooking pots, types CH 29–31.

Close modal
Figure 22.

Triangular-rimmed cooking pots, types CH 32–35.

Figure 22.

Triangular-rimmed cooking pots, types CH 32–35.

Close modal

Regarding bases (CH 41–52), the most interesting development is the abandonment of the molded feet typical of Roman types (Figure 23), which disappear in the fifth century, at which point rounded (as opposed to flat) bases become more common, as outlined earlier (Figures 24a–b). While there were Roman examples of rounded cooking pots, they tended to either start from a molded foot or spring at an angle from the base (CH 48–50), but in the sixth and seventh centuries bases tend toward the curve, making pots more globular (especially CH 42 and 46B). In this period, the corner of the pot also tends to be thicker, something that is preserved in later, Umayyad bases in the form of a slight lump around the base (CH 46A and 47). The rounder shapes of cooking pots in the early medieval centuries can be linked to the shift from built-up kitchens to cooking around hearths.

Figure 23.

Chronotypological diagram showing Casa Herrera cooking-pot bases.

Figure 23.

Chronotypological diagram showing Casa Herrera cooking-pot bases.

Close modal
Figure 24a–b.

Cooking-pot bases, types CH 41–52.

Figure 24a–b.

Cooking-pot bases, types CH 41–52.

Close modal

Jugs and bottles

The collection of tableware from Casa Herrera is completed with jugs and bottles. Handled jugs of varying sizes (CH 53–65) appear in all periods, although the most remarkable seem to be those of Umayyad date (Figure 25 and Figure 26). Some examples (CH 56, 62) can be linked to water jugs, while some of the ninth-century examples (CH 57) seem to be the characteristic drinking mugs of the Umayyad period with a straight lip and a rounded body (Figure 27).

Figure 25.

Jugs, rims, and handles, types CH 53–62.

Figure 25.

Jugs, rims, and handles, types CH 53–62.

Close modal
Figure 26.

Jugs, bases, types CH 63–65.

Figure 26.

Jugs, bases, types CH 63–65.

Close modal
Figure 27.

Chronotypological diagram showing Casa Herrera jugs.

Figure 27.

Chronotypological diagram showing Casa Herrera jugs.

Close modal

Caballero and Ulbert’s excavation of the necropolis produced a large number of complete small votive jugs, many of which are exposed in the National Museum in Mérida (Figure 28), but none of these late antique votive types seem to correspond to the forms identified in the excavation of storage/dwelling areas.27

Figure 28.

Votive juglet from the 1970s Casa Herrera excavations, currently in the National Museum of Roman Art in Mérida.

Figure 28.

Votive juglet from the 1970s Casa Herrera excavations, currently in the National Museum of Roman Art in Mérida.

Close modal

At least two types have been identified as bottles of late Roman date (CH 66–67), which follow the earlier Roman tradition of wide-bodied bottles with a narrow neck (Figure 29).

Figure 29.

Bottles, types CH 66–67.

Figure 29.

Bottles, types CH 66–67.

Close modal

Storage

The excavations also produced remains of many large vessels, probably linked to storage (types 68–70) of Visigothic and Umayyad date (Figure 30). These types of storage vessels substituted the dolia which had been the most common large ceramic container in the Roman period.28

Figure 30.

Storage vessels, types CH 68–70.

Figure 30.

Storage vessels, types CH 68–70.

Close modal

Amphora, glass, Samian

Beyond coarse-ware types, there were diagnostic sherds of glass, Samian ware, and amphora (Figure 31). The one amphora rim (CH 71) could be a fifth- or sixth-century Palestinian Late Roman Amphora 4.29 While fragments of late antique green glass have appeared in Areas 1 and 2, the only identifiable fragment is the neck of what could be a small unguentarium (CH 72). Lastly, Gaulish and Hispanic Samian sherds appear very worn and eroded in fifth-century contexts. While CH 74 could be identified with late Hispanic type Paz 83 (dated post-470s),30 CH 76 could be a Dragendorff 37. In any case, they all seem to be parts of bowls or cups, used along with the coarse tableware.

Figure 31.

Amphora, glass, and Samian forms, types CH 71–76.

Figure 31.

Amphora, glass, and Samian forms, types CH 71–76.

Close modal

Other Finds

Ceramic building material

The abundance of brick and tile on-site, especially in the abandonment layers of the buildings of Area 2, would confirm that the late antique and early medieval structures had tiled roofs, acknowledging also the fact that broken fragments of brick and tile appear to have been used in the construction of the walls of the basilica. The one thing of note is that the ceramic building material assemblage of Casa Herrera is an imbalanced mixture of early Roman tegulae (249 fragments), and more abundant new, late antique tegulae and imbrices (478 fragments). These later tiles are characterized by their uneven or irregular (ox-red) firing, heterogeneous thickness, irregular lips, and large mineral inclusions (mostly quartz) in the clay. Contextually, it is probable that these tiles are fifth or sixth century in date, but there is nothing in the typology that can refine that broad chronology (Figure 32).

Figure 32.

Late antique roof tiles from the 2007 excavations.

Figure 32.

Late antique roof tiles from the 2007 excavations.

Close modal

Coins

Four coins in total have been found in the 2012 and 2013 excavations, and all were small-denomination, late Roman, copper alloy coins. The presence of late Roman small change in later contexts is not unusual, but even to this day these numi or minimi are difficult to come across in rural contexts, as mostly have been found in urban areas, where monetary exchange was perhaps more common.31

Three of the coins come from the sealed contexts of Area 2:

  • 8265/2001: Roman emperor facing right with diadem, possibly a Valentinianic coin of the Gloria Romanorum series (364–375). Very worn. Copper alloy, weight 5.4 g, diameter 29 mm. Legend: [DN] VA[LENTINIANVS?]

  • 8285/2106: Illegible, but possibly a winged Victory with Dea Roma. Copper alloy, weight 2.6 g, diameter 17 mm.

  • 8265/2017: Illegible, but probably fifth-century numus. Copper alloy, weight 2.4 g, diameter 16 mm.

The only readable coin comes from Area 3 and was found in the abandonment layer that seals the spoliation trench.

  • 8265/3001: Antoninianus of Probus (276–282), Romae aeternae series. Obverse: IMP(ERATOR) PRO BVS AVG(VSTVS), emperor facing left with a radiate crown; reverse: ROMAE [AETERNAE?] with frontal view of a temple. Copper alloy, weight 2.7 g, diameter 28.1 mm.

The lack of proper deposits associated with the use and abandonment of the main structures makes it difficult to interpret the development of the site over the long run. Nonetheless, the 2012–13 excavations, modest as they were, confirmed the sequence of construction at the site: a first Roman phase represented by the aqueduct, two late antique phases (one dated to ca. 500 and the other to the late sixth century) that correspond to the basilica and its ancillary buildings (perhaps including the undatable structure on Area 1), and an early medieval (ninth century) phase of dismantling, abandonment, and possible prison reuse.

Despite the lack of a previous early Roman settlement at the site, the late antique complex of Casa Herrera was not built into a virgin landscape. From a broader perspective, the territory of Mérida had been centuriated and distributed in its entirety during the Augustan period.32 But, from a narrower angle, the complex was built on that hill because of the aqueduct and the road that had been there, probably since the Flavian period.

The importance of the aqueduct in the location of the Casa Herrera complex is twofold. First, it provided good quality building material on-site. The vault that sealed the conduit (which was accessible from ground level) was made of large voussoirs bound with mortar, and considering the mismatched and reused marble bases and columns of the basilica, it would not be surprising if the vault had been dismantled and recycled for the construction. Second, the dismantling of the vault gave access to the aqueduct’s water: the chronological overlap between the construction of the basilica (ca. 500) and the eventual siltation of the conduit (during the sixth century) suggests that water was still flowing when the vault was dismantled. By that time, it is unlikely that the aqueduct still fed the city, so usurping unkept public infrastructure was probably not much of an issue.33 In fact, the structures of Area 2 are built in parallel but beyond the aqueduct “exclusion zone” defined in Roman law, so it is possible that the city council had a say in the matter.34 Considering the large amount of water consumed in mortar construction (and there is a lot of mortar in the walls and floors of the basilica), this water would have been an asset at that stage.35 Then, once the construction was over, the aqueduct could have been turned into a water point for the community that settled on the site.

The basilica existed not as an isolated monument in the countryside but as part of a building complex, of which we have only a very partial view. There is little we can say about the razed wall of Area 1 other than it is pre-seventh century and most likely coeval with the first phase of the basilica. The same can be said about B1 and B3 in Area 2: their fifth-century chronology in an otherwise and until then open field suggest that they were built together with the basilica as part of a larger complex. In any case, the ongoing burials in the basilica and the extension of B1a with B1b would suggest that the complex was constantly in use during the sixth century. Later on, B3 was obliterated by the construction of B4 and B5, which have been dated to the sixth century, but because these are perfectly aligned with B1 and its extension, it is very likely that B1 was standing and in use during this second construction phase. This most likely coincided with the second phase of the basilica and the construction of the main building of Area 1. Furthermore, surface finds and satellite photography suggest that the buildings of Areas 1 and 2 formed a larger, L-shaped structure. This was not necessarily a single construction, though, but rather a series of buildings set along those main axes as a late-sixth century complex (Figure 33), and further excavations will help clarify this proposal.36

Figure 33.

Reconstruction of the Casa Herrera complex in the late sixth century, based on surface marks.

Figure 33.

Reconstruction of the Casa Herrera complex in the late sixth century, based on surface marks.

Close modal

The existence of these buildings next to the basilica suggests that Casa Herrera was a monastic complex, or at least that it became one during the second late antique phase.37

The presence of the basilica is enough evidence to rule out the identification of Casa Herrera as a village or a farmstead: post-Roman peasant communities (especially at this early date) were not focused on churches or parishes, which represent the presence of an urban, elite power unrelated to the type of peasant communities that have been identified elsewhere in the Peninsula.38 Funerary basilicas, furthermore, belong to urban burial traditions rather than rural ones, which by contrast favored row-grave cemeteries.39 The ancillary structures (especially B1) do have parallels in coeval, storage structures from peasant villages, but we must see in these an example of rural architecture, not something specific to peasant communities.

In its first late antique phase, the site might have been part of an aristocratic estate, on which a private funerary complex was built (perhaps linked to the familia of Asella or Sabinianus the charioteer, both named in the inscriptions), comprising a basilica, B1, and B3.40 The large size of the nave (nearly 190 m2) would have been part of the design, to allow for more burial space. It could, equally, have been built by the bishops (who were, in any case, landholding aristocrats) in order to keep a close eye on dispersed rural populations, adding a parochial purpose to the basilica without necessarily linking it to a specific village-like settlement.41

While the layout and structures at the site do not fit with later perceptions of landholding monasteries, recent research into the archaeologies of early monasticism in the West unequivocally underlines that, at the time of the construction of Casa Herrera, no firmly established building or settlement type could be labeled as “monastic.” In fact, it is clear from the sources that any religious complex could have an associated community, which includes pilgrimage sites, hospitals, and even episcopal complexes. Private funerary complexes with a secluded religious community also fall within this category. Communal monasticism in the fifth and sixth centuries was in a phase of experimentation, and in the West this was strongly promoted by urban aristocrats; this first phase could be one of these private, unregulated religious communities that acted as monasteries outside direct ecclesiastical control.42

The phasing is key here. There is no reason to suggest that the late-fifth-century basilica and the late-fifth-century surrounding buildings were not part of the same construction effort.43 Such a predesigned religious complex would fit perfectly as either a private or an episcopal foundation, following the patterns visible in North African monasticism (a movement that had a brief but notable influence in the Iberian Peninsula).44

This chronological sequence would also fit the historical context. In Mérida in particular, the late fifth century was a period of monumentalization after the civil wars and the Gotho-Suevic conflict. This is famously exemplified not only by the inscription that commemorates the reconstruction of the walls and the bridge by the Visigothic dux and the bishop but also by the second phase of the Saint Eulalia suburban complex and the construction of the rural basilica of San Pedro de Mérida.45 It is this political context that can explain the construction of Casa Herrera as an attempt to reinforce the control of the territorium of the city by the urban elites.

This argument for a private monastic enterprise is further supported by the nature of the second phase of the site. The expansion of the basilica in the late sixth century, the creation of new cult spaces, and the addition of a baptistery are all clear signs of direct episcopal control (only bishops could do baptisms). The expansion of episcopal power into the rus was done to consolidate the church’s control, integrating into the urban ecclesiastical hierarchy both rural populations and the various, private monastic enterprises that had emerged in the previous decades.46 This happens also in a period when the church is first seen directly managing rustic properties, many of which were obtained through wills and donations handing over properties to be managed by monastic communities.

The spatial organization of the second phase may be further evidence for its monastic nature. The ancillary buildings of the site do not only appear to form an L-shape; they also open toward the inside, away from the road. There is not enough evidence to speak of a cloister, and the date is also too early to label it as such (perhaps it is better to see this area as a half-enclosed courtyard), but there is, in any case, a separation between the access to the site (on the aqueduct road) and the buildings that were for the community. This separation already provides the degree of privacy and seclusion that is visible in other monastic structures of this period (in fact, this separation of public and private areas is one of the defining characteristics of early monastic communities).47 While a physical enclosure (a wall or fence) circling the site has not been identified, the visual separation between the inward-looking, L-shaped complex and the open road might have been enough to highlight this separation of spaces. During the second phase, moreover, a second entrance was added to the basilica, an entrance to the corridor that led to the baptistery and that gave an alternative access to the basilica. These could have been, hypothetically, an entrance for the monastics (entering from the north, from the open courtyard formed by the ancillary buildings, and around the counter apse) and another for the lay public (which would have run along the back of the ancillary buildings and directly toward the road). Inside the basilica, they would have been segregated already from the first phase, but the blocking of the intercolumnations in the second phase would have made this separation more evident.

This seclusion was also partly geographic. Casa Herrera is not in the wilderness (the “desert” being a key element in the idea of monasticism) inasmuch as it is in the rural and centuriated hinterland of Mérida, but it is off the main roads. The metaled road of Area 3 was linked to aqueduct maintenance; it was a minor path that branched off the suburban streets of Mérida, not one of the main highways.

It should be noted that the very experimental nature of monasticism in this period means that comparisons are of little use. Francisco Moreno’s detailed survey clearly shows this.48 In his catalogue, all but one of his archaeologically identified, fifth-century monasteries are urban or suburban, and the one rural example (Cap des Port in Menorca) consists of a cluster of buildings abutting a baptismal basilica with associated production sites.49 Other sixth-century sites of probable monastic nature and certain chronology, like Punta de l’Illa (Valencia) or the more recently excavated site of Els Altimiris (Lérida), on the contrary, share with Casa Herrera a number of separated buildings surrounding a basilica.50 Neither of these sites has a clear public/private division, but, on the other hand, they were located in far more remote locations with fewer lay visitors (and the lack of baptismal pools would indicate that they were not associated with a lay, rural population). The territory around Mérida was, in the sixth and seventh centuries, full of monasteries known through the written sources, but other than generic mentions of to storage rooms, gardens, cells, and a monastic school, these sources are not particularly helpful.51 Of course, it is from the seventh century onward when we find the most famous monastic sites of the early Middle Ages. But of those known archaeologically, many of them have completely different layouts (Los Hitos), and some of them are not fully known (Guarrazar) or are from much later (El Trampal, Melque).52

Lastly, we should mention the final transformations and abandonment of the site. The dismantling of the basilica and its transformation into a possible prison took place during the ninth/tenth century (as dated by the handwriting), but none of the domestic ceramic types identified in the building of Area 1 can be dated beyond the ninth. The correlation between these two events is not clear, but it is likely that the basilica and the building were dismantled at the same time. If the pottery under the debris is any indication, then the sixth/seventh-century building of Area 1 was used into the ninth century. By that period, the aqueduct had been backfilled long ago, but it is probable that the buildings of Area 2 were still in use, or that new buildings were built up there (B2). This last phase of Casa Herrera can be compared to the Minorcan site of Son Bou and to the basilica of El Gatillo (Cáceres), both of which were partially dismantled and reoccupied without religious purpose in the Umayyad period, the former by the late eighth century but the latter in the late ninth.53

It is thoroughly documented that monastic foundations continued to be active in the Umayyad period and that the conquest did not cause the end of monasticism. Considering this, Casa Herrera might have continued to function as a monastery into the ninth century. It is impossible to determine if there is a causal link between the abandonment, the quarrying, and the “prison” phase, but it is clear that these events all took place during the ninth century. Of course, in the early ninth century the city of Mérida was in political turmoil with constant local (Christian) rebellions against the increasingly stronger Umayyad authorities.54 The end of Casa Herrera as a Christian site does not necessarily imply an episode of religious repression, but it does reflect the urban political conflict and the turning tide for the local Christian aristocracy and the episcopal elites.

For almost a century, Casa Herrera has been a point of reference in what was once termed early Christian archaeology as one more example of basilical architecture and late antique burial customs. Recent excavations around the basilica demonstrate that the site was not just an isolated monument but a larger building complex with a rich history closely related to the old Roman capital of Hispania. From this perspective, Casa Herrera has much more to offer, and it can become a paradigmatic site to understand the development of both rural settlements in the late antique period and the development of early monasticisms in the West, and how these were connected to episcopal power. Its location outside a main urban center like Mérida adds to its relevance and explains its continuity. The site can, therefore, be used as a point of comparison not only for other cities in Visigothic Spain but also for other peri-urban sites of the post-Roman western Mediterranean.

Our interpretation of the site as monastic remains hypothetical (especially considering the limited area of the 2007–2013 campaigns), and it will only be confirmed with further excavations. Still, it is possible to re-create a full interpretative sequence for Casa Herrera, one in which it is established as a private funerary basilica managed by a religious community, one of the types of experimental and yet undefined monasticisms that dominated the late fifth century, established by a member of the urban elite. Subsequently, during the sixth century, the site and its community was likely transferred to the church of Mérida, the type of donation that wealthy families are seen to be doing in contemporary sources like the Lives of the Fathers of Mérida. The church, in turn, used Casa Herrera as one of its focal points to extend its control over the rural populations, as the presence of the baptistery would demonstrate. At any rate, it is almost certain that the site’s second late antique phase of construction was linked to a better-defined monastic community. Even if in that period there were still no standardized layout or form, the basic principles of exclusion and seclusion that inherently define monastic communities would potentially be there. The site ceased to function as a Christian complex after the Umayyad invasion, although this development most likely did not occur in the immediate aftermath of the 711 invasion but rather during the political turmoil of the ninth century.

In terms of material culture, Casa Herrera’s assemblage matches the sequence of the many urban excavations of Mérida dated to this period. The types and the fabrics clearly show a deep interrelation between the city and the site in a way that rural peasant communities elsewhere in the Iberian Peninsula do not. Casa Herrera might have been located in the countryside but it was part of Mérida’s urban world.

Overall, Casa Herrera is a unique example of the archaeology of the late antique and early medieval longue durée, and even if it is located in the countryside, it is a perfect reflection of the evolution of the city of Mérida. With a chronology that ranges from the early Roman into the Islamic period, it is a site that improves our understanding of the privatization and dismantling of public infrastructure in Late Antiquity, exemplifies the development of new nucleated rural settlements (potentially through monasticism), and illustrates the expansion of ecclesiastical power beyond its urban hubs. Many questions still remain unanswered, but we hope that Casa Herrera can feature in more discussions on urban-rural relationships in the post-Roman West.

This paper was completed as part of a María Zambrano project (NAHR - Nuevas Aproximaciones a la Hidráulica Romana), funded by the European Union-NextGenerationEU. We would like to thank Miguel Alba (Consorcio de Mérida) above all for the opportunity to dig at Casa Herrera and for his constant support and help during the field seasons (and beyond). We also need to thank Carlos Cabrera, Pedro Dámaso Sánchez, Gilberto Sánchez, Sara Rodríguez, Luis Hidalgo, and Rafael Sabio for their help at different stages of the project. The excavation would not have been possible without the funding of the Consorcio de Mérida, the Craven Committee, the Meyerstein Fund, Lincoln College, and the Spanish Foundation for Science and Technology (FECYT), or without the students and archaeologists who dug with us. Lastly, the editors, Louise Blanke, Harry Lidgley, and Carlos Tejerizo have read, commented on, and improved earlier versions of this paper. All errors and shortcomings remain our own.

1.

Daniel Osland, “Urban Change in Late Antique Hispania: The Case of Augusta Emerita” (PhD diss., University of Cincinnati, 2011); Osland, “Text and Context: Patronage in Late Antique Mérida,” Studies in Late Antiquity 3.4 (2019): 581–625; Isaac Sastre de Diego, Mérida capital cristiana: De Roma a Al-Andalus, Cuadernos Emeritenses 41 (Mérida: Museo Nacional de Arte Romano, 2015); Isaac Sastre de Diego and Miguel Alba Calzado, “Bajo la protección de la mártir: Mérida en la antigüedad tardía (siglos IV–VII),” in Urban Transformations in the Late Antique West: Materials, Agents, and Models, ed. André Carneiro, Neil Christie, and Pilar Diarte (Coimbra: University of Coimbra, 2020), 203–30.

2.

Tomás Cordero Ruiz and Isaac Sastre de Diego, “El yacimiento de Casa Herrera en el contexto del territorio emeritense (siglos IV–VIII),” in Espacios urbanos en el occidente mediterráneo (s. VI–VIII), ed. Alfonso García (Toledo: Toletum Visigodo, 2010), 211–18, at 211.

3.

Sastre de Diego, Mérida capital cristiana, 133.

4.

Luis Caballero Zoreda and Thilo Ulbert, La basílica paleocristiana de Casa Herrera en las cercanías de Mérida (Badajoz), Excavaciones arqueológicas en España 89 (Madrid: Ministerio de Educación y Ciencia, 1976), 73–104.

5.

Isaac Satre de Diego and Rafael Sabio González, “Escultura arquitectónica y litúrgica en la basilica tardoantigua de Casa Herrera (Badajoz, Mérida). Novedades aportadas desde la revisión de los fondos del MNAR,” Anas 33 (2020): 291–317.

6.

The marble quarries of Estremoz (Portugal) were opened under Augustus, and according to ongoing research by André Carneiro (University of Évora, personal comment), there was still some degree of quarrying in the Visigothic period.

7.

Sastre de Diego, Mérida capital cristiana, 130, fig. 36.

8.

Isaac Sastre de Diego, Los altares de las iglesias hispanas tardoantiguas y altomedievales, un estudio arqueológico, BAR International Series 2503 (Oxford: Archaeopress, 2013), 424–25, EX5, fig. 233. For other comparanda, see Matthew McCarthy, “Beyond Models and Diffusion, Centers and Peripheries: Religious Art in Roman Africa,” in Roma y las provincias: modelo y difusión, ed. Trinidad Nogales and Isabel Rodà (Mérida: Museo Nacional de Arte Romano, 2011), 439–48.

9.

Caballero Zoreda and Ulbert, La basílica paleocristiana. See José Luis Ramírez Sádaba and Pedro Mateos Cruz, eds., Catálogo de las inscripciones cristianas de Mérida, Cuadernos Emeritenses 16 (Mérida: Museo Nacional de Arte Romano, 2000), no. 18. Three other inscriptions were found in the basilica (Sádaba and Mateos, Catálogo de las inscripciones, 2, 51, 52) and are of similar chronology.

10.

Javier Martínez Jiménez, Isaac Sastre de Diego, and Carlos Tejerizo García, The Iberian Peninsula between 300 and 850: An Archaeological Perspective, Late Antique and Early Medieval Iberia 6 (Amsterdam: Amsterdam University Press, 2018), 168–78.

11.

Sastre de Diego, Mérida capital cristiana, 189; Javier Martínez Jiménez and Patricia González Gutiérrez, “Knowledge and Specialised Trades in the Late Antique West: Medicine vs Engineering,” Journal of Late Antique Religion and Culture 11 (2017): 38–58.

12.

Caballero Zoreda and Ulbert, La basílica paleocristiana.

13.

Carmen Barceló Torres, “Escritos árabes en la basílica paleocristiana de Casa Herrera (Mérida),” Madrider Mitteilungen 43 (2002): 299–315. Our English translation from Barceló’s Spanish version. We thank Edward Zychowicz-Coghill for his assistance with the Arabic text.

14.

Thilo Ulbert, “Nachunterschungen im Bereich der frühchristlichen Basilika von Casa Herrera bei Mérida,” Madrider Mitteilungen 32 (1991): 185–207.

15.

Cordero Ruiz and Sastre de Diego, “El yacimiento de Casa Herrera,” 212.

16.

Mercedes Gómez de Segura, Pedro Dámaso Sánchez Barrero, Nuria Sánchez Capote, and Isaac Sastre de Diego, “Las conducciones romanas de Mérida: nuevos datos para su conocimiento,” in Aquam perducendam curavit: captación, uso y administración del agua en las ciudades de la Bética y el occidente romano, ed. Lázaro Lagóstena Barrios et al. (Cádiz: University of Cádiz, 2011), 129–46.

17.

Burials were identified both by Serra i Rafols and by Thilo Ulbert and Luis Caballero Zoreda, and the anthropometric analyses were published in detail by the latter two (Caballero Zoreda and Ulbert, La basílica paleocristiana, 237–48), but it has not been possible to obtain a permit to obtain samples from the burials to carry out stable isotope analyses.

18.

Sastre de Diego, Los altares, 88–95.

19.

Sastre de Diego, M1, RJ1, P6.

20.

Alfonso Vigil-Escalera, “Primeros pasos hacia el análisis de la organización interna de los asentamientos rurales en época visigoda,” in La investigación arqueológica de la época visigoda en la Comunidad de Madrid, ed. Jorge Morín de Pablos, Zona Arqueológica 8 (Alcalá de Henares: Museo Arqueológico Regional, 2007), 367–72.

21.

Elena Sánchez López and Javier Martínez Jiménez, Los acueductos de Hispania: construcción y abandono (Madrid: Fundación Juanelo Turriano, 2016), 254–57.

22.

Sinter limescale was visible in the abandonment layer inside the conduit, though mixed in the clayey matrix—perhaps detached from other sections of the conduit further upstream.

23.

Miguel Alba Calzado and Santiago Feijoo Martínez, “Pautas evolutivas de la cerámica común de Mérida en épocas visigoda y emiral,” in Cerámicas tardorromanas y altomedievales en la Península Ibérica: ruptura y continuidad, ed. Luis Caballero Zoreda, Pedro Mateos Cruz, and Manuel Retuerce Velasco, Anejos de Archivo Español de Arqueología 28 (Madrid: Consejo Superior de Investigaciones Científicas, 2003), 483–504; Miguel Alba Calzado and Sonia Gutiérrez Lloret, “Las producciones de transición al mundo islámico: el problema de la cerámica paleoandalusí (siglos VIII y IX),” in Cerámicas hispanorromanas: un estado de la cuestión, ed. Darío Bernal Casasola and Albert Ribera Lacomba (Cádiz: Universidad de Cádiz, 2008), 585–613; Macarena Bustamante Álvarez, “Las cerámicas comunes altoimperiales de Augusta Emerita,” in Cerámicas hispanorromanas II: producciones regionales, ed. Darío Bernal Casasola and Albert Ribera Lacomba (Cádiz: University of Cádiz, 2012), 407–33. For comparison, see Raúl Aranda González, “Una aportación al conocimiento de las producciones cerámicas de época visigoda: el conjunto cerámico de la parcela R3 de la Vega Baja (Toledo),” Espacio, Tiempo y Forma: serie I 6 (2015), 381–450, and more generally, Iñaki Martin Viso et al., eds., Cerámicas altomedievales en Hispania y su entorno (siglos V–VIII d.C.) (Valladolid: Glyphos, 2018). For the Umayyad period, cf. José Cristobal Carvajal López and Peter Day, “Cooking Pots and Islamicization in the Early Medieval Vega of Granada (Al-Andalus, sixth to twelfth centuries),” Oxford Journal of Archaeology 32.4 (2013): 433–51, and Victoria Amorós Ruiz, “Entre ollas y marmitas: una reflexión sobre la producción cerámica entre los siglos VII y IX en el sureste de la península Ibérica,” Arqueología y Territorio Medieval 27 (2020): 11–36.

24.

See ahead for the dating of the ceramic building materials.

25.

Francesca Grassi and Juan Antonio Quirós Castillo, “Arqueometría y arqueología de la cerámica medieval en España: balance crítico y perspectivas de futuro,” in Arqueometría de los materiales cerámicos de época medieval en España, ed. Francesca Grassi and Juan Antonio Quirós Castillo (Bilbao: Universidad del País Vasco, 2018), 23–38.

26.

Martínez Jiménez et al., Iberian Peninsula, 232–33.

27.

Caballero Zoreda and Ulbert, La basílica paleocristiana, 229–31.

28.

Bustamante, “Las cerámicas comunes,” 421.

29.

Paul Reynolds et al., “Almagro 54,” in Roman Amphorae: A digital resource [data-set], (York: Archaeology Data Service, 2014), https://doi.org/10.5284/1028192.

30.

This is a bowl that imitates either Hayes 87A or 61A. Juan Ángel Paz Peralta, “Las producciones de terra sigillata hispánica intermedia y tardía,” in Cerámicas hispanorromanas: un estado de la cuestión, ed. Darío Bernal Casasola and Albert Ribera Lacomba (Cádiz: Universidad de Cádiz, 2008), 497–540.

31.

Ruth Pliego Vázquez, “The Circulation of Copper Coins in the Iberian Peninsula during the Visigothic Period: New Approaches,” Journal of Archaeological Numismatics 5/6 (2015–16): 125–60; Ruth Pliego Vázquez, “Rethinking the Minimi of the Iberian Peninsula and Balearic Islands in Late Antiquity,” Journal of Medieval Iberian Studies 12, no. 2 (2020): 125–54.

32.

Dio Cassius, Historia romana 53.26.1, translation from Dio Cassius: Roman History, ed. and trans. Ernest Cary, Loeb Classical Library 83 (Cambridge, MA: Harvard University Press, 1917); Agennus Urbicus, De controversiis agrorum, ll, 9–15, ed. Carolus Thulin, Corpus Agrimensorum Romanorum 1.1 (Leipzig: Teubner, 1913). Cf. Enrique Ariño Gil and Josep María Gurt Esparraguera, “Catastros romanos en el entorno de Augusta Emerita,” in Les campagnes de la Lusitanie romaine: Occupation du sol et habitats, ed. J.-G. Gorges and M. Salinas (Madrid: Casa de Velázquez, 1994), 45–66.

33.

Javier Martínez Jiménez, Aqueducts and Urbanism in Post-Roman Hispania, Gorgias Studies in Classical and Late Antiquity 26 (Piscataway, NJ: Gorgias, 2019), 197–201.

34.

Frontinus, De Aquis 2.127, Stratagems: Aqueducts of Rome, ed. and trans. Charles E. Bennett and Mary B. McElwain, Loeb Classical Library 174 (Cambridge, MA: Harvard University Press, 1925); Codex Theodosianus 15.2.1, The Theodosian Code and Novels and the Sirmondian Constitutions: A Translation with Commentary, Glossary, and Bibliography, trans. Clyde Pharr, (Princeton: Princeton University Press, 1952) = Codex Iustinianeus 11.43.1.

35.

Javier Martínez Jiménez, “Water in Ancient Construction,” in The Role of Water in Production Processes in Antiquity, Panel 3.19, ed. Elena Sánchez López, Archaeology and Economy in the Ancient World 21 (Heidelberg: Propylaeum, 2020), 11–20.

36.

The basilica is built following its own liturgic alignment: César González García and Juan Antonio Belmonte, “The Orientation of Pre-Romanesque Churches in the Iberian Peninsula,” Nexus Network Journal 17 (2015): 353–77.

37.

Cf. Tomás Cordero and Bruno Franco, “El territorio emeritense durante la Antigüedad Tardía y la Alta Edad Media,” in Visigodos y omeyas: el territorio, ed. Luis Caballero Zoreda et al., Anejos Archivo Español de Arqueología 61 (Madrid: Consejo Superior de Investigaciones Científicas, 2012), 147–69 at 157–58.

38.

Carlos Tejerizo García, Arqueología de las sociedades campesinas en la Cuenca del Duero durante la Primera Alta Edad Media (Bilbao: University of the Basque Country, 2017), 220. But cf. the approach for the Balearic Islands: Catalina Mas Florit, Miguel Ángel Cau Ontiveros, and Silvia Alcaide, “Buildings of Faith: Early Christianity in the Countryside of the Balearic Islands,” Journal of Roman Archaeology 33 (2020): 271–90.

39.

Andreia Arezes, O mundo funerário na Antiguidade Tardia em Portugal: as necrópoles dos séculos V a VIII (Lisbon: Centro de Investigação Transdisciplinar «Cultura, Espaço e Memória», 2017), 182–87; Martínez Jiménez et al., Iberian Peninsula, 213–16.

40.

See note 9.

41.

Cf. Mas Florit et al., “Buildings of Faith,” 286.

42.

Francisco Moreno Martín, “La configuración arquitectónica del monasterio hispano entre la tardoantigüedad y el alto medievo: balance historiográfico y nuevas perspectivas,” Anales de Historia del Arte, extraordinary issue (2009): 199–217. Cf. Albrecht Diem, “Merovingian Monasticism: Voices of Dissent,” in The Oxford Handbook of the Merovingian World, ed. Bonnie Effros and Isabel Moreira (Oxford: Oxford University Press, 2020), 320–43.

43.

Sastre de Diego, Mérida capital cristiana, 126–30.

44.

Sastre de Diego, 175, 197. Like Abbot Nanctus—Vitas Sanctorum Patrum Emeretensium 3.2, translation in Lives of the Visigothic Fathers, trans. A. T. Fear, Translated Texts for Historians 26 (Liverpool: Liverpool University Press, 1997), 55.

45.

Osland, “Text and Context”; Sastre de Diego and Alba Calzado, “Bajo la protección de la mártir”; Sádaba and Mateos, Catálogo de las inscripciones, 10.

46.

Moreno Martín, “La configuración arquitectónica,” 169; Sastre de Diego, Mérida capital cristiana, 173–77; cf. Merovingian Gaul: Pascale Chevalier, “Merovingian Religious Architecture: Some New Reflections,” in The Oxford Handbook of the Merovingian World, ed. Bonnie Effros and Isabel Moreira (Oxford: Oxford University Press, 2020), 657–92.

47.

Francisco Moreno Martín, “Espacios públicos y espacios de uso común en los primeros monasterios hispanos (siglos V–X),” in Oeuvrer pour le salut: Moines, chanoines et frères dans la péninsule Ibérique au Moyen Âge, ed. Amélie de las Heras, Florian Gallon, and Nicolas Pluchot (Madrid: Casa de Velázquez, 2019), 165–86.

48.

Francisco Moreno Martín, La arquitectura monástica hispana entre la Tardoantigüedad y la Alta Edad Media, BAR International Series 2287 (Oxford: Archaeopress, 2011). Cf. Mateu Riera Rullán, El monacat insular de la Mediterrània Occidental: el Monestir de Cabrera (Balears, segles V–VIII), Studia archaeologiae christianae 1 (Barcelona: Institut Català d’Arqueologia Clàssica, 2017), 76–77.

49.

Moreno Martín, La arquitectura monástica, 196; Mas Florit et al., “Buildings of Faith.”

50.

Moreno Martín, La arquitectura monástica, 207; Marta Sancho Planas, “Evidencias arqueológicas de un monasterio de los siglos VI–VII en el Prepirineo catalán: Santa Cecilia de Els Altimiris,” in Monastères et couvents de montagne: Circulation, réseaux, influences au Moyen Âge (Paris: Comité des travaux historiques et scientifiques, 2019).

51.

Vitas Sanctorum Patrum Emeretensium 2.8–14 (Fear, Lives of the Visigothic Fathers, 52–53). Cf. Isaac Sastre de Diego, Tomás Cordero Ruiz, and Pedro Mateos Cruz, “Territorio y monacato emeritense durante la antigüedad tardía,” in Monasteria et territoria: elites, edilicia y territorio en el Mediterráneo medieval (siglos V–XI), ed. Jorge López Quiroga et al., BAR International Series 1720 (Oxford: Archaeopress, 2007), 141–62.

52.

Moreno Martín, La arquitectura monástica. Cf. Luis Caballero Zoreda and Francisco Moreno Martín, “Balatalmelc, Santa María de Melque: un monasterio del siglo VIII en territorio Toledano,” in Lo que vino de Oriente: horizontes, praxis y dimensión material de los sistemas de dominación fiscal en Al-Andalus (ss. VIII–IX), ed. Xavier Ballestín and Ernesto Pastor, BAR International Series 2525 (Oxford: Archaeopress, 2013), 182–214; Juan Manuel Rojas Rodríguez-Malo, “El primer año de trabajos en Guarrazar: la confirmación de un importante yacimiento arqueológico,” in VII Jornadas de Cultura Visigoda (Guadamur: Ayuntamiento de Guadamur, 2013), 37–66; also, with caution, Jorge Morín de Pablos and Isabel Sánchez Ramos, eds., Los Hitos: Arisgotas—Orgaz, Toledo—De palacio a panteón visigodo (Madrid: self-published, 2015).

53.

Mas Florit et al., “Buildings of Faith,” 282; Luis Caballero Zoreda and Fernando Sáez Lara, “La iglesia de El Gatillo de Arriba (Cáceres): apuntes sobre una iglesia rural en los siglos VI al VIII,” in El siglo VII frente al siglo VII: arquitectura, ed. Luis Caballero Zoreda, Pedro Mateos Cruz, and Mª Ángeles Utrero Agudo, Anejos de Archivo Español de Arqueología 51, Visigodos y omeyas, 4: Mérida, 2006 (Madrid: Consejo Superior de Investigaciones Científicas, 2009), 155–84.

54.

Martínez Jiménez et al., Iberian Peninsula, 276–82.