Gaulish terra sigillata in Seville. New data, new perspectives

Estudos & Memórias UNIARQ series (Centre for Archaeology. University of Lisbon) Series coordination: Ana Catarina Sousa Founded by Victor S. Gonçalves 18. VIEGAS, C.; BUSTAMANTE-ÁLVAREZ, M. (eds.) (2021) – South Gaulish sigillata in Southwest Hispania: circulation and consumption. Estudos & Memórias 18. Lisboa: UNIARQ/FL-UL. 244 pp. Cover: Catarina Viegas and Rui Roberto de Almeida. Photograph of the cover: South Gaulish sigillata from the Guadiaro’s wreck, by Macarena Bustamante-Álvarez. Typesetting and layout: Rui Roberto de Almeida Printed and bound by Sersilito – empresa gráfi ca, Lda. 300 copies ISBN: 978-989-53453-4-2 / Depósito legal: 0000000000000 DOI: https://doi.org/10.51427/10451/50262 Copyright © 2021 the authors. Any reproduction of text and images is prohibited, without the express authorization of the author, under the terms of the current law, namely DL 63/85, of March 14, with subsequent amendments. Lisboa, 2021 This work was fi nanced by Portuguese funds through FCT - Fundação para a Ciência e a Tecnologia in the framework of the projects UIDB/00698/2020 and UIDP/00698/2020. This book was also fi nanced by the following projects: 'Corpus Vasorum Hispanorum. Analisis tipológico, cronológico y prosopográfi co de los sigilla en terra sigillata hispánica a partir de los centros consumidores. Parte I: Lusitania' (PGC2018-093478-A-I00) fi nanced by Convocatoria Proyectos de Excelencia – Plan Estatal de Generación de Conocimiento – Ministerio de Ciencia, Innovación y Universidades – España. And by the Project 'Aplicación de nuevas tecnologías para el desarrollo del Corpus Vasorum Hispanorum. Una herramienta para análisis tipológico, cronológico y prosopográfi co de los sigilla en Terra Sigillata Hispánica' fi nanced by Fundación BBVA in the context of the Program Logos Ayuda a la Investigación en Estudios Clásicos 2019. CONTENTS Prologue. Overview on the study of terra sigillata in Portugal Carlos Fabião ....................................................................................................................................................... 5 Foreword of the editors Catarina Viegas, Macarena Bustamante-Álvarez ............................................................................................ 13 The Mainz internet database of Names on Terra Sigillata (samian ware), Portugal and Spain and the collaborative samian research network Geoffrey B. Dannell, Allard W. Mees ............................................................................................................... 17 South gaulish terra sigillata found in Bracara Augusta (North Portugal) Rui Morais ............................................................................................................................................................ 27 South Gaulish terra sigillata in the forum of Aeminium (Coimbra, Portugal): a decisive component of the dating process Pedro C. Carvalho, Ricardo Costeira da Silva ................................................................................................. 47 South Gaulish terra sigillata in Ammaia (São Salvador de Aramenha, Marvão): typological and stratigraphical analysis José Carlos Quaresma ........................................................................................................................................ 57 Potters and vessels: consumers, markets and dynamics at play in the circulation of Gaulish terra sigillata: the evidence for the region under the influence of Olisipo (Lisbon) Rodrigo Banha da Silva ...................................................................................................................................... 75 Gaulish terra sigillata trade in Augusta Emerita (Mérida). Data from the sigilla in the Museo Nacional de Arte Romano (MNAR) Macarena Bustamante-Álvarez, José M. Jerez Linde ..................................................................................... 93 Preliminary results on the Gaulish terra sigillata from the fish-salting production centre at Tróia Ana Patrícia Magalhães, Inês Vaz Pinto, Patrícia Brum ................................................................................. 117 The circulation of South Gaulish sigillata in southern Lusitania: Alentejo region and the Algarve (Portugal) Catarina Viegas, Carolina Grilo, Catarina Alves ............................................................................................ 131 The Gaulish terra sigillata in Baelo Claudia. Analysis of the contexts from the 'Puerta de Carteia' Macarena Bustamante-Álvarez, Dario Bernal-Casasola, Alicia Arévalo González .................................. 151 Gaulish terra sigillata in Seville. New data, new perspectives Elise Arnold, Enrique García Vargas .............................................................................................................. 177 First approach on the distribution of amphorae from Gaul in Lusitania Rui Roberto de Almeida, Vitor Filipe ................................................................................................................ 211 SOUTH GAULISH SIGILLATA IN SOUTHWEST HISPANIA: CIRCULATION AND CONSUMPTION GAULISH TERRA SIGILLATA IN SEVILLE. NEW DATA, NEW PERSPECTIVES Elise Arnold [email protected] Archéologie Alsace Enrique García Vargas [email protected] University of Seville Abstract This paper deals with the study of three sets of Gaulish sigillata from contexts in the Roman port area of Seville. The chronological sequence of the assemblages, dated between the end of the Julio-Claudian and the end of the Flavian period, has allowed the evolution of the consumption patterns of this type of pottery in the city of Seville to be investigated. A statistical analysis of the volume of imported types and the size of the pieces obtained at different chronological moments, when compared to data from various contexts of reference throughout the Empire, have revealed the existence of a series of consumption patterns, and moreover ones specific to different areas within the Roman world. Resumo El presente artículo aborda el estudio de tres conjuntos de cerámica sigilata sudgálica descubiertos en contextos portuarios hispalenses. La sucesiva cronología de los lotes, fechados entre finales de la época julio-claudia y finales de la época Flavia, permitió realizar una aproximación a la evolución de las pautas de consumo de este tipo de cerámica en la ciudad de Sevilla. Un análisis estadístico del volumen de tipos importados y del tamaño de las piezas adquiridas en diferentes momentos cronológicos, confrontado a los datos de varios contextos de referencia a través del Imperio, han puesto de relieve la existencia de una serie de pautas de consumo propias a las diferentes zonas del mundo romano. Keywords Supply patterns; production modules; metrological study; evolution of consumption; marbled sigillata. Palavras-chave Pautas de aprovisionamiento; módulos de producción; estudio metrológico; evolución del consumo; sigilata marmorata. 1. Introduction The study of the ceramic consumption in Hispalis (Seville) between the Late Republic and the end of the Late Antiquity has already achieved a certain tradition concerning amphora production, which has been relatively well studied in recent years (Amores Carredano et al., 2007; García Vargas, 2007, 2009, 2012, 2015, 2016). Further, the repertoires of ceramics imitating imported fine wares, as a reflection of the impact on the local ‘markets’ of Mediterranean imports, have received attention recently too (García Fernández and García Vargas, 2014). However, the importation itself of these fine wares has been the subject of only a few studies (among them Arnold 2015, wich is a previous french version of this chapter), despite the fact that, in recent decades, the stratigraphical information for them has increased in parallel with the excavations brought about by the urban boom of the first decade of the 21st century. In this work, we present a qualitative and quantitative panorama, as broadly as possible, of the importation of Gaulish sigillata to the city as evidenced from three archaeological interventions with their several contexts: namely, i) the excavation of Alemanes St. 25-29, ii) that of the plots 1.1, 1.2 and 1.3 of the modified Plan Especial 12 of San Bernardo Neighbourhood, and iii) of the Patio de Banderas of the Seville Alcázar. In this way we offer data to increase our, at present, scarce knowledge of this imported ceramic class in the valley of the Guadalquivir (Fig. 1). ARNOLD, Elise & GARCÍA VARGAS, Enrique (2021) - Gaulish terra sigillata in Seville. New data, new perspectives. In Viegas, Catarina & Bustamante-Álvarez, macarena (eds.), South gaulish sigillata in SouthweSt hispania. Circulation and ConSumption. Estudos & Memórias, 18. Lisboa: UNIARQ - Centro de Arqueologia da Universidade de Lisboa, p.177-210. https://doi.org/10.51427/10451/50282 178 Gaulish terra sigillata in Seville. New data, new perspectives elise arnolD, enriQue garCÍa vargas 2. Sevillian contexts studied 2.1. 25-29 Alemanes St. In 2006, an area of about 100 square metres was excavated at 25-29 Alemanes St. (Vázquez Paz, 2006), in front of the Patio de los Naranjos of the Cathedral, which in antiquity corresponded to an area outside the walls, close to the historical course of the Guadalquivir river. The stratigraphy of the site begins with two overlapping deposits (SU 184 and 185) that served as a base for a wall composed of vertically set amphorae. The Punic amphorae of Cádiz, of type 7.4.3.3, were broken off below the rim and sunk in the ground to form the core of a mud wall, a kind of structure usual in artisanal and/or port areas. These basal layers contain fragments of amphorae of types 7.4.3.3.3, 9.1.1.1, Pellicer D, LC 67, Guadalquivir ovoid and Dressel 7-11. The Italic black-gloss tableware of the B circle (Lamboglia 1, 1-8, 4 and 33 b forms), the Italian common and cooking wares (Vegas 14 and Luni 1 forms), as well as the thin-walled ware (Mayet I form) date this ‘foundation’ to the amphora wall at the end of the Republican period, immediately before the arrival of the fi rst Italian sigillata, and so around the beginning of the third quarter of the 1st century BC. The abandonment levels of this ‘amphorae-wall’ (SU 179 and 164) show a similar amphorae repertoire, made up of the T.4.3.3., Pellicer D, Lamboglia 2 and Dressel 7-11 forms familiar in the Cadiz and Algeciras bays. A grooved fragment of an amphora from Ibiza (probably an 18 PE) from SU 164 completes the amphorae set that is accompanied by black-gloss (Lamb. 1, 3, 5 and 33b), thin-walled ware of the form Mayet I and imported common ware (pots with inverted and fl at rims), which gives a date similar to the previously mentioned levels, indicating that the construction and abandonment of the ‘amphoraewall’ occurred in a brief period of time, prior to the arrival of the fi rst Italic sigillata. The vase forms in the Turdetan tradition are kitchen wares with bifi d rims, painted urns with triangular- and rectangularsectioned rims and basins with ‘SS’ motifs in red, a set characteristic of the 1st century BC in the city. Seville Fig. 1 – Seville in the Iberian Peninsula. SOUTH GAULISH SIGILLATA IN SOUTHWEST HISPANIA: CIRCULATION AND CONSUMPTION 179 Above the abandonment level of the ‘amphorae-wall’, an irregular wall was built, the construction trench of which presents material largely similar to the previous ones, but now Italic sigillata, thin-walled ware, Peñaflor type ceramics, grey kitchen ware and Italic common ware of the Pompeian red-slip ware type, (sometimes represented by local imitations) are present. Also recovered are Gaulish sigillata of the forms Drag. 18a and 29a, that places the construction of the wall to within the Julio-Claudian period, allowing that the black-gloss ware of this context is probably residual. The Gaulish sigillata of SU 149 is clearly Neronian, and is associated with abundant fragments of Peñaflor type ceramics, especially the Martínez 14 form. Most of the Gaulish sigillata of these contexts comes from these abandonment levels connected with the wall of amphorae. They are recovered from SU 149 and 161 that still follow the angle of the riverbank, but gradually correct its slope. Towards the end of the 1st century AD, a wall on a N-S orientation is constructed over these Julio- Claudian units, of which only the foundations of stone have been documented. 2.2. Neighbourhood of San Bernardo (plots 1.1, 1.2 and 1.3 of the modified Plan Especial 12) At the southeast end of the city, the current San Bernardo neighbourhood is located in the middle of the historic course of the Tagarete river, one of the watercourses that, along with the Guadalquivir, bordered the small terrace (T13) on which the city stood. The 1st century AD was a period when the hydrological stability of these rivers was at its best and when the economic strength of the city was at its greatest. An archaeological intervention took place in 2012 (Valiente de Santis, 2015), in plots 1.1, 1.2 and 1.3 of the modified Special Plan 12 of the San Bernardo suburb between Juan de Mata Carriazo and Guadaira streets, right at the limit of the current San Bernardo suburb. This revealed an interesting operation to dry out and consolidate the area, prone to flooding, which was carried out in the second half of the 1st century AD. This work consisted in the creation of an artificial platform on the dried and stabilised river silts of the Tagarete flood basin. The platform was effected by means of a massive dump (SU 205) whose base was made up of materials of larger size such as tegulae, bricks, common wares and Haltern 70, Dressel 20, Dressel 7-11 or Beltrán IIb amphorae, and whose upper part contained fragments of Gaulish sigillata, thin-walled ware of the forms Mayet, XXXVIII and XLII, lamps with evolved volutes and common wares. The fill levels were topped by a compacted soil (SU 22), which constituted the walking surface of the platform. On this compacted ‘floor’ some very deteriorated structures were constructed, which should have been buildings related to the handling and storage of goods. These had walls of tegulae of around 0.45 m width. By them, one can identify structural alignments related to one other, likely from warehouses or buildings whose roof was apparently of a perishable material since there were no debris recovered. One of the alignments was interpreted as a jetty platform; others looked like channelling dykes or some such means of regulating the stream channel. Isolated and very incomplete, other structures were documented as fragments of laterculi pavements or the rim of a brick well. Most of the Gaulish sigillata that we include in this work from this site were documented in the preparatory landfills, especially in SU 205. 2.3. Patio de Banderas The excavations carried out in the square called the ‘Patio de Banderas’ of the Alcázar of Seville (Tabales Rodríguez, 2015) are part of a research project for an integrated study of these Islamic and Christian royal palaces from an archaeological point of view (including Archaeology of architecture). The Patio de Banderas is a ceremonial esplanade located inside the third of its historical enclosures (the only one preserved today). The archaeological surveys carried out in three campaigns from 2009 to 2014 covered 180 Gaulish terra sigillata in Seville. New data, new perspectives ElisE Arnold, EnriquE GArcíA VArGAs a total area of some 2000 square meters in the centre of the square, and have reached a total depth of some 5 m below its surface level. Traditional historiography attributed to this area a clear religious function prior to its palatial role, as it was supposedly the site since Late Antiquity of the main church of Hispalis. The finding of the marble covering of the tomb of Bishop Honorius (the authenticity of this funerary epigraph has recently been questioned) and the interpretation of a lacus excavated in the 1970s as part of a baptistery were considered to support this hypothesis. The 2009-2014 excavations, however, seem to definitively refute this interpretation: the supposed baptismal font is now placed within a set of similar structures in various buildings of a commercial and artisanal function dated to the Flavian period. These structures, like all those of the Patio de Banderas, are inserted in a complex stratigraphy that began around the 9th century BC, in the Iron Age I, and run on down to the present day, although not all the possible cultural phases are represented in the sequence, as long periods of stratigraphical hiatus were identified. The phases detected in the excavation of the Patio de Banderas are the following: . Phase 1. The oldest human occupation is represented by a series of small pits filled with the remains of ashes, ceramics and organic remains that have been interpreted as community kitchens, then filled with the remains of the activity carried out inside them. Radiocarbon (C14) samples date these structures to the 9th century BC. There already exist wheel-thrown ceramics, among them amphorae that correspond to the Phoenician colonization in the Far West. . Phase 2. After a prolonged hiatus, an important building activity is documented, dated between the end of the 2nd and the beginning of the 1st centuries BC. Partially lost due to the buildings of the later phase that are preserved, the remains of this Republican Phase I have been interpreted as belonging to a building with adobe walls on a stone base, developed around a central patio. The central open space was paved with opus signinum and the columns of the peristyle were made with quadrant-shaped bricks. . Phase 3. In the third quarter of the 1st century BC (50-25 BC) there was built in the southern part of the city a complex of public structures whose dimensions are larger than the space excavated in the Patio de Banderas. The whole we must assume to have been outside the citywalls. The construction of the new buildings eliminated all previous features: a port complex is created, divided into three zones, on a north-south alignment. To the north, is a building given elongated naves, more than 11 m in length and of variable widths. The central zone is a large open space that connects to the north and south sectors. In the southern area, large rooms have been documented that open onto a portico, in this case, a cryptoportico. All the structures are made of opus africanum and opus quadratum with calcarenite ashlar blocks. The function of these buildings is not clear, although the shape of the rooms to the north suggests the possibility that it was a large warehouse, arguably for grain. To the south, rectangular naves are documented, presumably also for storage, although some sumptuary elements allow us to think that reception or religious functions could apply to some of the buildings outside the excavation limits. Towards the change of era (in Augustan-Tiberian times) small modifications were carried out to the building, such as the raising of circulation levels in the northern sector of the complex. . Phase 4. The harbour buildings from the Late Republican period were renovated during the Flavian period. In some areas, the changes are profound: in the northern sector, the ground floors of the warehouses are terraced and some walls are redesigned using brick and, above all, tegula as a construction material for wall repairs. The central area presents some vats for ‘industrial’ activities, related to the abundant use of water and whose construction seems to belong to this period. SOUTH GAULISH SIGILLATA IN SOUTHWEST HISPANIA: CIRCULATION AND CONSUMPTION 181 On the other hand, all the southern areas remained largely unchanged until the beginning of the 3rd century AD, when most of the buildings collapsed. Most of the Gaulish and Hispanic sigillata documented in the excavation in their original contexts of deposition are from this phase. . Phase 5. Between AD 200 and 225, according to the C14 dates along with those obtained from the ceramic imports, the whole complex suffers a traumatic end. No more building was undertaken in this area until the end of the 5th century AD. The remains of debris layers alternate with strata of muddy layers with an aquatic origin, with fluvial-marine components that testify to a high-energy event responsible for the end of the use of this space. . Phase 6. As mentioned, during Late Antiquity, from the end of the 5th century, the space is occupied again. The construction of a large building took place: once again this extends beyond the excavated area in the Patio de Banderas. The floor plan is organized in rooms paved with bricks, around a large porticoed courtyard. The function of this building is not clear, but it was certainly not domestic. A palatial or monastic use is hypothesized. . Phase 7. After the great brick and masonry building at the end of the 5th century, the construction of a new dwelling structure (of which very few remains have survived) is documented. A wall with buttresses and the basis of what could have been a transverse arch indicate that we are in front of a building of some size. . Phase 8. The destruction of this building was followed by the flattening of the site and the beginning of an Islamic suburb of the 10th century, where houses with façades onto a street were excavated. . Phase 9. Structures were documented corresponding to an important suburb of the Caliphal-taifa period. What exists in the excavated area is articulated around a north-south street, following the previous Caliphate period layout. On both sides of it are a series of central courtyard houses, sometimes with small water tank. This neighbourhood seems to have started at the beginning of the 11th century and continues until the end of the century, when the first enclosure of the fortress (‘Alcazar’) was built. The street has at least four repaving layers. It is noteworthy that the waste-water treatment was effected by blind wells, except in the last repaving phase which included a large sewer that collected domestic waste. . Phase 10. The demolition of the taifa neighbourhood and the flattening of the land for the construction of enclosure I of the Alcázar. . Phase 11. Corresponds to the works of infrastructure, drainage system and pavements of the square in modern times. . Phase 12. Reflects the works that during the Contemporary period have turned this space into what we currently know as the Patio de Banderas. 3. Questions and methodology The standardised nature of terra sigillata production, when a limited number of forms were repeated almost identically in the various workshops of the Roman Empire, has raised the question of the 188 Gaulish terra sigillata in Seville. New data, new perspectives ElisE Arnold, EnriquE GArcíA VArGAs 1 4* 2 3* 5 65 cm0 Fig. 5 – San Bernado: Drag. 29 fonts (* pieces in marbled sigillata). SOUTH GAULISH SIGILLATA IN SOUTHWEST HISPANIA: CIRCULATION AND CONSUMPTION 189 1 2 3 4 8 7 * 5 6 5 cm0 Fig. 6 – San Bernardo: decorated forms. 1-3 and 6: Drag. 30; 4: Hermet 7 ?; 5: Knorr 78; 7-8: Drag. 37 (* piece in marbled sigillata). 190 Gaulish terra sigillata in Seville. New data, new perspectives ElisE Arnold, EnriquE GArcíA VArGAs corresponding to the transition between the two dynasties, such as Felix ii and Marsus, as well as a stamp attributable to Calvus ii and another of difficult attribution (M-II). The last, with its characteristic letters, nevertheless appears well represented in the consuming centres and could therefore have been dated by Hartley and Dickinson (2012) between 55 and 85 AD. Only two stamps, Salve tu and Scottius, have an earlier chronology and could correspond to residual pieces. Two tria nomina (C. Iulius Celer and M. Iulius Severus), four marks on the internal base of decorated vases of type Drag. 29 (Felix ii, C. Iulius Celer, Marsus and Murranus) and four stamps on pieces in marbled sigillata (Albinus, Aper – an apparently unpublished stamp – and two stamps of Secundus) are documented. The two tria nomina are interesting since they correspond to stamps until now unknown in centres of consumption and very little documented either in La Graufesenque (Genin, 2007). The most illustrated potter is Secundus, with five occurrences, followed by Albinus, Germanus and Pontus, each represented on three occasions. The potters Aper, Calvus ii, M. Iulius Severus, Patricius and Vitalis are documented twice and the rest only once. In the case of Pontus, Calvus ii, M. Iulius Severus and Vitalis, it should be noted that the stamps identified have also identical characters and appear to have been produced by the same stamp, on pieces of identical typology or modules (Fig. 4, 4 and 13, 16-17). One can therefore defend an interpretation of the whole assemblage as belonging once to a single and homogeneous cargo – perhaps partly produced at the same time in the workshop of La Graufesenque – which was not divided up for redistribution on arrival at the port of Seville. The pieces could thus correspond to an accidental loss, or to an intentional discard at the time of unloading (by noting their fragmentation pattern). 5 cm0 Fig. 7 – San Bernado: bowl Drag. 37. SOUTH GAULISH SIGILLATA IN SOUTHWEST HISPANIA: CIRCULATION AND CONSUMPTION 191 Fig. 8 – Stamps documented for all San Bernardo (* pieces in marbled sigillata). No. Reading Potter Form and comments 1 OF ALBIN[I?] Albinus Drag. 24/25 2 OF ALBI Albinus Drag. 15/17 3 ALBIN(I) Albinus Drag. 15/17 (marmorata) 4 APRI (retrograde) Aper Drag. 24/25 (marmorata) 5 OF A[…] Aper Drag. 18 6 AR[…] Ardacus? Drag. 15/17 (marmorata) 7 OF CA Calvus ii Drag. 27 (2 pieces with identical stamp) 8 OF FEICIS Felix ii Drag. 29 (internal base stamp) 9[G]ERMA Germanus Drag. 18 10 OF.GERM Germanus Drag. 18 11 GERMA[N] Germanus Undetermined cup (Drag. 24/25?) 12 IA[…] Undetermined Undetermined cup (marmorata) 13 OF CICELRS C. Iulius Celer Drag. 29 (internal base stamp) 14 M.IVL.SEVE M. Iulius Seve(rus) Undetermined cup 15 OF L.ABIO Labio Drag. 18 16 OF MA M-II Drag. 24/25 17 [MAR]SS[I]MA Marsus Drag. 29 (internal base stamp) 18 OF MAT.VGE Matugenus Drag. 15/17 19 MOM Mommo Drag. 24/25 20 [O]F.MVR[ANI] Murranus Drag. 29 (internal base stamp) 21 PASSIE Passienus Undetermined cup (Drag. 27?) 22 OF PATRIC Patricius Drag. 27 23 OF PATRICI Patricius Drag. 27 24 OF PONT Pontus Drag. 24/25 (2 pieces), Drag. 27 (1) 25 SALVETV (retrog.) Salve tu Drag. 15/17 26 OF SCOTI Scottius Undetermined cup (Drag. 24/25?) 27 OF [S]ECVND Secundus? Drag. 18 28 (S?)ECVN Secundus i? Undetermined cup 29 (S)ECVN Secundus i? Ritt. 9 (marmorata) 30 SECVN.F Secundus iDrag. 18 31 OF SECV[N] Secundus ii? Drag. 18 (marmorata) 32 [OF SE]VERI Severus ii Undetermined cup (Drag. 27?) 33 OF VITA Vitalis Ritt. 9 (2 pieces with identical stamp) 34 VI[…]V Virthus?Drag. 29 (internal base stamp) 35 VI[…] Virthus?Undetermined form (marmorata) 36 […](APV?)[…] Undetermined Undetermined plate (marmorata) 37 MA[…] Undetermined Drag. 24/25 38 […]NE Undetermined Drag. 18 1 2 3* 4* 56* 79 10 11 12* 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29* 30 31* 32 33 34 36* 37 35* 38 8 1 cm0 Table 2 – Identification of the stamps documented in the San Bernardo complex. 192 Gaulish terra sigillata in Seville. New data, new perspectives ElisE Arnold, EnriquE GArcíA VArGAs 4.3. Patio de Banderas (AD 75/80-90) This last horizon is composed of 120 individuals, among which 13 plain forms and six differently decorated forms have been identified. The repertoire of the plain forms is dominated, as is usually the case in the second half of the 1st century AD, by the two types of plates and the two most recurrent types of cups: the Drag. 18 type plates with 32 MNI (Fig. 9, 12-15) are well ahead of the other forms, followed by the Drag. 24/25 cups with 14 MNI (Fig. 9, 1-2), the Drag. 15/17 plates with 12 MNI (Fig. 9, 10-11) and finally the 1 2 3 4 5 17* 18* 10 11 14 12 13 9* 6 7 8 Out of context Out of context 15 16 10 cm0 Fig. 9 – Patio de Banderas: plain forms (* : pieces in marbled sigillata). SOUTH GAULISH SIGILLATA IN SOUTHWEST HISPANIA: CIRCULATION AND CONSUMPTION 193 Drag. 27 cups with 9 MNI (Fig. 9, 3-4). The Flavian shape Drag. 35/36, according to the chronology of the set, gains in importance, recording three plates of type Drag. 36 (Fig. 9, 8), three cups Drag. 35 (Fig. 9, 6-7) and three rims whose specific form could not be determined. The Ritt. 8 and 9 cups are still present, although to a lesser extent than in the previous group, with four and one individuals respectively. The features of the Ritt. 8 (Fig. 9, 5), namely flattened and thick-walled, also denotes a late production of the type. The rest of the forms, documented in single pieces, are constituted by the Drag. 16 and Drag. 4/22 plates, the Drag. 15A type (Fig. 9, 16), the bowl Drag. 33 and the cup Ritt. 14B (Fig. 9, 9). The last is of 4 5 6 3 2 1 5 cm0 Fig. 10 – Patio de Banderas: decorated forms. 1-3: Drag. 29; 4: Hermet 9; 5: Déch. 67; 6: Drag. 37. 194 Gaulish terra sigillata in Seville. New data, new perspectives ElisE Arnold, EnriquE GArcíA VArGAs 4 7 1 2 3 56 5 cm0 Fig. 11 – Patio de Banderas: Drag. 37 bowls. SOUTH GAULISH SIGILLATA IN SOUTHWEST HISPANIA: CIRCULATION AND CONSUMPTION 195 particular interest for the dating of the ensemble, since the start of production would be around AD 80/90 (Passelac and Vernhet, 1993; Genin, 2007). The marbled gloss of the documented specimen also seems to demonstrate the prolongation of the use of this type of gloss beyond the reign of Nero on at least until the end of the Flavian era, as proposed by M. Genin (2007). The decorated pieces, which this time represent 24,4% of the total number of Gaulish sigillata, are again marked out by the prevalent types Drag. 29 and Drag. 37, which account for 87,8% of the decorated forms. However, it is surprising to observe – although the form Drag. 37 logically gains in importance, with its 13 MNI – that the vessels Drag. 29 seem to predominate (even if but slightly) until the middle of the Flavian period, with 16 individuals recorded (Fig. 10, 1-3). Other typically Flavian forms make their appearance, although in a discreet way, being represented by individual examples of the Knorr 78, Déch. 67 (Fig. 10, 5) and Hermet 9 types (Fig. 10, 4), while the Drag. 30 form decreases in importance with only two individuals represented. The style of the decorations is now clearly that of the period of ‘decadence’ as described by Hermet (1934), with manufacturing defects and even overprints being noted in some cases (Fig. 10, 3). Both Drag. 29 and Drag. 37 are dominated by zoned decorations and compositions, alternating with palisades and overlapping panels. Hunting scenes abound (Fig. 10, 1; Fig. 11, 3-4), as well as areas decorated with strigils (Fig. 11, 4-7). Vegetation characteristic of the decadent style is also observed, positioned underneath animals and characters (Fig. 10, 6; Fig. 11, 3-4). Finally, anthropomorphic motifs appear, becoming increasingly important in later times, for example in a gladiatorial scene (Fig. 11, 2). Thirteen potters’ stamps can be attributed to 11 different workshops in this group (Fig. 12). Potter Sabinus ii and mould-maker C. Iulius Sabinus appear twice. The chronological scope defined by the potter’s stamps seems to be centred on AD 80. The date is determined for this particular context by the intradecorative marks documented on Drag. 37 by M. Crestio (Fig. 11, 7), C. Iulius Sabinus (Fig. 11, 3 and 4) and Patricius (Fig. 10, 6), which provide a post quem terminus of AD 75/80 for this context. It should be noted that the intradecorative mark of M. Crestio and the two marks of C. Iulius Sabinus, inserted in a characteristic No. Reading Potter Tipology and comments 1 OF CAT (retrograde) Catus/Catlus? Undeterminate plate 2 COS[O]IVRA Cosius Urapus Drag. 27 3 MCRESTO M. Crestio Drag. 37 (intradecorative stamp) 4 C.I.SA C. Iulius Sa(binus) Drag. 37 (intradecorative stamp, documented twice) 5 IVLLINI Iullinus Drag. 18 6 LABIO Labio Ritt. 8 7 OF MA M-II Undetermined bowl 8 MEDD[IL]LV[S] Meddillus Drag. 29 (internal base stamp) 9 [P]ATRIC (retrograde) Patricius Drag. 37 (intradecorative stamp) 10 QVIN Quintanus Drag. 27 11 OF.SABIN Sabinus ii Drag. 15/17 12 OF.SABI[N] Sabinus ii Drag. 15/17 13 Anepigraph Undetermined Drag. 29 (internal base stamp) 14 OF LI[…] Undetermined Drag. 27 15 MA[…] Undetermined Drag. 29 (internal base stamp) 16 OF SA[…] Undetermined Drag. 24/25 Table 3 – Identification of the stamps documented in the Patio de Banderas. Fig. 12 – Stamps documented on Patio de Banderas. 1 2 3* 4* 56* 79 10 11 12* 13 14 15 8 1 cm0 196 Gaulish terra sigillata in Seville. New data, new perspectives ElisE Arnold, EnriquE GArcíA VArGAs manner for this workshop in the frieze of ovolo in the upper part of the decorated area, have also been documented in the cargo of Culip IV, sunk near the Catalan coast around AD 70-80/85 (Nieto Prieto and Puig, 2001). Although the various scholarly works consulted agree to place the start of production of these last two workshops at a time barely before AD 80 (Mees, 1995: 75, 93; Hartley and Dickinson, 2008: vol. 3, 174; Nieto Prieto and Puig, 2001: 25-26), Patricius’ retrograde stamp could be attributed up to AD 85/90. The characters of the stamp as well as the style of the vessel indeed seem to correspond to the second decorative series defined by Mees for this mould-maker (Mees, 1995: fig. 164, n° 8-14). 5. The marbled sigillata, a typical Mediterranean tableware? With a total of 35 MNI (67 fragments) documented within the three interventions studied, in all of the chronological contexts seen, the marble productions represent on average in Seville 6,5% of the total of the fragments of Gaulish3 sigillata. This percentage therefore places the entrance to the city among the best supplied centres in sigillata marmorata within the western Mediterranean, ahead of even the 3. Even if this figure is absolutely unrepresentative (since it does not take into account the restricted chronological scope covered by marbled productions), it nevertheless serves to establish comparisons with the majority of Iberian sites, with their often earlier publications, for which this figure appears usually calculated on the total number of fragments of Gaulish sigillata documented on the site and without any chronological considerations. Site/Assemblage Bibliography Period % of marmorata Pompey Pucci, 1977 AD 40/79 16,8% Pompey Huchin, 2014 AD 40/79 10,3% San Bernardo assemblage - AD 65-75 10,2% Ostia Martin 1994 40/end of the 1st cent. AD 10% (c.) Seville (average) - AD 60-90 6,5% Mérida (puticuli) Pérez Maestro, 2004 AD 40/70 6% Algarve (Castro Marim, Faro, Balsa) Viegas 2003b, 2011 1st cent. AD 5,2% Valeria Sánchez-La Fuente Pérez, 1985 1st cent. AD. 5,2% Baelo Claudia Bourgeois and Mayet, 1991 30-end of 1st cent. AD 4,4% Ampurias Aquilué et al., 2005 Neronian context 4,3% Cartagena Castellano Castillo, 2000 AD 60/70 4,3% Narbonne (shop) Sabrié and Sabrié, 2006 AD 70/90 4% Patio de Banderas assemblage - AD 75/80-90 2,5% Alemanes St. assemblage - AD 60-70 2,4% Tossal de Manises Ribera i Lacomba, 1988-89 1st cent. AD 2,1% Valencia Ribera i Lacomba and Poveda, 1994 1st cent. AD 1,4% Chãos Salgados (Mirobriga?) Quaresma, 2003 1st cent. AD 1,4% La Graufesenque (all periods) Genin, 2007 AD 15-170 1,3% Alcaçova de Santarém Viegas, 2003a 1st cent. AD 0,8% La Graufesenque (Gallicanus pit) Genin, 2007 50-70 d.C. 0,6% Conimbriga Delgado et al., 1975 1st cent. AD 0,3% Braga Morais, 2005 1st cent. AD 0,2% Table 4 – Percentages of marbled pieces documented in the total number of Gaulish sigillata, in different sets and deposits within the Iberian Peninsula, Italy and Narbonensis. SOUTH GAULISH SIGILLATA IN SOUTHWEST HISPANIA: CIRCULATION AND CONSUMPTION 197 capital of Lusitania, which to date had the highest figures documented for the Iberian Peninsula (Table 4) (Mayet, 1978) 4. This proportion is even more significant in the San Bernardo district as a whole, where the marble pieces reach 10,2% of the total, thus matching the figure calculated in the port of Ostia or in Pompeii (Huchin, 2014). Agreeing with the observations of Genin, we also note that the highest percentages, ranging between 4 and 5,2% of the total of Gaulish sigillata, are all to be found in important Mediterranean cities, in most cases at ports 5. This weighted distribution of marbled sigillata in the Mediterranean basin and particularly in Italy (Pucci, 1977; Martin, 1994; Huchin, 2014) contrasts with its extreme rarity in the northern provinces (Genin, 2007: 159), as well as its scant representation on the Atlantic façade of the Iberian Peninsula. In noting this concentration of marbled pieces in Italy, Baetica and Lusitania, Genin proposed interpreting them as vessels of a ‘Mediterranean taste’, clearly consumed mostly by Italian customers or those strongly Romanized such as in the southern Hispania (Genin, 2007: 161). In this sense, the strong concentrations of sigillata marmorata documented in the port of Narbonne, which contrast with their scarcity in the centres near La Graufesenque, as well as in the workshop itself – where it is estimated that the marbled pieces represented only 1,3% of the production (Genin, 2007: 155, 158), support the theory of a set of tableware destined for exclusive export to Mediterranean markets. This hypothesis seems to be fully confirmed in the sets studied in Hispalis, where the important assemblage of marbled pieces uncovered in the district of San Bernardo has few published parallels in the Iberian Peninsula in terms of the number of pieces, as well as their state of conservation. 4. Unfortunately, the figure of the total MNI of Gaulish terra sigillata documented in the puticuli of Mérida studied by C. Pérez Maestro in 2004 is not published. Thus it is therefore impossible to calculate the weight represented by the 81 fragments of marbled pieces documented with respect to red gloss productions. 5. In the case of Mérida, Genin explains the strong concentration of marbled production by the site’s likely role as redistribution centre for the rest of Lusitania (Genin, 2007: 161). Forms San Bernardo Patio de Banderas Alemanes St. Total Drag. 27 8 3 - 11 Drag. 24/25 7 - - 7 Drag. 15/17 6 - - 6 Drag. 4/22 4 1 - 5 Drag. 29 3 2 - 5 Drag. 35/36 5 - - 5 Drag. 18 3 - - 3 Drag. 16 - 1 1 2 Ritt. 9 2 - - 2 Ritt. 8 1 - - 1 Ritt. 1 1 - - 1 Ritt. 14B - 1 - 1 Herm. 18 1 - - 1 Drag. 37 1 - - 1 Drag. 30 - 1 - 1 Knorr 78/Herm.9 - 1 - 1 Bowl with rim similar to Halt. II - 1 - 1 Undetermined 9 3 1 13 TOTAL 51 14 2 67 Table 5 – Numbers of fragments and typological distribution of the marbled pieces documented in each of the three archaeological interventions studied, including pieces in residual contexts. 204 Gaulish terra sigillata in Seville. New data, new perspectives ElisE Arnold, EnriquE GArcíA VArGAs 6.3. Evolution of sizes Since research on sigillata began, the existence of several manufacturing modules for the different forms of the repertoire has been recognized. Most of the plates and cups have thus been produced in two or three standard sizes, precisely calibrated, with a view to optimising their stacking both in the kilns and during export transport. Forms of different typologies, but of identical modules could thus be stacked together (Vernhet, 1981). However, this calibration seems to have been more strictly applied for certain forms in the repertoire than for others (Polak, 2000). Alongside the main production modules, a small number of forms with non-standard measurements have been frequently remarked upon, probably manufactured in response to specific demands or uses. Too few metrological studies on sigillata pieces have been published to provide a clear explanation for these phenomena. However, since Bémont’s first work on the Gallicanus pit (Bémont, 1987) and the important studies carried out by both Polak on the material of Vechten (2000) and Genin on the sets of La Graufesenque (2007), we now have clear data that shows that the dimensions used by the potters in the manufacturing process evolved over time. Based on these studies, calculations of the dimensions of the pieces have been carried out, for greater representativeness, using the two most voluminous Sevillian ensembles, namely those of the neighbourhood of San Bernardo and that of the Patio de Banderas. The study focused mainly on the rim diameters of the four main plain forms – the Drag. 24/25 and 27 cups and the Drag. 15/17 and 18 plates – as the best represented in the different sets and therefore more conducive to a statistical study. For each of these shapes, the existence of two main sizes or manufacturing modules is known, as well as some atypical pieces, omitted from the study as unrepresentative. Given the absence of publications mentioning the size of the pieces, the only sets that permit of a comparative chronological study of the sizes are those of La Graufesenque (Genin, 2007), of the Culip IV cargo (Nieto et al., 1989) and in some cases the assemblage from the port of Narbonne (Fiches, Guy and Poncin, 1978)14. Thus, a number of observations could be made. The curves in the graphs obtained from the average measurements of the diameters of each module show, as previously noted by Polak and Genin, first a tendency towards a progressive dimimishing of the diameters for both cups and plates, and then a slight increase again in the 2nd century AD (Fig. 19 a-b). The sizes of the different types of plates are methodically identical for module 2 in each period – not so the case for module 1, while for the different types of bowls, the diameters have slightly different measurements according to type, the Drag. 27 cups presenting in general diameters some 10 mm wider than those of the Drag. 24/25. The measurements in module 2 of the Drag. 24/25 type show the most noticeable changes, being reduced by almost 30 mm between the reign of Claudius and the Flavian era, while those of the other bowls seem to decrease by only 5 to 10 mm in the same time frame. The diameters of the plates of both modules show a slightly more marked change, dropping by about 20 mm during the same time interval. Modules 1 and 2 of the bowls of each type have quite different sizes and are systematically distinct (35 as opposed to 50 mm), while the plate modules show a less dramatic difference, with only 17 to 30 mm (20 mm in most cases) between each. Finally, in the case of plates, it seems that during the 2nd century AD, even perhaps from the end of the 1st century AD, the small module pieces gradually began to disappear. This change could lie behind the absence of plates of the type Drag. 15/17 of module 1 in the cargo of Culip IV, and is arguably confirmed at La Graufesenque in the great kiln, where only the plates Drag. 18 of module 2 are represented. 14. We have chosen not to use here the measurements calculated by Polak in his study, since they correspond to pieces whose dating was determined by the average chronology provided by their stamps, without mentioning their specific contexts. SOUTH GAULISH SIGILLATA IN SOUTHWEST HISPANIA: CIRCULATION AND CONSUMPTION 205 These observations, for now somewhat summary as based on the little material available and the few existing reference publications, would clearly deserve to be investigated further by complementary studies and juxtaposed against comparable data from other standardized forms of the repertoire. All in all, it seems evident that the progressive decrease in the diameters corresponds to an evolutionary trend common to all the main plain forms, which fact should therefore be taken into account as an additional chronological marker in dating of South Gaulish productions. It would then be interesting to see if the same reduction in the size of the pieces can be observed in other Gaulish production centres, as well as for Hispanic sigillatas or those of Italic origin. Detailed studies on the subject could perhaps help to define the causes of these changes in the production method that occurred between the Julio-Claudian and Flavian periods, things once defined by a stricter calibration of the pieces and less diversity of sizes from AD 70 onwards, sizes outside the standards established for the main modules tending then to disappear. 6.4. Preferences for small or larger pieces The existence of different standard sizes for pieces of identical typology seems to indicate that parts of different modules were intended for different uses. It was therefore worth quantifying in what proportions the main forms of small and large module cups and plates had been consumed over time. When comparing the graphs obtained, it can be seen first that the volume of cups and plates of the same module evolves in a strictly parallel way (Fig. 20 a-b and 21 a-b). This seems to reflect the existence of a precise selection criterion, which responds to changing needs. Fig. 19 – Evolution of the average rim diameter of the main plain forms in the La Graufesenque, Culip IV and Seville ensembles, classified in chronological order (see Table 1 for dates and bibliography). 206 Gaulish terra sigillata in Seville. New data, new perspectives ElisE Arnold, EnriquE GArcíA VArGAs In general, small cups (module 1) and large plates (module 2) predominate in the repertoires. However, towards the beginning of the Flavian era there was a progressive preference for cups of a larger module (which did, however, decrease in actual size at the same time), which were then quickly abandoned in favour of smaller cups at the end of the 1st century or during the 2nd century AD. On the Fig. 20 – Quantitative evolution of the volume of cups in each module in the La Graufesenque, Culip IV and Seville ensembles, classified in chronological order (see Table 1 for dates and bibliography). Fig. 21 – Quantitative evolution of the volume of plates in each module in the La Graufesenque, Culip IV and Seville ensembles, classified in chronological order (see Table 1 for dates and bibliography). SOUTH GAULISH SIGILLATA IN SOUTHWEST HISPANIA: CIRCULATION AND CONSUMPTION 207 other hand, in the case of plates, a peak in consumption of small module pieces is evident towards the end of the Julio-Claudian or at the beginning of the Flavian period: this type suddenly became scarce shortly afterwards and had practically disappeared by the end of the 1st century AD. However, this observation is based mainly on the figures from the assemblage of San Bernardo, and it is thus difficult to know whether it is an isolated case or whether it reflects a generalised preference in the city of Hispalis or in Baetica at the beginning of the Flavian era. In the northern provinces, Polak testifies to a general shortage of small-module dishes: ‘The rarity of small dishes appears to be characteristic, not only of Vechten, but of the entire northwest of the Roman Empire’ (Polak, 2000: 76). In the later ensembles of the Patio de Banderas and the Culip IV cargo, destined for the Hispanic provinces, the dishes of the greater module predominate. A final decision here is therefore still pending, waiting the publication of new data. 7. Conclusions: towards a definition of local consumption patterns and regional markets The diversity and abundance of sigillata of South-Gaulish origin documented in Hispalis confirms the central role played by the city in the redistribution of those products into the south of Hispania. A major consumer of these ceramics, the southern coast of the Iberian Peninsula largely favours the productions of La Graufesenque for its table services, probably from the reign of Claudius until the beginning of the 2nd century AD; local Hispanic products do not compete with them. The comparative study carried out on assemblages from successive chronological contexts demonstrates progressive changes in the characteristics of the imported tableware, as well as in the specific composition of the repertoires. In this way, consumption patterns can be defined as to what looks specific to either the Iberian Peninsula or the Mediterranean world. They define the existence of specific demands and preferential markets for some forms of the repertoire of the Gaulish sigillata on a regional basis within the Empire. These preferences had some influence on the trade in sigillata, and therefore on the products made. The adaptation of the La Graufesenque workshop to meet customer demand is particularly noticeable in the case of the marbled sigillata, almost exclusively intended for Mediterranean markets. This selectivity implies that the manufacture of these particular items was in response to demand from these regions, and that there was a specific sector of the workshop dedicated to the products intended for the southern export routes. Nieto’s observation about the fact that ‘the diffusion [of sigillata sets] does not depend on the demand of the consumer, but on that of the transporter and, consequently, the consumer limits himself to acquiring those ceramic types that the transporter brings to the market’ (Nieto Prieto, 1993: 81), is therefore only partially true. If the populations have probably not been able to choose between buying a Gaulish sigillata or one of Italic origin, it seems in exchange that the transporters, first interested in satisfying the consumer to sell their goods, have been concerned to respond to the tastes and needs of the people with whom they intended to trade. These importation preferences are expressed in the Iberian Peninsula by a predilection towards certain forms, such as the Drag. 24/25 cups, heavily present in the repertoires until the end of the Flavian period, and perhaps also for the Ritt. 8 and 9 cups. It does not seem coincidental that these latter forms represent the Gaulish equivalents of the three Italic sigillata types of cups most consumed in Pompeii at the time of its destruction, namely the cups Consp. 34, Consp. 36 and Consp. 26/27 (Ettlinger et al., 1990; Huchin, 2014). This correspondence, together with the strong representation of plates and marbled products in the Hispanic repertoires, as well as the increase in the proportion of decorated pieces during the Flavian period – when these same vessels were in strong demand in central Italy, reveals the existence of a strong Italic influence on Iberian consumption patterns. All this therefore 208 Gaulish terra sigillata in Seville. New data, new perspectives ElisE Arnold, EnriquE GArcíA VArGAs arguably reflects the adoption and assimilation in Hispania of Italic culinary customs, probably the more strongly rooted in that region due to its early Romanization. Other variations, such as those observed in the calibration of the produced modules, may, in turn, reveal changes in the internal organisation of the workshop, and are probably linked to the standardisation of production techniques or transport conditions. However, considering the foregoing observations, it is possible that the general decrease in the size of the table service during the Flavian era is in some way related to cultural issues or changes in culinary practices, which have come to influence the production process at its very root. In short, to the question posed by Genin in her conclusions: ‘existe-il différents types d’approvisionnement dans telle ou telle région ou province, autrement dit, avons-nous affaire à des 'circuits' commerciaux traduisant des marchés préférentiels ?’ (2007: 348), to which a study on the diffusion of potter’s stamps could not then provide a satisfactory answer15, we can now proclaim an affirmative answer. Masked behind an apparent uniformity and ‘globalization’ of the trade of the sigillata in the Roman period, discrete local tendencies may yet be detected, whose closer study could advance our knowledge of the identities the local populations adopted or maintained in the process of their acculturation into the Roman world. Bibliography AMORES CARREDANO, F.; GARCÍA VARGAS, E.; GONZÁLEZ ACUÑA, D. (2007) – Anforas tardoantiguas en Hispalis (Sevilla, Espana) y el comercio mediterraneo. In BONIFAY, M. y TRÉGLIA, J.-M. (eds.) – LRCW 2: Late Roman Coarse Wares, cooking wares and amphorae in the Mediterranean : archaeology and archaeometry. BAR. Int. Ser. 1662. Oxford, p. 133-146. ARNOLD, É. (2015) – Trois lots de sigillée de La Graufesenque découverts dans le port d’Hispalis (Séville, Espagne) : nouvelles perspectives de recherche pour sa diffusion et l’évolution de sa consommation en Bétique. SFECAG, Actes du congrès de Nyon, Marseille, p. 411-433. AQUILUÉ, J.; CASTANYER, P.; SANTOS, M.; TREMOLEDA, J. (2005) – Presencia de producciones sigillatas sudgálicas en las excavaciones realizadas en el foro de la ciudad romana de Empúries (L’Escala, Alt Empordà). In NIETO, J.; ROCA ROUMENS, M.; VERNHET, A.; y SCIAU, P. (dir.) – La difusió de la Terra Sigillata sudgàl.lica al nord d’Hispania. (Monografies 6). MAC-Barcelona, p.163-177. ATKINSON, D. (1914) – A hoard of samian ware from Pompei, Journal of Roman Studies, 4, London, p. 27-64. BÉMONT, C. (1987) – La fosse Malaval 1 (La Graufesenque). Traitement numérique. Rei Cretariae Romanae Fautorum Acta, 25/26, p. 331-342. BOURGEOIS, A. ; MAYET, F. (1991) – Belo VI, Les Sigillées. Fouilles de Belo, Madrid: Casa de Velázquez. CASTELLANO CASTILLO, J.-J. (2000) – La terra sigillata sudgálica del área del anfiteatro romano de Carthago Nova. Saguntum, 32, p. 151-167. DELGADO, M.; MAYET, F.; ALARCÃO, A. M. (1975) – Fouilles de Conimbriga IV, Les sigillées. Paris: E. de Boccard. ETTLINGER, E.; HEDINGER, B.; HOFFMANN, B.; KENRICK, P.M.; PUCCI, G.; ROTH-RUBI, K.; SCHNEIDER, G.; VON SCHNURBEIN, S.; WELLS, C.M.; ZABEHLICKY-SCHEFFENEGGER, S., (1990) – Conspectus formarum terrae sigillatae italico modo confectae, Bonn : R. Habelt. FICHES, J.-L. ; GUY, M. ; PONCIN, L. (1978) – Un lot de vases sigillés des premières années du règne de Néron dans l’un des ports de Narbonne, Archaeonautica, 2, p. 185-219. GARCÍA FERNÁNDEZ, F. J.; GARCÍA VARGAS, E. (2014) – Comer a la moda. Imitaciones de vajilla de mesa en Turdetania y la Bética Occidental durante la Antigüedad (S. VI a, C. – VI d. C.). Col.lecció Instrumenta 46. Barcelona. 15. The most documented of the exported stamps seem in effect to be repeated in the different centres of consumption without any apparent pattern. For example, 13 of the 16 most frequently exported workshops are registered in Seville (Genin, 2007). SOUTH GAULISH SIGILLATA IN SOUTHWEST HISPANIA: CIRCULATION AND CONSUMPTION 209 GARCÍA VARGAS, E. (2007) – Hispalis como centro de consumo desde época tardorrepublicana a la Antigüedad Tardía. El testimonio de las ánforas, Cuadernos de Arqueología Cordobesa, 18, p. 317-360. GARCÍA VARGAS, E. (2009) – Las ánforas republicanas en Hispalis Sevilla) y la cristalización del repertorio anfórico provincial, in FERRER, E.; CRUZ-AUÑÓN BRIONES, R. (coords.) – Estudios de Prehistoria y Arqueología en homenaje a Pilar Acosta Martínez. Sevilla, p. 437-464. GARCIA VARGAS, E. (2012) – Hispalis (Sevilla, España) y el comercio mediterráneo en el Alto Imperio Romano. El testimonio de las ánforas. In KEAY, S. (ed.) – Rome, Portus and the Mediterranean. Archaeological Monographs of The British School at Rome 21, p. 245-266. GARCÍA VARGAS, E. (2015) – Ánforas vinarias de los contextos severianos del Patio de Banderas de Sevilla. In AGUILERA ARAGÓN, I.; BELTRÁN LLORIS, F. y DUEÑAS JIMÉNEZ, Mª J.; LOMBA SERRANO, C. y PAZ PERALTA, J. A. (eds). –De las ánforas al museo. Estudios dedicados a Miguel Beltrán Lloris. Zaragoza, p. 395-412. GARCÍA VARGAS, E (2016) – Amphora Circulation in the Lower Guadalquivir Valley in the Mid Imperial Period: the Lusitana 3 Type. In PINTO, I.V.; ALMEIDA, R.R.; MARTIN, A. (eds.) –Lusitanian Amphorae: Production and Distribution. Roman and Late Antique Mediterranean Pottery 10. Oxford, p. 285-298. GENIN, M. (2007) – La Graufesenque (Millau, Aveyron). Vol. II. Sigillées lisses et autres productions. Bordeaux : Fédération Aquitania. HARTLEY, B. R.; DICKINSON, B. M. (2008-2012) – Names on Terra Sigillata. An index of makers’ stamps and signatures on gallo-roman terra sigillata (samian ware), Institute of classical studies. London : University of London, 9 vol. HERMET, F. (1934) – La Graufesenque (Condatomago). Paris : Librairie Ernest Leroux., 2 vol. HUCHIN, R. (2014) – La céramique sigillée à Pompéi en 79 apr. J.-C. L’exemple du mobilier de l’insula I.9, SFECAG, Actes du congrès de Chartres, Marseille, p. 677-702. LEBLANC, O. (1994) – La sigillée gauloise à Saint-Romain-en-Gal (Rhône), SFECAG, Actes du congrès de Millau, Marseille, p. 143-164. MADRID FERNÁNDEZ, M.; COMAS I SOLÀ, M.; PADRÓS I MARTÍ, P. (2005) – Etude archéologique et analyse archéométrique de la céramique sigillée sud-gauloise documentée à la ville romaine de Baetulo (Badalone, Barcelone). In NIETO, J.; ROCA ROUMENS, M.; VERNHET, A.; y SCIAU, P. (dir.): La difusió de la Terra Sigillata sudgàllica al nord d’Hispania, Monografies 6. MAC-Barcelona. p. 179-197. MARTIN, A. (1994) – Nouvelles observations sur la sigillée sud-gauloise en Italie, SFECAG, Actes du congrès de Millau, Marseille, p. 115-126. MAYET, F. (1978) – Les importations de Sigillées à Mérida au Ier siècle de notre ère. Conimbriga, 17. Coimbra, p. 79-100. MEES, A.W. (1995) – Modelsignierte Dekorationen auf südgallischer Terra Sigillata, Stuttgart: Theiss. MORAIS, R. (2005) – Autarcia e Comércio em Bracara Augusta. Contributo para o estudo económico da cidade no período Alto-Imperial, Bracara Augusta, Escavações Arqueológicas 2. Braga: Unidade de Arqueologia da Universidade de Minho. NIETO PRIETO, J. (1993) – De la Dragendorff 29 a la Hayes 8: consideraciones sobre el comercio marítimo en los siglos I y II d.C., Cypsela, X, Girona, p. 77-85. NIETO PRIETO, J.; JOVER ARMENGOL, A.; IZQUIERDO TUGAS, P.; PUIG GRIESSENBERGER, A.M.; ALAMINOS EXPOITO, A.; MARTIN MENENDEZ, A.; PUJOL HAMELINK, M.; PALOU MIQUEL, H. y COLOMER MARTI, M. (1989) – Excavacions arqueològiques subaquàtiques a Cala Culip I. Girona : Centre d’Investigacions Arqueològiques de Girona, Sèrie Monogràfica 9. NIETO PRIETO, J.; PUIG, A. Mª (2001) – Excavacions arqueològiques subaquàtiques a Cala Culip. 3. Culip IV: la Terra Sigilata decorada de la Graufesenque, Monografies del CASC 3, Girona. PASSELAC, M.; VERNHET, A. (1993) – Céramique sigillée sud-gauloise. In PY, M. (dir.) – Diccionaire des céramiques antiques en Méditerranée nord-occidentale, Lattara 6, Lattes, p. 569-580. PÉREZ MAESTRO, C. (2004) – Análisis de la terra sigillata marmorata documentada en un vertedero de época Altoimperial de Augusta Emerita, Mérida. Excavaciones Arqueológicas, Memoria 7, Mérida, p. 361-369. 210 ElisE Arnold, EnriquE GArcíA VArGAs Gaulish terra sigillata in Seville. New data, new perspectives POLAK, M. (2000) – South Gaulish terra sigillata with potter’s stamps from Vechten. Rei Cretariae Romanae Fautorvm Acta, Svpplementvm 9, Nijmegen. PUCCI, G. (1977) – Le terre sigillate italiche, galliche e orientali. In CARANDINI, A. (dir.) – Instrumenticum domesticum di Ercolano e Pompei nella prima eta imperiale, Quaderni di cultura materiale, Rome, p. 9-21. QUARESMA, J. C. C. (2003) – Terra sigillata sudgálica num centro de consumo: Chãos Salgados, Santiago do Cacém (Mirobriga?), Trabalhos de Arqueologia, 30. Lisboa. RIBERA I LACOMBA, A. (1988-1989) – Marcas de terra sigillata del Tossal de Manises, Lvcentvm, VII-VIII, Valencia, p. 171-204. RIBERA I LACOMBA, A.; POVEDA, A. (1994) – La comercialización de la Terra Sigillata Sudgálica en el País Valenciano, SFECAG, Actes du congrès de Millau, Marseille, p. 95-102. SABRIÉ, M.; SABRIÉ, R. (2006) – Un magasin de céramiques sigillées à Narbonne? In MAUNÉ, S. y GENIN, M. (dir.) – Du Rhône aux Pyrénées : aspects de la vie matérielle en Gaule Narbonnaise (fin Ier s. av. – VIe s. ap. J.-C.), Archéologie et Histoire Romaine, 15. Montagnac : Monique Mergoil, p. 131-135. SANCHEZ-LA FUENTE PÉREZ, J. (1985) – Comercio de cerámicas romanas en Valeria. Cuenca : Diputación Provincial. SANCHEZ, C.; SILVÉRÉANO, S. (2005) – Le port de Narbonne et la diffusion des sigillées de la Graufesenque: étude préliminaire de la collection Bouscaras. In NIETO, J.; ROCA ROUMENS, M.; VERNHET, A.; y SCIAU, P. (dir.): La difusió de la Terra Sigillata sudgàllica al nord d’Hispania, Monografies 6. MAC- Barcelona, p. 163-177. SOUSA, E. M. (1996) – Terra sigillata marmoreada proveniente de Tróia de Setúbal, Conimbriga, 35, p. 207-215. TABALES RODRIGUEZ, M. A. (2015) – Excavaciones arqueológicas en el Patio de Banderas del Alcázar de Sevilla. Memoria de investigación (2009-2014). Sevilla. VALIENTE DE SANTIS, J. A. (2015) – Informe-memoria de la intervención arqueológica en los solares 1.1, 1.2 y 1.3 del P.E. Arrabal de San Bernardo. Documento inédito depositado en la Delegación Provincial de Cultura de la Junta de Andalucía en Sevilla. VÁZQUEZ PAZ, J. (2006) – Informe preliminar sobre la excavación de urgencia en la calle Alemanes nos. 25, 27, 29 de Sevilla. Documento inédito depositado en la Delegación Provincial de Cultura de la Junta de Andalucía en Sevilla. VERNHET, A. (1981) – Un four de La Graufesenque (Aveyron) : la cuisson des vases sigillés, Gallia, 39, p. 25-43. VIEGAS, C. (2003a) – A terra sigillata da Alcáçova de Santarém. Cerâmica, economia e comércio. Trabalhos de Arqueología, 26. Lisboa. VIEGAS, C. (2003b) – Les sigillées du sud de la Gaule à Castro Marim et Faro (Algarve, Portugal), SFECAG, Actes du congrès de Saint-Romain-en-Gal, Marseille, p. 641-646. VIEGAS, C. (2011) – A ocupação romana do Algarve. Estudo do povoamento e economia do Algarve central e oriental no período romano. Lisboa: UNIARQ. Digital sources Samian Research Database, Römisch-Germanisches Zentralmuseum Mainz (RGZM). http://www.rgzm.de/ samian (accessed 20.2.2019).