Ancient Theatres in Modern Spain: A Structuralist Interpretation

Zeynep Aktüre

The hypothesis The appearance of performance buildings in any provincial context of the Roman West has traditionally been interpreted as an indication of ‘Romanisation’, which denoted until recently an evolution in the provincial society towards civilisation through an adoption of the conqueror’s culture (Alcock 1997: 1). In this dominance model, architecture is pictured as a common, imposed language that was spoken in the same dialect in every city of the Empire with regional differences in minor decorative details and some construction techniques (Ball 2000: 247). Part of the vocabulary is an idealised type of Roman Theatre that has a cavea ‘on high subconstructions from level ground with a rich façade, a colonnaded gallery, and sometimes shrines on top’ (Bieber 1961: 189). The occurrence, in Roman period theatres, of some characteristics that are attributed to an equally idealised Greek Theatre (such as the construction of the cavea on the natural slope of a hill and as detached from the stage building in such a way as to unify visually and physically with the landscape, as in some examples from Asia Minor) has been interpreted as the survival of the local Greek-Hellenistic theatre-building tradition well into the Roman period, as a form of resistence against Roman cultural domination (e.g. Frézouls 1982: 396–409; Bernardi 1990: 133; Sear 2006: 24).

A systematic evaluation of the examples in the Iberian Peninsula (Fig. 1) would reveal, however, that the majority have a cavea ‘built against the hillside, and therefore have no outside façade’ (Bieber 1961: 189). Therefore, they meet the ideal of the Greek Theatre but in the absence of a Greek-Hellenistic theatre-building tradition (Fig. 2). The two examples featuring a cavea over high sub-constructions from level ground, as per the idea of the Roman Theatre, are among the largest theatres so far unearthed in the Peninsula. They therefore support the explanation provided by Pierre Gros (1994: 61) for the invention of the system of large and complex arches and vaults that forms the substructure of the Roman theatre cavea as a solution to the problems posed by the necessity of accommodating huge numbers of spectators (Fig. 3). These two theatres are located in the provincial and conventus capital Corduba (modern Cordoba) and in the conventus capital Caesaraugusta (modern Saragossa), the former of which was strategically, and the latter economically, important for the Roman control over the Peninsula (Fig. 4). There would seem to exist a correlation between the construction technique applied in the cavea of a Roman theatre and its size, while its size would appear to be correlated with the rank of the settlement in which it was located within the Roman administration, and the rank of a settlement within the Roman administration would have been largely dependent on its strategic importance in economic, administrative, or military terms.

In pioneering works that have adopted systems approach in classical studies (e.g. Woolf 1997; Alcock 1993; Rizakis 1997), the emergence of an urban network that was articulated by frequent communications and exchanges of all kinds is argued to be the most obvious indication of Roman rule in the provinces of the Early Empire. The reference here is to the systems theory suggested by I.M. Wallerstein, which has been adopted as a model for centreperiphery relations in Roman studies due to its stress on the essential interrelation of the individual parts of larger systems such as an empire (Alcock 1993: 5; 1989: 87). As a starting hypothesis, a classification of ancient theatre remains by their size and by the construction technique applied in their cavea would overlap the hierarchies in the urban network that was produced by the system of socio-economic and administrative structures established under Roman control over the Mediterranean basin. This paper will attempt to test the validity of this hypothesis in the light of the rest of the data pertaining to the theatre buildings in the Roman provinces of Hispaniae.

Figure 1: The state of preservation of the ancient theatres in the Iberian Peninsula. The data

Although Roman involvement in the Iberian Peninsula dated to the Punic Wars of the third century B.C., the establishment of such an urban network as mentioned above had to wait for the Augustan reform of the Empire, when Cantabria, Asturias and Galicia still remained outside it (Curchin 1991: 52). The Augustan re-organisation was a project undertaken in the years 26/25 B.C. when, having fallen ill in the Cantabrian Wars, Augustus stayed in the Colonia Tarraconensis (modern Tarragona). ‘It was in fact while he was staying in Tarraco in the winter of 26/25 B.C. and using it as his headquarters that a mission from the east had sought him out, asking that he accept public cult; and with this moment, everything had begun, at least in the west’ (MacMullen 2000: 77). This made Tarraco not only the capital of the province of Hispania Citerior but also temporarily the centre of the power and political decisions of the whole Empire immediately after the establishment of monarchy in 27 B.C., as the first instance when the imperial seat was translated from Rome (Alföldy 1988: 20; 1991: 35, 38). In addition to stylistic evidence coming from some of its Corinthian capitals and an altar of the Imperial Cult dedicated to the numen of Augustus (Mar et al. 1992: 14, 16), the Augustan dating of the Theatre of Tarraco appears to be largely informed by this imperial presence (Fig. 5).

The monument is one of the seven Augustan period theatres of the Iberian Peninsula, four of which are notably located in the early or later provincial capitals, including Tarraco. The oldest-dating among them appears to be the one at Corduba, the capital first of the province of Hispania Ulterior and then of the senatorial province of Baetica after its carving out of the former during the Augustan re-organisation. The construction of the monument is attributed to M. Claudius Marcellus, the patron of the city, largely on the basis of its similarities with the Theatre of Marcellus in Rome (Ventura 1999: 62–3). The Cordoba theatre was followed by the one at the veteran colony of Emerita Augusta (modern Merida), the capital of the province of Lusitania that corresponded to the rest of the Hispania Ulterior after the creation of Baetica. Here, the donation of the monument in 16/15 B.C. by Marcus Vipsanus Agrippa is evidenced in an inscription panel (CIL 2.474). Likewise at Carthago Nova (modern Cartagena), the once Punic stronghold which appears to have served as the capital of Hispania Citerior before its Augustan translation to Tarraco, the reconstruction of two highly fragmented inscription panels has enabled an identification of the donors of the Theatre as Caius and Lucius, the adopted grandsons of Augustus who were designated as his successors in 6 B.C. after the death of their father, Agrippa (Ramallo and Ruiz 1998: 125–41). Except the one at Tarraco, these four theatres are among the largest of the Peninsula in terms of cavea size and/or spectator capacity, conforming to the hypothesis of a correlation between the size of a Roman theatre and the rank of the settlement in which it was located within the Roman administrative network (Fig. 3).

An analysis of the geographical location of these four monuments, on the other hand, would reveal the dependence of settlement rank on strategic importance in economic, administrative, or military terms. Tarraco and Carthago Nova were harbour towns, the latter of which was of Phoenician origin as one of the best natural ports of the Mediterranean that was suitably-placed for contact with Carthage. This had made it a primary target in the strategy of the Scipio brothers to control this vital link between Hannibal and his supply bases in Hispania during the Second Punic War (Richardson 1998: 17, 27). The other two provincial capitals, Emerita Augusta and Cordoba, not only served as a bulwark against the marauding Lusitanian tribes of the area between the Guadiana and the Guadalquivir (Knapp 1983: 8), but also as important fluvial ports respectively on these two among the major rivers of the Peninsula. Other important fluvial ports of the Peninsula included Hispalis (modern Seville) on the Guadalquivir and Caesaraugusta on the Ebro (Curchin 1991: 12). Remains from an ancient theatre of yet unattested size are being investigated in Seville, and Saragossa has remains from a large intramural theatre. As in the very large theatre at Corduba, the cavea of this latter monument rises over high sub-constructions from level ground except in its lowest portion, as in the common idea of a Roman Theatre, apparently because both settlements extended over terraces along a moderate slope. This would highlight the impact of site topography along with size in the selection of an appropriate construction technique for the cavea of Roman period theatres. All dating from the Tiberian period, as does the one at Caesaraugusta, the other large theatres of the Ebro valley and northern meseta would seem to support this conclusion.

Figure 2: Methods applied in the construction of the cavea in ancient theatres of the Iberian Peninsula.

While the Baetis (Guadalquivir) valley and the surrounding hillsides present the largest concentration of ancient theatres evidenced in architectural and epigraphic remains, the Ebro valley appears to have the largest concentration of large theatres in the Peninsula. The overall concentration of theatre remains in this area is comparatively low though. This uneven distribution would seem to be due to structural differences between various parts of the peninsula, possibly reflecting the absence, in the Ebro valley, of middle-ranking settlements that had adopted Classical town and religious life. While the lowlands of the Baetis valley were characterized by a high degree of pre-Roman cultural sophistication, a long-standing urban tradition, and a probably landed native elite (Mierse 1999: 121), a common feature of the northern meseta were the Celtic castros (hill-forts). This type of fortified Iron Age village was also a common feature of the Galician littoral and of the northwestern Iberian Peninsula (Naveiro and Pérez 1992: 91) where remains from ancient theatres are remarkably not evidenced. These would seem to conform to the argument that ‘[I]n Baetica, Rome was much more successful in the lowlands, and along the rivers, than on the plateaux… In the mountainous northwest of Spain with the added difficulty of distance, Rome penetrated late on and with little success’ (Braudel 1995: 34). The later Flavian date of the theatre of Segobriga, which is among the places distinguished by the Celtic suffix –briga meaning hill-fort (Curchin 1991: 16–7), would seem to support this interpretation.

An evaluation of the overall geographic distribution of ancient theatres in the Iberian Peninsula by their size and by construction date would suggest a progression: from the Late Republican coastal theatres around the southern tip of the Peninsula to the Augustan ones along the favourable lowlands of the Guadiana and the Guadalquivir, from there to the Tiberian examples of the Ebro valley further to the north and finally to the lesser examples in the highlands of the meseta. These latter appear to have waited for the funds allocated to them with the Vespasian grant of municipal status (ius Latii) for the monumentalization of their urban scene in the Flavian period (Keay 1998: 78). Therefore, the structural inequality within the Roman urban network would seem to have been reflected onto the theatre-building activity in the Peninsula. As we go from the lowlands of the river valleys to the highlands of the meseta, carving of the cavea into the natural rock as in the Tiberian theatre of Bilbilis Augusta becomes the rule rather than an exception (Fig. 2). A notable example in this respect is the Tiberian theatre of Clunia, which is among the large examples within the study area that features a cavea carved into the natural rock. Hence, it seems not possible to conceive the technique of constructing the cavea on high sub-constructions from level ground simply as a solution applied for the problems posed by the necessity of accommodating huge numbers of spectators, as argued in the starting hypothesis of this study in reference to Gros (1994). In all, the data coming from the ancient theatres of the Iberian Peninsula would seem to support an interpretation of some of their architectural characteristics, such as the size or the construction technique applied in their cavea, as a product of the urban network that was established under Roman rule over the Mediterranean, and a reflection of the hierarchies intrinsic in that network. This would hint at the possibility of approaching the problem of classification in studies on ancient theatre architecture from a geo-historical perspective to include each and every example so far documented in archaeological remains, adopting the trifocal model in Fernand Braudel’s The Mediterranean and the Mediterranean World in the Age of Philip II (1995).

The theoretical framework Initially completed in 1947, The Mediterranean would rank high on a list of works that set Mediterranean historiography onto a new course, especially after the first publication of its revised edition in English in 1972. This influence has been mainly due to Braudel’s approach to the past along three planes of historical time. These are the almost timeless history of the relationship between the individual and the environment that is called ‘geo-history’; the gradually-changing history of economic, social and political ‘structures’; and the fast-moving history of ‘events’. Braudel’s main argument is that the history of ‘events’ would be ‘unintelligible without the history of structures, which is in turn unintelligible without the history of the environment’ (Burke 1990: 40). When set against this framework, the formation of seating rows over a natural slope in ancient theatres, and usually by carving into it, appears to be part of the almost timeless history of the relation of the individual with the environment. This trifocal conception of time is reflected onto the organisation of the book in the form of a division into three parts. Followed by one on ‘Collective Destinies and General Trends’ to render intelligible the last part on ‘Events, Politics and People’ that includes none less than the deeds of Philip II, the first part of The Mediterranean, which is also the longest, is dedicated to ‘The Role of the Environment’, on the basis of the idea that geography ‘helps us discover the slow unfolding of structural realities, to see things in the perspective of the very long term’ (Braudel 1995: 23). Indebted heavily to the French geographical school represented by Vidal de la Blanche and also to one of the founding fathers of the Annales journal, Lucien Febvre, this conception of the ‘very long term’ (longue durée) proved to be the most powerful aspect of The Mediterranean. Braudel’s most influential essay was also dedicated to the same concept (Braudel 1980a). Among the many volumes inspired from this emphasis on the longue durée, two notable compilations from the field of archaeology are those edited by Bintliff (1991) and Knapp (1992). This aspect of the Braudelian paradigm has also triggered the most radical of all criticism his work has received: that of geographic determinism. This has resulted, according to some critics, in a history without the individual, who is described by Braudel (1980a: 31) as prisoner of his/her mental framework and the physical environment. By placing them in context in their milieu wherein they appear as ‘surface disturbances, crests of foam that the tides of history carry on their strong backs’ (Burke 1990: 35) Braudel emphasises the insignificance of events, which he likens to waves, beneath which one has to dive to understand history in the stiller waters that run deeper and are concerned with ‘the history of structures – economic systems, states, societies, civilizations and the changing forms of war’ (Burke 1990: 34–5). This medium-term history moves at a slower pace than that of events, in generations or even centuries so that contemporaries are scarcely aware of it.

Figure 3: Cavea dimensions in ancient theatres of the Iberian Peninsula.

From an epistemological point of view, all works of architecture (including ancient theatres) are obviously ‘events’ along Braudel’s third plane of historical time, and have often been analysed as such (e.g. Isozaki and Asada 1999; Rajchman 1999; Tschumi 1999). The essentiality of theatres for the Greek and Roman town and religious life, on the other hand, would enable a conceptualisation of their construction as a product of a change in Braudel’s ‘structures’, that was brought by the establishment of Roman control over the Mediterranean that arguably resulted in the emergence of an urban network. For Braudel, the heart of European history was the Mediterranean, and the heart of the Mediterranean was its cities that formed a pumping urban system in control of the pulsation of the region’s parts (Kinser 1981: 75). An example for the part played by geography in the formation of that system may be found in the location of the largest theatres of the Iberian Peninsula in the conventus capitals that were fluvial (Augusta Firma Astigi [modern Ecija], Corduba, and Hispalis) or sea (Gades [modern Cadiz]) ports of an agricultural hinterland along the river Baetis, which had a pre- Roman Carthaginian and Greek urban past (MacMullen 2000: 50, 54), and ‘was fairly rapidly to become the heart of Roman Spain, a garden of towns that soon became too brilliant and overpopulated’ (Braudel 1995: 82).

The comparison, under the previous heading, of the geographical distribution of the largest ancient theatre remains so far attested in these towns of the plains with those on the central Iberian plateaux and along the Ebro valley further to the north may, in fact, be seen as an attempt to integrate data on each individual example, which is often uneven in scope and amount. If accomplished, such an attempt may contribute in a reconstruction of the formal evolution and chronology of ancient theatres by bringing into attention ‘the more dilapidated or less well preserved monuments that are sometimes an indispensable link in the chain’ (Frézouls 1994/95/96: 60) and by allowing inter-regional evaluations that may reveal the existence of local building types through a comparative study of each and every monument, as suggested by Edmond Frézouls (1994/95/96: 60) in his contribution in the census, edited by Paola Ciancio Rossetto and Giuseppina Pisani Sartorio (1994/95/96), of all the ancient theatres so far documented in architectural and/or epigraphic remains, and/or in written record. This attempt may be likened to Braudel’s in The Mediterranean whose tripartite format has been interpreted by some critics as an attempt to combine the study of time with that of the complex interaction between the environment, economy, society, politics, culture, events, and individuals, which the author integrates into a ‘total history’ (Burke 1990: 40–2). Indeed, for Braudel ‘history is the total of all possible histories—an assemblage of professions and points of view, from yesterday, today, and tomorrow’ (1980b: 34). Hans Kellner (1979: 217–22) points to the impossibility of this project of ‘total history’ by highlighting a linguistic dilemma rooted in it: ‘… as the post-structuralists have repeatedly pointed out, structuralist activity is its own dismantling; for the relationship of structuralism to its structures may also be viewed structurally – ironically – thus displacing infinitely the central notion from which structuralism proper, the Saussurean tradition, proceeds. […] All texts contain their own deconstruction: this deconstruction is not done to them – any more than the structuralist creates the structures that he identifies – it is always already there’ (Kellner 1979: 220).

According to Kellner, while displaying this dilemma, Braudel’s The Mediterranean also offers a linguistic solution for it and, thus, cries out desperately to be noticed and understood ‘formally’. Hence, Kellner characterises The Mediterranean by a flood of anatomical details that apparently offer an analysis of the issue into its constituent parts in three clear-cut volumes. Samuel Kinser (1981: 89) similarly highlights Braudel’s strategy against the structuralist dilemma, of viewing the Mediterranean and the Mediterranean world in the age of Philip II both as an organic whole in itself and as a mechanical agglomerate of constituent parts, as that of displacing the problem of conceptual coherence to one of representational coherence by skilfully achieving an integrated text. Essential to the idea of ‘organism’ is the notion of a ‘whole’ that includes and orders ‘parts’ in a given group of phenomena, which act and react in accordance with the organisation of the inwardly-centred whole that possesses autonomy as it operates outward from the heart, brain, or soul. ‘An organism’s evolution is processive, connecting change in one part with change in all others; thus, an organism evolves toward greater complexity, greater inclusiveness, and ever grander totalization.’ (Kinser 1981: 72) By contrast, one part of a ‘mechanism’ may be set into motion without involving other parts as it possesses only adjacent surfaces that interact but no centre and, hence, no permanent ‘inside’. Mechanical evolution is usually conceived as occurring in a sequential order determined by adjacency in the sense that ‘what happens next to a mechanically articulated aggregate is conditioned by the sequence of what has happened before to the mechanical or organic entities with which it is in contact’ (Kinser 1981: 72).

The distinction would appear highly promising to overcome the difficulty of conceptualising centre-periphery relations in such a vast empire as the ancient Roman one, which surfaces in studies on ancient theatre architecture in the form of a trap between the extremes of uniform centralism and heterogeneous regionalism (Mariner 1982: 17). On the one hand is the temptation for uniformity by considering each and every colonia or municipium, regardless of their location and period, as a small Rome that was apt for applying what is known to have occurred contemporaneously in the capital. On the other is the tendency of postulating a provincial stance, especially in the case of provinces relatively distant from the Italian Peninsula, by projecting over provincial theatre ideas of decentralisation, which have been accredited as successful in other cultural aspects such as language or religions, allowing for the assumption of their maintenance of their pre-Roman structure. Regional heterogeneity would seem to have an appeal for scholars like Martín Almagro Basch, who has studied a number of ancient theatres remains in Hispania in detail to opinion that ‘each case is distinct’ (Mariner 1982: 23) while a great majority of the general works on Greek and Roman theatre and theatre architecture, including classics like Margarete Bieber’s The History of The Greek and Roman Theatre (1961), seem to reflect the uniform centralist view. This enables the use of early permanent theatre architecture in the city of Rome, and chiefly the Theatres of Pompey and Marcellus, to draw analogies for provincial examples, as exemplified by Ramsay MacMullen (2000: 78) who argues that ‘theatres in Spain followed the model of Pompey’s in Rome’ in featuring a sacellum in summa cavea while this component is evidenced archaeologically only in a few examples (i.e. definitely at Bilbilis and Saguntum, and possibly at Carthago Nova and Corduba). A review of theatre-building activity in the provinces of Hispaniae within this framework would significantly suggest the dominance of a mechanical pattern rather than an organic one operated from the city of Rome, however with equally significant connections to the city of Rome that throw considerable light on the processes of ‘Romanisation’ in this part of the Empire.

Figure 4: Judicial districts (conventus) of the Roman provinces of Hispaniae (after Mayer and Rodà 1998: 222).

Discussion To start with the senatorial Provincia Hispania Ulterior Baetica, its capital Corduba was also the seat of a judicial conventus that extended up to Emerita Augusta and Malaca (Bernier 1975: 21). It forms one of four such subdivisions or juridical districts in the province, alongside Hispalis, Astigi, and Gades, all of which would have risen to such prominence thanks to their important trading positions and administrative functions (Cuenca 1993: 13; Aguilar 1995: 18). Constituting an active unity between the two fundamental circumscriptions of province and civitas, these juridicial districts are known to have existed only at Ilirico and Asia within the context of the Empire, alongside the three provinces of Hispaniae (Beltrán 1985: 81). Ancient theatre remains, of yet unattested size, are being hypothesised in Hispalis and Astigi on the basis of the actual urban morphology while those revealed beneath later layers in Corduba and Gades constitute the largest two theatres so far unearthed in modern Spain, enabling a presumption of the former two at least among the very largest in the country. After the Augustan organisation, the rest of the Ulterior that corresponded approximately to modern Portugal now formed the province of Lusitania that had its capital at Emerita Augusta (Baird 2000; Fear 2000: 31). Emerita Augusta was also the capital of one of the three judicial districts of the province, with the other two having been located at Scalabis (modern Santarém) and Pax Iulia (modern Beja) now in Portugal (Keay 1988: 61). While Emerita Augusta features one of the larger theatres in the Peninsula, no ancient theatre remains have been reported so far in these latter two conventus capitals that may have had similarly large theatres. The rarity of theatre remains in the area corresponding to Lusitania may perhaps be partially explained by Fear’s (2002: 21) suggestion that, after the establishment of the Pax Romana, Rome would have actively discouraged urbanisation in this area. This is attested in Appian’s mention of Ti. Semponius Gracchus’ prohibition on the Celtiberian tribes building further poleis or fortifying their existing ones. Further evidence is provided by Strabo’s remark about Rome’s reduction of the majority of the poleis in northern Lusitania to mere villages while improving a few and settling there themselves, as in the case of Metellinum.

The remainder of the Iberian Peninsula that once formed Hispania Citerior was often referred to as Hispania Tarraconensis after the Augustan re-organisation and had its capital at Tarraco (Fear 2000: 31). Hispania Tarraconensis was one of the largest, if not the largest, among the provinces of the Roman Empire (Alföldy 1988: 20). It consisted of seven conventus centres at Tarraco, Carthago Nova, Caesaraugusta, Clunia, Asturica Augusta (modern Astorga), Bracara Augusta (modern Braga), and Lucus Augusti (modern Lugo), the nucleus of whose administration was Tarraco where the 300 municipalities of the province meet once a year (Arce 1976). The formation of a military, political, economic, and administrative infrastructure immediately after the establishment of control over this region through the creation of such colonies as Asturica Augusta, Bracara Augusta, and Lucus Augusti is generally attributed to Agrippa. He would have been involved very directly also in the administrative re-organisation of the Peninsula in general after Octavian fell ill during the Cantabrian campaign (Jiménez 1992: 17; Vincent and Stradling 1995: 35). He was helped by his famous Orbis pictus depicting the geographical characteristics of the Peninsula for a better assessment of possible sites for future colonies along important communication roads (Bendala 1990: 38–9). The juridical capitals of the province, however, were not created before Claudius (Palol 1978: 19). Remains from ancient theatres have been revealed in the former four of these latter conventus capitals. These are all among the larger theatres of the Peninsula except the one at Tarraco. As to the remainder of the conventus capitals in Tarraconensis, the examples of Carthago Nova, Caesaraugusta, and Clunia may render it likely that large theatres would have existed in them and a theatre or amphitheatre was indeed reported at Bracara Augusta (modern Braga) in the eighteenth century (Curchin 1991: 116–7). Asturica Augusta, Bracara Augusta and Lucus Augusti, however, were all located in ‘the least Romanised part of Hispania during the early empire’ (Keay 1988: 62) that was characterised by an almost total lack of public buildings and certainly nothing as advanced as a theatre, temple or circus, especially in Galicia (Curchin 1991: 183). Here, the establishment of Roman coloniae at these apparently pre-existing native sites was to deprive the local tribes of their potential mountain strongholds and make them more vulnerable to Roman policing rather than ‘civilising’ or ‘Romanising’ them (Fear 2002: 24). So, it may be unlikely that theatre remains would be found in these settlements.

As such, our current state of knowledge on the theatre-building activity as a whole in the Roman provinces of Hispaniae would seem to support the earlier observation of structural differences between various parts of the peninsula, encouraging the adoption of the mechanical model in its interpretation. Hence we have proceeded from one province to the adjacent one, marking out their differences rather than their similarities and links especially with the city of Rome, except for the imperial patronage in the theatres of the provincial capitals. The notable exception of the theatre of Tarraco, smaller in size in comparison to the theatres in the other provincial capitals, would appear as a reflection of inter-regional differences within the Empire. ‘Tarraco may have been the capital of a province, but it was not a large city, perhaps ca. 60 ha, the same as Narbo. It was larger than Emporiae or Barcino, but smaller than Corduba at 70 ha or Carthago Nova at 80 ha and certainly smaller than the Gallic capital of Lugdunum at 140 ha’ (Mierse 1999: 233). The example of Tarraco would conform to the remark quoted earlier from Kellner that ‘structuralist activity is its own dismantling’ (1979: 220), through an identification of the structures on the basis of which to proceed with decontructionist activity.

Figure 5: Initial construction dates of ancient theatres in the Iberian Peninsula.

The deconstruction should continue with the early date of the Theatre of Gades, which is also among one of the two very large theatres in the study area, whose initial construction predates the Augustan re-organisation. While the huge size of the edifice may be explained in terms of its location in the most flourishing Phoenician, and later Roman, harbour and stronghold in the Western Mediterranean, its early construction date would seem to be connected with the pro-Caesarian history of the city of Gades. It won for Caesar the final victory over Pompey’s successors during the final stages of the Late Republican Civil Wars that were fought in the Iberian Peninsula between 49–44 B.C. The initial construction of the theatre of Gades is dated to 43/42 B.C. in the immediate aftermath of this victory, and attributed to Balbus the Younger, of the Latin-speaking equestrian Balbii and the nephew of Balbus the Elder, Caesar’s banker whose Roman citizenship, received from Pompey the Great in 72 B.C., was famously defended by Cicero in 56 B.C. (Rodríguez 1973: 105; 1980: 35, 41– 2, 55, 170). Balbus the Elder became consul in Rome in 40 B.C., the first consul from the provinces. His nephew became the questor of Asinion Polion in 44 B.C. and was later elected as quattuorvir of Gades (Rodríguez 1973: 267). Ramón Corzo Sánchez (1993: 135) argues that decades before the comprehensive building programme carried out by Augustus in Rome, for remodelling the city as the new imperial capital, the ideas shaping that programme had been put into practice by Balbus in Gades. The theatre stood as a symbol of the new imperial mentality and organisation, with its twenty different types of marble coming not only from local quarries but also from such distant places as Carrara (Italy), Syria, and Anatolia (Esteban et al. 1993: 155). After this accomplishment in his native city, Balbus would support the Augustan programme in Rome through the construction of the Theatre of Balbus and the annexed Crypta Balbi shown in the Forma Urbis and noted by Suetonius and Pliny for its four onyx columns (Rodríguez 1973: 273; Corzo 1993: 135; Bieber 1961: 184). This was an honour granted to him by the Roman Senate for his triumph over the Garamantes in North Africa in March 19 B.C. (Blázquez 1976). This was ‘a unique distinction for one not born a Roman citizen and an exceptional honour in that thereafter traditional military triumphs were not granted to those outside the imperial family’ (Beacham 1999: 119). A contextualisation of the construction of the theatre of Gades as such would both support and challenge Braudel’s portrayal of the deeds of individuals as waves over the deeper stiller waters consisting of ‘structures’. After all, the Balbii of Gades were integrated into those structures thanks to ‘the openness of the upper class to provincials that enables Rome, alone of the preindustrial empires in this region, to avoid the construction of local dynasties through the regular circulation of governors and to create a competitive atmosphere between cities and within cities, for which the governor was the referee’ (Potter 1999: 9).

Another equally significant exception comes with Tiberian patronage in the medium-sized theatre of the administratively middle-ranking town of Italica. Imperial patronage would appear to be the rule for the theatres of the high-ranking provincial and conventus capitals of the provinces of Hispania during the Augustan period; these are also among the largest of the Peninsula in proportion to the importance of their donor. The case of Italica may be explained perhaps as a re-establishment of the historic ties the city had with Rome. These date back to the foundation of Italica near modern Seville, as the only city of the Iberian Peninsula with an exclusively Roman name that gives reference to Italy. The foundation was for the wounded soldiers of the decisive Battle of Ilipa fought in the Guadalquivir valley in 206 B.C., in which the Roman army under the command of the young Publius Cornelius Scipio Africanus expelled the Carthaginians after five years in a state of war. Italica seems to have re-stated this special historical link with Rome some time in the second half of the first century B.C. when it started to develop its public image in line with the programmes of Imperial propaganda that were being developed in contemporary Rome. Perhaps this later link would explain also the choice of the technique of constructing its cavea over manmade sub-construction, as in the examples in the city of Rome, however after cutting the natural slope of a hill in the case of Italica. This may have been a solution against the alluvial residues that were brought by the constant flooding of the nearby Baetis. Later in its history, Italica would have closer ties with the city of Rome as the patrida of Trajan’s Ulpii and Hadrian’s Aelii families. Hadrian would honour the city with the creation of a new district (nova urbs), featuring a Traianeum, to the south of the existing one (vetus urbs) (Luzón 1982: 187). Unfortunately, the construction of this new district over an alluvial clay layer would result in its rapid decline, from the end of the second century A.D. onwards, and its final abandonment in the third century A.D. (Cepas Palanca 1997: 218). As such, the cases of Gades and Italica are a reminder of the organic, rather than mechanical, nature of the essential interrelation between individual parts of larger systems such as empires, as suggested by the systems theory.

The theatre constructed at Gades by Balbus the Younger, on the other hand, would provide a positive answer to the question ‘whether some events at least might not modify structures, rather than simply reflect them’, which was raised by Peter Burke (1990: 91) against Braudel’s firm conviction that ‘events’ resulting from human action were simply mirrors reflecting the history of ‘structures’, i.e. reactions or responses to structural change. The three-partite cavea of the edifice rests over galleries in its lower and middle part. Excavations in the lower gallery revealed that its vault was out of opus caementicium in independent horizontal layers consisting of blocks that rest over one another like the stones of an arch, instead of benefiting from the plastic quality of concrete to form a continuous bending surface (Corzo 1989: 211–2; 1993: 135). Well-documented in examples from Tyre and the Phoenician zone of Cyprus, the early evidence for this construction technique in the theatre of Gades would conform to the argument, both of Vitruvius and of Varron, that the technique of construction with masses of mortar rolled between moulds of timber had a Hispanic and Carthaginian origin. As suggested by the excavator of the edifice, Corzo Sánchez (1989: 212), the layering of mortar would have been an originally Egyptian technique diffused across the Mediterranean by the Phoenicians and the construction principles of concrete should have been known and implemented in Gades long before the construction of the theatre by Balbus the Younger. Although the limitation of the remains from the later Theatre of Balbus in Rome to scanty segments below the modern urban texture deprives us of a comparison between the two monuments, the idea is still useful as it forces us to reconsider our starting dominance model in the evolution of Roman theatre architecture. Examples such as the theatre of Gades would justify the attempts of some experts (such as Roldán, Bendala et al. 1998) who have started an extensive and detailed examination of architectural remains within provincial contexts. Their intention is to trace the incorporation, into those already existing before the establishment of Roman control, of construction materials and techniques that were disseminated from other parts of the Empire, as a key to a better understanding of the ‘Romanisation’ process that would have taken place afterwards.

As a conclusion When taken all together, the theatre-building activity in the Roman provinces of Hispaniae may be observed to conform to the comparatively recent conception of ‘Romanisation’ as an umbrella term that may refer to a variety of processes including acculturation, accommodation, and resistance, depending on the peculiarities of the affected local societies (Keay 2001: 122). The examples mentioned so far fit the model of direct involvement on the part of the patroni (as in the construction of the theatres of Tarraco, Cordoba, Carthago Nova, Emerita Augusta, and Italica), of self-Romanisation (as presumably at Segobriga where a theatre was built after the Vespasian grant of municipal rights and finances), or the elite model (as in the construction of the theatre of Gades). These would have been only a few among the multitude of separate processes and a multiplicity of responses to cultural contact concealed under the umbrella term of ‘Romanisation’ in the past two decades (Woolf 2000: 7, 15). A recent overview of some commonly used models of ‘Romanisation’ is offered by Curchin (2004: 12–14; also underlined by MacMullen 2000: 78). This new conception challenges the long-standing associations of the dominance model, on the basis of new research in the past three decades that has revealed each provincial territory to have had ‘its own physical environment, its own indigenous culture, and its own history of relations with Rome, all of which affected and modified the Romanization process’ (Curchin 2004: 2, 8).

Histories of theatre architecture based on this new conception of ‘Romanisation’ have already started to come out. One example is the three-volume catalogue of all archaeologically and historically documented Greek and Roman theatres that have been classified by the editors of the publication according to the modern state in which they are located (Ciancio Rossetto and Pisani Sartorio 1994/95/96). Their introductory article in the first volume includes the regional Greco-Roman and Gallo-Roman types of theatres alongside the Greek and the Roman type of theatre and the odeon, providing a description and axonometric drawing for all except the first. The presentation of the catalogue in Frank Sear’s Roman Theatres – An Architectural Study (2006: 117–424) comes even closer to the new conception of ‘Romanisation’ in its integration of data on each individual example through a classification by Roman provinces rather than by modern states, after some introductory chapters that would seem to adopt the uniform centralist approach on issues like theatre and audience, finance and building, or Roman theatre design. These two comprehensive publications form a firm basis for interpretations such as the present one, which would hopefully contribute in the establishment of a link between typological studies in ancient theatre architecture and ‘Romanisation’ studies at large.

Department of Architecture, Izmir Institute of Technology

Acknowledgements This paper is based on research conducted mainly in the library of the Faculty of Philosophy and Letters of the Autonomous University of Madrid, and complemented with working trips to the major sites with the 2001 Summer Grant of the Spanish Ministry of Exterior Affairs that was distributed through the Agencia Español del Cooperación Internacional and the Turkish Ministry of Education. It is the first publication on the topic, out of my Ph.D. in Architecture dissertation entitled A Typology of Ancient Theatres in Modern Spain and Greece: A Geo- Historical Approach, which was completed in May 2005 at the Middle East Technical University in Ankara, Turkey. I wish to thank the reviewers and editors of this volume for their insightful comments, which led to a re-structuring of the initial draft of the paper into this much more integrated present version.