A Critical Approach to the Concept of Resistance: New ‘Traditional’ Rituals and Objects in Funerary Contexts of Roman Baetica

Alicia Jiménez

Introduction In this paper I would like to explore the usefulness of the concept ‘resistance’ as a tool to interpret the consumption of ‘archaic’ looking objects and the performance of ‘traditional’ rituals, employing as an example the funerary contexts of Roman Baetica. Traditional approaches to this notion often portray an ‘inverted’ image of the controversial ‘Romanization’ theory in those places where native communities ‘failed’ in the process of becoming Roman. In some contexts, the concept of ‘resistance’ has also been used to symbolically connect contemporary populations – subjected in the recent past to European rule – with a glorious history of opposition to the Roman invader. In this respect postcolonial theory provides us with new tools to stress the links between the present and the interpretation of the past and to overcome binary oppositions such as conqueror:conquered, Roman:native, civilization:barbarism and domination:resistance.

1. Resistance as a colonial concept Resistance per se is not a new concept in archaeology, but appeared as early as the thirties in the work of scholars working on acculturation. In 1936 R. Redfield et al. claimed that one of the possible results of intercultural contact was the development of movements that would try to counteract the inferiority imposed on a certain group by referring to prestige elements associated in the collective imagination to a phase predating the first contacts (Redfield et al. 1936: 152). Vogt et al. (1954: 987) described this kind of processes as a ‘reactive adaptation’ that lead to the reassertion of native ways, the restoration of ancient cults and the emergence of isolationist programmes or nationalist movements. Similarly Gruzinsky and Rouveret (1976: 199–204) argued that a failure in a process of acculturation sometimes may produce examples of ‘resistance’ or ‘counter-acculturation’.

The idea of ‘resistance’ has also been especially important for historians and archaeologists born in countries decolonized in the recent past. If French or British writers made a connection between the ‘Romanization’ of regions like North Africa and the ‘civilizing mission’ of certain European nations, North African researchers, like their ‘conquerors’, stressed the link between ancient and contemporary occupations, assimilating modern national movements of liberation with the fight for freedom of their ‘ancestors’ (Mattingly 1996: 49, 57; Hingley 2005: 40). This connection between the present and pre-Roman ancestors, interestingly enough, has also been made by European scholars. Hingley (2000: 72–85) has studied some English instances (Boadicea, Caratacus), Hessing (2001) the equivalent ‘Batavian myth’ in Dutch historiography and García Moreno (1988: 83), Jimeno and de la Torre (2005) the role played by Viriatus and the ‘heroic’ example of Numantia in Spanish literature, to mention only a few examples. As a consequence, several nations obtaining political independence during the 19th and 20th centuries were involved in a process of creating new narratives that were intended to portray an ‘authentic past’- drawing on a pre-colonial History. In many cases, the colonial period was presented as a phase of rupture with the real essence of the homeland, which would have been only recovered after the wars of independence against European powers. The native intellectual must ‘make an aggressive response to the colonialist theory of pre-colonial barbarism’, wrote Franz Fanon in 1961, ‘[A]s for we who have decided to break the back of colonialism, our historic mission is to sanction all revolts, all desperate actions, all those abortive attempts drowned in rivers of blood. […] While politicians situate their action in actual present-day events, men of culture take their stand in the field of history’ (Fanon 1961:198–199). The important contribution of Bénabou to the study of the ‘Romanization’ of North Africa needs to be understood in this historical context. Bénabou cast doubts on the supposedly spontaneous ‘Romanization’ of natives willing to replicate the Roman way of life and defined two types of ‘résistance africaine’: military resistance -the reaction of a human group to foreign domination- and cultural resistance – the refusal of innovations and the desire to preserve the native personality against the influence of others- (Bénabou 1976: 369). However, this way of understanding history – in opposition to the history imposed by the colonizers – shares in some instances important traits with the type of narratives imposed by Europeans, implying, ironically, that to some extent a certain colonisation of the past has also already taken place (Rowlands 1994: 135).

‘Resistance’ is, therefore, a notion that has deep roots in contemporary experiences of colonialism and as well as an interpretative model derived from a colonial view of reality that requires the existence of two sets of distinctive cultures: Roman and native, with roles of dominance and rebelliousness respectively (van Dommelen 1997: 306). Resistance is, in this sense, a perfect mirror of the traditional vision of ‘Romanization’. As Greg Woolf has pointed out, understanding the adoption of Roman culture essentially as a side effect of Roman power has led to the interpretation of any ‘failure’ to adopt it as an example of ‘resistance’, equating rejection on the cultural plane to military rebellion. As Greg Woolf claims, the birth of nationalism made easier to explain a conflict between peoples as a conflict between cultures, but not even in the earlier phases of the conquest is it possible to put the adoption of Roman culture on an equal footing with the acceptance of Roman rule (Woolf 1997: 340; 1998: 20–22). Military resistance in Gaul, for example, was not linked to a rejection of Roman culture, since the ancient sources have recorded instances of rebellions by Gauls who were not only Roman citizens but had also held positions of responsibility in the imperial administration. Another case in point is Arminius, later to become a ‘national hero’ in Germany. He is most famous for his victory as leader of the Cherusci over the Roman legions in the Teutoburg Forest in 9 A.D., despite the fact that he was, as the very least, a Roman citizen (Struck 2001: 98). In these cases the elites posed ‘resistance’ to the Roman threat using ‘Roman culture against the Romans, to win power in the new order’ (Woolf 1998: 22).

It is clear that using Roman material culture does not lead to becoming Roman nor can ‘failing’ to adopt it be always interpreted as a clear act of resistance. In fact, ‘resistance’ appears to be only in some analyses a useful concept to contrast the acceptance of Roman culture by the native elites and the resistance to it by the low strata of conquered peoples. Of course, I am not the first to point out that Roman culture was not simply an elite culture and that finding the thin line between high and low culture in Rome is, on the whole, simply not possible (Woolf 1998: 18). What is interesting here, however, is that the cliché of uncultivated peoples rejecting ‘Romanization’ is not restricted to contemporary literature. In the ancient accounts of the conquest we also find references to fierce rebellions of savage inhabitants of distant or mountainous regions who could not adapt to the life in the city. Opposing Roman culture is, in this sense, opposing ‘civilisation’ or ‘humanitas’ in the language of the classic writers. Humanitas is to a certain extent an ambiguous term originally used as a synonym of the Greek word paideia. It allowed the speaker to make a distinction between educated men and crude individuals. Humanitas means improvement of the self as much as human control over the surrounding world in the form of cities, stone houses, public buildings or cultivated fields (Tacitus Agr. XXI, 1–2). In this sense it bears a close resemblance to the modern concept of ‘civilisation’ and makes clear why, from this point of view, some men are wild, like plants, while others ‘have evolved’ thanks to culture (Caesar BG I, 1, 3; Veyne 1991: 398; Gros 1998: 144–145; Woolf 1995: 15; Woolf 1996: 370– 371; Woolf 1998: 54–60). At this point we should ask ourselves if the interpretation of the past through sets of binary oppositions (Roman:non-Roman, humanitas:barbarism, colonist:colonised, assimilation:resistance) is only a product of a present-day interpretation of ‘Romanization’ as a civilising process or whether we should also recognize the intrusion of ancient discourses in contemporary readings of the past, that, traditionally, have been taken at ‘face value’ compared with ‘subjective’ interpretation of the information offered by material culture . Ancient authors made use of a sophisticated discourse about humanitas to justify the wars of conquest (bellum iustum) (Lomas 1996: 46; Dueck 2000: 115–122). According to some ancient authors (e.g. Aristotle Pol. 5; Cicero De re publica 3, 36; De officiis, 2, 26), Nature shows us that the stronger rule over the weaker. From this point of view, the government of the best is seen as good for the conquered, while the bravery of brutal natives is stressed when necessary in order to enhance the merit of the conquerors. Peoples that are labelled as ‘warlike’ in commentaries on the conquest by Polybius are considered only ‘bandits’ in texts by Strabo, Livy or Diodorus describing the Empire at the turn of the era, in accordance with the propaganda about the peaceful reign of Augustus (pax augustea) spread through different media at this time. Is it a nineteenth century depiction of social evolution signalling that the noble but innocent natives of Iberia are shepherds and survived thanks to banditry, while those who were farmers and lived in cities were highly developed? Or is it a romantic view that shares some points in common with the myth of the ‘noble savage’ present in Stoic philosophy and read by those explaining ‘Romanization’ in the nineteenth and early twentieth century?

2. New ‘traditional’ rituals and objects in necropolis of Roman Baetica We might have over-simplified the complex problem of ‘resistance’ in our analyses, but it is obvious that ‘resistance’, from armed rebellions to the so-called ‘silent resistance’ (Scott 1985; Scott 1990), did occur in diverse contexts and at different levels and, therefore, it must be taken into account in our studies of the changes in power structures that followed the Roman conquest. Colonization creates a tension between unification of cultural traits and discrepant identities – as defined by Said (1993: 35–50) – and studied recently through the archaeological record by Mattingly (2004: 22; 2006: passim), giving shape to different versions of Roman and non-Roman identity in the provinces of the Empire. But cultural bilingualism also allowed different individuals to present in different contexts a ‘Roman’ or a ‘local’ face (Mattingly 2006: 527). To the latter nuances we should add the fact that false compliance can be sometimes interpreted as ‘Romanization’ and, conversely, integration of ‘traditional items’ into the hybrid Roman identity of the Empire can be understood as resistance. This is particularly problematic because links with the past can be used discursively for very different purposes, such as, for example, integrating native narrations in a local Roman myth about the origins of the group. I will take as an example of these difficulties certain funerary contexts of southern Iberia and analyse some of those ‘native survivals’ that fit uneasily in our description of Roman Baetica. The south of Spain was one of the first regions to be conquered at the end of the third century B.C. Iberia was then divided into two provinces. The south took initially the name of Ulterior and was renamed Baetica during Augustus’ reforms. Roman Baetica has traditionally been considered the homeland of highly developed native peoples, a prosperous territory where cities flourished well before the Roman conquest. According to Strabo, the Turdetanians and ‘particularly those that live about the Baetis, have completely changed over to the Roman mode of life, not even remembering their own language any more. And most of them have become Latins, and they have received Romans as colonists, so that they are not far from being all Romans’ (Strabo Geography III. 2. 15, Loeb Classical Library. trans. H. L. Jones, London, 1960). However, it is not until two hundred years after the beginning of the Roman conquest that significant changes can be noticed in funerary contexts of the province. It is true that there were important similarities before the conquest between the funerary rituals of extensive regions in the Mediterranean (Fuentes 1992: 600). Apart from certain cities where the Punic roots were especially strong, native populations in southern Iberia usually cremated their dead, like the Romans, and placed the remains inside urns, even though some rituals, the location of cemeteries and the monuments located in them were quite distinctive to the region.

For the first time, in the last decades of the first century B.C. the funerary space of some cemeteries is planned according to a ‘Roman’ concept, and for the first time tombs are aligned along the roads, close to the city walls. This new way of placing the dead in relation with the living is highly relevant. It implies a new concept of the religious limits of the city (pomerium), the establishment of a different system of land division in funerary plots (recorded by epigraphy as well as archaeological remains) and a novel way of interacting with ancestors and with the history of the community every time people left or entered the town. In the same period the necessity of marking the tomb with the name of the deceased spread for the first time in the province and some of the first funerary epigraphs could be found in the cemeteries. In parallel with these developments, some decided, for the first time, to build for themselves or for their families funerary monuments resembling those that were built at the time in Italy or the Urbs itself (Jiménez 2008: 315–348). The coexistence of alternative ways of honouring the dead is only given full significance in the context of these ‘new’ cemeteries ‘resembling’ their Roman counterparts. Corduba was one of the most ancient colonies in Turdetania (Fig. 1). The city was the capital of Ulterior, first, and Baetica and the conventus Cordubensis, later. At the beginning of the Roman Empire monuments imitating the actual tomb of Augustus were built at the western entrance of the city (Fig. 2). There were funerary streets, divided into fenced plots, according to the Roman system of measurement (in agro pedes, in fronte pedes) (Fig. 3) and the passer-by would contemplate a variety of funerary constructions decorated in a Roman fashion (Fig. 4). One of the highest concentrations of gladiatorial epitaphs outside Rome has been unearthed in Colonia Patricia along a Roman road that led to the recently discovered amphitheatre (Vaquerizo 2001; Vaquerizo 2002a; Vaquerizo 2002b; Murillo et al. 2002; Márquez 2002). In the same period, next to one of the western roads, a group of people decided to bury their ashes inside containers resembling the shapes and painted decoration of pre-Roman pottery (Fig. 5) (García Matamala 2002). Those pots, which have been found in small quantities in the contemporary levels of the ancient city, had the highest ritual importance, since they were meant to receive the funerary remains of the deceased. Even the poorest people, who could not afford any kind of funerary goods, were at least provided with a funerary urn of this type. The important thing to stress here is that they were not copies of the pottery used in the region for four hundred years before Augustus, but Roman wares freely inspired by local shapes and decorations, present in the region for several centuries, in order to confer a ‘traditional’ appearance on these recipients. In most of these fenced plots few ‘Roman’ monuments have been found. However, several of these customary cinerary urns covered by a humble bowl were present, as was the case in many local cemeteries before Roman colonisation.

Figure 1. Map of the Iberian Peninsula including the provincial and conventual borders after Augustan reforms and the capitals of the conventus (Corduba, Astigi, Gades and Hispalis). Location of the three main sites cited in the text: 1. Baelo Claudia, 2. Castulo and 3. Colonia Patricia/Corduba.

Figure 2. Colonia Patricia. Western necropolis. Funerary monuments restored in Paseo de la Victoria (Photo: A. Jiménez).

Figure 3. Colonia Patricia. Northern necropolis. Funerary street with fenced plots (Vaquerizo 2002b: fig. 3). Figure 4. Colonia Patricia. Northern necropolis. Hypothesis of reconstruction of the funerary monument found in calle Adarve (Vaquerizo, 2001: figs. 11b y 11c).

Figure 5. Evolution of one type of traditional pre-Roman urn into Imperial times (after García Matamala, 2002: fig. 8).

Interestingly, one of these enclosures protected a funerary chamber discovered at the beginning of the 20th century (Fig. 6). To find something similar we would have to travel to Punic necropolis, such as Villaricos (Almería), where the burial chambers built in the 6th–5th century B.C. were still in use in the early Empire (Astruc, 1951; Almagro Gorbea 1984), or to Carmona (Seville) a city with deep Punic roots. In Carmona we can find one of the best examples of Punic ‘nostalgia’. Outside the city – again from Augustan times on – several funerary chambers similar to those located in Punic North Africa were built and decorated with painting on the inside, making use of Roman designs (Fig. 7). Nevertheless, they followed ritual patterns that are reminiscent of ancient Punic customs like sealing the entrance of the chamber tombs with ashlars that would only be moved when it was time to bury a newly deceased family member (Fig. 7) (Bendala 1976; Bendala 2002). A similar device has also been found in funerary chambers of Cordoba (Fig. 8). This is in contrast to the Roman custom of visiting the tomb on several occasions: festivities of the dead – parentalia, lemuria – and the birthday of the deceased (dies natalis). The chambers in Cordoba and Carmona share, however, ritual cavities to pour libations during the year. Strikingly the Roman necropolis of Carmona had very little sigillata to offer to archaeologists. The exclusion of sigillata from the funerary goods is a trend that has been also noticed in certain necropolis of southern Spain (sectors of Cordoba, Baelo Claudia, Castulo, etc.) (Bendala 1991: 184–186; Jiménez 2002). In my opinion perhaps this absence should not be explained only as a matter of distribution or commercial networks, as some have claimed, because sigillata has been found in many of these cities. It would also be difficult to fall back on the argument of the difference in wealth in examples like Baelo Claudia, where even in monumental areas of the necropolis the sigillata is not present (Fig. 9). The study of the funerary rituals of Castulo (Linares, Jaén) is of especial interest from this perspective. Castulo is another city with an important pre-Roman history, mentioned by the ancient sources as the centre of a mining district. Several cemeteries are known to have existed around the settlement between the 8th century B.C. and the first century A.D. (Fig. 10), providing the opportunity to study the graves of the grandfathers and greatgrandfathers of those buried in the ‘Roman’ tombs at the beginning of the first century A.D. Once more, the research at Castulo produced ambivalent results. The site has yielded one of the biggest collections of Roman funerary reliefs (Fig. 11) of the upper Guadalquivir valley (Baena and Beltrán 2002) and, at the same time, examples of imperial necropolis where the pre-Roman past seems to have been brought back to life through the usage of traditional items (Blázquez and Molina 1975; Canto 1979).

Figure 6. Colonia Patricia. Western necropolis. Roman funerary chamber known as ‘the Great Tomb’. The cuts in the door indicate it was blocked with ashlars. Possible ritual cavity for libations (on the right) (after Romero de Torres, 1941: fig. 42).

Figure 7. Necropolis of Carmona. ‘Mausoleo cuadrangular’. Ritual cavities for libations (after Bendala, 1976: fig. 29).

Figure 8. Colonia Patricia. Northern necropolis. Funerary chamber found in calle de la Bodega (Vaquerizo 2002a: fig. 17).

Figure 9. Baelo Claudia. Eastern necropolis. Mausoleum number 496 (Paris et al. 1926: fig. 43).

In the cemetery placed on the road leaving the city from the northern gate some cremations were placed inside a fenced plot in the first century A.D. The funerary urns might have looked quite traditional at the time, if we compare them with those found in the pre-Roman necropolis (Fig. 12). The funerary recipients were covered by a bowl as they were in ancient times. Since the Phoenician red gloss cups that were traditionally used as funerary goods were not available in the early Empire, the inhabitants of Castulo chose to use probably for the same ritual purpose similar vessels, like Roman ‘eggshell’ ware cups or local versions of ancient Phoenician drinking cups. All these traditional devices were nothing but a display of things that looked old-fashioned in the first century A.D. because of their links with pre-Roman local material culture. Yet if we study the pre-Roman necropolis of the town, things looked the same ‘but not quite’, to use the famous words of Bhabha (1994: 86). By the north gate, virtually all the tombs followed the ‘traditional’ Iberian ritual of placing the ashes in a container covered by a bowl. However, in the ancient necropolis of the site some cremations lay directly on a hole in the ground, not inside an urn; in other cases the remains of the body were left to rest on the burial pyre accompanied by several objects, and finally, as in later times, only in some cases were the bones and ashes selected after the cremation to be buried elsewhere inside an urn. Whilst in earlier cemeteries the sides of the burial hole were covered with clay, throughout the imperial period people used the body of an amphora for the same end. The ancient inhabitants of Castulo included some personal items (arms, rings, pins) that are absent in Roman times, but used pretty much the same ritual objects (urn/bowl, drinking pots). Nevertheless, before the Roman conquest, people living in Castulo never placed all these objects intact in the grave. Things needed to be ritually killed, broken after the funerary banquet, to reach the underworld with the deceased. During the early Empire, the dead person kept their burial goods in perfect condition for the journey. In these contexts, we should probably attach some meaning to the fact that in the imperial necropolis of Castulo, a Roman city, once again, the tombs of the northern gate were not provided with sigillata (Jiménez 2008: 131).

Figure 10. Castulo. Location of ancient cemeteries (after Blázquez 1985: fig. 1, modified by the author).

3. A postcolonial critique to the colonial concept of ‘resistance’ If recent studies about ‘Romanization’ have dismissed essentialist views of Roman society it is important not to apply an essentialist model to native culture. G. Spivak has pointed out that nativist ideas only reproduce a European fantasy about European origins through the image of a lost pure culture, the culture of the Other and criticized several instances of this kind of nostalgia (Spivak 1988a: 129; Spivak 1988b: 128; Vega 2003: 299; van Dommelen 2007). The use of binary terms such as conqueror:conquered, Roman:native or domination:resistance in our analysis of the local culture of the Baetica province – at the time a territory under Roman control for two hundred years – are fundamentally flawed (van Dommelen 2001a: 81; van Dommelen 2001b: 141; Rowlands 1998: 328). The relation between culture and resistance is clearly more complex. Rome was resisted by individuals making use of elements of Roman culture and, as the research about examples of ‘mimicry’ in different colonial situations has showed, the use of Roman traits cannot be equated with mere ‘emulation’ (Bhabha 1994: 85–92; Webster 2003: 30–40). Similarly, not every regional variant can be interpreted as an example of resistance, just as not every instance of use of Roman material culture can be read as a symbol of submission to the Empire. And, of course, local objects and rituals did not keep their meanings intact for centuries, either before or after the Roman colonisation (van Dommelen 2007).

Figure 11. Fragment of a funerary monument decorated with masks (Baena and Beltrán 2004: fig. 30.1).

In any case, archaic items that have been interpreted as ‘native survivals’ are not copies of the past, but, as we have seen in the case of pottery shapes or funerary rituals, a re-elaboration of what was possibly believed to be the past by a certain community. New ‘traditional’ items of this type allow change in culture through a supposedly faithful repetition of customary ways of ‘doing’ things. In this kind of settings, ‘truth’ and ‘authenticity’ involve a continuous repetition of an apparently ‘original’ way of being (Rowlands 1988: 44–45). The past forms of objects help to create a path for originality and, in this sense, ‘also give rise to categories of thought and notions of sensibility’ (Gosden 2006: 440). ‘Traditional’ material culture and rituals are especially meaningful when found inserted in a dialogue with changes caused by colonisation. I suggest that they might have been, to some degree, statements about perceived links with the past in one of the most powerful settings for the negotiation of identity: funerary rituals. Lineages project an image of the place we come from; they provide us with a perfect stage to make claims about who we are through a representation of what we were. Recently, Hingley (2008) has described the Roman Empire as an entity comprising different groups that were able to retain the core of inherited identities in a heterogeneous society. Performing ancestor cult in a certain fashion was quite probably one way of integrating different local pasts present in the hybrid communities of southern Iberia.

Figure 12. Examples of persistence of traditional ceramic types in the necropolis of Castulo. a. vase from the pre-Roman necropolis of Los Baños de la Muela (Blázquez, 1975: fig. 43.3), b. vase from the Roman necropolis of the northern gate (Blázquez, 1975, fig. 146.3), c. Funerary urn from the Roman necropolis of the northern gate (Blázquez, 1975: fig. 137), d. Funerary urn from the pre-Roman necropolis of Los Patos (Blázquez 1975: fig. 43.1).

‘Archaic’ objects and rituals are testimonies of the existence of several discourses in a settlement about the meaning of ‘being Roman’ in a given time and context. They can be understood, to a certain extent, as alternative forms of self/group definition to those we can read in Roman literature or the iconography of the forum. It is quite feasible that archaic objects were seen as traditional features that linked the group with its pre-Roman past and possibly, at the same time, allowed the integration of local memories into new narratives about the meaning of romanitas in the provincial setting. All these elements were constituents of a hybrid identity formed after a colonisation process, in which probably different layers -individual, family, city, province- coexisted without being perceived necessarily as contradictory. The metaphor of bilingualism and codeswitching advanced by Wallace-Hadrill (1998) to understand the process of hellenization in Rome can be useful in this respect. Like bilinguals who are able to use different languages discursively depending on the context, the inhabitants of Roman Corduba might have been equally competent to read and relate to the imagery of Aeneas in the forum and the traditional burial rituals in the necropolis. The interesting thing here – as in the problem of the hellenization of Rome – is that we are not facing a simple opposition between the public and the private sphere in the provinces, because the dominant or hegemonic language in the cemeteries of Baetica seems to be quite variable and sometimes almost idiosyncratic of a given city. The importance of studying cities and their surrounding territory as contextual units and avoiding broad generalizations has also been stressed, for example, by Terrenato (1998) in his study of the city of Volaterrae. He also pointed out the contrast between large-scale urban architecture and the consumption of imported prestige goods (usually interpreted as massive acculturation) and examples of continuities displayed, for instance, in elite funerary rituals. On the other hand, individuals buried with imported items in buildings inspired by ‘Roman monuments’ displaying Latin inscriptions, sat alongside supposedly pure ‘Iberian’ tombs. In reflecting and contrasting each other, both made a statement about different types of social identity – ancestry, wealth, gender, age group, occupation – including local memories about how ancestors must be properly commemorated, despite the fact that both of them can be seen as new forms of expressing the meaning of being Roman in the provinces at the beginning of the Empire.

Not using certain imports – like sigillata – as funerary goods, or following traditional patterns in funerary rituals or architecture did not undermine the power of the emperor or the provincial elites, as we would expect they would do, if we were looking for traits of resistance according to the traditional concept, but we should probably attribute to these performances in relation to ancestors at least the same importance in our studies as to the discourse about other types of Roman identity spread through different media (the forum, coin iconography, literature) by the state, studied in the case of Augustan Rome by Zanker (1987). This is especially true if we take into account that the path of cultural transformation in Rome and in some conquered areas seems to be very similar at that time (Wallace-Hadrill 2000: 293; Woolf 1997: 346). It would be naïve to think that the only message to be read, in both the ‘official’ discourse and the ‘private’ discourse, is one merely related to grand ideological narratives of Empire and resistance, because all these discourses were deeply interwoven into power and social relations at different scales, symbolized through metaphors of ‘possible pasts’ in the case of contexts related with the cult of the ancestors. Consequently, if the study of alien items of material culture that helped colonial elites to create power relations is meaningful, so too should be the analysis of the ‘reaction’ to such representations by certain communities (Hingley 1997:88) or, at the very least, we should acknowledge the existence of alternative representations of collective identity that might have stressed the mixed origins of certain groups of population inside the Roman Empire. Richard Hingley (2005: 70) wrote recently that ‘Imperial discourse was a generalized ideology – not fully conscious’. It is likely that the same could be said about certain aspects of ‘resistance’ understood in this new form. Is evasive behaviour always a conscious form of resistance? One of the most difficult questions to tackle is to what extent this type of ‘discrepant experiences’ or humble acts of ‘everyday resistance’ – if we want to use the label proposed by J. C. Scott – were perceived as such by individuals that had, at the same time, the need to use the ‘cultural vocabulary’ of the dominant group (van Dommelen 1998: 27–28). But, in any case, these alternative discourses -or ways of reaffirming local ways of ‘doing’ certain things- integrated into the Roman provincial society, coexisted in the 2nd–1st centuries B.C. with different forms of opposing Roman rule: from military revolts to other types of the so-called silent resistance, like tax evasion, pilfering or poaching (van Dommelen 2007).

We could then understand these examples, following Foucault (1976: 92–96) not as a dominant and a rejected discourse, but as a multiplicity of discursive elements that are able to act in different strategies in the context of the continuous re-negotiation of Roman identity and power relations in a colonial setting. Certainly to consider all these behaviours as some form of resistance we have to understand power not as institutions, superstructures, individual strength or an allencompassing opposition between rulers and ruled. According to Michel Foucault (1976: 94):

‘Power comes from below, that is, there is no binary and all-encompassing opposition between rulers and ruled at the root of power relations’. Power relations are immanent in every type of relationship, in this view and, therefore ‘[W]here there is power, there is resistance, and yet, or rather consequently, this resistance is never in a position of exteriority in relation to power’. The Roman Empire could be seen, in that case, as constituted by a ‘less dichotomous and more intricate pattern of inequality’ where heterogeneity was an important force to perpetuate the imperial order (Hingley 2008).

Finally, it is important to recognize, too, that resorting to archaisms – or not adopting ‘Roman novelties’ – in the context of the ‘cultural revolution’ of the early years of the Empire, is not a feature of ‘uncultivated’ peoples or ‘popular’ culture (Hingley 1997: 95). Archaisms should not be explained as a characteristic of low strata and therefore deprived of meaning. Examples to support this point can be easily found in Rome itself, where even Augustus made use of them in his iconographic program, or in Italy, where the well-to-do employed all kinds of archaisms, especially in contexts related with the cult to ancestors. Ancient authors devoted considerable time, however, writing about resistance to Roman power in lands inhabited by barbarians and the lack of it in civilised regions such as Baetica. We have inherited from Roman historians a colonial discourse in our analyses of ‘resistance’, a line of argument based on the notion of humanitas –a word closely related with modern concepts such as ‘civilisation’- that was extremely successful in imposing the colonizer view not only over the past, but also over the present.

Centro de Ciencias Humanas y Sociales, Instituto de Historia. Consejo Superior de Investigaciones Científicas (CSIC). SPAIN

Acknowledgements Research funded by the European Social Fund (Contrato Postdoctoral I3P). Research Project SP6.E.58/03. I am very grateful to Richard Hingley and Peter van Dommelen for their comments, their bibliographical suggestions and for giving me access to unpublished material. The remarks made by an anonymous referee and Corisande Fenwick, helped me to clarify some important points of the paper. I am also in debt to Lyra Monteiro and Michael Rowlands-Hill for improving the English of this text.