Before and After Rome: The Incised Contours Technique in the Art of Gallia Narbonensis

The art of ancient Gallia Narbonensis constitutes a coherent corpus thanks to a number of shared visual characteristics. Prominent among them is a carving technique that consisted of deeply incising the contours of figures in sculpted reliefs, a trend that was rare in Roman art elsewhere. This paper investigates the presence of incised contours in monuments erected between the late first century bc and the early decades of the first century ad at sites such as Orange, Carpentras, and Glanum. I argue that we can explain this peculiar feature by establishing continuities with pre-Roman monuments created in the region, which reveal a similar aesthetic solution. Relying on visual analysis and employing the lens of two postcolonial paradigms, creolization and hybridity, I offer a reconsideration of well-known provincial monuments, expanding our understanding of the role of visual culture as an essential marker of the colonial transformation of Gaul.


Introduction
The art of Gallia Narbonensis, one of the earliest Roman provinces in the West, is characterized by a remarkable visual coherence. Besides the natural visuality that local building materials give to the monuments, certain iconographic motifs and technical traits define the art produced in this region as a distinct group. Prominent among them is a carving technique that consisted of deeply incising the contours of figures in relief sculpture, a trend that while rare in Roman art elsewhere, appears systematically in Narbonensis. The application of this technique resulted in a strong contrast of light and shadow that enlivened figures and added dynamism to the compositions. This paper investigates the presence of the incised contours in several monuments erected during the late first century bc and the early decades of the first century ad at sites such as Orange and Carpentras (in modern Vaucluse), and Glanum (modern Saint-Rémy-de-Provence). I trace the origins of the motif to pre-Roman monuments and explore the implications of its persistence for the expression of Gallic identities in areas under Roman control. I argue that after the Roman conquest the art of Narbonensis retained technical and aesthetic features that signal the active involvement of local artisans versed in traditional ways of making. Elements of native traditions that were already present before Roman intervention were assimilated and repurposed to create a new form of artistic expression. Relying on visual analysis and employing the lens of postcolonial theory, with an emphasis on creolization and hybridity, I reconsider a group of well-known Gallo-Roman monuments, expanding our understanding of the role of visual culture in the colonial transformation of Gaul.
Several studies have brought attention to the incised contours in the art of Gallia Narbonensis (Hatt 1951;Toynbee 1953;Rolland 1970;Kleiner 1977). In his exhaustive exploration of Gallo-Roman funerary monuments, La tombe Gallo-Romaine, Jean-Jacques Hatt suggested that the contours were the result of marks left in the stone by the hand of inexperienced artisans who worked under the supervision of Greek or Italian sculptors (Hatt 1951: 127-131). 1 Fred Kleiner later revised this suggestion, noting that many of the grooves were made after the figures had been carved (Kleiner 1977: 682). Following Toynbee (1953: 95), Kleiner suggests that the contours were a deliberate aesthetic choice (1977: 683). 2 He hypothesizes the existence of a single workshop of traveling artists who completed many of these monuments in the region, at sites such as Glanum, working under the direction of Italian masters (1977: 679-680). As he rightly notes, artistic production intensified after the conquest, in particular after the foundation of Arles, and workshops might have benefitted from the coordinated work of native and Roman artisans; indigenous artists brought an expertise in carving local materials (mainly limestone) and foreigners provided new iconographical themes desired by Italian patrons. Although the idea of cooperation between local and Italian artists is compelling, it does not go far enough to explain the pervasiveness of the contours as a visual strategy, which appears repeatedly and almost exclusively throughout Narbonensis. 3 This essay builds upon this previous body of work to explore the possibility of a continuity between pre-and post-Roman art.

Evidence and Theories
One problem that arises when investigating the art of pre-Roman Gaul is the scarcity of the evidence. Archaeology provides essential information about life in this region, but many of the monuments survive only in fragments now scattered through regional museums. 4 Literary sources offer only partial clues about the geography, history, and population of ancient Gaul. Greek and Roman authors including Posidonius, Strabo, and Diodorus Siculus, recorded observations about the landscape and the people of Narbonensis. Their accounts naturally reflect a Rome-centric point of view, and they usually construct an othering image of the inhabitants of Gaul; many of these authors do not seem to have visited the region. 5 Whenever available, epigraphy offers useful information on the identities of individuals commissioning architectural monuments and commemorative structures, especially after the Roman conquest (Hatt 1951;Duval 1953: 22 and passim;Goudineau 1996: 467-469, 496;Mullen 2013: 122-143). Because of the limited and problematic nature of the available evidence, this study relies on a combination of diverse types of data-visual, archaeological, literary, and epigraphical-understood through postcolonial theory. Previous studies of the incised contours have primarily followed the model of Romanization, the longest-standing paradigm in the study of Roman expansion and colonization. Although recent scholarship has attempted to recover its advantages in a more nuanced dialogue (Mullen 2013: 9-14;Woolf 2014), Romanization has largely prioritized a hierarchical top-down approach, focusing primarily on the experiences of elite individuals-both Roman and native-at the expense of less visible groups, such as non-elites, women, and slaves. 6 A different way to approach the incised contours issue is to look at traces of the same technique that seem to have survived in monuments created before the Roman conquest. Although necessarily speculative, this methodology allows us to envision a continuity between pre-and post-Roman visual production, presupposing the merging of old and new traditions in the creation of a new, regionally specific artistic style. To tackle this issue from a postcolonial theoretical perspective, I turn to two interconnected paradigms, creolization and hybridity, which are particularly pertinent to describe the production of material culture in a colonial context.

Creolization
Creolization was first employed to describe the merging of languages in the formation of new dialects in the colonies of the Caribbean (Abrahams 1983). Scholars soon employed creolization to describe the process of creation of a new cultural code that blended African, European, and indigenous beliefs and behaviors in the Caribbean and other parts of the Americas (Braithwaite 1971;Chaudenson and Mufwene 2002;Stewart 2007). For example, studies of the material manifestations of slavery in the United States have revealed how artifacts that were introduced by European slave owners were used in ways that reflected distinctly African customs and traditions (Ferguson 1992;Yentsch 1994). From the perspective of material culture, creolization proposes the blending of elements from two or more pre-existing cultures, as opposed to the replacement of one culture by another. The process is conditioned by the power imbalances that are inherent to relationships between colonizers and colonized, and the result is a clearly ambiguous assemblage of material culture that reveals the adoption of diverse cultural codes according to shifting historical circumstances.
A few studies have applied the principles of creolization to the study of Roman cultural history, with special emphasis on the Western provinces (Webster 2001;2005;Carr 2003;. Jane Webster has championed creolization as a way to understand the creation of a new identity in central-eastern Gaul after the Roman conquest, looking at how the iconography of native deities, such as the equine goddess Epona, seems resistant to syncretization with Roman gods (Webster 2001). Noting how creolization helps explain the high degree of ambiguity that characterizes the material record of Roman Gaul, she proposes a sliding scale of appropriation of Roman culture, or what she terms an ' adaptive synthesis' of Roman and native traditions (Webster 2001: 217). Creolization offers a counter narrative to Romanization because it allows us to analyze the process of cultural formation from the ground up, as opposed to understanding new cultural products as the result of imitation of Roman prototypes by non-Roman elites (Webster 2001: 218-219).
The model of creolization has not gained currency in Roman studies. Critics have noted that it presupposes the existence of distinct cultures that develop independently before mingling in a colonial situation ( van Dommelen 1998: 31). Some have brought attention to the anachronisms inherent in extrapolating linguistic concepts to anthropological phenomena in vastly distant geographical contexts (Mullen 2013: 66;also, Palmié 2006). Yet creolization offers a valuable avenue for understanding certain aspects of the colonization process in Gaul. One area where this paradigm could be expanded is in an exploration of the material manifestations of the conquest with an emphasis on artistic production. A handful of studies have addressed these issues, if only superficially (Webster 2001;2005;Jiménez 2011). Webster, for instance, deals with representations of deities that became popular or originated after the Roman conquest, but her analysis involves minimal observations on the iconography, materiality, and sensory aspects of the works discussed. In this essay, I test the usefulness of creolization for understanding the creation of a new artistic style in a provincial setting, looking in detail at a distinctly local carving technique.

Hybridity
A second theory that offers a path for understanding ancient colonial visual culture is hybridity, in particular as formulated by Homi Bhabha as a means to describe the postcolonial world (Bhabha 1988;1990;. Bhabha has referred to hybridity as a permanent and continuous aspect of culture(s), a 'third space' that allows the emergence of new cultural positions from pre-existing ones. That third space in turn, '… displaces the histories that constitute it, and sets up new structures of authority, new political initiatives, which are inadequately understood through received wisdom' (Bhabha 1990: 211). Hybridity is a process that generates new meanings from the disparate meanings of old traditions; the resulting cultural products are generally characterized by a high degree of ambiguity. Critics of hybridity have raised concerns about the use of the model in a manner that reinforces notions of cultural purity (Gilroy 2000: 250-251;Palmié 2013: 464-465). They have also pointed to hybridity's implicit Eurocentrism, noting that it tends to reveal more about the interpreter than about the supposedly hybrid subjects it seeks to understand (Liebmann 2015: 320-321). Yet hybridity can productively be adapted to the study of ancient material culture, especially when considering specific cultural phenomena at a micro geo-historical level.
For example, Peter van Dommelen (1997) has employed hybridization as a corrective to dualistic, Westerncentric views of the Carthaginian colonization of Sardinia between the fifth and the third centuries bc. Looking at artifacts such as terracotta figurines found in various localities throughout the island, he demonstrates how they reflect indigenous interpretations of diverse technical and stylistic traditions originating in Sardinian, Punic, and Italic contexts; the resulting assemblages represent a new, hybridized cultural practice (1997: 318). Similarly, Alicia Jiménez has explored the concept of hybridity for the study of material culture in the Iberian Peninsula (2008;. Her analysis shows how hybridity can illuminate the processes of cultural change brought about by the Roman colonization, with an emphasis on the production and reception of visual representations. Analyzing Iberian sculpture of the late Republican and early Imperial periods, Jiménez demonstrates how these works show clear signs of the intentional hybridization of native and Roman attributes (2011: 105-112). For instance, some of the sculpture she discusses, likely portraying local worshippers, combine traditionally Iberian attire with Roman togas and armor. The contexts in which these images appear-mostly suburban sanctuaries and cemeteries-suggest a continuity with native funerary practices (Jiménez 2011: 106-107). Applied to these cases of visual production, hybridity subverts previous interpretations of the Ibero-Roman sculptural corpus, which tended to see such works as poor imitations of Roman art. Instead, hybridization signals an active reception of Roman sculpture by local artists who highlighted those elements they considered representatively Roman, while retaining desirable indigenous features. In this much, hybridity constitutes a hermeneutical tool that facilitates the reception and repurposing of cultural products ( van Dommelen 2006: 138-139;Jiménez 2011: 117).
Within contemporary discourses of postcolonial theory, creolization and hybridity offer productive avenues for understanding the complex cultural products resulting from colonial interventions in diverse geographic contexts. Both paradigms attempt to explain instances of cultural interaction from the bottom-up, rejecting a unilateral idea of acculturation and revealing multi-vocal experiences of the past. Applied to studies of material culture, they help illuminate aspects of the production, uses, and reuses of artifacts that are often poorly understood. Such an approach is especially suitable to assemblages that remain only in a fragmentary state, such as the sculptural corpus of pre-and post-Roman Gaul. These monuments, moreover, reflect a high degree of ambiguity in their conception, likely responding to the hybrid identities of those who commissioned and created them. Looking at the incised contours technique through the dual theoretical lens of creolization and hybridity allows us to deal with such elusive aspects of the art of ancient Gaul, going beyond the overt messages the monuments convey through their form, decoration, and inscriptions. Creolization sheds light on particular aspects of the making of the monuments, revealing an active, generative process driven by the mixed identities of artists and by changing patronage patterns. Hybridity, in turn, reveals itself as a useful interpretive model that allows us to explore the ancient viewership of the monuments, recuperating the hybrid gazes of ancient individuals with diverse cultural horizons, and who are mostly silent in the literary record. In combination, these two models offer an alternative to Rome-centric understandings of the art of Gallia Narbonensis both before and after the conquest.

Gallo-Roman Monuments in Narbonensis
The province that the Romans called Gallia Narbonensis, in southern France, was delimited by the Mediterranean Sea to the south, the Alps to the east, and the Cévennes and Pyrenees mountain ranges to the north and southwest respectively; at its heart was the valley of the Rhône River. Before the arrival of Roman settlers, the lands of Narbonensis were home to a confederation of Gallic tribes known as the Cavares (Strabo,Geographia 4.1.11). The region was under the powerful cultural sphere of Massilia (modern Marseille), a late seventh century bc colony founded by Phocaean migrants. The Greek city was an active and diverse cultural center that extended its influence into the surrounding settlements (Strabo, Geog. 4.1.5;Benoit 1969: 7-8;Duval 1977: 108-109;Bromwich 1993: 197). 7 Permanent Roman presence in the area began in 125 bc after the conquest of Massilia and the campaign against the Salluvii by M. Fulvius Flaccus. His successor C. Sextius Calvinus founded the colony of Aquae Sextiae (modern Aix-en-Provence) in 122 bc, and in 118 bc Narbo Martius (Narbonne) became the capital of the new province (Livy, Periochae 61; Velleius Paterculus, Historiae Romanae 1.15;Strabo,Geog. 4.1). Both Caesar and Octavian founded or re-founded colonies at sites such as Vienne, Valence, and Nîmes (Goudineau 1996: 473). There are indications that many indigenous sites-either oppida or larger urban centers-were annexed to or absorbed into the new colonial foundations (Strabo, Geog. 4.1.12;Goudineau 1996, 474-475). The area became populated by a combination of indigenous Gauls, Roman veterans, immigrants from the Italian peninsula, and Greeks or Greek descendants from Massilia and its outposts. The Romans saw Narbonensis as a gateway to both Iberia and northern Gaul and the construction of an impressive network of roads gave easy access to the newly conquered territories.
With the Roman conquest, Narbonensis saw the introduction of new building types, such as amphitheaters, aqueducts, honorific monuments, and Italic-style temples (Bromwich 1993;Woolf 2000;Anderson 2013). These new monuments served, among other things, as permanent markers of the Roman conquest in the landscape. While in their functions and commissioning processes these structures were unmistakably Roman, other aspects of their making possibly involved local input. This included the builders themselves, the materials employed, the style of the applied decoration, the repetition of certain figural motifs, and technical elements such as the focus of this study, the practice of carving figures with deeply incised contours. Several surviving monuments from this region feature prominently this technique in the reliefs that decorate their surfaces.
One of them is a freestanding arch that stands in the small city of Carpentras, ancient Carpentoracte (Figure 1 ;Picard 1960;Turcan 1983;Kleiner 1985: 44-45;Gros 1996: 68-69). Built in the early first century ad, and now moved from its original location, the structure is composed of two robust piers that frame a narrow bay. Fluted Corinthian pilasters and three-quarter columns articulate the exterior. The interior is plain, and the imagery is concentrated on the two wide lateral facades. The reliefs show a pair of male captives standing beside a tree-trunk trophy. The captives are over life size and the attention paid to their features and attire suggests the depiction of specific individuals. The captive on the left wears a tunic tied at the waist and a long cape trimmed with tassels that falls below his knees. The one on the right is clad in a simple animal skin. Based on their clothing scholars have suggested that they represent non-Gallic captives, possibly Parthians and/or Germans, but the evidence for such identification is inconclusive (Picard 1957: 283-284;Kleiner 1985: 45;Silberberg-Peirce 1986: 318). The figures are carved in low relief, and the artists used deeply incised lines to highlight the contour of their bodies. Additional details, such as the weapons that appear next to the captives' legs and parts of their clothing, are simply outlined in the stone background.
Another well-known monument that employs this regional technique is the Arch at Orange, in what was the ancient city of Arausio (Figure 2; Amy et al. 1962;Anderson 1987;Bellet 1991: 44-61). The triple-bayed monument was likely built during the early reign of Tiberius (ad 14-37) to commemorate the crushing of an indigenous rebellion, although the date continues to be debated (Anderson 1987;Rodríguez 2018: 19-22).
The monument presents a complex visual program composed of reliefs that depict spoils, trophies, scenes of battle, captives, and allegorical figures. These panels employ the technique of incised contours with a great degree of variation. For example, it appears in the frieze, running around the four sides of the arch, to outline the pairs of Roman and barbarian fighters engaged in single combat (Figure 3). It also features prominently in the reliefs that occupy the spaces above the lateral arches, depicting heaps of spoliated weapons and armor. In those, the technique was used to add decorative details and to label some of the objects represented, such as the names of Gallic chieftains inscribed on the shields. Elsewhere on the monument, especially on the reliefs that display naval spoils on the first attic, the technique is used much more  aggressively to outline a great number of the artifacts shown, including ship prows, anchors, and tridents. 8 In these panels, the artists took advantage of the combination of three-dimensional modeling and line contouring to maximize the amount of information shown, a strategy that possibly saved time and effort. In the battle scenes that decorate the central section of both sides of the attic, the contours were also used in combination with three-dimensional sculpting to show the mélange of Roman and barbarian soldiers taking part in a fierce battle. The style of these reliefs evokes Hellenistic precedents, such as battle-themed sarcophagi and paintings. 9 The combination of a Hellenistic style to depict a Roman subject matter, and executed in what is presumably a local technique, exemplifies the polysemic and multicultural character of the art of Gallia Narbonensis.
This combination is even clearer on the so-called Mausoleum of Glanum, likely erected between 30 and 20 bc, in modern Saint-Remi-de-Provence (Figure 4; Rolland 1970;Kleiner 1973;1977: 662-666). The rectangular carved panels that decorate the four sides of the base employ incised contours to convey the illusion of spatial depth and to confer a vibrant dynamism to the battle scenes. Besides the contouring of figures, accentuating their detachment from the flat background, there are again certain elements, such as weapons and body parts, barely suggested with lines incised in the stone. The artists were skilled in conveying the illusion of movement, aided in part by the contouring of elements that would otherwise overcrowd the space and obstruct legibility.
The relief panels that decorate these monuments illustrate the aesthetic advantages of the incised contours technique, which allowed artists to accomplish several things simultaneously. One was to accentuate the figures, highlighting the contrast of light and shadow and making the low relief clearly visible from a distance. The effect would have been even more pronounced if, as it seems likely, the figures were painted (Kleiner 1977: note 106). This technique also afforded the flexibility to control the level of detail for individual elements, stressing the principal motifs while economically suggesting secondary ones. As mentioned above, interpretations of this phenomenon have assumed the involvement of local artisans working under the guidance of foreign masters. Yet looking at this problem from a postcolonial perspective offers an alternative explanation. Considering a series of fragmentary pre-Roman monuments, now scattered throughout regional museums, I propose that the artists who completed these works were functioning within a well-established, ancient artistic tradition with roots in Gallic and Hellenistic modes of representation.

Pre-Roman Gallic Monuments
One example that suggests such continuity is an impressive bust of a warrior found in the small locality of Sainte-Anastasie, approximately fifteen kilometers north of Nîmes ( Figure 5). It was carved from local limestone and has been dated to the third century bc (Pobé and Roubier 1962: cat. 40;Duval 1977: 108-109;Py 2011: 29-33). 10 The bust formed a pair with a second one and seems to have been purposely defaced and deposited in a ritual burial. The figure wears a large headpiece that likely represents a leather helmet with spiral metal ornaments. The lower part of the bust follows the rectangular shape of the stone, where incised lines create the illusion of a breastplate. Below the simple geometric elements that suggest the shape of the pectoral, a horizontal register features three galloping horses roughly incised in the stone. Pigment analysis has revealed that the bust was painted in bright red ochre (Barbet 1991: 63). The face of the warrior is merely sketched with round, deeply carved eyes, linear mouth, and a flat nose that merges with the brow. The somber features seem to underscore the funerary character of the monument. The use of lines incised in the stone to suggest decorative details in combination with three-dimensional sculpting, echoes their use in later Gallo-Roman reliefs.
Also dating to the third century bc is a small relief showing the heads of four horses on a plain background (Figure 6). It was found at the so-called sanctuary at Roquepertuse, near Aix-en-Provence, where it possibly functioned as a lintel for a gateway that gave access to a shrine or elite burial (Pobé and Roubier 1962: cat. 13;Duval 1977: 113;Barbet 1991: 72;Py 2011: 97-98). 11 The carvers employed incised lines to minimally suggest the outlines of the figures, yet the grooves vary in depth, emphasizing certain elements, such as the lower jawlines, and adding volume to the figures. The artists also conveyed the illusion of spatial depth, and even movement, with the use of scale hierarchy. As with much of the sculpture discussed here, the relief was originally painted in red ochre and blue (Barbet 1991: 72). The motif of a series of galloping horses recalls the breastplate of the Sainte-Anastasie warrior bust. Horses were an important part of Gallic belief systems and rituals (Barbet 1991: 77;Green 1989: 146-149;Webster 2011: 220-223). This motif seems to prefigure the representations of horses in motion that are common in later artwork, including those that decorate the Arch of Orange and the Mausoleum of Glanum (see Figure 4).
Further examples that employ incised contours to highlight details in three-dimensional sculpture include the numerous seated warriors that have been found throughout Provence at sites such as Glanum and Roquepertuse (Pobé and Roubier 1962: cats. 35, 36;Barbet 1991;Boissinot 2011;Py 2011: 85-90, 103-118;Kruta 2015: 78). Two of them, also from the sanctuary at Roquepertuse represent headless male figures sitting in a rigid, vertical posture with hands resting on the knees (Figure 7). 12 The figures sit on stone pedestals measuring 50 centimeters in height. Their bodies are rendered in a geometric fashion, although there   is an attempt at portraying anatomical details and musculature. Although now severely faded, the figures exhibit elaborate costumes consisting of a breastplate and a square shield that hangs at their backs. The costumes were drawn in the stone with incised lines that compose geometric patterns, brightly painted in red and black. The meaning of these figures remains elusive, but their context suggests the funerary effigy of a pair of warriors or ancestors. 13 Once again, incised lines were combined with three-dimensional modeling, signaling a distinct regional style. A final example is a sculpture known as La Tarasque de Noves, found in an oppidum a few kilometers southeast of Avignon (Figure 8; Pobé and Roubier 1962: cat. 31;Cavalier 2004;Py 2011: 186). 14 Dating to the second half of the first century bc, it represents a fantastical creature-part lion part wolf-in the process of devouring a human figure; the arm of the unfortunate victim protrudes from the animal's open maw. The heads of two other victims appear below, shown as trophies beneath the paws of the beast. The body of the creature has been sculpted in the round, but the artist used deeply incised lines to represent anatomical details, such as the ribcage, the facial features, the musculature of the limbs, and the scales that cascade down its back. The same technique conveys details of the severed human heads, including facial features and beards. This terrifying figure likely had an apotropaic function similar to the many images of lions found throughout the ancient Mediterranean world, which protected city gateways and tombs. Comparisons can be established, for instance, with the Lion Gate at Mycenae, the Avenue of the Lions in the island of Delos, and even closer, with multiple similar figures found in Gaul and Iberia (Gaul: Pobé and Roubier 1962: cat. 64;Iberia: Chapa 1985;Pérez 1999;Jiménez 2011: 109 and Plate 5.) The motif of displaying severed heads as trophies, by contrast, was distinctly Gallic. The practice is widely attested in the literary and archaeological records (Diodorus Siculus, Bibliotheke 5.29.4-5;Strabo, Geog. 4.4.5;Livy, 10.26.10-11;23.24.11-12;Dedet 2011). For example, the sanctuary at Roquepertuse included pillars with niches especially carved to display severed heads (Duval 1977: 111). Seated stone figures holding severed heads have been found at sites such as Entremont (Py 2011: 145-151, 184-185). La Tarasque de Noves exemplifies the creolization of pan-Mediterranean and Gallic ritual practices and visual traditions during the early decades of the colonization of Narbonensis.

Models of Continuity
This brief survey of monuments created before the Roman conquest of Narbonensis reveals how the combination of three-dimensional sculpting, following Greek models likely imported from Massilia, with the use of incised lines to render contours and highlight details, was a defining aspect of the visual production of the region. Despite the fragmentary state of the available archaeological evidence, and the chronological gap that exists between these early works and the later Gallo-Roman monuments, it is possible to suggest a continuity when focusing specifically on the incised contours technique. Signs of such entangled aesthetic and technical relationships appear elsewhere. Coins, for instance, provide a parallel for a similar phenomenon. Before the Roman conquest, locally minted coins circulated in Gaul, where Greek denominations were combined with typically Gallic iconography (Duval 1972;Nash 1987;Gruel 1989;Gruel and Haselgrove 2006: 126-129). Some centers continued to produce coins with indigenous designs, even after Roman authorities took control over the minting process (Duval 1977: 170-171;Cassibry 2016: 138, 151). One example of this appears in a gold stater likely minted by the Parisii tribe (Duval 1977: 170-171). 15 The obverse depicts a head with a straight nose and flowing hair reminiscent of coins bearing portraits of Hellenistic kings (Figure 9; Duval 1972: 635-636). The reverse shows the figure of a horse in a stylized design that recalls the linear quality of Gallic sculpture, including the Sainte-Anastasie warrior bust and the horses lintel at Roquepertuse (see Figures 5 and 6). A second stater, likely produced in Britain, also uses the head motif, this time rendered in even more schematic fashion. 16 On the obverse, a large central head is surrounded by four other heads that seem to float around it, tied with ropes to the main one. The composition evokes the practice of displaying the severed heads of enemies discussed above. These coins offer a point of comparison to understand how Gallic artists absorbed Greco-Roman ideas and repurposed them in the creation of a new form of visual production. They provided an opportunity for Gallic artists to experiment with a new medium, allowing them to explore forms of representation-both pictorial and narrative-hitherto unknown in the region (Cassibry 2016). These coins can be understood as a case of creolization in which a foreign cultural product was adapted to respond to symbolic associations that were ingrained in local custom, and which propelled a regional, creolized visual style. 17 A second parallel where indigenous practices merged with Hellenistic styles and Roman monumental needs, appears in what scholars know as trophy monuments. These are the architectural expressions of the tradition, common throughout the Mediterranean, of erecting tree trunks dressed in the armor of defeated enemies at the site of battles, often with piles of weapons and other military paraphernalia deposited at the base (Picard 1957;Ibarra 2009;Kinnee 2018). These trophies had apotropaic connotations and, especially for the Romans, functioned as a way to mark the landscape of conquered territories. Although traditionally understood as a typically Roman practice, the form seems to have had roots in a Gallic military ritual. In his commentaries of the Gallic Wars, Julius Caesar describes this tradition noting that the Gauls gathered great numbers of captured enemy weapons at the location of battles, consecrating the place with sacrifices (Caesar,Bellum Gallicum 6.17.4.). Archaeological and literary evidence demonstrates that it was in Gaul where the first trophy monuments appeared in architectural form. The earliest known were those erected by the Roman generals Q. Fabius Maximus and Cn. Domitius Ahenobarbus at the location of victories over Gallic tribes in 121 bc (Strabo,Geog. 4.1.11;L. Annaeus Florus,Epitome 1.37.3). Other examples include those built by Pompey at the site of Panissars (71 bc), and a triple trophy monument that stood in Lugdunum Convenarum (modern Saint-Bertrand-de-Comminges), both in the Pyrenees (Pompey: Strabo, Geog. 3.4.1;Pliny, Naturalis historia 7.96;Cassius Dio, Roman History 41.24;Lugdunum: Espérandieu 1915: 7653-8, 7665;Picard 1957: 272, Figure 6;Schenck-David 2003).
Also in Gaul is the trophy monument that still stands in the small Alpine town of La Turbie, overlooking Monaco (Figure 10). Now heavily reconstructed, the trophy was a senatorial dedication to honor Augustus after the culmination of the Alpine Wars. It stands on a large square base measuring 47 meters wide by 15 meters tall; as reconstructed, the structure reached a height of 49 meters (Formigé 1949). Above the base is a large platform supporting a colonnaded tholos. The structure is topped by a conical roof that likely supported a statue of Augustus. At the base of the monument, a large inscription conspicuously named the forty-four tribes brought under Roman control as a result of the war (Pliny, HN 3.136). The Augustan Trophy at La Turbie illustrates the intersection of Gallic rituals, Hellenistic architectural forms, and Roman commemoration practices. The structure stands as a symbol of the Roman domination in Gaul, but it borrows elements of a visual vocabulary that was typically Mediterranean. The style of the architecture exhibits Greek precedents, especially clear with the incorporation of the tholos. The purpose of this monument was twofold: on the one hand, it blazoned the Roman victory in the face of the defeated local population, perhaps as a means to contain the constant threat of native insurrections. On the other hand, the monument participated in a visual language that was familiar for local viewers, addressing especially those who had embraced Roman rule (Rodríguez 2018: 179-184). The reception of the monument by native viewers was mediated through a hybrid experience of Gallic, Greek, and Roman monumental traditions and ritual practices. Understood in the light of the entangled histories of minting practices and trophy monuments, the incised contours in the art of Narbonensis emerges as a fundamental sign of the cultural change brought about by the Roman conquest. This process took multiple forms, but in its visual manifestation resulted in a new style that was neither exclusively Roman, nor native, but that can be more accurately described as Gallo-Roman . The dual lens of creolization and hybridity provides a useful framework to understand the interdependent processes of production and reception of the new monuments respectively. The dedication of commemorative structures, such as the Arch at Orange or the Mausoleum of Glanum, can be envisioned as a collaborative enterprise between local artists and elite patrons, likely Romanized locals or Italian settlers. Patrons possibly controlled important aspects of the commissions, including the selection of the artists/workshops, the location, materials, and overall design of the monuments, as well as details of dedicatory inscriptions and statues. Other particulars of the production, however, such as technical and stylistic details, were possibly left to the discretion of artists according to their specific abilities and levels of expertise. Workshops might have been well-known for mastering certain skills, likely accumulated over centuries of tradition. In this scenario, it is plausible that traces of indigenous artmaking, such as the incised contours, could have survived in the art of later centuries. The final products, architectural forms such as triumphal arches and monumental tombs, expressed Roman ideas rendered in Hellenistic styles and executed in what is presumably a local technique. In this creolized production process, old customs blended with new aesthetic needs to give rise to a new form of artistic expression.
At this point it is pertinent to return to the agency of local artists, asking, to what extent were they truly involved in the creative process? This question is hard to answer given the absence of direct evidence. 18 Previous scholarship has assumed that Gallic artists were insufficiently skilled to undertake such complex monumental projects, or have presupposed the involvement of foreigners in various stages of the creative process (Hatt 1951: 129;Rolland 1970: 49;Kleiner 1977: 683). Yet native artists involved in the design and making of the monuments might have had sufficient space to express their technical competencies and creativity even when working within patron-imposed constraints. 19 Clues of this appear in the subtle ways in which local identities are visible on monuments such as the Arch at Orange, where the names of known Gallic chieftains are conspicuously carved on some of the shields in the spolia panels (see Figure 3). They were also prominently represented as the Gallic fighters fiercely resisting Roman forces in both the frieze and the attic reliefs on the same monument. At Carpentras, the clear individuality of the captives, who command the space with their robust bodies and proud stances, seems to betray a similar intention (see Figure 1). Local viewers were possibly aware of these cases of subtly defiant visibility, even when the monuments underscored the figures' vanquished condition.
In addition to these instances of iconographical visibility, local viewers could have also recognized how local traditions were embedded in technical details, such as the incised contours. Although freestanding arches and other creolized monuments were not part of the artistic horizons of Narbonensis before the Roman conquest, viewers would have been familiar with the technique through other media such as small reliefs and freestanding sculpture, found in domestic and funerary contexts. 20 Establishing connections between the old and the new monuments by identifying the motif as a commonality was possibly a natural response. The incised contours were bearers of meaning as much as iconographic themes and monumental design were. In what can be characterized as a hybrid viewing process, Gallic viewers, both Roman and native, actively constructed those meanings as they encountered the new monuments according to their pre-existing visual literacy. Those viewers likely recognized how artists were astutely merging old techniques and motifs with imported artistic forms in a process of cultural hybridity that, as Bhabha argues, '… gives rise to something different, something new and unrecognizable, a new area of negotiation of meaning and representation ' (1990: 211).

Conclusions
Postcolonial theory such as creolization and hybridity allows us to reconsider the incised contours technique in the art of Gallia Narbonensis as a continuation of pre-Roman technical conventions. Similarly to the ways in which coins and trophy monuments reflect entangled ideological and aesthetic concerns, the production of stone monuments in this region was the result of the blending of Gallic, Greek, and Roman ideas and ways of making. Previous interpretations of this phenomenon have ignored potential precedents in the indigenous artistic record, implying a visual vacuum for the creation of Gallo-Roman art, and obscuring the agency of colonial individuals. There is ample evidence for the active involvement of local elites in the monumental project propelled by the conquest; they used their dedications to assert alignment with Roman interests through cases of private commemoration and public munificence (Frézouls 1984;Woolf 1998: 123-126;Webster 2001). Yet elite patrons were not the only ones participating in this process. Native artists seem to have quickly adapted to new artistic demands incorporating old practices in the crafting of a visual third space that came to define the arts of the region as a whole. Viewers might have responded to the monuments with engaged, hybrid gazes that actively generated new meanings from old traditions.