Family Values
ART AND POWER AT GHIRZA IN THE LIBYAN PRE-DESERT
ART AND IMPERIALISM
This chapter concerns the intersection of art and power in the Roman world, but it is also about how we study Roman art. A common emphasis on the formal qualities of so-called Romanized art has created the perception of art as symbolic of the success of Rome and of the acquiescence of indigenous peoples to its rule. From this perspective, provincial art can easily be dismissed as an often inadequate imitation of Roman style and images. In turn, this tendency has also led to greatest emphasis being placed on recording and displaying those works that attained the highest technical and stylistic affinities with examples from Rome or Italy. There is thus a strong centrist tendency in art historical studies of the Roman world and it depends to a large extent on the assumption that people in the provinces would have wished to emulate the art of the colonial power regardless of the underlying symbolism.1 Such a view may have some truth, some of the time, in elite circles, but I suspect that the degree of direct and uncritical artistic emulation was overall rather small. Much less examined is the extent to which elements of Roman iconography or style were appropriated to serve indigenous agendas, rather than simply to mimic Roman culture.2 If we accept this line of argument, then it is clear that while people might have amalgamated imperial imagery and elements of Roman art, the process was often critically selective and it also involved the redefinition and reinvention of native traditions of art.
There seems to me to be a further double weakness with the approach of many art historians to the Roman world; first, it is commonly assumed that aesthetic values in the past were the same as those of today, and second, that this cultural “gold standard” was universally accepted throughout the empire and at all levels of society. Both of those ideas are flawed by their acceptance of a level of consensus that cannot be shown to have existed. Art and material culture are frequently hard to disentangle in past societies and the possibility that the message was frequently as important or more so than the medium should also give us pause for thought.
A specific attribute of much art is its use by different groups in society to express power relations. Recent work on “official” art of the Roman Empire has emphasized the degree to which imagery was shaped and exploited in support of the emperors.3 Yet, similar power-based dialogues were conducted at many different levels in society through the medium of display and consumption. Much more work is certainly needed on the social significance and power connotations of imagery that was not publicly sanctioned. In addition, recent discussions of the operation of power in the Roman world have suggested that we must deconstruct the discourse of modern imperialism if we are to improve our understanding of the ancient situation.4 Jane Webster’s work on the impact of colonialism in the religious sphere is particularly pertinent in that it demonstrates perfectly the way in which iconography can become blurred, depending both on the contexts of its use and the audience.5
An alternative approach to the study of Roman provincial art would embrace a series of discreet perspectives and consider different underlying themes in the imperial dialogue (including resistance, imitation, and adoption). In this context, I am interested in ways in which the adoption of so-called Romanized style also facilitated the continuation of indigenous traditions. These then are the sorts of issues I shall explore with a case study from the Libyan pre-desert in the fourth century AD.
THE GHIRZA TOMBS
The site of Ghirza lies about 250 kilometers southeast of Tripoli (see fig. 9.1). The settlement was one of the largest in the Libyan pre-desert zone, comprising a “village” of about 40 buildings clustered together on the west side of the Wadi Ghirza, around 10 kilometers south of its junction with the Wadi ZemZem.6 The buildings centered on several major fortified structures and there was also a substantial temple, arguably of far wider significance than the village of Ghirza itself.7 There were two cemeteries with monumental tombs at Ghirza (see figs. 9.1–9.2); the first lay a short distance south of the settlement on the west side of the wadi (the North Tombs); the second lay on the east bank about 1.5 kilometers south of the main site (the South Tombs).8 The north group (see figs. 9.1–9.4) comprized six ashlar tombs (North A–F) and the south group another six (South A, C–G). There was an additional monument in each cemetery (North G and South B) whose function is more enigmatic. For convenience, I shall refer to the various funerary structures as NA, NB, SA, SC, and so on. The Ghirza group mostly conforms to a single architectural form, the temple type, with a series of colonnaded arcades arranged around a central masonry core, or cella (see figs. 9.3–9.4). In some cases the cella was decorated with a false doorway, with the burials normally being placed in a subterranean chamber below.9
FIGURE 9.1. General plan of Ghirza, showing the relation between the settlement and its cemeteries. Inset: Map of Libya, showing location of Ghirza. (After Brogan and Smith 1984, with modifications.)
The heyday of the site (and the period to which most of the tombs relate) appears to be the late third to fourth centuries AD. The mausolea constitute the single largest concentration of elite funerary monuments in the region, suggesting that Ghirza had some special status within the pre-desert zone.10 Yet reaction to these tombs and their iconography has been quite variable:
FIGURE 9.2 Plans of North and South Cemeteries. (After Brogan and Smith 1984, with modifications.)
And although I had not allowed my imagination to rise at all in proportion to the exhilarating accounts I had heard, I could not but be sorely disappointed … I found them of a mixed style, and in very indifferent taste, ornamented with ill-proportioned columns and clumsy capitals. The regular architectural divisions of frieze and cornice being neglected, nearly the whole depth of the entablatures was loaded with absurd representations of warriors, huntsmen, camels, horses, and other animals in low relief, or rather scratched on the freestone. … The human figures and animals are miserably executed, and are generally small, though they vary in height from about three feet and a half a foot in height, even on the same tombs, which adds to their ridiculous effect.11
On me n’avait pas exageré la beauté de ces monuments. Elle a dépassé de beaucoup mon attent. Ces ruines sont les plus belles de toute la Tripolitaine … On ne trouve nulle part en Afrique des tombeaux comparable à ceux-ci par la richesse de la sculpture et par les proportions.
The beauty of these monuments has not been exaggerated. They surpassed by far my expectation. These ruins are the most beautiful in the whole of Tripolitania. One cannot find anywhere in Africa tombs comparable with these for the richness of the sculpture or for their proportions.12
FIGURE 9.3 Aerial view of North Cemetery looking north with tomb NA on the left. (Photo: ULVS.)
One group of scholars, then, has celebrated the art of these tombs as a sign of the “Romanization” of the pre-desert communities, a point enhanced by the fact that several of the tombs bore Latin inscriptions. Some went so far as to claim that Italian colonists must have been responsible (though the names on the tombs are Libyan).13 The ambivalence or dissatisfaction of a second group of scholars is in part a product of the “otherness” of these tombs when judged by the highest standards of Roman decor and representation. We are dealing with a very regional style of art and symbolism. Failing to assimilate it easily with the mainstream, the tendency among some nineteenth- and early-twentieth-century scholars was to devalue it, or more insulting still, to claim that it must be Byzantine in date as it was so debased in style!14
Both of the approaches described above, however, share a common perception that the art of the tombs was fundamentally attempting to emulate Roman culture. The difference of opinion is essentially a question of judgment concerning the relative success achieved. We may note in passing that higher standards of representational sculpture were achieved on elite estates nearer to the coast—as in the case of the well-known plowing scene from a mausoleum at Tigi.15 Looked at from a Roman perspective the debased sculptural norms of the Ghirza reliefs are suggestive of a trickle down effect of Roman culture on the remoter fringes of empire.16
Yet can the tombs and their art support another type of reconstruction? There is no doubt that these sculptures have some unusual characteristics when judged against mainstream Roman art. Key stylistic elements here are the horror vacui and multiple horizons of many scenes, leading to some unusual and distinctly un-Roman compositional fields (see figs. 9.5–9.9).17 These standards go back to Punic times in North Africa and find their counterpart in other areas of Romano-African art—perhaps most notably in mosaic composition.18 Yet, when we explore the African context of the iconography and style, a remarkably different picture of the art emerges. In particular, the artistic themes of the Ghirza tombs have many parallels in the corpus of funerary art of other parts of Tripolitania and with Libyan and Punic sacred art.19 I want to emphasize at the outset that these tombs must be read not simply as monuments to the dead but as structures that had a continuing religious significance for the living. The architecture of the tombs fits into a long tradition in Roman Africa, fusing Hellenistic, Punic, and Roman traditions with African ritual needs and ideology.20 This cultural complexity—indeed, ambiguity is a better word—runs through art, religion, and material culture in Roman Africa.21 We ignore at our peril the Libyan component in all this. As it happens, the closest parallel for the architectural form of the predominant temple tombs at Ghirza comes from even farther out into the desert, from the oasis center of Ghadames (see fig. 9.1 for location).22
FIGURE 9.4 Detail of temple tomb NC, showing original position of reliefs on the arcade aches and on the frieze above. The surviving fragment of the frieze shows (left) the harvesting of cereals and (right) a hunting scene. (Photo: ULVS.)
THEMES IN THE ART OF THE TOMBS
In her brilliant and perceptive discussion of the tombs and their art, Olwen Brogan broadly divided the sculptures into three groups: (1) symbolic images, (2) scenes of daily life, and (3) religious scenes.23 I have summarized these data in tables 9.1 and 9.2 and commented in another paper on some aspects of their significance.24
In the rest of this chapter I shall focus on imagery relating to the family and to the power networks constructed around both living and dead members of the principal families at Ghirza. The existence of two major fortified structures at the heart of the settlement and of two discreet monumental cemeteries suggests that there were two main elite families, or perhaps two branches of a single family, at Ghirza. An alternative would be to see the cemeteries as clan-based, but in Libyan society such units were in any case normally based around extended family units.25 As a working hypothesis, then, it seems reasonable to infer the presence of two dominant families.
Symbolic sculpture, fertility imagery, and mythological themes on the Ghirza tombs
Source: Data extracted from Brogan and Smith 1984, with some additions.
Note: Tomb types are abbreviated as follows: Te = Temple type; Ar = Arcaded temple type; OB = Obelisk type.
Scenes of daily life, hunting, ceremonial and ritual activity on the Ghirza tombs
FIGURE 9.5 Chieftain scene from Tomb NB. (Lepcis Magna Museum.)
The family links are given added emphasis on the tombs in the North Cemetery that feature detailed inscriptions (NA, NB, NC), where the importance of lineage and of the maintenance of tombs and rituals associated with them are stressed. The named individuals on these three tombs were success stories in the Roman imperial equation. They bear Roman nomina alongside Libyan cognomina and were able to pay for expensive and locally highly impressive funerary monuments. The Latin of these texts (albeit far from perfect) would at first sight appear to add credence to the view that these people were above all interested in the emulation of Roman elite behavior. However, a number of factors suggest a different conclusion, not least the fact that several others of the tombs bore texts in Punic (so-called Latino-Punic) and Libyan.26
IMAGERY OF POWER AND STATUS
There is abundant imagery on the tombs that speaks of individual power over others, and this may be understood to be a prime function of the tombs’ iconography.27 The so-called chieftain scenes on tombs NB and NC (and echoes of chiefly regalia also on some of the South Cemetery tombs—for example, SF) are the prime evidence of this (see figs. 9.5–9.7). They appear to show the bestowal of certain items of regalia or equipment (scepter, scroll, wine cup, jug, quiver) and other gifts to an individual whose power is emphasized both by his scale in the scene and his position, seated on a folding chair.28 The scene is paired on both NB and NC by another, more disturbing one in which two individuals appear to inflict punishment (perhaps even execute) a third man standing in submissive pose between them.29 There are numerous other scenes showing individuals bearing arms and in several cases (tombs NA, NB, SF) these feature scenes of human combat.30 From this we may conclude that the chieftains at Ghirza held substantial powers, and it has been plausibly suggested that these were delegated to them by Rome as part of the increasingly informal late Roman frontier arrangements.31
FIGURE 9.6 Chieftain scene from Tomb NC.
Another group of scenes depicts farming practices (see fig. 9.8). We see the entire process of the cultivation of cereals from ground clearance, to ploughing and sowing, to harvesting, threshing, and winnowing.32 Two other scenes show camel caravans and another carving depicts an amphora being carried by a man in front of a camel bearing two further amphorae on its back and pulling a two-wheeled cart.33 We are perhaps entitled to consider these latter scenes as illustrating the marketing of the produce of the wadi farms (grain and liquid products such as oil and wine). These economic scenes have generally been read simply as scenes of daily life. An alternative would be to see them as symbolic of the economic power of the elite families. To take this a stage further, they could also be read as images of things requiring the protection of the ancestors honored by the tombs.
FIGURE 9.7 Scene with chieftain and regalia, tomb NC. (Jamahariyah Museum, Tripoli.)
There are many depictions of men on horseback, an obvious indication of status—especially in an age when regular army units on the frontier were having their mounts withdrawn for reasons of imperial cost-cutting, and though some of these scenes show hunts, others have a more ceremonial look (for example, see fig. 9.7).34 Sleek dogs feature not only in the thick of the action in numerous hunting scenes (see fig. 9.4, scene on frieze of tomb NC), but are also represented in other scenes.35 Both horses and hunting dogs can be read as clear status symbols.36
Status is commonly denoted by subtle differences in the dress of the figures, which is in all cases very simply and schematically indicated. Most of the figures depicted in the reliefs wear very simple tunics, gathered in at the waist by belts, with heavy vertical lines presumably signifying folds in the cloth rather than texture or pattern of the weave (see fig. 9.8). The elite figures are emphasized by reason of their exaggerated size, by their central positions in ceremonial scenes, and by the fact that they are almost invariably depicted with more complex costumes (see figs. 9.5–9.7). Most commonly there are indications of outer garments worn over the tunic. In many cases for both men and women this is indicated as a generously cut length of cloth wrapped diagonally around the body in a manner more akin to the Arab burnoose than the toga. Although in the case of portraits (see fig. 9.9) it could be argued that this represents a funeral shroud rather than an everyday garment, there are examples of full-length figures in “scenes of everyday life” who are picked out by such differentiation in costume. Indeed, diagonal lines in the depiction of dress on the Ghirza tombs can almost invariably be taken as a sign of status. In other instances (e.g., NC), men are depicted with cloaks and women with ornate dresses. In the two scenes with seated chieftain (NB and NC), the principal figure is shown wearing an ornately decorated cloak or garment with horizontal bands of ornamentation (see figs. 9.5–9.6). Other symbols of female elite status include the depictions of items of jewelry and turbanlike headresses. Overall, these costumes do not appear to emulate Roman models.37
FIGURE 9.8 Cultivation scene, tomb SC. (Jamahariyah Museum, Tripoli.)
IMAGERY OF THE FAMILY
Several tombs feature portraits of the deceased—this is certain for NA, SA (where a man and woman feature both as busts and statues), SC (see fig. 9.9), SD, and SE.38 The naive style of these portraits can easily be mistaken for poor craftwork, but there are ample parallels for this style of representation in Roman Africa and it seems to have been deliberately maintained even at times and places when stonemasons were producing far more sophisticated work for other types of use. Typically, the style is associated with the so-called Saturn stele and with funerary monuments. Characteristic of this indigenous tradition of representation (which I shall describe as the Libyco-Punic tradition) is the frontality, the large and dominating eyes, and the differential scaling of figures in terms of importance in the scene.
FIGURE 9.9 Family portrait, tomb SC. (Jamahariyah Museum, Tripoli.)
As has already been noted, ancestral links are established and supported by the architectural setting and epigraphic statements of the tombs. While the Libyan notables clearly appropriate some Roman imagery, it can hardly be claimed that they aimed simply to emulate Roman culture. To summarize thus far, the Ghirza tombs are laden with powerrelated imagery. The question is, at whom was it aimed, and for what purpose?
THE NATURE OF THE LIBYAN ANCESTOR CULT
I have noted earlier that many images depict aspects of life requiring the protection of ancestors. That support was maintained through the observation of elaborate rites long after the death and burial of individual ancestors. The visual display of the earthly powers of ancestors could serve to remind their spirits to look to the continuance of such power in later generations. Such reverence of ancestors and their tombs was certainly an important trait of Libyan culture, and there is absolutely no need to seek explanations outside Africa for this.39 The essential counterpart of permanent visual prompts to one’s ancestors was a formal ceremony at the tomb on a regular basis. Another element of the Libyan ancestor cult was the practice of sleeping at a family tomb for the experience of visionary responses. This is attested across a broad time frame as demonstrated by the two examples below from the fifth century BC and first century AD:
For divination they take themselves to the graves of their ancestors and, after praying, lie down to sleep upon the graves: by the dreams that come to them they guide their conduct.40
The Augilae consider the spirits of their ancestors as gods, they swear by these and consult them as oracles, and, having made their requests, treat the dreams of those who sleep in their tombs as responses.41
The popularity of offering tables, altars, and libation spouts on tombs in the Maghreb supports the view that specific rituals were offered to the dead throughout the Roman period, and all these features are found at Ghirza. The tombs themselves evolved as sacred architecture, and the predominant adoption of the temple form at Ghirza is striking in this regard. However, the most convincing evidence of an ancient ancestor cult is provided by the two least spectacular structures in the monumental cemeteries. Both the North and South cemeteries at Ghirza contain a seventh monumental structure of near identical plan and unadorned with reliefs (NG and SB). These structures comprised a semisubterranean mortuary chamber that interconnected with a small chamber accessible from the surface and containing a bench along the back wall facing the tomb chamber (see fig. 9.10).42 Given the lack of architectural decoration, it is possible that these structures represent some sort of vision house, where one could commune on specific occasions with the mortal remains and spiritual “vibes” of Grandad Fydel or Grandma Thesylgum. Unlike the other tombs, where there was no provision for visitors to spend time in the subterranean burial chambers, these two structures appear to be purpose-built as chapels or dream houses in which the living could pass time in close proximity to the dead.43 The two structures occupy pivotal space within both cemeteries, set apart from but faced by the other tombs (see fig. 9.2). A case can be made, then, for arguing that the ancestor cult was particularly strong at Ghirza, linked in part to Ghirza’s status as a regional religious center (of which more will be mentioned shortly).
FIGURE 9.10 Comparative plans of the possible “dream houses” NG and SB. (After Brogan and Smith 1984, with modifications.)
The tomb inscriptions from NB and NC contain injunctions to the children and grandchildren of the deceased to continue to visit and observe rites at the tombs. Libation spouts were provided for the pouring of liquids down into the funerary chambers of the tombs, and an extraordinary Latin inscription found in close proximity to tomb NA records the celebration of an event called in the text parentalia, but which I doubt bore the slightest resemblance to the Italian festival of the same name.44 The Italian ceremony was generally a commemoration of their close kinsfolk by small family groups, accompanied by simple offerings. It is likely that the Latin term is employed at Ghirza simply as the closest one to convey the concept of an ancestor cult. The Ghirza text informs us of a great sacrifice held to celebrate this parentalia, apparently with fifty-one bulls and thirty-eight goats being killed.45 As family celebrations go, this was some barbecue party—there would have been enough meat here to feed several thousand people!46 Invitations to this event thus extended well beyond those actually living at Ghirza and this was probably a tribal affair, in part linked to the regional power of the settlement’s leading families and in part to Ghirza’s status as the probable cult center of the Libyan god Gurzil.47 The number of animals slaughtered is ample testimony to the wealth and power of the Ghirza elite, though I suspect that in fact many of these animals were brought specially for the occasion by those attending from outside Ghirza. It is debatable whether there was enough grazing land in the immediate vicinity of Ghirza itself to support a herd of several hundred cattle, and many of these animals most probably arrived on the hoof, brought as gifts or tribute by participants in the ceremony. In this context, the imagery reflecting the personal prestige of the principal families has a clear ongoing function—after all, it would be seen at such parentalia celebrations by hundreds or thousands of people and would help to persuade them to continue to respect the power of the elite at Ghirza and to bestow “tribute”—gifts for the sacrifice. The scale of the sacrifice matches the impressive funerary architecture and iconography produced to honor and appease those ancestors.
Although the tombs may be considered to have been ostensibly Roman in influence, to a larger extent the art and iconography described above is a product of Punic and Libyan culture, and one may suggest that in part it reflects cultural resistance to Rome.48 Much of it was of course susceptible to different or multiple readings, but the selection of images was not accidental, and virtually everything must have had significance within the cultural milieu of the Libyan pre-desert. Tombs were not simply empty memorials; they served to demonstrate the social standing of descendants and they had an important religious and sacred value in Libyan society. We may note that many standard elements of pagan Roman funerary iconography are absent or poorly represented here—scenes from classical mythology in particular. Those that are present may have been selected because they had a certain resonance in Libyan society. This role would seem to have overlapped with other areas of religious belief—a point particularly emphasized by the many parallels between the symbols on the tombs and those represented in the thousands of Saturn stelae from Tunisia and Algeria. Saturn (the Roman version of Baal Hammon) was arguably the most popular divinity in pre-Christian Africa, but his role was taken on to a considerable degree in Tripolitania by Ammon (or Jupiter Hammon), the god of the desert who seems to have shared many of Baal Hammon’s attributes.49 Many mundane decorative elements that might at first appear redolent of Romanitas can be shown to have special significance in Libyan religion (see figs. 9.4 and 9.7). Rosettes, roundels, vines, grapes, the date palm, pomegranates, lions, sheep, goats, livestock, peacocks, birds (especially the cockerel), and farming imagery can all be found on the Saturn stele. Lions and peacocks had associations with Tanit/Caelestis, as did fish. The grape vine, bunches of grapes, the cantharus can be seen as symbols of Punic Shadrapa (over time adopting much of the iconography of Liber Pater or Dionysos), and this was another cult popular among African communities (the wine jugs and cups present in the chieftain scenes on the tombs may have been for sacred rather than for profane use). We need to review much more closely these Punic and Berber cultural affiliations of the iconography of Romano-African art.50
In this context, the heraldic bull heads on tombs SC, SD, and SG; the bull sacrifice on NA; and the bullfights on NB and NC take on another connotation in connection with a Libyan god called Gurzil, who features largely in the sixth-century poem of Corippus.51 Gurzil was the bullheaded progeny of Ammon and clearly had a cult center in the Libyan pre-desert region. The similarity of the name with Ghirza is not of course conclusive (though we know from El-Bekri that there were tombs at a place called Gurza that were (still) worshipped in the tenth century—showing that the toponym is likely to be a very early one).52 On the other hand, the importance of the bull iconography and the extraordinary scale of the bull sacrifice held at Ghirza, perhaps on an annual basis, are to my mind compelling arguments for the identification. So, my alternative interpretation of the Ghirza tombs links them to a distinctly Libyan cultural agenda, with a family exercizing regional power in part through the close association between their own ancestor cult and a significant religious center. The artistic repertoire of the family tombs served to legitimate and enhance their power in this society.
The late Neolithic rock art of the Sahara is another possible source of inspiration for the relief art that has as yet been unexplored by classical archaeologists. Although separated from the Ghirza reliefs by at least 3,500 years, the rock engravings and paintings do detail the emergence of a hierarchical society with increasingly complex religious rituals.53 Of particular interest in the light of the Gurzil cult is the prevalence of images of bovines in the Saharan rock, some with decorated horns, solar discs, and the like.54 Numerous scenes show theophoric figures, with human bodies and animal heads or masks.55 Gurzil presumably developed in part from this tradition. Future studies, then, could perhaps try to compare and contrast the image repertoire of Neolithic and Roman art in Libya.
CONCLUSION
The study of art and power merits much greater attention. Here I refer, of course, not simply to power in society (there is ample evidence from these tombs to demonstrate that function of the art), but in a broader sense to the relationship between an imperial system and the artistic production within its frontiers. Roman imperial art provided an image pool to support the legitimacy of power in provincial society. But the cultural agenda followed need not be one of simple cultural emulation (as often understood by the term Romanization). The art from Ghirza seems to borrow Roman ideas about the representation of power, but to deploy them alongside a set of family values and religious reference points that were preeminently un-Roman.56 To that extent, Romanization seems a very inadequate and inaccurate term for the cultural process discussed in this chapter.57
This chapter was originally published as Mattingly 2003b (in Scott and Webster 2003), having been presented as a paper at the second RAC conference in Nottingham, England, in 1997. I am grateful to the Cambridge University Press for permission to reprint a reedited and slightly emended version here and to the Department of Antiquities for permission to publish my photographs from the National Museum in Tripoli. I have made a few minor emendations to the text and some additions to the notes.
1 Scott 2003.
2 Webster 2003.
3 Walker and Burnett 1981; Zanker 1988.
4 Webster and Cooper 1996; Mattingly 1997a.
5 Webster 1997a, 1997b.
6 Brogan and Smith 1984; Mattingly 1995, 197–207; cf. Barker et al.1996b, 118–20.
7 Brogan and Smith 1984, 40–92.
8 Ibid., 119–226. More recent discussions of the reliefs from the tombs include Fontana 1998; and Mattingly 1999, 2003b.
9 Barker et al. 1996a, 144–49.
10 See, for example, Brogan 1965, 47–56; Barker et al. 1996a, 144–49; and Mattingly 1995, 162–67.
11 Captain Smythe, quoted in Beechey and Beechey 1828, 504–12.
12 Mathuisieulx 1912, 71–72.
13 Goodchild 1976, 8.
14 Merighi 1940, 177.
15 Brogan 1965; cf. Ferchiou 1989b.
16 The important article by Fontana (1998) appeared after the early drafts of this paper was written (and its essential shape achieved) and his analysis, though different in many respects, shares an emphasis on power and is broadly supportive of the approach adopted here.
17 Di Vita 1964, 73–79.
18 See the interestingly similar ideas expressed in Fantar 1999.
19 See Brogan 1954, 1965, 1978 for Tripolitanian funerary reliefs.
20 See, for example, Coarelli and Thébert 1988, 761–818; see also the important collection of essays in Stone and Stirling 2007.
21 Mattingly and Hitchner 1995, 204–9.
22 Coro 1956; Mercier 1953.
23 Brogan and Smith 1984, 207–12, 215–24.
24 Mattingly 1999b.
25 Brett and Fentress 1996, 200–270; Mattingly 1995, 17–24.
26 Brogan and Smith 1984, 181, 262–63.
27 Ibid., 217; Camps 1989, 11–40. This is the theme that is most strongly picked up on in Fontana 1998.
28 Brogan and Smith 1984, plates 63, 78.
29 Ibid., plates 63, 79.
30 Ibid., plates 55, 61, 123–24.
31 Ibid., 227–32; Mattingly 1995, 194–200; cf. Rushworth 2000, 2004 on the correlation of late antique Berber kingdoms with limes sectors along the African frontier.
32 See, for example, Brogan and Smith 1984, plates 64–67.
33 Ibid., plates 67, 82, 110.
34 Ibid., plates 62, 69, 70.
35 Ibid., plates 70, 79, 82.
36 Daumas 1968.
37 Mathuisieulx 1912, 73.
38 See, for example, Brogan and Smith 1984, plates 53, 99, 103.
39 Mattingly 1995, 207.
40 Herodotus, History, 4.172.
41 Mela, De Situ Orbis, 1.8.45.
42 Brogan and Smith 1984, 178–80.
43 Camps 1986, 151–64; Luni 1987.
44 Brogan and Smith 1984, 135, 161, 261–62.
45 Ibid., 162.
46 Holloway 1975, 59–62.
47 Mattingly 1995, 206–7.
48 Mattingly 1996b.
49 Brouquier Reddé 1992b, 255–66.
50 Brett and Fentress 1996, 17–31; Camps 1980; Stone 2007.
51 Diggle and Goodyear 1970; Mattingly 1983; cf. Encyclopédie Berbère 38–39, s.v. “Laguatan.”
52 El Bekri 1913, 31; Mattingly 1995, 212–13.
53 Le Quellec 1998; Mori 1998; Muzzolini 1991; cf. Lutz and Lutz 1995; Gauthier and Gauthier 1997.
54 Lutz and Lutz 1995, 114–28.
55 Ibid., 145–64.
56 Mattingly 1995, 160–70, 202–13; see also Grahame 1998.
57 Indeed, it seems an excellent example of discrepant identity in action on the fringes of the empire.
