This much-anticipated publication of the 1997–2004 excavations at the Bir Messaouda basilica at Carthage tells an admirably clear, succinct, and accessible story; it is a model for making the most of a stratigraphically challenging site. Aside from archaeologists, the report is likely to pique the interest of other readers, especially historians, for its framing of fragmentary and arcane archaeological evidence within larger local and regional concerns of urban topography and land use, as well as ecclesiastical politics and history. From the introductory first chapter, Richard Miles makes this his primary goal.

Miles's introduction (9–23) to Late Antique Carthage puts the basilica of Bir Messaouda, constructed in the 530s and substantially rebuilt in the 570s, into the historical and archaeological context of the city. He sets aside traditional preoccupations with linking churches named in texts to archaeologically attested sites and focuses instead on how basilicas and other Christian buildings transformed the topography and functioning of the city. The large number of Christian buildings renovated or newly constructed by Justinian according to Procopius's De aedificiis are traditionally part of a narrative of imperial restoration after a century of Vandal disruption. The archaeological record of 12 churches, three chapels, two “monasteries,” and a martyrium built or rebuilt in the Byzantine 6th c., if not in Justinian's reign, generally confirms Procopius's claims, but it also supplies a corrective to textual sources that emphasize the rupture of civic and religious life under the Vandals. The Bir Messaouda basilica makes a good case for the civic continuity fundamental to an urban renewal that began in the late Vandal period and continued under Byzantine administration.

In Chapters 2–6, Miles and Simon Greenslade give a concise account of the three phases of the important but poorly preserved basilica of Bir Messaouda. These chapters are copiously illustrated with clear, detailed plans, annotated photographs, and axonometric reconstructions. It is a pity that the 35 color plates at the end of the volume could not have been integrated into the text. In particular, the phased plans and detail photographs of mosaics and paintings associated with Chapters 2–8 would have added vitality to the austere presentation of some of the basilica's most remarkable features. The numerous photogrammetric plans in the volume are difficult to read at the scale presented, whether in black and white or in color, especially in contrast to the clarity of the drawing of the phase-1 mosaic (62, fig. 3.33).

Chapter 2 is devoted to the Punic and Roman periods of the site and discusses primarily an insula of 3rd to 4th-c. Roman houses between Cardines 9 and 10 east that featured opus sectile, mosaic, and cocciopesto flooring and included shops on the Cardo 10 frontage. These houses together with a little-explored second insula on a terrace above Cardo 9 suggest a residential neighborhood. In keeping with the goal expressed in the introduction, the razing of these dwellings in the late 5th to early 6th c. is the first known transformation of urban space at Bir Messaouda; the abandonment and destruction of the basilica in the 7th c. and later (Ch. 6) is the last. Between these two phases of activity are phases 1–3 of the basilica complex. Given Miles's emphasis on urban change, it seems counterproductive to assign phase numbers only to construction and not to destructive processes.

Phase 1 begins with the early to mid-6th-c. construction of a hall, oriented north–south (technically, northeast–southwest with the city grid) and dating to the very late Vandal or very early Byzantine period (TPQ of 530–34), which is the focus of Chapter 3. This large space (17.7 m × 8.25 m) was divided into three aisles and 12 bays, and an elongated northern apse enclosed by side rooms; it was entered through an inconspicuous side entrance from Cardo 10 east. Too large for a private audience hall and devoid of both the liturgical arrangements and on-axis front entrance customary in churches, the authors suggest that it may have been a meeting hall for an association or guild. By wondering whether associations large enough to require this space really existed in the 6th c. and by offering the first phase of the Dermech 1 basilica as a structural parallel to the Bir Messaouda hall, they signal that it may have been a church despite appearances to the contrary (on which more shortly).

Structurally and in terms of urban transformation, a more evocative parallel to Bir Messaouda 1 than the Dermech 1 basilica in its first phase is the phase-2, early 5th-c. elongated hall at Carthagenna, just two city blocks to the south. The so-called “monument à colonnes” was a somewhat smaller hall, aligned northwest–southeast and divided into three aisles and 10 bays; at its southeast end was a rectangular area flanked by side rooms, of which the southerly one was an entrance courtyard from Cardo 10. The space of the same size at the northwest end was occupied by three rooms, the northernmost a narrow corridor providing access to Cardo 9 and the other two rooms.Footnote ¹ Liliane Ennabli concluded that the phase-2 structure (adapted from a 4th-c. porticoed courtyard of a house) was a church; she interpreted the two rooms (A and C) at the northwest end of the hall as a potential presbyterium and sacristy. Part of her argument is that the site retained its ecclesiastical character when the Vandal-period church was replaced by a Byzantine basilica. Ennabli's argument is worth discussing in light of parallels between the “monument à colonnes” and the Bir Messaouda hall. This opportunity was missed by Miles and Greenslade.

Chapter 3 also recounts the transformation of Bir Messaouda's aisled hall into a Christian basilica from the 540s or 550s. At this time a chancel was installed in the nave; its 11 surviving postholes suggested two possible arrangements (fig. 3.42). This alteration to the interior of the hall was in tandem chronologically with the installation of the chancel at the Dermech 1 basilica in the first half of the 6th c., which connected the old and new altars, and, one might add, was contemporaneous with the construction of the Byzantine basilica at Carthagenna.Footnote ²

Chapter 4 describes phase 3, the remarkable refashioning of the basilica in the 570s into a southeast–northwest-oriented, five-aisled transept basilica with an enclosed apse and a dome over the altar at the intersection of the transepts (figs. 4.50 and 4.51). The church was entered through a triple arcade from Cardo 9; its narthex occupied the first bay of the basilica but was clearly distinguished from it by a columnar screen and the north–south orientation of the pattern of its mosaic floor. A later addition to the basilica was a solea down the center of the nave, including an ambo. It is not clear why the authors consider the installation of a screen in the phase-1 hall at Bir Messaouda a new phase, while the installation of the solea and ambo in the phase-3 basilica is not. These features are key to the organization of the basilica's liturgical space and potential clues to the form of the liturgy. Integral to the phase-3 basilica complex were a roughly square baptistery southwest of the nave and a martyrial chapel or crypt to its northeast, discussed at length in Chapter 5. Both structures are outstanding examples of creative adaptation of available urban space to contemporary needs.

At 17.70 × 17.90 m, the baptistery lay along, and evidently could be accessed from, Cardo 9 east, as well as from the south side aisle of the contemporary basilica. Its irregular double-shelled plan included no fewer than 47 columns. The hexagonal, marble-revetted font was on a platform above the level of the baptistry floor, perhaps to accommodate a pre-existing drain leading to a soakaway tank (a repurposed Punic cistern). In the absence of a canopy over the font, the inner set of columns may have been adapted to serve as a visible barrier. The authors might have suggested here that the notable elevation of the baptistery floor over that of the basilica, beyond practical considerations, might have signaled its importance; this was the case at Bir Ftouha, where the floor of the baptistery was higher than that of the adjacent peristyles.Footnote ³ Not mentioned in this chapter is the small circular font inserted later into the hexagonal one (see Ch. 14, 306), a significant omission considering that the reduction of a font's size may have signaled a change in ritual practice.Footnote ⁴ It is a pity that the authors do not consider alterations to the late 6th-c. basilica a fourth phase, which might have dated to the 7th c. The changes are significant and affect many parts of the complex; they include a new solea and ambo, new mosaics and a door or gate in the southern aisles, and repairs to the narthex floor (Ch. 4, 94–96).

Though the plan of the Bir Messaouda baptistery is idiosyncratic, it is one of a family of baptisteries at Carthage added to extant churches in the Byzantine period (Carthagenna, Dermech 1, Damous el Karita) or built new as part of a Byzantine complex (Bir Ftouha). Many others in smaller churches around Tunisia are known from their elaborately shaped, often spectacular mosaic-clad baptismal fonts of Byzantine date (at Demna, Hammam Lif, Sidi Jdidi, and Bekalta, to name just a few).Footnote ⁵ Simply put, baptisteries were characteristic of Byzantine churches in North Africa; their ubiquity leaves no doubt that the sacrament of baptism, whatever form it took, played a prominent role in the restoration of the Homoousian church.

Two Roman cisterns on the north side of the basilica were converted into a shrine, which may have been entered from the north transept wing; its floor was considerably elevated over that of the narthex at the bottom of a set of steps, by which visitors probably exited.Footnote ⁶ The shrine itself consisted of a waiting area (the leveled and filled eastern cistern), which preceded a short flight of steps down to a screened opening onto a tomb or reliquary crypt (the western cistern) adorned with paintings, discussed in Chapter 7 by Claudia Goodbrand. The striking arrangement, considerable size, and elevation of the baptistery and shrine over the five-aisled basilica suggests a clear hierarchy of space in the late 6th-c. complex.

After the abandonment of the basilica, and the fire that damaged it, the crypt was modified into living quarters for a brief period (Ch. 6). The end of this activity came with the infilling of the crypt and a linear feature nearby, the fill of which included pottery of the last quarter of the 7th c. Perhaps during this period, two graves were cut into a Roman mortar floor that underlay the southern end of the south transept. One of these graves contained at least three individuals. From approximately 700 CE, the church complex was used as a quarry for building materials and marble, at least some of which was made into mortar in a lime kiln on the site.

Before discussing the finds reports (Chs. 7–14), it is appropriate to emphasize that all but one of their authors grappled with the same challenge: to make sense of fragmentary and often unimpressive disjecta membra, undatable or residual material that was mostly not in situ. The responses to that challenge range from chapters intended to give readers a sense of the textures of the site in comparison with other sites, to a dry, technical report disengaged from the site and the city of Carthage.

The most insightful and wide-ranging of these chapters is Ben Croxford's study (Ch. 14) of the stone finds. Its merits include accounting for stone from all phases of the site: everything from sculptural fragments, part of a sundial, and the remnants of cube-screens (fig. 14.28) likely to be from the underlying Roman houses, to the marble types and sources represented in the flooring, wall veneers, and architectural elements of the basilica. The discussion of marble, in addition to types and their sources, includes some thoughts on stone recycling (318–20), like an unusual mosaic foundation of white marble chips (321), potentially derived from material on hand to fit recycled elements into new settings. Despite the fragmentary character of the architectural elements recovered from robbing fills, Croxford proposes a decorative scheme for the phase-3 basilica based on distribution patterns. The columns of the nave of the five-aisled church were probably of bianco e nero antico on white marble bases with recycled Imperial-period white marble Corinthian capitals. Exactly how the local Keddel limestone capitals may have fit into this scheme is unclear. One wonders if there were any traces of stucco on these Keddel fragments that may have given them a smooth white finish to blend in with the marble capitals. Black columns, including spiral fluted ones, were also associated with the east–west portion of the basilica. For the floor, Croxford posits opus sectile in the center of the nave, around the altar, and in front of the apse. The post settings and channels of the phase-3 screen, although an addition to the phase-3 structure, were clearly distinguishable from the phase-2 screen in the earlier church.

Jane Chick's Chapter 7 describes Bir Messaouda's rich collection of 6th-c. mosaics and offers a cogent interpretation of the later 6th-c. basilica's mosaic program. The description of the mosaics is often quite difficult to follow from the detailed photographs of the fragments, and the reader would have been well served by an overall drawing of the various mosaic panels in the context of the church. A visual representation of how the various patterns and panels complemented each other, distinguished different parts of the basilica, and directed traffic through it would have come close to reconstructing an ancient pilgrim's experience of the mosaics, which was the goal of the interpretative section of Chick's chapter (143–50). These pages offer stimulating insights into the directional character of the mosaics. The north–south pattern of the narthex pavement directed the faithful into the west–east pavement of the north or south outer aisles rather than directly into the nave. The nave and inner north and south aisles were screened (virtually) by the geometric panels in intercolumniations. Little remained of the nave pavement except its rather conventional enclosing border patterns, but the non-directional scallop-shell and pelta pavements of the inner aisles were surprisingly austere by comparison with the richness of those in the outer aisles. It seems as if this center of the basilica was meant for the business of regular congregants and the clergy rather than to impress and instruct pilgrims in the same way as the floors of the outer circuit. The fact that the west-aisle mosaic of the north transept wing echoed the iconography of paradise and salvation in the narthex suggests that the two areas served similar functions and constituencies. For mosaic comparanda, however, the author turns chiefly to Libya, all but ignoring the Bir Messaouda mosaics’ closest kin among the Byzantine churches in Carthage and elsewhere in Proconsularis, Byzacena, and Numidia.

Sylvia Funfschilling's report (Ch. 9) on the glass assemblage derived from baptistery and basilica contexts is a marked contrast; here, vessels that range widely in date are compared mostly to Carthage assemblages. More closely associated with the basilica itself in this assemblage was cylinder-blown window glass and two fragments of imitation-marble glass opus sectile inlays (cat. nos. 245 and 229, pl. 9.2), which have exact parallels in a 7th-c. context of the Crypta Balbi in Rome (163).Footnote ⁷ Even more interesting is the (catalogued but unillustrated) assemblage of glass from the post-basilica fills of the crypt (187–91) and the perhaps associated linear feature, which includes vessels of the light-blue glass typical of the late 6th and 7th c. and lamps with solid, slender stems, again typical of the late 6th and especially the 7th c.

Claudia Goodbrand's pottery report (Ch. 10) is thorough, and her characterization of the assemblage is made accessible through graphs; the raw data behind these is tucked into tables and an appendix. Although the assemblage is large, it is not helpful in dating the phases of the basilica complex, specifically because much of it is presumed to have been imported to the site in fill. As with the glass, the most interesting pottery derived from the closed fill of the crypt, which dates to the last quarter of the 7th c. This deposit included sherds of LR7 amphoras that date from the late 4th through the end of the 7th and into the 8th c., and flanged bowls (Hayes 109) with a date range into the late 7th c. and perhaps later, along with a few Islamic sherds. The pottery from the fill of the nearby and possibly related linear feature is similar, save for the absence of Islamic material (203–8).

Other specialists’ reports were limited by the quantity or quality of the material. Ralf Bockmann's modest chapter (257–62) is concerned mostly with metal building and construction materials; its highlight is the 57 nails in a burnt layer, all that remained of a charred door between the five-aisled basilica and its baptistery. Stefan Krmnicek's catalogue of the shockingly small number (385) of coins lacks any numismatic discussion or comparison with other assemblages at Carthage. However, the author provides a helpful, if rudimentary, list of contexts with coins and pottery dates, something that was not provided in the stratigraphic accounts in Chapters 2–6. Indeed, a full accounting of excavated contexts of the site as a whole, arranged by phase, would have been welcome for reference purposes, even if only in an appendix. Philip Mills's report on ceramic building materials (Ch. 13) gives these their due in a dense report that will be mostly read by fellow specialists. Raw data is presented only partially digested, perhaps because the author failed to take full advantage of comparable assemblages at Carthage, especially at Bir Ftouha.Footnote ⁸ Readers interested in the site of Bir Messaouda and Carthage will find a chapter that contributes little to either.

In the concluding chapter (Ch. 15), Miles returns to the broader themes of the introduction. He draws three main conclusions from Bir Messaouda. First, its strategic location on the lower eastern slopes of the Byrsa hill made it part of an intermural belt of churches, built in the 5th–6th c. in residential quarters for the use of local congregants, that extended from Carthagenna in the south to Dermech in the north. Second, this belt was associated with the increasing importance of the Maritime Forum, the late Roman administrative and commercial center that had superseded the city's traditional civic center on the Byrsa. Here, Miles draws on the work of John Whitehouse and Sami Harize, available to readers in an Appendix (365–84), which offers a cogent argument for locating the Maritime Forum between Cardines 13 and 15 east and Decumani 1 and 3 south. Third, Miles concludes that this belt of churches around the Maritime Forum was an urban renewal project begun in the Vandal 6th c. and carried forward in the Byzantine period, driven by African lay and ecclesiastical elites with the goal of reinventing their city as a pilgrimage center.

Miles's discussion of urban development in 6th-c. Carthage includes a seemingly minor hypothesis that the phase-1 aisled hall at Bir Messaouda probably was a church, all appearances to the contrary (352). He expands on Liliane Ennabli's cautious hypothesis that the phase-2 “monument à colonnes” at Carthagenna was a church by proposing that Bir Messaouda 1 may have been a newly constructed Arian church that lacked the fixed altar and screened chancel by which scholars are accustomed to recognizing Christian churches. He suggests rather that these features at Bir Messaouda, perhaps like its baptistery, are characteristic of its Byzantine successors. This hypothesis deserves application to other sites in North Africa. Might the off-center entrance, absence of a fixed altar, and screens in basilica 3 at Sidi Jdidi, for example, be an indication that it was an Arian church?Footnote ⁹ If the evidence from other sites bears out this hypothesis, it could shed some welcome light on the appearance of Arian churches, about which virtually nothing is known from hostile textual sources; that would be groundbreaking for early Christian archaeology. The Bir Messaouda report, in short, demonstrates the value of taking a historical view of archaeological minutiae to drive interdisciplinary research forward. That is no small feat.