As part of the prestigious series ‘Internationale Archäologie’ (no. 92), an important monograph by J. Eingartner about the templa cum porticibus in Northern Africa appeared in 2005 (Eingartner Reference Eingartner2005). Temple buildings grouped under this umbrella-label fit within an architectural typology – combining local Punic traditions and Italic features – which became very popular and widespread in that area, especially during the second and third centuries AD. Hence, such temples are usually referred to as ‘Romano-African’: i.e., monumental temples on podiums derived from Italic prototypes, enclosed by a surrounding portico and thus impressively dominating the ahead-set piazza (for a detailed insight into the typological issue, its origins and variations, see Aiosa Reference Aiosa2013, 163–88).

However, the work by Eingartner, whose catalogue includes almost all the buildings attested in Africa Proconsularis, Numidia, and Mauretania Caesariensis, has neglected – quite surprisingly – some already known temples, such as the Serapeion at Lepcis Magna (Di Vita et al. Reference Di Vita, Pugliese Carratelli, Di Vita-Evrard, Lazzarini and Turi2003, with previous bibliography). Even more puzzlingly, the whole architecture of Mauretania Tingitana has been left aside. Three so far acknowledged temples were built in this far west African province and fit in this considered architectural typology: the so-called Temple C at Volubilis (Euzennat Reference Euzennat1957; Majdoub Reference Majdoub, Mastino and Ruggeri1994), the Capitolium in the same ancient city (Luquet Reference Luquet1964; Domingo Magaña Reference Domingo Magaña2012b, 386–89, with previous bibliography),Footnote ¹ and the Capitolium of Sala – this latter being the most ancient among the three temples in this small, Tingitanian group. For these reasons, the here-reviewed study by Stefano Camporeale, recently published as the third issue of the series ‘Mediterranean Archaeology Studies’, is very welcome: focusing on the Capitoline temple of ancient Sala, it provides the opportunity to reconsider the topic of Romano-African temples and fill up the gap – at least partially.

The volume is divided into six chapters; they are preceded by abstracts in French and English,Footnote ² acknowledgements (again in both languages), a list of captions, and a useful introduction. This latter provides readers who are less confident about the topic with some general coordinates, which guide them in understanding the peculiarities of Mauretanian architecture (31–37). Data about the history of the investigations at Sala and, more specifically, its Capitolium are presented in chapter 1, along with some notes about the urban development of the city and the new architectural survey of the building, which was carried out through use of current methodologies and digital techniques (39–52). A detailed description of the building and its structures follows in chapter 2 (53–89), which also includes some notes about the Capitolium's dedicatory inscription (86–88).

Chapter 3 focuses on architectural decoration (91–104). It is a brief, but useful catalogue (the architectural members had previously been published in Mugnai Reference Mugnai2018), which gathers and presents 51 architectural elements, regrettably not accompanied by drawings – those by Hallier, published in Appendix C, cannot be considered as proper representations, being mainly dimensioned sketches. Both in situ members and isolated items are included, the latter not being always attributable with certainty because of spoliation, reuse, and continuous displacement over time. The catalogue is organized according to the location of use for each element in the building – if known – and, subordinately, to their typology. The label ‘basi quadrangolari di lesena’ (adopted after Mugnai Reference Mugnai2018, 143 and 270), used for the first pilaster and doorjamb bases from the west on the south façade, creates some misunderstandings; since it refers to the squared profile of their mouldings, ‘basi di lesena a profilo semplificato’ (i.e., simplified pilaster bases) would have been more suitable.

In chapter 4, Camporeale deals with a detailed analysis of building issues related to the construction site, focusing on the Roman phase (105–47). The proposed reconstructions are carefully argued; some connected aspects such as the adopted length-unit standards, the employed building materials and techniques, and the working traces (including some masons’ marks) are also well addressed. Noteworthy is the wide use of the opus africanum masonry. Appendix D, presenting a catalogue of building techniques, is to be read as strictly connected to this chapter (285–94); the catalogue is structured according to the criteria of the ACoR Project (‘Atlas des techniques de la Construction Romaine’),Footnote ³ of which Camporeale himself is one of the editors.

Chapter 5 is devoted to the reconstruction of the later phases of the building (149–58). First, probably during the fourth century AD, the area was entirely transformed into a dump and a cemetery. Later, between the thirteenth and the fourteenth centuries (Islamic phase), once the temple had been completely dismantled, a building with a courtyard arose in the area of the former temple pronaos, while the underlying tabernae were partially transformed and reused. Finally, in chapter 6 (159–71) the author drafts up some conclusions, bringing the Capitolium of Sala into the current discussion about temple architecture in Mauretania Tingitana; rightly, he also emphasizes its role within the urban frame of the ancient city. Various apparatuses conclude the volume: four appendices (173–294), plates preceded by a list of their captions (295–304), and a substantial bibliography (305–17).

The volume is enriched by images, all of which are of very good quality, and in colour. However, the photos are often paginated away from their references in the text: this forces the reader to move – quite uneasily – from one page to the other and back; the same applies to the tables in chapter 2, which are placed at the end of the chapter. Moreover, in chapter 3, it would have been useful to reproduce the photographs of the architectural members at the same scale to make comparison easier. There are only few misprints throughout the text.

A short description of the main contents of the book follows below, along with some comments intended to foster scholarly discussion.

The Capitolium Novum of Sala – as the building is supposed to be recorded in its dedicatory inscription (IAMLat suppl. 861 = AE 1991, 1750), if one accepts the proposal by Boube (Reference Boube1990) – occupies the area of a whole block within the urban plan of the Roman city. Its construction, commissioned by C. Hosidius Severus (a local, equestrian-class citizen included in the imperial administration ranks), responds clearly to a precise urban programme, thus belonging to a scale – i.e., the one which includes the urban frame – wider than the bare architectural one. Moreover, it is now acknowledged that the building arose over a more ancient, but hitherto unreconstructible palimpsest; in contrast, its excavators had considered this area as previously unoccupied by constructions. Here an architect/urban planner (possibly indigenous and probably trained by Roman masters) built the temple complex – which was itself a remarkable result – and possibly designed the regular street grid surrounding it as well. If, expanding the frame of his work, we attribute to him also the construction of the nearby public monumental buildings (Buildings G1 and G2, Temples B and C), and if – further on – we assign to him the overall design of the forum piazza south of the Capitolium, this needs to be carefully considered: even looking quite believable, for the time being this is nothing more than a fascinating hypothesis, probably destined to remain an unsolved question.

The forum piazza south of the Capitolium has commonly been referred to as the main decumanus of the city. However, a few years ago, Niccolò Mugnai (Reference Mugnai2018, 141), now followed by Camporeale, convincingly argued that it should preferably be acknowledged as a forum adiectum: in fact, a previous monumental district functioning as a forum area had been already set in the easternmost part of the citadel (Mugnai Reference Mugnai2018, 136–40, with previous bibliography), impressively dominated by the Mauretanian, five-cellae Temple A.Footnote ⁴ If one refers to the hitherto accepted reconstruction of the Capitolium's inscription (supra), the forum might also be intended as a forum novum;Footnote ⁵ hence, the case of Sala would fit with that ‘multiplication of fora’ phenomenon (or ‘polycentrism’), which is already acknowledged as typical for the North African Roman cities (Gros and Torelli Reference Gros and Torelli2007, 392).Footnote ⁶ Together with Building D,Footnote ⁷ the contemporaneous Nymphaeum by the piazza's south edge, and the monumental honorific arch (possibly of Antonine date: Camporeale Reference Camporeale and Papiforthcoming) marking the entrance to the open area from the west, the forum's general plan eventually found its physiological completion.

The site's orography, sloping from north to south, led to build the porticoed enclosure on a high terrace; this latter was supported at the bottom by a row of nine tabernae and, further to the north, by an artificial earth-filling contained by the rear wall of the tabernae themselves. Almost all of them opened onto the forum, within an imposing pilastered façade. On the ground floor, the pilasters are set on Attic bases and crowned by smooth-leafed, Corinthian capitals. As for their entablature, the author suggests presence of cornices with small, square dentils hanging from an upper taenia: these elements (catalogue nos 17–18) are currently stored in the so-called Western Building. As for the tabernae's function, this remains still unidentified; widely intended, it can be recognized as commercial, financial and/or administrative, if not connected with banking activities; rebus sic stantibus, their prototypes can be recognized in the Markbauten of Asia Minor and in the templa cum tabernis typology (see, for instance, the Flavian Temple at Lepcis Magna: Dolciotti et al. Reference Dolciotti, Mighetto, Dell'Era and Loglio2013–14a, 133–40; Reference Dolciotti, Mighetto and Limoncelli2013–14b). At any rate, it looks likely and plausible that, being rented, their management granted some form of income for the sanctuary's coffers.

The entrances to each of the rooms open up in the intervals between one pilaster and the one next to it. The doorjambs reproduce the same architectural order of the main pilasters, but at a lower scale. According to the author, the jambs were supporting flat arches, to which it is tempting to attribute at least one erratic keystone found in the area. Camporeale reconstructs tympanums above them, in the form of aediculae; the suggested tympanums are alternately triangular and semi-circular, being modelled on those of the Building of Eumachia at Pompeii and the hemicycle of Trajan's Markets in Rome. However, lacking any evidence, such reconstruction – even if somehow persuasive – turns out to be just an assumption.

Above, at the same level of the porticoed piazza, a similar pilastered façade is also suggested. It might have been crowned by an attic wall: here, one would place the imposing inscription, carved on joint slabs of Parian marble. As a matter of fact, the analysis of both the overall dimensions of the epigraphic text and the height of the letters – indeed corroborated by effective comparisons – leads the author to new conclusions about its placing. The hypothetical position above the north entrance door, as originally suggested by Boube, looks quite far from the truth; therefore, it may be challenged and rejected in favour of an arrangement at a well-elevated height from the observer – and what better location than the one overlooking the forum area, as in the proposed reconstruction? This attic wall might be completed with small pilasters, possibly aligned with those of the lower storeys: they would have enlivened the otherwise flat surface of the wall, framing the inscription – which was centred with respect to the forum and not to the façade – and defining its epigraphic field.

The upper piazza of the Capitolium – which was accessible from the north as well as from the east – was enclosed by a porticus triplex. Remains of the portico are preserved along its northern and eastern sides: respectively, the stylobate to the north, with the two westernmost columns in situ, and its foundations to the east. Postulating the presence of a third side of the portico to the south does not create any difficulty:Footnote ⁸ it can be assumed that its colonnade was laid exactly on one of the ashlar arches which were spanning within the opus caementicium barrel vaults of the underlying tabernae and, as such, working as static ribbons. The western half of the encased squareFootnote ⁹ is occupied by a tetrastyle temple on podium; at the bottom one could see a canonical cyma recta, evidenced by a short profiled section in the west part of the north wall. Along the longitudinal axis, the altar would have been found: only its foundations are preserved, but they are sufficient to outline its general layout.

To the north, a small temple obliterates the area between the major sacred edifice and the west end of the north portico. This small temple belongs to a later phase, which the author attributes to the second century AD; a lack of elements prevents any attempt of proposing a more precise chronology. The idea that it might have hosted the cult of the Domus Principis is convincing, thus strictly connected with that of the Capitoline triad. Coeval – or shortly later – is the paving of the square, in local limestone: the paving slabs lean against the marble veneering of the first step of the staircase.

As for the main temple, it presents a symmetric plan, mirrored with respect to the axis separating the deep pronaos from the cella. Camporeale proposes to attribute a pilaster capital – once again of the smooth-leafed type and found immediately north of the building – to the east end of the north wall, instead of a canonical anta. The suggested solution is not unknown, being also adopted, for instance, in the Temple of the Gens Septimia at Cuicul, which is mentioned by the author himself. This creates a slight misalignment between the axis of the wall and that of the columns of the pronaos, which becomes visible on the entablature and roof as well.

One fragmented, corner Ionic epistyle and a few smooth ashlar blocks have been so far attributed to the temple's entablature. The ashlar blocks, presumably used for the frieze, are moulded at the top and show joist recesses on their rear. Fortunately drawn by Gilbert Hallier, to date, they seem to be no longer traceable. Finally, to the monumental entrance of the cella, Camporeale assigns ashlar blocks of a flat arch, moulded at the top with a cyma recta, a fillet, and a fascia. The blocks of a round arch, crowned with the same mouldings and thought to be set in a presumed relieving arch, are attributed to the same door as well.

The portico's columns also employed the same types of Attic bases and smooth-leafed Corinthian capitals which are used in the rest of the building: a strict coherent conception, a formal as well as building homogeneity (as a result of the adopted masonry techniques, being opus africanum widely used), and a substantial design unity are clearly resulting from this. On the north stylobate, the columns’ rhythm is evidenced by the levers’ traces, used for laying the columns; quite puzzlingly, empolia are lacking. A slightly irregular rhythm for the columns on the north stylobate must be acknowledged; differently, the interval between the columns of the south side can only be reconstructed as a regular one.

What emerges once again is the high degree of difficulty experienced in the study of this building, which S. Camporeale restores to new life, presenting it to the wider community and not only to specialists. Such a result is accomplished by combining: (i) the direct study of the ruins, rigorously guided by specific methodologies, i.e., the ‘archaeology of architecture’ and ‘archaeology of construction’; (ii) archival research, as it emerges from the analysis of the vast, hitherto unpublished documentation – both photographs and drawings – kept at the ‘Camille Jullian’ Centre in Aix-en-Provence and now available in Appendices B and C; (iii) a wide bibliographic research; (iv) a direct, long-experienced knowledge of Roman North African architecture. The sober, but elegant 3D renderings – gathered in Appendix A and made by Rossella Pansini – complete the study.

Some final notes concern the local scale of the construction issues. Apart from marble, possibly Lusitanian and sometimes adopted for the veneering, the employed lithotypes (calcarenite, conveniently stuccoed, but also limestone from two neighbouring quarries), the dimensional modules and length-units (Roman foot for the plan and Punic cubit for the elevation, indeed combined for the modular grid of the overall plan), and the features of the architectural ornament unmistakably refer to the autochthonous identity of the craftsmen who worked at the construction site of the Capitolium. The catalogue presented in chapter 3 (see supra) provides firm data about decorations, which find their natural framework in the broad North African panorama and not only within the borders of Tingitana. This is the case for the well-expanded tori of the Attic bases, which are also identifiable – apart from some specific peculiarities – on artefacts from Carthage, Bulla Regia, Mustis, and Thuburbo Maius, all of them supposed to be Hellenistic in date (Ferchiou Reference Ferchiou1989, 25, no. I.I.A.3, 29, no. I.II.A.6, 47, no. I.III.A.2, 48–49, no. I.III.B.2, figs 1.c–d, 2.a, 3.b, 9.b). This also applies to the atrophied scotia carved on the pilaster base which is set by the north-western access to the porticoed piazza: mutatis mutandis, the same profile is attested on bases at Banasa (Mugnai Reference Mugnai2018, 245–52, passim, and related photos in pls 20–25), as well as on those of the Temple of Roma and Augustus at Lepcis Magna (Livadiotti and Rocco Reference Livadiotti, Rocco, Di Vita and Livadiotti2005, 271–74, nos B1–9 and figs 2.96–97, 205, figs 2.37–38), and on early imperial-period bases from Gigthis and Assuras (Ferchiou Reference Ferchiou1989, 31, no. I.II.B.1.1, 33, no. I.II.B.1.5, 35, no. I.II.B.2.1, figs 4.c and 5, pls III.b and IV.a). Finally, this is the case for the Corinthian capitals with smooth leaves, quite widespread in Mauretania (Mugnai Reference Mugnai2018, 221–28 and figs in pls 9–12, 254–58 and figs in pls 27–29, 277–81 and related figs in pls 37–39; Pensabene Reference Pensabene2011, passim).Footnote ¹⁰

As for these capitals, Camporeale comments on the lack of carving of their acanthus leaves; even if a layer of stucco was applied to the stone – perhaps providing a marble-like appearance – this would not have achieved the detailed representation of a more naturalistic shape. The author argues that they derive from the architectural trends dictated by Carthage across the whole of Africa Proconsularis, although in a less elaborate version. At the same time, following Pensabene (Reference Pensabene and Giardina1986, 387–94; Reference Pensabene2007, 394), he traces the origin of such capitals as connected with simplification and standardization processes – the same processes that are justified by the need to export half-finished marble items from the quarries, intended to be completed at their final construction site. Therefore, the capitals from the Capitolium at Sala would represent – indeed, reproduced in local lithotypes – an early application within the framework of ‘una specifica organizzazione di cantiere, un'esigenza di celerità della costruzione o una scelta del committente’ (94) – circumstances, those implied by the author, which are used to explain both the decorative and construction homogeneity of the building. However, smooth-leafed acanthus capitals deriving from the described process and used uncarved, because formally conceived as finished, are not attested before the late Antonine and Severan periods; mainly in the form of composite capitals, they became very popular especially in Rome during Late Antiquity (Pensabene Reference Pensabene and Giardina1986, 288–90, 324–33; Reference Pensabene2007, 465–68; Pensabene and Panella Reference Pensabene and Panella1993–94, 120–21). On the other hand, smooth-leafed capitals had already acquired their own typological identity in previous periods, which has nothing to do with a supposedly unfinished stage of workmanship: as for this, some notes about Sabratha, its Iseion and Temple of Liber Pater, written by Antonino Di Vita (Reference Di Vita2017, 63–93) some decades ago – but only recently published because of the tortuous paths that our lives sometimes experience – may help to foster reflection.Footnote ¹¹

Overall, Camporeale's book represents a new, valuable opportunity to reconsider the multifaceted connections that were established between Rome and the periphery of the Empire from an architectural viewpoint. He has given us the possibility to learn about an important building at the edges of the Roman world, which, to date, was almost unknown and regrettably underestimated. The data provided on the architectural decoration, building techniques, and the related aspects of the construction yard broaden the picture of our knowledge of Roman architecture in Northern Africa, which still needs further investigation.