First works about the arch and the inscription

The first report of the arch, isolated and with hardly any information, is provided by M.D. Shaw (Reference Shaw1743, 227), who visited the site in 1727 and was able to see it in a ruinous state. Otherwise, the first announcement of the inscription comes from Camillo Borgia, a Neapolitan cardinal who toured Tunisia between 1815 and 1817 with the purpose of exploring antiquities and conducting a series of excavations. According to the editors of the eighth volume of the CIL, there is a copy of fragment D of the inscription among his manuscripts.

The first publication is by Greenville T. Temple (Reference Greenville1835, 351, n. 177), who copies the same block without any attempt at reconstruction or providing information on the place and conditions of the find (Figure 2). The same is done by Pellissier (Reference Pellissier1848, 395) who presents fragment D (Figure 2), although with some reading errors which were solved five years later in another work (Pellissier, Reference Pellissier1853, 253).

Figure 2. Readings of the main inscription of the Mustis arch according to different authors.

In 1858, as a result of one of his trips to Tunisia and with the intention of preparing Salammbô, G. Flaubert (Delavoye, C-M., Reference Delavoye1999, 133–34) describes the remains of the arch, mentioning the existence of a second, less well-preserved, arch which we can identify with the one dedicated to Commodus.

Barely four years later, Guérin (Reference Guérin1862, 100–102) provided the first scientific description of the remains of the monument, stating its measurements and locating at its base the four fragments compiled in CIL VIII, 1577 (Figure 2).

As early as 1881, G. Wilmanns, editor of the first part of the eighth volume of the Corpus Inscriptionum Latinarum, dedicated to the Inscriptiones Africae Latinae, presents in number 1577 the reading and reconstruction of fragments B, C and D,Footnote ⁵ together with a fourth text, called A, engraved on a block originally placed in the upper part of the arch (Figure 2). Ten years later, in number 15572 of the first volume of the supplement, R. Cagnat and I. Schmidt proposed a different solution for the end of the second line. This suggestion, with minor variations, has prevailed in the scientific literature ever since. The first person to compile this proposal was J.C. Rockwell in his PhD dissertation (Reference Rockwell1909: 21, n. 88).

Shortly after the CIL, the architect H. Saladin (Reference Saladin1892, 547–49) published a sketch of the arch and proposed a restitution of its original form following the classical patterns of this type of building. As we shall see, this proposal was incorrect and was corrected in the subsequent restoration project for the monument.

Restoration and the discovery of the fourth fragment

In 1965, a team of workmen began to demolish the south end of the arch on the orders of President H. Bourguiba. However, in 1966, the project to restore the building began, transforming it into its present appearance (Vérité Reference Vérité1988, 54–55; Reference Vérité2021, 27–28). The project was awarded to J. Vérité by M. Fendri, Director of the Service du Monuments Historiques de Tunis. Beschaouch (Reference Beschaouch1969, 273) refers to this work when he mentions the context of the discovery of the fourth fragment. Although he does not specify where it was found (Beschaouch Reference Beschaouch1969, 273–74; Reference Beschaouch1981, 108), we can assume that it was at the foot of the monument, as were almost all the blocks that formed part of its structure (Vérité Reference Vérité2021, 5).

The project to restore the arch was explained by the architect responsible for its execution in a little-known publication (Vérité Reference Vérité1988). However, years earlier, N. Ferchiou (Reference Ferchiou1985) analysed and assessed the work carried out, concluding that certain parts of the monument, such as the cornice, were poorly restored. In particular, the author proposes that the fragments containing war-themed decorative motifs should be placed on the east side of the arch, towards Karthago, while the flowers and fruits would be oriented towards Mustis (Ferchiou Reference Ferchiou1985, 111). In this work, the author mentions for the first time another inscription related to the arch and dedicated to the Genius Mustis Augustus (Ferchiou Reference Ferchiou1985, 102).

The study by J. Vérité (Reference Vérité1988; Reference Vérité2021) rejected Saladin's proposal. In particular, he suggests that the dedicatory inscription was placed on the frieze and not on the attic (Ferchiou Reference Ferchiou1985, 101). Furthermore, J. Vérité (Reference Vérité2014, 3) rectified part of his own work, proposing that the head of the arch was crowned by four supports, not two: fragment A of CIL VIII, 1577 – our inscription no. 1, which would have supported a statue of Gordian III; AE 1981, 866, whose text alludes to the C(oloniua) I(ulia) K(arthago); another anepigraphic block; and the inscription, already mentioned by Ferchiou, dedicated to the Genius Mustis Augustus – our inscription no. 3. According to Vérité (Reference Vérité2014, 14), the inscriptions and statues of Gordian III and the Genius of Mustis would have been placed on the east side of the attic, welcoming passers-by entering the city. On the west side, on the other hand, would be placed the text referring to the city of Karthago and the anepigraphic block. As Vérité recognises, this new layout would fit in with the proposed placement of the cornice enunciated years earlier by N. Ferchiou (Vérité Reference Vérité2021, 18). The dimensions of the four blocks, ca. 50 cm high, fit with the waiting beds documented in the upper part of the attic (Vérité Reference Vérité2014, 2–4). However, Beschaouch (Reference Beschaouch1981, 108–109) identifies the block with the inscription C.I.K as a boundary marker of the territory. It was found in the vicinity of the monument, but Beschaouch does not clearly suggest its original position in relation to the arch. In any case, it seems that the monument dedicated to Gordian III served as a boundary marker for the territory of Mustis with respect to the pertica Carthaginiensium (Beschaouch Reference Beschaouch1981, 109; Ferchiou Reference Ferchiou1985, 111; Vérité Reference Vérité2021, 22), just as that of Commodus marked the eastern limit of the nucleus with respect to the territory of Sicca Veneria (Beschaouch Reference Beschaouch1981, 111–14).

The information provided by Beschaouch and the subsequent work

Four inscriptions were associated with the arch and the present study will focus on three of them. Below, we will call inscription no. 1 the dedication in favour of the emperor Gordianus – Block A of CIL VIII, 1577; inscription no. 2 the main text of the arch, divided into four blocks – three of them included in CIL VIII, 1577, with the letters B, C and D; and inscription no. 3 which alludes to the Genius Augustus of the city, not included in any other repository or published work (Table 1).

Table 1. Table of comparisons of the inscriptions to be analysed.

The information provided by Beschaouch (Reference Beschaouch1969) is not accompanied by photographs, drawings or a transcription of the full text of inscription no. 2 and, although the block B has been known since 1969, a clear picture of its text has never been published so far. Beschaouch limited himself to mentioning that the unpublished fragment B ‘donne une idée exacte de la disposition des lettres et des lignes et fait rejeter les restitutions proposées par l’éditeur du Corpus’ (Beschaouch Reference Beschaouch1969, 18). He then lists that: (1) the arch can be dated to 239 AD; (2) that its erection was promised by a perpetual flamen of Mustis; (3) that it was completed by his heir, a native of Zama Regia; (4) that the sum employed amounted to 50,000 HS; and (5) that, thanks to the expression SVPERPOSITIS STATVIS, we know that several effigies were originally placed on the monument.

The first researcher to echo this news is Duncan-Jones (Reference Duncan-Jones1974, 91), who values the cost of the arch at 50,000 HS, without further details. However, in Ramírez Sádaba's tabulations (Reference Ramírez Sádaba1981, 238), the cost is listed as ‘more than 5,000 HS’, as in a publication by Duncan-Jones (Reference Duncan-Jones1962: 80, n. 37) prior to Beschaouch's report (Reference Beschaouch1969). Beschaouch (Reference Beschaouch1981, 108–109) mentions the discovery of the unpublished block again in another work, but does not provide any further details about it.

Subsequently, Wesch-Klein (Reference Wesch-Klein1990, 150, n. 11) copies the transcription proposed by the editors of the CIL, but includes, in line 6, the expression SVPERPOSITIS STATVIS. He is followed by Fagan (Reference Fagan1999, 274, n. 40), who gives a partial reading of the whole text, only highlighting those parts that are most interesting for his study. Saastamoinen (Reference Saastamoinen2010, 524, n. 550) also follows the reading of the CIL, broadly following Wesch-Klein's proposal. On the other hand, C. Blonce (Reference Blonce2015) does not recognise the variants enunciated by Beschaouch. She intends to analyse the decorative motifs of the honorific arches and their meaning based on the terms of the inscriptions, but in this case she relies solely on the CIL's proposal which, as we shall see, is erroneous in many of its points.

Finally, we must mention the unpublished work by Vérité (Reference Vérité2021) which, although focused on the restoration project of the arch, provides in annex 2 a drawing of inscription no. 2 in which, surprisingly, appears the block B enunciated by Beschaouch (Reference Beschaouch1969) but never published.Footnote ⁶ In this same work, the architect mentions that ‘Je n'ai pas trouvé de restitution valable (celle de l'Université de Heidelberg est erronée), je donne ici ma copie de la dédicace de la frise’ (Vérité Reference Vérité2021, 18).

Inscription 1

The so-called fragment A of the edition of CIL VIII, 1577 is an independent inscription (Table 1). It was located during the restoration of the structure at its foot on the east side (Vérité Reference Vérité2021, 14). It is a partially damaged limestone block, ca. 81 cm long, 77 cm high and 59 cm depth (Figure 3), containing the remains of the negatives of the anchor of a possible statue of the emperor (Vérité Reference Vérité2021, 20). Traditionally, the reading proposed by Guérin (Reference Guérin1862, 100) and followed by the editors of the CIL has been accepted (Figure 2). In contrast, the unpublished drawing by Vérité (Reference Vérité2021, 14) allows the inclusion of a first line, partially fragmented, which alludes to his position as emperor (Figure 3):

Figure 3. On the left, Vérité's drawing (Reference Vérité2021, 14) showing the location of the dedication to the Gordian III and, on the right, the digitisation of the drawing of the artifact.

[I]mp(eratori) / Gordiano / Aug(usto)

The small size of the title is in keeping with the dimensions of the pedestal. In addition, the high height at which it was placed would make it unnecessary to provide further details about it.

Inscription 2: The dedication of the arch

The remaining elements of CIL VIII, 1577 are from the dedicatory construction inscription. The CIL editor's proposal does not correspond to the Vérité drawing (Figure 4), which in turn confirms that most of the facts announced by Beschaouch (Reference Beschaouch1969) are correct. However, there are more details that need to be commented on in depth.

Figure 4. Drawing of Vérité (top) and its digitisation (bottom). This shows our proposed reading. In green the missing characters and in red those that could be removed.

If we assume the reliability of J. Vérité's drawing, which is plausible in view of the care taken in representing the scale, the breaks in the piece and the photographs of the arch available (Figure 5), we can propose the following transcription of the text:

Figure 5. On the left, photograph showing the state of the arch in 1986 (Vérité Reference Vérité2021, 12). On the right, detail of CIL VIII, 1577 in its present state. (EDCS © - Museum RGM Köln).

Pro salute Imp(eratoris) Caes(aris) M(arci) Antoni Gordiani Pii Felicis Aug(usti) p(atri) p(atriae) trib(unicia) pot(estas) II co(n)s(ul) [desig(nati) II vel iter(um)] arcum quem / [---------] Ianuarius flamen per(petuus) inlata legitima summa ex HS V n(ummum) quae [patriae] suae promiserat / C(aius) Corne[lius] [Vi]talis vel [Na]talis Maior Zamensis Regius vir egr(egius) in memoriam quod soceri sui [et adfe]ctionem Mustitanis / [multipl]icata summa ex HS L m(ilibus) n(ummum) superpositis statuis solo publico coepit et[iam] dedicavit datis / [sportulis dec]urionibus et epulo c[ur]iis omnibus et gymnasio univ[er]sis po[pul]aribus

As Beschaouch (Reference Beschaouch1969, 18) seems to suggest in the sentence quoted above, (Vérité 1921, 18), the distribution of the text in the blocks collected in the CIL does not fit with the actual layout and, therefore, neither with the drawing by Vérité (Reference Vérité2021), especially as far as the left end of fragment C is concerned. According to this new graphic material (Figure 4), the block contains more text than was represented at the end of the nineteenth century. We believe that this may be due to a misinterpretation of the ellipses placed in the drawings of the other authors. For example, just as Guérin (Reference Guérin1862) did not place any in fragment D because he was able to read all the letters present, those in fragment C do not indicate a break, but a series of characters that the author was not able to transcribe. On the other hand, the editor of the CIL interprets them as the fracture of the stone, which leads to the above-mentioned misunderstandings in comparison with Vérité's drawing, which we consider more reliable.

In this way, we can safely complete practically the entire contents of line 1, which records the emperor's title. Thus, we conclude that the term Invicti is not among his designations. This is logical considering that such an epithet, although in use since 238 AD, was not officially bestowed on him (Kienast et al. Reference Kienast, Eck and Heil2017, 187). Moreover, we believe that we can postpone the date of the epigraph's production to AD 240, instead of the AD 239 proposed by Beschaouch (Reference Beschaouch1969: 273). The break in block C at its right end, in contact with block D, leaves a space that inevitably had to be filled with some element of the imperial titulature, so we believe that a formula such as desig(nati) II or desig(nati) iter(um) is plausible, a condition that Gordian III reached in July 240 AD (Kienast et al. Reference Kienast, Eck and Heil2017, 187).

The second line indicates that the construction project began thanks to a promise ob honorem flamonii perpetui, made by a personage whose cognomen is Ianuarius. The summa legitima was paid to the local public treasury after attaining priestly office, thus complying with the law, which stipulated that any pollicitatio ob honorem, i.e. made in an electoral context, had to be fulfilled after attaining that office (Melchor Reference Melchor Gil1994, 201). Unfortunately, we do not know the full name of the donor. Guérin (Reference Guérin1862, 100–102) read [---]VI[---] in the second line of fragment A. Because of their arrangement, the characters could be part of the character's nomen. If we assume that the reading is correct, the few characters correspond to too many variants. This prevents us from proposing a safe option, but it could correspond, for example, to the nomen [Fla]vi[us]. However, the nomen Cornelius is typical in Mustitan epigraphy also as a tribe, to which its inhabitants were assigned (Lepelley Reference Lepelley1981, 147). In view of this, perhaps the characters read by Guérin are not complete, lacking the initial stroke of the N and the horizontals of the E, so that it can be transcribed as [Cor]ne[lius]. We have left out the mention of the tribe [Cor]ne[lia] due to the limited space available to also insert the character's praenomen and nomen.Footnote ⁷

Thanks to the unpublished block, we know the full name of the son-in-law, whose cognomina could have been [Vi]talis Maior or [Na]talis Maior. We are inclined towards the first option, both because of the space available and because the first anthroponym is more common in the provinces of North Africa.Footnote ⁸ As Beschaouch (Reference Beschaouch1969, 273) has already stated, the heir financed the construction project, even though he came from another city, Zama Regia. Moreover, he is described as vir egr(egius), which is logical if we consider that his relative attained the perpetual flaminate. This priesthood may have served to honour important wealthy local families, whose members had played a good part, if not all, of the municipal career (Melchor Reference Melchor Gil1992, 90), and could thus be promoted to the equestrian rank.

The reconstruction of the end of the third line must also be modified with respect to the CIL proposal, since as is the case in all the lines of the text, the editors assumed the existence of a fifth fragment placed between C and D, which, judging by the fit of the two, should never have existed. Therefore, the text in this part, in all the lines, must be smaller than the one proposed at the end of the nineteenth century.

The transcription of the fourth line must be altered both in the distribution of the characters and in its content. Instead of recording the inclusion of all kinds of ornaments for the structure (Figure 2), it merely states that the arch was paid for, after multiplying the money, with 50,000 HS. This sum also included the statues placed on the attic of the arch, which, as we have seen, were to represent the emperorFootnote ⁹ and the Genius of the city. We believe that the drawing by Vérité (2020) shows more characters than those present at this point, since it expresses an unusually precise amount of 50,050 sesterces with a strange solution … …ex HS L m(ilibus) L n(ummum)… … Perhaps the architect, in his zeal to copy every character present on the piece, duplicated both numerals, thus causing the text to be more crowded than usual in his drawing. We therefore consider it more likely, in the absence of photographs, to be the presence of an expression such as ex HS L m(ilibus) n(ummum). Since Beschaouch (Reference Beschaouch1969) announced a figure of 50,000 HS, we consider our assumption about Vérité's failure on this particular point as reliable.Footnote ¹⁰

The entire structure was erected on public land, a concession which necessarily emanated from the decurions of Mustis after the express request of the evergeta. Moreover, C. Cornelius began the works after Ianuarius had passed away. After its completion, he also dedicated the monument -coepit e[tiam] dedicavit-.

It should be mentioned that Guérin (Reference Guérin1862, 100–102) drew, at the beginning of the fourth line, the characters [---]DI[---], which the editor of the CIL reconstructed as [a]di[ec]t[is]. However, the presence of the word [multipl]icata, which is only present in African epigraphy, is obvious. This has traditionally been interpreted as an imprecise extension of the original quantity, unlike the more specific terms duplicata or triplicata, also typically African (Saastamoinen Reference Saastamoinen2010, 344). The space left at the beginning of the fourth line could perfectly accommodate the term proposed in CIL, but it would imply the first attestation of both past participles in succession, whether or not they were separated by the et particle. This makes such a solution highly improbable. Perhaps the T corresponds to the first character of this type in the word [multipl]icata and the I could have been confused with the beginning of one of the strokes of the M or the L. Unfortunately, we have not found a plausible solution for the reading of the D. If we take into account the general ordinatio of the text, arranged without following a central axis of symmetry, we consider that before the term [multipl]icata there was no other word.

Furthermore, the missing text at the right end of block C, as in the previous lines, could not have been so extensive, so we propose the sole presence of the et particle or, given that the modulus of the letters and their limited separation in the rest of the text suggest the existence of more characters, a reconstruction as et[iam]. Although it is rare, we have also documented it in Africa Proconsularis.Footnote ¹¹

The last line of the text includes the ephemeral donations that the evergeta offered to the different sectors of the population in the context of the dedication of the building. In this way, the munificent would increase his existimatio and popularity in the population centre, enhancing the day of the inauguration of the monument and ensuring the greatest possible attendance at the ceremony (Gómez-Pantoja and Rodríguez Ceballos Reference Gómez-Pantoja, Rodríguez Ceballos, Rodríguez Neila and Melchor2006, 350; Melchor Reference Melchor Gil1992, 225–26). The editor of the CIL proposes the celebration of ludi for several days, the distribution of sportulae to the decuriones, gymnasia and the celebration of a banquet for the populares. However, thanks to Vérité's drawing (Figure 4), we know that these donations consisted of the distribution of sportulae to the decurions, the holding of a banquet for all the curiae and the distribution of gymnasia for all the populares. Thus, although all the population groups benefited in the same way, the destination of the distributions is different, the duration and the specific quantity of the donations are not indicated, and the organisation of public games is not present.

Therefore, the translation of the whole text would be:

For the health of Imperator Caesar Marcus Antonius Gordianus Pius Felix Augustus, father of the fatherland, during his second tribunician power, in his first consulship, already designated for the second. Once the legitimate sum of 5,000 sesterces promised to his homeland by … Ianuarius, perpetual flamen, had been delivered; Caius Cornelius Vitalis Maior, originally from Zama Regia, vir egregius, in memory of his father-in-law and out of appreciation for the people of Mustis, began the arch after multiplying the sum to 50,000 sesterces, with superimposed statues and on public land. He dedicated it after offering sportulae to the decuriones, a banquet to all the curiae, and gymnasio to all the populares.

Inscription 3: Dedication to the Genius Augustus at Mustis

We have previously discussed that there must have been at least two more inscriptions on the arch, one dedicated to Gordian IIIFootnote ¹² and another consecrated to the municipal Genius. This latter inscription, engraved on a square limestone block measuring 60 cm in height, 87 cm in length and 71 cm in depth, was located, according to Vérité's plans (Reference Vérité2021, 14), at the base of the arch on its east side (Figure 6). Its upper part contains the negatives of a possible statue attachment (Vérité Reference Vérité2021, 20). Furthermore, thanks to the new graphic apparatus (Figure 6), we know that the text is distributed over two lines as follows:

Figure 6. On the left, Vérité's drawing (Reference Vérité2021, 14) showing the location of the dedication to the Mustitan Genius and, on the right, the digitisation of the drawing of the artifact.

Genio Mu[s]/tis Aug(usto)

Although known since 1966, it has never been published individually in a scientific journal. Several researchers have mentioned its discovery, such as Ferchiou (Reference Ferchiou1985, 102 and 111), without providing any photographs or drawings of the find. Only Vérité (Reference Vérité2021, 14) offers a drawing of the discovery. Beyond this, we know that this deity was identified with Pluton Frugifer (Beschaouch, Reference Beschaouch1973, 104) and was probably placed at the top of the arch, alongside the dedication to Gordian III, welcoming those who entered the city (Vérité Reference Vérité2021, 20). This aligns with the references to the so-called African Pluto on the arch itself, contrasting with the warlike motifs associated with Carthage (Ferchiou Reference Ferchiou1985, 111). Thus, this epigraphic support would serve, along with the rest of the structure, as a boundary marker for the city's territory and would represent the second known dedication to the Mustitan Genius.Footnote ¹³