I INTRODUCTION TO POMPEII'S SAFAITIC GRAFFITI

As Pompeii's Safaitic graffiti are little-known to most classicists, I begin with a brief introduction. The texts were inscribed on the north wall of the theatre corridor (VIII.7.20), a passageway connecting the Via Stabiana to the east with Pompeii's theatre complex.Footnote ⁴ Although the texts were recorded as early as 1832, they remained a mystery until 1987, when Calzini Gysens published her editio princeps; an editio altera followed three years later.Footnote ⁵ Since Calzini Gysens’ breakthroughs, the texts have received almost no attention, with two notable exceptions. First, Varone included five photographs of the Safaitic in his 2012 photographic survey of the graffiti published in CIL 4.Footnote ⁶ Second, Al-Manaser and Macdonald included Pompeii's Safaitic in the Online Corpus of the Inscriptions of Ancient North Arabia (OCIANA), which is now the standard database for Safaitic.Footnote ⁷

As published originally by Calzini Gysens, the Pompeian Safaitic graffiti were construed as nine inscriptions, but two of these texts have been split in OCIANA, resulting in a total of eleven distinct texts.Footnote ⁸ The texts are written right to left, and are well incised into the plaster that covers the north wall of the theatre corridor, though damage and loss of plaster have affected some of the inscriptions. All are clustered in close proximity to one another.Footnote ⁹ As for their date, in her ed. alt., Calzini Gysens suggested 80 b.c.e.–c.e. 62, on the basis of proposed archaeological time stamps for the construction and use of the corridor, and following Maiuri's hypothesis that the theatre complex ceased to be used after the 62 earthquake.Footnote ¹⁰ However, according to more recent archaeological work, it seems instead that use of the larger theatre and the corridor continued, and thus the terminus ante quem for the corridor's graffiti can be extended to the date of the eruption itself.Footnote ¹¹

In terms of content, all of Pompeii's Safaitic graffiti record simple, commemorative signatures by their authors (‘by X’) or very brief patrilineal genealogies (‘by X, son of Y’). Such texts are the most common type of Safaitic inscription in the Ḥarrah, where these genealogies can extend up to a remarkable twenty generations.Footnote ¹² At Pompeii there are, at most, two generations represented (author and father). Moving across the wall from east to west and from top to bottom, I reproduce here the texts following OCIANA's editions and translations. Note that editorial conventions for Old North Arabian epigraphy differ considerably from those of Latin and Greek texts: so, for instance, curly braces indicate doubtful readings (like the Leiden underdot), while four hyphens indicate one or more unreadable characters.Footnote ¹³

While not all of the readings are completely secure, what is clear enough are the names of twelve individuals who were once present in the theatre corridor. Two of the inscriptions include genealogies, and in the case of CGSP 9, we seem to have two brothers, Ṣhb and (possibly) Hʾhys¹. Among the names that are fairly securely read, most are well attested elsewhere in the Safaitic corpus from the Ḥarrah. For example, querying the OCIANA onomastics database for ʾnʿm, the author's name in CGSP 6, returns hundreds of hits.Footnote ¹⁶ On the other hand, several names are rarer. For example, kʿz, the name of the father of ʾnʿm in the same inscription, otherwise appears only once in the corpus.Footnote ¹⁷ ʾtbt, read only with difficulty in CGSP 1, is so far unique to Pompeii.Footnote ¹⁸ Thus even if individual readings cause some difficulties, the identification of the graffiti in the corridor as Safaitic is secure. Yet the very existence of these texts at Pompeii raises more questions. Who were Tm, Ṣhb, and their companions? How did they end up at Pompeii?

Outside Pompeii, Safaitic graffiti are almost exclusively found cut into the basalt rocks of the Ḥarrah, and scholars studying this corpus have been able to reconstruct a great deal about the lives of their authors.Footnote ¹⁹ These were pastoralist nomads, whose livelihoods came from breeding camels, sheep and goats, and who migrated across the Ḥarrah according to the seasons, the availability of water and the rhythms of life in the black desert. So far, only Calzini Gysens has attempted to explain how these nomads might have made it to Pompeii, though she offers ‘only hypotheses … with the greatest caution’: she tentatively suggested that the nomads might have been ‘slaves, political hostages, or just travellers, probably coming from nearby Puteoli’.Footnote ²⁰ In the absence of any particular evidence for the authors being slaves or political hostages, Calzini Gysens highlighted the possibility that these men were eastern merchants, and this seems to be the best working explanation hitherto.Footnote ²¹ Against this, it might be said that we have no evidence for nomad involvement with trade in Puteoli — or, in fact, with trade of any kind, not even the caravan trade in Syria.Footnote ²² But, after all, Calzini Gysens advanced her hypothesis only exempli gratia. Still, with so much recent progress on the study of ancient graffiti, both Pompeian and Safaitic, I believe that it is now possible to do better.

II THE THIRD LEGION: A NEW HYPOTHESIS

The starting point for a new hypothesis is the north wall of Pompeii's theatre corridor. Though published in isolation, Pompeii's Safaitic graffiti have many neighbours.Footnote ²³ The theatre corridor was — and is — a densely inscribed space, whose plaster walls seem to have served as an inviting epigraphic canvas for passers-by.Footnote ²⁴ Roughly 150 additional graffiti have been identified here, and recent restudy of this material by Benefiel in her Ancient Graffiti Project (hereafter AGP) has made these inscriptions more accessible than ever.Footnote ²⁵

The majority of non-Safaitic graffiti in the corridor are, in fact, figural. These include a wide variety of drawings, from boats to gladiators to animals, including horses, fish and the occasional unidentifiable quadruped.Footnote ²⁶ Among textual graffiti, Latin dominates, but there are also seven inscriptions either in Greek or appearing with a combination of Greek and Latin characters.Footnote ²⁷ As for Latin, there are around fifty-seven inscriptions that exhibit a considerable variety in kind. Many — not unlike the Safaitic — simply record names. Among others, Felix, Gaius Ovius, Faustus and Fadius Nasso all commemorated their visits to this space, the last apparently providing a self-portrait.Footnote ²⁸ Some graffiti are more complex. The enslaved woman Methe, for example, recorded a prayer to Venus Pompeiana on the corridor's south wall.Footnote ²⁹ Elsewhere, a trio of men — self-styled as the Geryones — left behind a pair of graffiti, one of which records (with a precise date) their liaison with a certain Tyche on 22 November 3 b.c.e.Footnote ³⁰

Regarding the mystery of the Safaitic, however, two Latin inscriptions demand particular attention. Both are on the north wall, and are located close to the Safaitic, just c. 1.5 m to the east, heading toward the Via Stabiana, only three or so paces away.Footnote ³¹ But neither inscription is entirely straightforward. I present both texts below, according to recent updates by AGP.Footnote ³²

1. (1) CIL 4.2415, cf. p. 223 = AGP-EDR161769Footnote ³³

1. (2) CIL 4.2421 = AGP-EDR166493Footnote ³⁴

Both texts exhibit the sort of variation in orthography that is common in Vesuvian graffiti. The loss of the final –i in Tertiani, for example, should probably be understood as an elision before hic, with the initial h of that word perhaps not pronounced (it is, indeed, unwritten in the verb form that follows: <h>abitaru<n>t).Footnote ³⁵ In <h>abitaru<n>t, the loss of n before t is stranger, but can also be paralleled in the Pompeian corpus, while the loss of a geminated consonant like l in fel<l>as is more common.Footnote ³⁶

Turning to content, the pair are similar. Each begins with an announcement of the presence of the Tertiani, ‘the men of the Third’, whose identity I will return to shortly. On the basis of palaeography, it may be the case that the Tertiani sections of each inscription were written by two different hands. In (1), for example, the writer uses E capitalis, whereas in (2) we find a two-stroke E (||) (though such variation is also possible in the same hand). Following this name, each text offers a deictic hic, referring passersby to the corridor itself. Next comes a form of the verb habitare, here meaning simply ‘to be present’.Footnote ³⁷

Following the Tertiani section, each inscription includes a phrase involving vale. In (1), the word preceding verpa is a crux and the extant traces on the wall admit no secure solutions, though one might expect a name here. The valediction that follows addresses someone referred to as a verpa. This is an example of an anatomical term used pars pro toto either as a term of general abuse, or, as is possible with verpa, as a disparagement of the addressee that insinuates that they are a pedicator or irrumator.Footnote ³⁸ The farewell to Rufa in (2) offers more detail. The citation of her prowess at fellatio has landed her on McGinn's register of ‘possible prostitutes at Pompeii’, where she shares the company of Tyche, mentioned above; on the other hand, whether this reference to Rufa is actually evidence for prostitution or simply verbal abuse seems to me far from obvious.Footnote ³⁹ Either way, since Zangemeister's edition in CIL, it has been suggested that Rufa's valediction in line 3 should actually be severed from the Tertiani altogether and be read separately as its own inscription, a suggestion that has had its followers.Footnote ⁴⁰ The same thing might also be the case for the final vale-phrase at the end of (1), although the palaeography is indecisive. Be that as it may, the reason that these two inscriptions matter for the present argument is the presence of the Tertiani.

As Zangemeister noted in his addenda to (1) in CIL, the reference to the Tertiani, ‘the men of the Third’, seems to relate these graffiti to the Third Legion. While there were several legions in the Roman army numbered ‘the Third’, Zangemeister made a crucial suggestion as to which one could have been responsible for these graffiti. Tacitus records that Legio III Gallica was in Capua during late 69 and early 70. Given that we know of no other opportunities for members of legions numbered ‘the Third’ to visit Pompeii, Zangemeister suggested that this moment likely provided the window of opportunity for the men of the Third to leave their marks here.Footnote ⁴¹

Zangemeister was, of course, not familiar with the then-indecipherable Safaitic graffiti when he made this observation, but his connection between the Tertiani and III Gallica matters a great deal, because it places in this corridor members of a Roman legion that had recently arrived in Italy from Syria, the distant homeland of writers of Safaitic. I argue that this connection provides the crucial link between Pompeii and the Ḥarrah. Gallica had been stationed in Syria for nearly a hundred years prior to 68, when it was moved, first to Moesia just prior to Nero's death, and then into Italy, at the head of the Danube legions, marching towards Rome on Vespasian's behalf.Footnote ⁴² Once Flavian forces had secured the City, shortly after 20 December, Gallica was billeted in Capua as punishment for the Vitellian sympathies of the Capuan aristocracy during the civil war.Footnote ⁴³ Thus, the Third, previously a mainstay in Syria, found itself in Campania. Its stay was short-lived, and Gallica was sent back to its long-time home sometime in January 70.Footnote ⁴⁴

Gallica would have drawn manpower from its occupation zone in the years prior to its travels to Campania. This manpower could well have included nomads from the Ḥarrah — and, in fact, there is evidence, which I discuss below, that suggests that the Ḥarrah was a recruiting ground for the Roman army. Building on Zangemeister, I argue that it was during this narrow timeframe that the authors of the Safaitic left their marks at Pompeii. Rather than looking for Safaitic-literate merchants, I believe the army was the most probable source of this most improbable epigraphic moment. But how exactly might III Gallica have incorporated the nomads?

III FROM THE ḤARRAH TO POMPEII

There are at least two ways that nomad recruits might have fallen in with the legion. First, and simply, the nomads may have served as legionaries themselves in the Third. Alternatively, the nomads could have been auxiliaries that marched with Gallica. In the final analysis, the evidence is not decisive. But what emerges is a reminder of the flexibility of the Roman army and the diverse ways that it could incorporate local populations. The issue is not that there were no opportunities for nomads to be included in the army; rather, there are too many alternatives to provide a firm conclusion here.

To start with the legionary option: Rome's legions became increasingly provincial under the Empire, from about fifty per cent provincial under Claudius and Nero until, under Trajan, legionaries of provincial birth outnumbered Italians 4:1 or 5:1.Footnote ⁴⁵ In the case of III Gallica, we actually have explicit testimony regarding the effect of provincial recruitment on the unit — and for precisely the time period of interest. About two months prior to the arrival of III Gallica at Pompeii, the men of the Third sowed some productive confusion at the onset of the Second Battle of Bedriacum on 24 October. Tacitus records the episode:

Thus, according to Tacitus and consistent with shifts in legionary demographics, III Gallica contained a significant element that was culturally Syrian.Footnote ⁴⁷ Aside from this passage, we largely lack evidence for the origines of the Tertiani during our time period, but it should be noted that recruitment of legionaries from areas in contact with Safaitic-literate populations is quite well documented later. In the third century, for example, III Cyrenaica drew a significant number of troops from ‘Semitic-speaking populations of Auranitis, Trachonitis, and the area around Bostra’.Footnote ⁴⁸ Ancient Auranitis included the Ḥarrah, and inscriptions that suggest nomad contact with Bostra are also known.Footnote ⁴⁹ Finally, in the light of Seyrig's study of heliolatry in Syria, the religious detail mentioned in Tacitus’ passage above should instill confidence in the historian's report.Footnote ⁵⁰ In sum, the extant evidence seems compatible with the hypothesis that Gallica included local, Syrian recruits among its ranks as it marched into Italy in 69 — and these men could have included nomads from the Ḥarrah.Footnote ⁵¹

The other possibility is that the authors of Pompeii's Safaitic were auxiliaries, who might have fought alongside III Gallica.Footnote ⁵² Though there was no standing policy of moving locally recruited auxilia over long distances, pressing security threats — in our case, conflict on the Danube, then civil war — could spur their relocation, and examples of long-distance redeployment of auxiliaries are indeed known from the civil war of 69.Footnote ⁵³ Further, Tacitus himself mentions auxilia fighting with the Tertiani, though he does not specify a particular, named auxiliary unit.Footnote ⁵⁴

There is also considerable evidence for recruitment of auxiliaries within Syria that makes the nomads good candidates, so to speak, for serving in such units. As Kennedy has noted, ‘with few exceptions the Syrian auxiliary regiments came from that section of the population which lived along the desert's edge and in the more isolated and difficult terrains of the Syrian desert itself and the lava country of the Hauran and Trachonitas’.Footnote ⁵⁵ The evidence from within Syria's ‘lava country’ is even more suggestive. In 2014, Macdonald argued that a number of challenging terms in Safaitic inscriptions point to the incorporation of nomads within the Roman army, serving in a cavalry unit, for example, or patrolling near Roman outposts in the greater Ḥawrān.Footnote ⁵⁶ Macdonald suggested no specific named alae or cohortes, and it should also be emphasised that the texts involved cannot be dated much better than within the first four or so centuries c.e. But he did suggest that such nomads might have been organised in some less formal ‘ethnic unit’ (natio, ἔθνος or ms¹rt in Safaitic) or perhaps even a cohors equitata.Footnote ⁵⁷

A number of Greek inscriptions from the region also relate to this question.Footnote ⁵⁸ These texts mention, for example, a ‘general of units of nomads’ (στρατη[γ]ὸς παρε[μ]βολῶν [ν]ομάδω[ν]), and another names an ‘ethnarch, general of nomads’ (ἐθνάρχου, στρατηγοῦ νομάδων).Footnote ⁵⁹ According to Macdonald, phrases like παρεμβολὴ νομάδων were most likely ‘Roman administrative terms for military units raised from the nomads’.Footnote ⁶⁰ Another text mentions a certain στρατηγός of the Αουιδηνῶν, identified by Sartre with the nomadic tribe ʿAwidh, which is well known from the Ḥarrah.Footnote ⁶¹ According to Sartre, the epigraphic evidence demonstrates ‘sans aucun doute’ that the Romans recruited nomads as ‘troupes auxiliaires’.Footnote ⁶² Finally, a recently published assemblage of graffiti from the Ḥarrah has been said to provide ‘the first concrete evidence for the activities of Roman auxiliary military units raised from the nomadic tribes of the ḥarrah’.Footnote ⁶³ The texts (five Safaitic inscriptions and one Arabic-Greek bilingual) appear to reveal that a certain Ẓʿn son of Kḥ s¹mnʾ served in the military and, importantly, that one of his comrades was gyṣ ḏ ʾl rm, ‘Gaius of the people of Rome’.Footnote ⁶⁴

While the evidence for nomad auxiliaries thus seems to be mounting, it is, admittedly, difficult to assess how securely this can be related to 69 and the Tertiani at Pompeii. Overall, the legionary option would be the simplest explanation, while, for auxilia, Macdonald's suggestions of a national numerus of nomads or some similar irregular equestrian unit, or perhaps a cohors equitata, are as close to definitive as the present evidence permits.Footnote ⁶⁵ In the final analysis, there were multiple pathways for incorporating Rome's subjects into her army. One can hope that further clarity will be forthcoming from the black desert, given that — even 34,000 inscriptions later — ‘the majority of the Ḥarrah remains unexplored’.Footnote ⁶⁶