The transfer of an imperial capital seems a drastic step, but conceptually simple. Indeed, we know about capitals. They are permanent, having been there if not forever, then for many centuries: London, Paris and Lisbon, for instance. Others, such as Washington, DC, and Brasilia, have been created in more recent periods. They could be transferred from one city to another, as in the case of Bonn and Berlin, St Petersburg and Moscow. Or things may be a little more ambiguous: Amsterdam is capital of the Netherlands, but The Hague (‘s Gravenhage) is seat of the government and, most of the time but not always as a matter of principle, royal residence. In the case of Rome, things are even more complicated.

Rome was no state; it was no people; it was not the capital of an Empire. It was a city, but it was also far more than that. Rome was the Empire, and that is what it remained until the fifteenth century, for the Byzantine Empire is a modern appellation. Those we often call ‘the Byzantines’ called themselves citizens of the Roman Empire, and the Byzantine Emperors were called Roman Emperors until the fall of Constantinople in the fifteenth century.Footnote ¹ In its classical age there was hardly a straightforward name for the Roman state, for civitas was a fluid term that could simply indicate any tribe or city, but could also refer to its citizens. The expression that most often is used for the ‘Roman state’ in Latin is nomen Romanum, the Roman name. It indicated all that was Rome: the people, the state, the Empire and its reputation.

Cicero is the first to use this term frequently: ‘For so great is the dignity of this empire, so great is the honour in which the Roman name is held among all nations.’Footnote ² Here it can still be interpreted as merely a term for ‘reputation’. That is no longer the case when Cicero says: ‘Who has such a hatred, one might almost say for the Roman name, as to despise and reject the Medea of Ennius or the Antiope of Pacuvius, and give as his reason that though he enjoys the corresponding plays of Euripides he cannot endure books written in Latin?’Footnote ³ Another clear instance is: ‘Plans have been formed in this state, O judges, for destroying the city, for massacring the citizens, for extinguishing the Roman name.’Footnote ⁴ Here civitas, urbs, cives and nomen Romanum are used in one and the same sentence in four different meanings, all designating essential aspects of Rome.Footnote ⁵ The expression nomen Romanum occurs quite frequently in Augustan literature, notably in the work of Livy, where I count more than twenty instances. It is found again in phrases that may indicate ‘the Roman reputation’ or ‘fame’,Footnote ⁶ but also in the sense of ‘power’: ‘Go, and with the help of the gods, restore the unconquerable Roman name!’Footnote ⁷ It is encountered frequently as well as a term for ‘the Roman people’: ‘But, they added, the immortal gods, taking pity upon the Roman name [i.e. the people], had spared the innocent armies.’Footnote ⁸ It can be used in a more abstract sense for ‘the existence’ or ‘identity’ of Rome: ‘the Volsci, their ancient foes, had armed for the purpose of extinguishing the Roman name.’Footnote ⁹ In the fourth century the Isaurians are described as inhabiting a region ‘in the middle of “the Roman name”’.Footnote ¹⁰ Here, of course, it refers to the Roman Empire in a territorial sense.

All the same it is telling that there was no more concrete term for the Roman state or the Empire. Rome was eternal, of course.Footnote ¹¹ It still is. However, nowadays we think of Rome as a city being eternal. Cicero, when he called Rome eternal, thought of the Roman Empire being eternal, but he never distinguished between city and Empire.Footnote ¹²

And yet, the capital of the Empire was transferred from Rome to Constantinople in the fourth century. That might seem a conceptual impossibility, but it happened. It was part of a major eastward shift of power, economic and military. That, however, is not the subject of this short chapter, which discusses the historical background to the transformation of the city of Byzantium into Constantinople.

When this happened there had been in fact a tradition of about four centuries suggesting that it might occur. There was a rumour to that effect at the time of the rule of Julius Caesar:

The fact that there was such a rumour does not mean, of course, that Caesar actually contemplated it, but it proves that the idea, or rather the fear, existed. It was thinkable to transfer the essence of what was the city of Rome to another city.

The second case arose not long after Caesar’s death: few stories are as familiar as that of Antony and Cleopatra, at least in one seventeenth-century version. There were rumours among the Romans that Antony, if victorious, intended to bestow their city upon Cleopatra and transfer the seat of power to Egypt.Footnote ¹⁴ Here we might notice already that the element of an East–West antagonism is clearly perceived, an antagonism which ultimately, more than four centuries later, led to the division of the Empire into an eastern and a western half. It is remarkable to note that Horace was worried when Augustus financed buildings in Ilion.

Rumours like those in the times of Caesar and Antony were spread as well during the reign of Caligula (AD 38–41):

Again, in the reign of Nero, in 68, there were reports that Nero planned to kill the senators, burn down the city and sail to Alexandria.Footnote ¹⁷ At a less sensational level Tacitus reports what Nero actually did: ‘Eager to make a brilliant name as learned and eloquent, Nero successfully backed Ilium’s application to be exempted from all public burdens, fluently recalling the descent of Rome from Troy and of the Julii from Aeneas and other more or less mythical traditions.’Footnote ¹⁸

In the early third century, after the death of Septimius Severus there is said to have been a plan to divide the Empire into two parts, the West going to Caracalla with Rome as capital; the Propontis would serve as boundary and Byzantium as military frontier city. Geta, the younger son, would obtain the East with Antioch or Alexandria as capital and Chalcedon as frontier city.Footnote ¹⁹

There was then a definite anxiety concerning the central role of Rome as the essence of its Empire attested from the early years of the principate onward. In most cases these rumours focused on rulers who were hated (or slandered) in particular and who died a violent death: Antony, Caligula, Nero, but not, for instance, Augustus or Claudius. At another level, Tacitus’ observation has been quoted often:

For clarification of these long-standing ideas about the place of Rome within the empire, we might recognize traditions about the origins of the city of Rome in the far distant past. The foundation myths of Rome emphasized a Trojan origin, most famously celebrated by Vergil in the reign of Augustus, but already attested in poetry of the third–second centuries BC.Footnote ²¹ Not only Rome but the house of Caesar also claimed to be of Trojan origin.

Well before the reign of Augustus, in the days of the republic, Troy already played a special role as the ancestral home of the Romans. Pyrrhus, who claimed to be a descendant of Achilles, regarded this tradition as an indication that he would defeat Rome, a Trojan colony.Footnote ²² Scipio Africanus visited Troy in 190 BC, emphasizing that the Romans were descendants of the Trojans.Footnote ²³ A century afterward, Sulla honoured the city. Subsequently, Caesar visited it:

Augustus visited Troy,Footnote ²⁵ and afterward Claudius again honoured the city in a manner typical of the astute, but somewhat pedantic historian that he was.

It will be clear from this little survey that there are two kinds of reports: those concerning historical leaders who were regarded as bad news and who are described as intending to abandon Rome and move power eastward. Then there are the more responsible characters who are reported as having merely paid their respect to the old city of Roman origins.

The rumours did not disappear. In the third–fourth centuries the Christian author Lactantius (c. 250–c. 325) predicted that what previously wicked rulers had only planned eventually and inevitably would happen in fact.Footnote ²⁷

We may note, incidentally, that Lactantius was close to Constantine. It will be appropriate here to refer to a text written more than four hundred years earlier, namely Polybius’ History, where he recalls Scipio Aemilianus’ famous tears at the destruction of Carthage.Footnote ²⁸ Scipio is said to have mentioned the fall of earlier empires and then, crying, to have quoted, ‘The day shall be when holy Troy shall fall and Priam, lord of spears, and Priam’s folk.’Footnote ²⁹ When asked ‘what he meant by these words, he did not name Rome distinctly, but was evidently fearing for her, from this sight of the mutability of human affairs’.Footnote ³⁰ Thus we find ruminations about the end of Rome in two authors as far apart as Polybius and Lactantius. Rome may fall one day like Troy in the past, besides other empires – the ones mentioned by Scipio, according to Polybius, are mostly in the East: Assyrians, Medes, Persians, besides Macedonia. The explicit idea that power will return to the Orient did not occur, of course, to Polybius, who wrote before Rome conquered the Near East. The idea of an east–west clash developed after Rome expanded farther east, to Anatolia, Syria and Egypt.

Eventually it happened. In 330 Constantine founded Constantinople.Footnote ³¹ Yet there are reports that the choice of Byzantium for the purpose had been far from certain. There were rumours that the first location had been closer to the traditional location of Roman origin, namely at Troy. We see this in Zosimus (c. 500):

Also, extensively in Sozomen (400–50):

Constantine is reported first to have thought of Serdica (Sofia),Footnote ³⁴ ChalcedonFootnote ³⁵ and perhaps even Thessaloniki. It is probably significant that the two really major cities of the East are not mentioned: Antioch and Alexandria. There is disagreement about the historical truth of these reports, but the least that can be said is that they reflect an idea that had been around for over almost four centuries.

In this connection it is important – as has been observed frequently – that in the third century the emperors frequently resided for long periods in various cities. According to Herodian, referring to the second century, ‘Rome is wherever the Emperor is’ (addressed to Commodus).Footnote ³⁶ Diocletian, Maximian and the other tetrarchs instead travelled between – and resided in – a series of smaller but still sizable cities nearer the frontiers, which were aggrandized with major public building projects such as palaces and hippodromes (often side by side as though they were little Romes), basilicas and baths: these included Trier, Milan, and Aquileia in the west; Spalato, Sirmium, Thessalonica and Serdica in the Balkans; Nicomedia and Antioch in the east. It is therefore conceivable that Constantine at first saw his foundation as belonging to this category.Footnote ³⁷

Why Byzantium was chosen over more important cities is a matter of speculation. Before Constantine made his decision, there was no particular reason to believe Byzantium was destined for this particular greatness and centrality.Footnote ³⁸ It has been called ‘the key to the Pontus’.Footnote ³⁹ It is well known that the site had important strategic advantages, but also problems, described briefly and to the point by Polybius: ‘The Byzantines occupy a site, as regards the sea, more favourable to security and prosperity than any other city in the world, but as regards the land it is in both respects more unfavourable than any other.’Footnote ⁴⁰ He then describes how it controls access to the Black Sea and occupies fertile land. The currents in the Bosporus supplied it with excellent fish.Footnote ⁴¹ However, it was hard to defend on the west side. The site was being surrounded by Thracians with whom the Byzantines were engaged in constant warfare. The prevailing winds made it difficult to approach and supply by sea from the South. It had no natural fresh water supply. A factor not emphasized in the ancient sources, but undoubtedly important as well and relevant for Constantine’s choice, was its controlling position on one of the two land bridges between Europe and Asia Minor.Footnote ⁴² The history of the city before the fourth century is a subject of great interest but less immediately relevant for the present discussion.Footnote ⁴³

Recent studies have suggested modifications in the traditional interpretation of claims, made first by Christian authors, that Constantine’s city was conceived from the start as the new or second Rome and also as a purely Christian city.Footnote ⁴⁴ Regarding its Christian character, it should be noted that Byzantium before Constantine had no distinguished history as a Christian city.Footnote ⁴⁵ It is not even mentioned in the New Testament. These later Christian authors offer their own perspective which ignores the fact that there were both secular and non-Christian sacred buildings in the city.Footnote ⁴⁶

The claim that, from the start, it was intended to replace Rome is not conclusively confirmed, given the relatively slow development of Constantinople as imperial capital. There is no contemporary evidence that Constantine always conceived his new foundation as the eastern capital of the empire, or that he intended that it should replace Rome.Footnote ⁴⁷ It has been argued that Constantine may not have had a fully formed plan already in 324.Footnote ⁴⁸ True, there are claims that Constantine decided in 324 to found it as ‘the new Rome’Footnote ⁴⁹ or ‘the second Rome’,Footnote ⁵⁰ but these sources are of a later date. In fact, the name that stuck was Constantinople.Footnote ⁵¹ The city actually took time to develop into an imperial capital.Footnote ⁵² A praefectus urbi as existed in Rome was not appointed until 359.Footnote ⁵³ The city became the undisputed centre of the Late Roman Empire only in the reign of Theodosius I. This has been demonstrated by an analysis of the subscriptiones of imperial constitutions, preserved in the Codex Theodosianus, showing that over 240 laws had been issued at that time, out of a total of more than 600 included in this compilation, which were enacted from Constantinople.Footnote ⁵⁴ Half a century after its foundation Constantinople formally became ‘the New Rome’,Footnote ⁵⁵ even though Rome, in the fourth century and during the first half of the fifth, retains a privileged position in the literary documents of the time that mention it, especially those produced by pagan, Latin-writing authors.Footnote ⁵⁶ Eutropius is a relatively early historian who mentions the ascent of Constantinople:

That still is a somewhat ambiguous, rhetorical assertion. Eutropius had a vested interest, for he was magister memoriae in Constantinople in the second half of the fourth century.

Conclusions

Reality trumped ideology again and again. Old fears became reality.

Rome was not a capital city. It was not supposed to be one. It was the state, the Empire, the collective citizenship, all in one. In theory it could not be duplicated or transferred. Paradoxically, there was fear and there were rumours for centuries that precisely this would happen, even at a stage when there was no such danger. Eventually it happened indeed, and when it did – another paradox – this may not have been the intention from the start. Before Constantinople was founded, there had been fears for more than three centuries that a new capital was to be located at or near the mythic cradle of Rome: Troy. These fears were combined with a pronounced view of an East–West antagonism which, somehow, was translated over time into the actual split of the Empire into an eastern and a western part, a Latin-speaking and a Greek-speaking half. Not so long before this took place, the new capital was planted on a suitable site without Roman historical significance, replacing an old Greek city.

Although the sources have been interpreted along different lines, it now seems very likely that Constantine did not declare from the start that he was founding a ‘New Rome’ or ‘Second Rome’, a city to replace Rome. Also, not having declared that he was in fact doing so, it did not become a reality in his days, but only under his successors. Like so many foundations, at first it was just another city on an important site, although an old and venerable one, renamed after the man who refounded it. Yet Constantine’s decision had drastic results: it initiated the transfer of the centre of the Empire eastward, to the Greek-speaking part of the Roman Empire, even though it took more than a generation before this became a reality, whatever had been intended. It is conceivable, although a matter of speculation, that Constantine, in selecting Byzantium for his new city of Constantinople, consciously avoided the antagonism that would have been the result if he had truthfully declared that he was establishing a New Rome or an alternative Eastern Rome on or near the site of Roman origins. Christians later claimed he had founded a Christian city – that was only partially true. Again, it is possible that Constantine avoided fierce conflict by refraining from establishing a fully Christian, second Rome.

The fear that Rome would come to an end never materialized (Rome is still there),Footnote ⁵⁸ but the shift of the centre of power eastward took place, not as the result of a one-time decision by a ruler, but through historical dynamics and preceded by centuries of suspicion and tension. The decision rather reflected a development that occurred, inevitably, over time. Emperors who called themselves Roman continued to reign in the East over subjects who called themselves Romans, although Greek was their language, until the fifteenth century, a thousand years longer than in the Western Roman Empire, where the last Emperor, Romulus Augustu(lu)s, abdicated in 476.Footnote ⁵⁹

Philip V of Macedon already recognized one of the decisive political strategies through which Rome strengthened the basis of its rule.Footnote ¹ In a letter to the inhabitants of the city of Larissa he talks of the fact that the Romans have recognized that there is an advantage in not being possessive about their citizen rights; rather, they try to expand the number of citizens by accepting foreigners, in this particular case freed slaves, and thereby to increase Rome’s power. This characteristic feature of Roman politics is referred to also by Dionysius of Halicarnassus, according to whom this practice was implemented in Rome as early as the days of King Servius Tullus.Footnote ² Philip V was probably too idealistic in his view of Rome’s attitude, but he recognized one of the basic principles of Roman politics. In his eyes, Rome did not see membership in the res publica populi Romani as exclusive; rather, Rome saw the necessity of accepting other people from the outside into the Roman citizen body – of course under certain conditions and for the advantage of the res publica.

This basic axiom can be traced in different ways in Rome’s history, starting in the days of the Republic; it becomes more prominent under Caesar, and even more so from Augustus on, when a single individual could make the essential political decisions in Rome.Footnote ³ The goal of this political and legal openness was primarily to create a strong commitment to Rome and its ruler; the policy was intended to strengthen the loyalty of people to Rome, or to create integration, as we would probably say today.

Not all members of the political ruling class saw this granting of Roman citizen rights as appropriate and useful for Rome. Moreover, the effect was not always what those who granted the civitas had hoped for. What took place in AD 9 in Germania east of the Rhine, namely the clash with the Cherusci and the annihilation of the Roman army under the command of Varus, was not normal. Arminius, the chieftain of the Cherusci who led the Germanic uprising, had not only received Roman citizenship from Augustus but even had become a member of the equester ordo – nevertheless, he broke his loyalty and friendship with Rome.Footnote ⁴ But in most cases the effect of granting the civitas Romana was what one could expect, at least, so far as can be known from our sources. The integration was strengthened, and the affiliation to Rome and its fundamental interests became more or less natural, for the majority of the new citizens and even more so for their descendants. When the soldiers of the Danube armies saved the Empire during the severe crisis of the second half of the third century AD, it was partly a consequence of the continuous granting of civitas Romana to hundreds of thousands of soldiers and their descendants, mainly from the Balkan provinces. We can see this process from the time of Claudius to the beginning of the third century through numerous military diplomas.Footnote ⁵ Long before the Constitutio Antoniniana, the extensive granting of Roman citizenship to the auxiliary soldiers, mostly born in the Danube provinces, created the decisive requirements for a profound integration of an essential part of the local societies. The affiliation to Rome via Roman citizenship led by and large to stability in the provinces and ultimately contributed to the fact that at the end of the third century, after the long crisis, the former stability was temporarily restored. The positive effect of this integration through Roman citizenship is thus evident.

The integration of the formerly conquered peoples, however, did not occur only through service in the Roman army. Perhaps more important was the integration of the ruling classes of cities and tribes. Participation in public matters in these societies, being usually timocratic and hierarchical, was open in general only to the leaders and not to all members to the same degree. But via the hierarchical structures the majority of the lower population was also integrated even if they could not gain independent political stature. Many ruling families of the Roman commonwealth obtained the Roman citizenship at an early stage, through viritane grants, through procedural actions like the granting of the ius Latii in some provinces of the West, or by establishing municipia, in which the previous socioeconomic leaders of single cities were incorporated into the Roman citizen body. Even in the new Roman colonies, local principes were often integrated into the new citizen body.Footnote ⁶

The road to integration, however, did not end there; rather, it led in different ways from the provincial cities to Rome, via families whose members became part of the two empire-wide groups, the equester ordo and finally the ordo senatorius. This process had already begun before Caesar; the first nonethnic Roman who became consul was L. Cornelius Balbus, who received Roman citizenship from Pompeius after the war against Sertorius and advanced to a consulship in 40 BC.Footnote ⁷ That was still unusual at that time and was also an effect of a time of massive changes; however, it marked very clearly what could be achieved through the decision of the Roman political elite and the ambition of the people, who until that point were excluded from the inner core of Roman society. But even after the turmoil of the civil wars, in which the loyalty to a powerful figure was the decisive criterion by which one could be accepted into the political elite, the admission of former peregrini to the ordo senatorius or the equester ordo did not end.Footnote ⁸ During his more than forty years of rule, Augustus developed a policy that was decisive also for his successors.Footnote ⁹ Claudius made it very clear in his speech in AD 48 regarding the request of the primores Galliarum to be admitted to the Senate with the following famous words:Footnote ¹⁰ sane novo m[ore] et divus Aug[ustus av]onc[ulus] meus et patruus Ti[berius] Caesar omnem florem ubique coloniarum ac municipiorum bonorum scilicet virorum et locupletium in hac curia esse voluit. This statement did not refer only to the coloniae and the municipia in Italy but also to Romans already living in coloniae and municipia in the surrounding provinces. Hence, Claudius continued by saying: quid ergo non Italicus senator provinciali potior est? iam vobis, cum hanc partem censurae meae adprobare coepero, quid de ea re sentiam, rebus ostendam, sed ne provinciales quidem, si modo ornare curiam poterint, reiciendos puto. The rhetorical question that Claudius addressed to himself shows that not all members of the Roman ruling class were in favor of expanding their sociopolitical group with people from the provinces; rather, they opposed the intention of the emperor more or less openly. However, Claudius’ decision, which he secured by a senatus consultum, overruled this resistance.

Claudius’ fundamental approach was a constant feature of all later emperors. And naturally, to the criterion si modo ornare curiam poterint a second criterion was added: loyalty toward the ruling emperor. An example is the conduct of Vespasian, who after Nero and the civil wars was forced to restore the decimated Senate to its former size.Footnote ¹¹ He included especially people who joined his side during the civil wars of 69/70. Among the senators who entered the Senate for the first time under his rule, one can find C. Caristanius Fronto from Antioch in Pisidia, consul suffectus in the year 90;Footnote ¹² Ti. Iulius Celsus Polemaeanus from Sardis/Ephesus, suffectus in the year 92; and C. Antius A. Iulius Quadratus from Pergamon, suffectus in 94.Footnote ¹³ Catilius Longus came from Apameia in Bithynia; he had served as prefect of an auxiliary unit in Iudaea under Vespasian, just like Caristanius Fronto.Footnote ¹⁴ From the Iberian Peninsula came L. Baebius Avitus, L. Antistius Rusticus and Q. Pomponius Rufus; the last two were consuls under Domitian just like Polemaeanus and Iulius Quadratus.Footnote ¹⁵ The homeland of Iavolenus Priscus was probably the province of Dalmatia,Footnote ¹⁶ and the two brothers Pactumeius Fronto and Pactumeius Clemens were born in Cirta in Africa.Footnote ¹⁷

This short list shows that already from the Flavian period the Senate was occupied by individuals from quite a few provinces. This trend intensified under the subsequent emperors, so that at the end of the second and the beginning of the third centuries, the senators whose original homeland was in the provinces already formed a clear majority in the Senate. The second volume of the congress “Epigrafia e Ordine Senatorio” (EOS) from 1981, published in 1984, presents all the information regarding the background of the members of the Senate known in that year.Footnote ¹⁸ That information is still valid and nearly complete, as is evident from the two new volumes of the “Epigrafia e Ordine Senatorio 30 anni dopo,” published in 2014.Footnote ¹⁹ Therefore it is possible to use the results of 1984 in analyzing the background of the senatorial families from the first to the third centuries.

The proportion of senators of Italian origin was never negligible during the imperial period, but that number declined slowly at first, during the first half of the first century, and then more sharply from the beginning of the second century. From the articles in EOS we know that, apart from the Italian senators, the members of the Senate came from thirty provinces. Only a few provinces are missing in this list (i.e., provinces that did not provide any senators at all) – at least as far as we now know; new inscriptions may change this picture in the future. To this group of missing provinces belong Britannia and Germania Inferior, from which we know of only one senator who came possibly from either of the two provinces.Footnote ²⁰ The three small provinces of the Alps did not have representatives in the Senate before the later third century or perhaps only from the beginning of the fourth: The two Moesian provinces were not represented in the Senate even in the later period, unless one wishes to take into account some of the so-called soldier emperors, who have little to do with the normal process of entering the Senate.Footnote ²¹ The island of Cyprus records no senator, nor does Sardinia.Footnote ²² The same is true regarding Iudaea/Syria Palaestina, although since 1984 more members of the equester ordo from there have become known.Footnote ²³ At least one senator, M. Valerius Maximianus, came from Pannonia; he was born in Poetovio and was appointed consul suffectus around 186.Footnote ²⁴ From Egypt, namely from Alexandria, we know of two senators, a P. Aelius Coeranus and his homonymous son, both of whom were consuls after the death of Septimius Severus.Footnote ²⁵ Altogether, the senators from the end of the first century BC to the third century AD who can be identified with some certainty as provincial came from cities located in the following provinces:Footnote ²⁶

The list shows that the majority of provinces of the Imperium Romanum sent members to the central governing organ of the empire. But the implementation was not equal across these provinces. The number of provincial cities from which we can identify senators with reasonable certainty is just over 200.Footnote ²⁷ This is only a fraction of all the autonomous cities in the provinces. The number of cities is divided very differently among the separate provinces. Not in all provinces, however, are the numbers of senators known to us always representative. Only four senators altogether are known from the sixty-four communities of the tres Galliae with their large territories. This is a very minimal representation, whereas Belgica, according to our current knowledge, is completely missing. This low number, however, probably does not reflect reality but can be explained by the epigraphic documentation, or paucity thereof, in the province of Aquitania, the Lugdunensis and the Belgica. The number of senators and their home cities from these provinces may have been in fact higher, especially since Gaul provided the requisite economic base for senators.Footnote ²⁸

The number of cities in the two Hispaniae, the Tarraconensis and the Baetica, is quite high, but the many known senators from these provinces come from a very limited number of communities: Eleven cities from the Baetica and fourteen from the Tarraconensis sent senators to Rome according to our current knowledge. The highest number of cities represented in the Roman Senate belong to the provinces that also occupy the top places in a senatorial career: Africa (proconsularis) and Asia. About thirty communities from Asia and at least forty-seven communities from Africa sent senators to Rome.Footnote ²⁹ A few cities are notable for the number of senators coming from them. In Asia, Pergamon and Ephesus stand out in this regard;Footnote ³⁰ Pergamon also had its first consul, C. Antius A. Iulius Quadratus, by the end of the first century and hence it is represented among the leading groups of the Senate.Footnote ³¹ Surprisingly, the first consul from Ephesus took office around the end of the second century;Footnote ³² Ti. Iulius Celsus Polemaeanus, who is almost automatically associated with Ephesus because he was buried there, in the basement of his library – hence the name Celsus-Library – was originally from Sardis. But the number of senators from the capital of Asia is the highest we know from all cities in the provinces.Footnote ³³ In Africa there are also several cities that stand out in this respect: Bulla Regia, Lepcis Magna, Hippo Regius.Footnote ³⁴ However, it appears that until the end of the third century Carthage, the caput provinciae, is missing among these cities known for their high number of senatorial families – notwithstanding that Carthage had many aristocratic families. That seems to have changed in the fourth or even the fifth century, when we know many viri clarissimi from that city, attested by newly found inscriptions from Carthage.Footnote ³⁵ They got senatorial rank and title through their occupation in the officia of the imperial administration, but they most probably had nothing to do anymore with the Senate in Rome.

Not a few families were represented in the Senate by members over the course of several generations. We know of the first senator from the familia Silia from Lepcis Magna as early as the beginning of the rule of Marcus Aurelius, and the last known member of this family was the governor of Germania Superior in 240.Footnote ³⁶ The earliest consul from the Cuspii of Pergamon was appointed in AD 126 under Hadrian;Footnote ³⁷ therefore, their senatorial status goes back at least to the time of Trajan. The last known member of this family was one of the consules ordinarii of the year 197 during the Severan period.Footnote ³⁸ Among other families, which however came from Italian cities, the membership in the Senate was even much longer. The Neratii from Saepinum had a seat in the Senate from the second half of the first century, and they were represented there at least until the end of the fourth;Footnote ³⁹ the same is true for the Bruttii Praesentes, who came from the region of Brutti in the south of Italy.Footnote ⁴⁰

From the middle of the first century AD, the Senate, while it met in Rome, appears to have turned more and more into a mixed assembly, when one considers the provinces of the empire from which the individual members and their families came. From a legal point of view and with respect to origo, there was no difference among the individual senators, since all of them had the same origo: Rome itself. Ideologically, it could not have been otherwise: whoever represents Rome must be a Roman, in all respects. The fact that Rome was the origo of all the senators is stated very clearly in the liber primus of the Sententiae of Paulus:Footnote ⁴¹ Senatores et eorum filii filiaeque quoquo tempore nati nataeve, itemque nepotes, pronepotes et proneptes ex filio origini eximuntur, licet municipalem retineant dignitatem. Senators lost, so to speak, their old origo and were associated with it at the most so far as they kept the dignitas, the reputation and the rank that they had acquired in their homeland. Other sources confirm this legal bond to Rome.Footnote ⁴² The members of the ordo senatorius were not connected anymore to their original community. Moreover, a senator and his relatives did not have any legal duties regarding their place of origin, and no one could oblige them to munera there. Rome was to be their base.

Beyond the legal aspects, the separation of the senators from their homeland was a matter of course also for practical reasons:Footnote ⁴³ They had their duties in Rome itself; above all, they had to take part in the meetings of the Senate and in other gatherings, for example the meetings of the fratres Arvales, which are well attested.Footnote ⁴⁴ This separation from the homeland was reinforced by official tasks required of officeholders. These official tasks were carried out partly in Rome, but mostly outside the capital in different provinces and sometimes in Italy. The emperor sent a senator to his home province very seldom, for a special reason or from carelessness.

If senators wanted to visit their home province,Footnote ⁴⁵ apart from Sicily and the Narbonensis, then they needed specific permission, which was issued at first by the Senate but soon by the emperor.Footnote ⁴⁶ This rule of requesting leave (commeatus) did not apply for Italy; only senators who were bound to Rome due to a magistracy had to ask. When Pliny the Younger was praefectus aerarii Saturni, he asked Trajan for a month’s leave so that he could settle some matters of one of his estates near Tifernum Tiberinum, located more than 150 miles from Rome.Footnote ⁴⁷ For most of the senators, however, regular personal contacts with their home cities in the provinces were not so simple, even under the relatively good travel conditions of the early and high empire. We can assume that many senators who accomplished many discernible things in their home city did so through letters.Footnote ⁴⁸ It is enough to consider Cuspius Rufus, a senator from Pergamon in the period of Pius, whose ties to his family’s home city were extremely close.Footnote ⁴⁹ Nevertheless, when we consider all the factors and circumstances of senatorial life, it is not surprising that many scholars have thought that the contact of most of the provincial senators with their home cities was broken off quite abruptly. This seems to be confirmed by the fact that senators were relatively seldom magistrates in their hometowns, despite the fact that it was still legally possible for them, as Hermogenianus stated in his liber primus iuris epitomarum written at the end of the third century.Footnote ⁵⁰ In practice, of course, this was almost impossible to do in persona, since no senator could have been absent from Rome for a whole year for nonofficial reasons. But it would have been possible for a senator to take care of the financial duties of a local office while the concrete functions were taken care of by a local praefectus.

On the other hand, we know of not a few examples of senators who time and again – at least when they reached the age of 65 and were no longer obliged to take part in Senate meetings and normally had no tasks as magistrates – retired to their home communities or at least made sure that they were present there.Footnote ⁵¹ When the city of Carthage wanted to honor Minicius Natalis Quadronius Verus, cos. suff. in AD 139, with a quadriga, after his proconsulate in Africa, he decided that the monument would be set up not in Rome but in Barcino in Tarraconensis, the city where the Natalis family originated. The quadriga with the senator’s statue stood in the center of the baths, which had been built by the senator and his father around the year 123.Footnote ⁵² Such an act is understandable only if the aging senator himself had retired to Barcino, or at least felt himself emotionally connected with the city.Footnote ⁵³ Similar is the case of C. Antius A. Iulius Quadratus, who saw to it that a great number of honorary statues would be set up for him in his home city, Pergamon. He seems actually to have returned to Pergamon personally as well.Footnote ⁵⁴ On the other hand, we can see that the home cities of senators in the provinces were rarely selected for honorary statues, which were erected by provincial communities for their former governors. Ti. Claudius Candidus, a legate of Pannonia Superior in the Severan period, was honored in his hometown Rusicade in Africa by a prefect of a fleet who served under his command at the middle Danube.Footnote ⁵⁵ Likewise P. Iulius Geminius Marcianus, legate of the province of Arabia under Marcus Aurelius, saw to it that some cities of his administrative district would erect statues in his honor in his hometown Cirta, even with inscriptions in Greek.Footnote ⁵⁶ Apart from the cities of Italy, such honorary statues in the provincial home cities of governors are only sporadically attested.Footnote ⁵⁷ This could indicate a weak relationship of many senators to their natural origo.

Less frequent are the cases in which senators were buried “at home,” like Celsus Polemaeanus, whose burial place in the basement of his library in Ephesus has already been mentioned.Footnote ⁵⁸ C. Iulius Quadratus Bassus, consul suffectus in 105, who died in Dacia at the beginning of Hadrian’s rule during a battle against the empire’s enemies, was buried in Pergamon, in his home province Asia, by an order of Hadrian.Footnote ⁵⁹ Herodes Atticus lived during the last years of his life in Marathon on the east coast of Attica because of tensions between him and Marcus Aurelius. He found his last resting place in Attica.Footnote ⁶⁰ The majority of senatorial graves, without regard for where the senators came from, have been found near Rome or cities in the vicinity,Footnote ⁶¹ in other words where normally the nonofficial life of the senators and their families took place.

All these observations seem to show, as indicated earlier, that the former home communities of the Senate members did not play an essential role in their lives and that the contacts with them were reduced or even broken off. Through this interpretation of our relevant material one decisive aspect of senatorial existence is not being taken into account: the economic basis of all these families. Each senator had to prove that he had a minimum fortune of one million sestertii, although many, if not the majority, had much more than that.Footnote ⁶² Otherwise many of them would have been classified as pauper senator. Plentiful evidence attests to far greater fortunes and respective incomes, not only for Seneca but also for many other people like Q. Vibius Crispus.Footnote ⁶³ The larger part of these fortunes consisted, as was general in Roman society, of landed property, which produced income directly through farming or by other forms of land use, from which the Roman families could make a living.Footnote ⁶⁴ We know that in the time of Trajan, each new senator had to invest a third of his fortune in Italian landed property, because – as Pliny writes in one of his letters – in this way it could be guaranteed that the senators would not look at Italy merely as hospitium aut stabulum.Footnote ⁶⁵ Under Marcus Aurelius this Italian requirement was evidently lowered to one quarter.Footnote ⁶⁶ This means two things: On the one hand, the relationship to Rome – at least among the homines novi – was still weak, while the connection to the home communities was stronger. On the other hand, this indicates more than anything else that the larger part of the productive land, despite the partial transfer to Italy, still remained in the provinces. If the majority of the landed property of the senatorial families stayed in their former homelands, then the members of these families were in one way or the other necessarily present there, via administrators (procurators) or maybe also via close relatives, who represented the interests of the senator at home. Moreover, the revenues from the lands must have gone to Rome, in kind or as cash value. Such close connections to the home cities are illustrated, for example, by the letter of Septimius Severus from AD 204, by which he conveyed to an unknown recipient – perhaps a proconsul from Asia – that a senator populi Romani according to a senatus consultum should not tolerate any billeting in his house against his will. Nine copies of this letter on marble slabs have been found in the territory of the provincia Asia; senators probably fixed the text publicly on their properties, in order to protect their houses in the cities of the province.Footnote ⁶⁷ The same is evident from a decree of Valerian and Gallienus,Footnote ⁶⁸ who verify in writing to a senator named Iulius Apellas that he should not tolerate any billeting in his house, which was probably in Smyrna. All the senators who posted such imperial documents not only had property in their home cities but also were carefully trying to protect the value of their property. These documents signify that the provincial senators – just like their colleagues from Italy – had to maintain economic contact with their places of origin. This contact could consist not only in impersonal connections involving merely the “transfer” of the revenues to Rome but also perforce in a real relationship with the citizenry of the communities, who were directly involved in the senators’ transactions; as a result, in turn, the senators retained knowledge of the situation in their cities, quite the opposite to the prevailing opinion about the relations of provincial senators to their places of origin.

In addition, many senators were also the patrons of their home cities, to whom the inhabitants, especially from the ruling classes, could turn and seek consultation.Footnote ⁶⁹ Cornelius Fronto, a senator from Cirta in Africa and a rhetoric teacher of Marcus Aurelius, is a clear example of continuing commitment to one’s home city,Footnote ⁷⁰ just as Pliny the Younger is with respect to his patria Comum. We can therefore assume that many senators, if not all, were kept informed about the situation of their home city and hence also about the situation of the province of their city.Footnote ⁷¹ It is surely no accident that Pactumeius Fronto is called consul ex Africa primus on an inscription of his daughter from Cirta.Footnote ⁷² The Superaequani said of their fellow citizen Q. Varius Geminus: primus omnium Paelign(orum) senator factus est.Footnote ⁷³ A centurio of the legio III Augusta honors his commander Ti. Claudius Gordianus and points to the fact that he, the commander, came from Tyana ex Cappadocia.Footnote ⁷⁴ Other texts also mention occasionally the origins of senators.Footnote ⁷⁵ The connection between a senator and his provincial city or an entire province was seen therefore as something absolutely real and accepted.

When we look at the origins of individual senators, it is undeniable that the Senate was the center of a multinational empire. Multinational is, of course, understood here not in the modern sense of nations but in the sense of different patriae, with which specific citizen rights were connected, which stood originally on the same level as Roman citizen rights. One can assume that the members of the Senate were aware that they came from different regions of the empire and that their cultural background was not the same; but most of them did not see these differences as multinational, namely as reflecting different patriae, but rather in a much simpler way: senators from Italy versus senators from the provinces. This is at least what the rhetorical question in Claudius’ speech assumes: non Italicus senator provinciali potior est.Footnote ⁷⁶ It seems that by this expression Claudius meant the notion of a majority in the Senate, with which of course he did not want to agree. In any case, it is hard to doubt the fundamental awareness of their diverse origins among most members of the Senate. These observations raise a further question: Did the knowledge and the consciousness of the different geographical and hence diverse cultural backgrounds have implications, either in the collective decisions of the Senate itself or in the decisions of individual senators? The question is legitimate and interesting, but answering it will require a wide-ranging investigation, which will be carried out in the future.Footnote ⁷⁷

The Nature of Idioms and Proverbial Expressions

This study is a search for cultural insights through common linguistic structures. More specifically, it is an attempt at gleaning some knowledge of ancient Classical views of foreigners from Latin idioms and proverbial expressions. Let us simply start with the basic definition of idioms as words or phrases that have a figurative meaning which is different from the literal meaning of the individual word(s) composing them. The characteristics of idioms, their social and linguistic functions, and their research are closely related and partially similar to those of proverbs.Footnote ¹ Like idioms, proverbs, in any culture, contain generalizations or approximate truths originating in real-life experience. Often they turn actual and factual realities into exaggerated and inaccurate assertions. In all their aspects, proverbs, proverbial expressions and single-word idioms, usually emerge from popular experience and are based on impressions of ordinary people. These become maxims reflecting common, sometimes prejudiced, concepts of situations, people, places and other details related to human encounters with the world. The humble origin of idioms and proverbs does not always mean that they do not hold facts and verifiable truths or that they cannot express educated ideas, but rather that they are not the result of scientific and informed observations. While their origin and initial transmission are both oral, they often penetrate learned and literary texts so that, specifically for ancient societies, we have some record of this popular and oral world of ideas preserved throughout generations.

The Greek paroimia, gnome or apophthegma and the Latin proverbium, sententia, elogium or dictum, were usually prose or metric phrases, concise, witty, and sometimes enigmatic or allegorical. They expressed experience and common sense, universal truths and popular wisdom. At times they held even moral and didactic values. In our present study we broaden the scope to include also more flexible proverbial expressions and single-word idioms. As we shall see, unlike strict proverbs in their traditional sense, not all of the idioms and phrases we discuss have moral implications as lessons to be learned or implicit advice or guidance, but all carry ethnic connotation and meaning.

Because of their popular origin, proverbs and idioms regularly reflect ancient times and preserve traces of past events and periods. Many sayings refer to specific regions and peoples. Accordingly, place names or mere ethnonyms, even when detached from a whole sentence as single-word idioms, become proverbial for particular situations, human characteristics and natural conditions. Furthermore, unique events and individual character traits are generally applied and often exaggerated to denote collective local and personal types.

The popularity of idioms and proverbs and their antiquity turn them into a first-rate source of interest for historians and anthropologists. They preserve various facts sometimes through generations after the original circumstances have long changed or after a certain phenomenon has disappeared. They are remains of the past undamaged by time. Their historical value is especially precious because they contain unbiased information and not such that was consciously and deliberately inserted for tendentious historiographical purposes. In this sense they are pieces of ‘oral archaeology’Footnote ² in the same sense as material archaeology is straightforward and unequivocal. Still, being an oral testimony, these linguistic phrases may have gone through changes and adjustments in the course of their transmission. But, unlike lengthy and detailed pieces of originally oral evidence, such as tales, fables and poems, which are more flexible and more prone to changes, it seems that proverbs and idioms are more stable. This has to do with their encapsulated and dense nature. As we shall see later, sometimes we can tell the approximate age of idioms through the dates of the written sources that quote them. Based on theories of the social and linguistic nature of proverbs and idioms, it is assumed that such phrases penetrate literary sources when they are already prevalent for long. Accordingly, they are generally older than the definite date of the quoting source we possess. If, then, we are fortunate enough to have evidence for another, later, use of the same proverbial expression, then we can be rest assured both of its old age and of its consistent durability over the ages.

Finally, the value of proverbs and idioms increases because they are one of the rare avenues that enable an access to the knowledge and concepts of popular, uneducated and illiterate sectors of Greek and Roman societies. Both their simple origin and their oral transmission emphasize their significant role within the public domain, unlike many written sources, which were kept within the narrower sectors of the educated elite.

Latin Ethnic Idioms

The following discussion intends to offer a brief study of mainly Latin ethnic idioms, meaning bywords and phrases that apply collective names of ethnic groups and associate them with fixed attributes. In analyzing the ethnic aspect of these idioms it is not always possible to separate places from their inhabitants and vice versa, because certain geographical regions prescribe in the popular mind certain human traits. In the present study this is apparent, for instance, with regard to Abdera, its air and its effect on the inhabitants’ stupidity, or to Campania, its riches and the resulting arrogance of its residents (see following discussion). Despite this occasional overlap between ‘geographical’ and ‘ethnic’ proverbs, I have tried to focus mainly on proverbs that apply strictly to ethnic denominations.

Such proverbial phrases in their specific ethnological aspect usually derive from an initial encounter with foreigners that frequently becomes exaggerated and distorted through oral transmission and rumour. One may easily imagine how someone – a merchant, a soldier, an administrator – visited a certain region where he met local inhabitants or where he met people from other places; a first impression was made by physical looks or on the basis of an act or unique behaviour; this impression was shaped – immediately or eventually – into a generalization pertaining to all local inhabitants or to all people belonging to the same ethnos or region. In this way, proverbs and idioms repeatedly transmitted stereotypic concepts – not necessarily bad ones – in the perhaps unfair belief that all people with a particular characteristic or ethnic origin were the same.

To begin this study I have used the modern collection of A. Otto, Die Sprichwörter und sprichwörtlichen Redensarten der Römer.Footnote ³ This collection is arranged alphabetically and includes some references to the main occurrences of each phrase and a brief explanation. On the basis of this valuable selection, and with frequent consultation of the CPG (Corpus Paroemiographorum Graecorum),Footnote ⁴ I have expanded the discussion by including parallel issues and by directing the inquiry toward the topic of ethnographic beliefs and prejudices. The subsequent survey demonstrates the theme by dividing the extant Latin proverbials into groups according to types of ethnic traits displayed in them.

Physical Traits

The primary encounter with foreign and unknown nations is clearly and always made through sight. Even if one does not talk to, or trade with, or fight, or approach, other people, a visual impression is made. Accordingly, we find several proverbial expressions related to physical appearance. In Plautus’ Poenulus (‘the little Punic’) Antamonides, a soldier in love with one of two Carthaginian girls, exclaims:

Egyptians thus are presented as a standard for blackness, even if the image is based not on an actual visit to Egypt but on the appearance of Egyptians who were brought to Rome and performed or worked in the circus. Perhaps these implied circumstances emphasized even more the physical difference between locals (Roman city dwellers who attended the theatre) and foreigners (Egyptian slaves). But Egyptians were not the usual symbol of dark complexion. Based on what we have available in writing, other North Africans were more commonly used as proverbial illustrations of black or dark skin.

In the so-called Priapic erotic epigrams, a certain very repulsive girl is said to be ‘no whiter than a Moor’ (non candidior puella Mauro) (46.1). In another Priapic epigram the Moors represent elaborately curly hair when mocking a feminine male who ‘primp[s] his hair with curly irons so he’d seem a Moorish maiden’ (ferventi caput ustulare ferro, ut Maurae similis foret puellae) (45.2–3).Footnote ⁶ The Latin MauriFootnote ⁷ sometimes referred specifically to the inhabitants of the region defined in ancient geographies as Mauritania, or Maurousia in Greek, which is more or less parallel to parts of modern Morocco and Algeria.Footnote ⁸ However, we often find the same terminology applied, especially in poetic works, to Africans in general.Footnote ⁹ Accordingly, the proverbial association of Mauri with dark skin could be understood as pertaining to the inhabitants of north-western Africa or to the inhabitants of the continent as a whole. It seems that even if the crowds had no precise geographical idea of peoples and places, the popular notion of certain groups who have black skin must have been established and transmitted.

The Latin references to Egyptians and Mauri as people with a darker complexion combine to form the traditional and most well-known use of Aethiops as the symbol of black skin already in Greek proverbial applications. The very etymology of the Greek word Αἰθίοψ, denoting a ‘burnt face’ (αἴθω, ὄψ), as well as the Greek idiom ‘to wash an Aethiops white,’Footnote ¹⁰ must have fixed this image in the minds of the crowds, even those who had never met any person from the relevant African regions. This is quite clear, for instance, in Juvenal’s contrast between ‘white’ and ‘Aethiops’ (derideat Aethiopem albus, Juv. 2.23).

Another unique physical trait was proverbially associated with the people of the island of Myconos. According to this popular notion, all people on the island of Myconos were bald. Strabo commented that ‘some call bald men Myconians (Μυκόνιοι), from the fact that baldness is prevalent in the island’ (10.5.9 and cf. Plin. HN 11.130). Lucilius through Donatus on verse 440 in Terence’s Hecyra, alluding to a person from Myconos, also says: ‘all young men in Myconos are bald’ (Myconi calva omnis iuventus).Footnote ¹¹ This notion, or image, perhaps explains another early Greek proverb related to Myconos. When Plutarch discussed the sitting order in a symposium, he commented that it is not rational to ‘make no difference in their seats, at the first dash making the whole company one Myconos (μία Μύκονος) as they say‘ (Quaest. Conv., Mor. 2.616b), meaning, in this context, treating all as one, all alike. Why and how did Myconos of all places gain such an attention in Greek and Latin proverbs? There seems to be no historical reason, and we must leave it at that.

Character Traits

Beside ethnic proverbials for physical appearances, the largest group of Greek and Latin ethnic idioms clearly and perhaps unsurprisingly relates to character traits, and mostly unfavourable ones. So enter the typically stupid.

First, and already in earlier Greek tradition,Footnote ¹² is Abdera in Thrace, which became typically proverbial for its foolish inhabitants. In several Latin contexts there is not even a need for explanation, and the mere locality indicates its prejudiced reputation; for instance, when Cicero says, Hic Abdera non tacente me (Att. 4.16.6), while reporting on some commotion in the senate when Cicero could not hold his tongue in front of what he considered sheer stupidity. Or when he says, using the Greek adjective, id est Ἀβδηριτικόν (Att. 7.7.4), commenting on some silly intentions of Pompey and his advisors. Similarly, Martial bluntly snaps at one Mucius and says: ‘You have the intelligence of Abdera’s rabble’ (Abderitanae pectora plebis habes) (Mart. 10.25.4). And Juvenal says of the philosopher Democritus of Abdera that ‘his wisdom shows us that men of high distinction and destined to set great examples may be born in a dense air, and in the land of fools’ (Juv. 10.48–50). This fixed image prevailed even in the supposedly scientific works of Galen, who commented that ‘[i]n Scythia there has been only one philosopher, but in Athens many; in Abdera there are many stupid people, but in Athens few’ (Scripta Minora 2.79). This prejudice toward the Abderitans originated in Greek discourse, but there is no hint of any special place Abdera had in Roman life. Therefore, while Abdera was not a central point on Roman routes and in imperial activities, this is a clear case of inherited proverbial expression, which carries with it inherited prejudice. The contexts and the meaning remained the same for both languages and societies.

The Boeotians were also typed as quite thick. In Cornelius Nepos’ biography of Alcibiades we find this comment: ‘all Boeotians devote themselves to body strength more than to mental power’ (omnes enim Boeotii magis firmitati corporis quam ingenii acumini inserviunt) (Nepos, Alc. 11.3). And when Horace discusses popular taste and judgment, he says:

Along similar theories of deterministic environments, this idea wears a pseudoscientific robe in Cicero’s On Fate:

The idea of stupid Boeotians features also in the earlier Greek expression of ‘Boeotian ear’ (Βοιώτιον οὖς), or, in other variations, ‘Boeotian mind’ (Βοιώτιον νοῦς) or ‘Boeotian pig’ (Βοιωτία ὗς), all pointing at a foolish behaviour.

The ear appears also in the Latin ethnic expression of ‘Batavian ear’, referring to the ethnos inhabiting the region of the modern Netherlands. Martial describes the reaction of a person he accidentally met in the street:

In this context, this means everyone who is not dull, simple, unrefined and graceless.Footnote ¹³

In the collective and popular mind of the Romans the Abderitans and the Boeotians were stupid, but the Gauls were not so bright either. They were depicted mainly as naïve, gullible and unsophisticated. An epigram by Martial addressed at the emperor conveys this idea:

Here, too, it seems that this Gallic credulitas was based on actual experience, as witnessed, for instance, in Julius Caesar’s record of his Gallic campaigns:

Such characterization of the Gauls as fickle, impulsive and gullible was not new. Already Polybius alluded to the frivolous nature of this ethnos.Footnote ¹⁴ If so, perhaps by Caesar’s time this image has become a literary topos or indeed a proverbial stereotype.Footnote ¹⁵ Yet Caesar clearly relied on his experience in Gaul. How then can one separate reality from a literary topos? It seems to me that Caesar’s description was indeed based on actual encounters with the local inhabitants. But, the interpretation of their behaviour was perhaps influenced by a set of preconceived notions deriving from current ideas which were already delivered, for instance, by Polybius. Even if Caesar has not read these sections in Polybius’ Histories, this probably common prejudice possibly prompted him to notice and emphasize these specific traits among the Gauls.

Then there are the typically arrogant people. Plautus comments on a certain mercenary:

So were also the people of Capua:

And again in another speech of Cicero:

Why Praeneste and Capua? The last reference ties this trait with the city’s riches and luxury, and both are known to have been prosperous; this probably was explained in the popular concept the arrogance of their residents.

Campanian arrogance was also proverbial. Cicero speaks of illa Campanorum arrogantia (De leg. agr. 2.33.9), of the Campanum supercilium (De leg. agr. 34.93) and of the ‘always proud Campanians’ (Campani semper superbi) (De leg. agr. 35.95). This specific image and stereotype perhaps resulted from the proverbially very fruitful region of Campania, as it was coined in the proverb denoted as vulgo dictum:

All these arrogant people – the inhabitants of Praeneste, Capua and Campania – lived on the Italian peninsula not very far from Rome; but Latin proverbials typed also the people of Rhodes as symbols of arrogance, perhaps again due to their high economic status. Accordingly, Cato the Elder seems to have relied on such common notions when he commented that ‘they say that the Rhodians are proud’ (Rhodienses superbos esse aiunt) (Gellius, 6.3.50). The aiunt here emphasizes the inauthoritative and probably popular provenance of this characterization. Note that Cato went on to say: ‘but in what does their pride affect us? Would it become us to impute it to them as a crime that they are prouder than we are?’, meaning that in this case Cato has not succumbed to negative notions associated with prejudice. Still dwelling on the arrogance of the Rhodians, we see that in Plautus the association of a certain pompous person specifically with Rhodes seems to be not just a geographical indication. The context reveals that the essence of Rhodius contributes to the overall description and portrayal of the man:

In Latin proverbs there were nations typed as thieves and frauds. First and foremost are the Punics. There is no need to expand here on Punica Fides – whole chapters are devoted to stereotyped profiles of this ethnic group in both Isaac’s The Invention of Racism and Gruen’s Rethinking the Other.Footnote ¹⁶ But the Punics were not alone. The people of Crete were conceived as liars and cheaters:

Clearly, these two Ovidian citations apply the people of Crete as supporters of truth. But it is the very use of them as a standard that proves the inherent prejudice.

To the extent that an historian can justifiably use the slippery concept of personal identity, it is usually conceded that an amalgam of various social roles and recursive human behaviors is involved. The problems then quickly multiply.Footnote ¹ How many of these roles or behaviors were available to an individual in any specific circumstance? And how were the given or selected roles mobilized and in what contexts? To begin to answer these questions, we must attempt to determine the range of inherited and arbitrary items assigned to a person as opposed to the number of more voluntarily adopted and assumed cultural roles and resources – elements out of which an individual formed his or her own identity and had it shaped by others. Even where the choice of one element was possible – for example, an adult who embraced the new faith of Christianity – one is still faced with decoding the circumstances governing the salience of this identity over any others. In what circumstances might the new Christian choose or not choose to forefront his or her new religious affiliation? We are then compelled to explain why the particular salience exists in that circumstance.Footnote ² In many cases, especially in complex ones that traverse significant lengths of time, an individual was not always essentially person ‘x’ or person ‘y’ but rather, to use one possible example, an adult, a man, a father, a Roman (citizen or not), a Gaul, a Trevir, a soldier, a Christian, a veteran, a farmer, a municipal magistrate, or, more likely, some combination of these by turn. What is being considered here is not some high-flown Barthian theory about ethnic boundaries. What is being envisaged, rather, is a series of more tangible aspects of social existence that allow persons to define themselves and others to identify them.Footnote ³ Among these items, a recent historical analysis relevant to our time has listed the following ones: language, arms and modes of fighting, costume, bodily styles (e.g. hair arrangements), cuisine, and similar cultural attributes.Footnote ⁴ One can easily think of other less material items such as traditional occupation and religious adherence. Given precise contextual factors, only some of these are properly construed as ethnic in nature. Even of this limited number, most usually converge in a configuration that identify one as a specific kind of person, like a centurion in the Roman army as opposed to one who has a linguistic-kinship-locational ethnic identity, like Numidian or Gaetulian. But the two could easily reinforce each other, as when a Musulamian man served in the First Flavian Cohort of the Musulamii. Such restrictive conventions are complicated by the liberal use of metaphor. Christians, for example, conceived of themselves as a ‘new race’ or ethnos.Footnote ⁵ For many persons in the Roman Empire beginning in the later first and early second centuries, but not before, a new potential identity had been created. Further to complicate the metaphor, men and women who were or became Christians began deploying familial models of power and a broad kinship terminology to express their relationship to an all-powerful god who was a father to his children. Christians as Christians became persons who were one another’s brothers and sisters. So even salience has problems with it. A restricted emphasis for a person – ‘I am (in essence) a Musulamus’ – can work if he can front or parade certain aspects of personhood while telling other ones to get lost or at least to hide in the closet for a while. Some given aspects of our personhood, however, are so durable that doing this is difficult. They might not accept the repudiation.

As has been perceptively noted, ‘in a multinational empire whose makeup was multiple, heterogeneous, unequal, and sometimes hostile and badly integrated, the identity of each individual was inherently complex’.Footnote ⁶ In making these remarks, Veyne suggests that the forming of personal identity, including civic or ethnic identities, was complicated by the very existence of the Roman Empire. An exemplary case has been provided for Africa by the interrogation of a witness before Zenophilus, the Roman governor of Numidia, in the year 320. Court appearances, after all, were one of the contexts that hailed forth assertions of who one was. Asked to identify himself, the man declared: ‘I am a teacher of Roman literature, a Latin grammarian; my father is a decurion here in the city of Constantina, my grandfather was a soldier who served in the imperial comitatus, and our family is descended from Maurian blood.’Footnote ⁷ So: occupational profession, inherited civic status, inherited military status, and ethnic lineage; each of them was an element configured by Roman imperial power. The variations and permutations necessarily proliferate. I would therefore like to focus on a single case of ethnicity that illustrates some of the problems. The man’s life is significant because his ethnicity was strongly implicated in the various identities that were created and offered to individuals by the Roman imperial state. His career has already received some attention but, I believe, it still poses a series of interpretive problems that make him deserving of more. His life is a manifest instance where the Roman imperial state, a complex and powerful institution, helped, by the use and application of its cultural and administrative categories, to create new possible identities. Let us first consider the bilingual Latin/palaeo-Tamazight inscription on our man’s gravestone found at the town of Thullium (modern Kef beni Feredj), directly north of Madauros in the proconsular province of Africa (see Fig. 4.1).Footnote ⁸

Figure 4.1 Tombstone of Gaius Julius Gaetulus / Keti son of Maswalat.

From Chabot, Recueil des inscriptions Libyques, no. 146.

Latin Text

Palaeo-Tamazight Text

The Latin text on the gravestone set up for our man tells us that the deceased named in in the epitaph, Gaius Julius Gaetulus, was a decorated veteran of the Roman army who returned to his home town where he held the high-ranking priesthood of flamen perpetuus in its municipal hierarchy. Gaetulus’ military decorations reveal that he received some of the imperial army’s most prestigious awards. The dona militaria of armbands and neck torcs were awarded only to Roman citizens.Footnote ¹¹ Almost certainly a citizen from birth, as indicated by his tria nomina, our man probably served in one of the legions of the imperial army, perhaps (but not necessarily) the Legio III Augusta in Africa itself.Footnote ¹² In the other text on the same stone, which is inscribed in the palaeo-Tamazight script, this same man is called KT’i son of MSWLT, Keti son of Maswalat, an ‘imperial servant’ or ‘soldier of the emperor’ from the people of the Misiciri, from the sub-people of the S’RMMi.Footnote ¹³ His personal name and his larger community identity in the African language are completely different from his public face in the Latin text on the same stone. About when did Keti die and to when does his gravestone date? Some think as early as the late first century. Given the rate at which novel elements in the formal language and abbreviated elements in funerary epitaphs developed and then penetrated the more remote highland zones, however, it seems more likely that we are considering a date in the early to mid second century.Footnote ¹⁴

Manifestly Gaetulus’ identity involved a number of locational factors that can be specified. First among them was his home town of Thullium. Then followed the larger region of the Cheffia, the highland lying to the southeast of Hippo Regius in which Thullium was located (see map Fig. 4.2). Further encapsulating both the Cheffia and Hippo was a larger region lying west of the frontiers of the old republican province of Africa which, for convenience, we might call either eastern Numidia or western Proconsularis.Footnote ¹⁵ Parts of the latter large region were far western extensions of the Khoumirie (Kroumirie) highlands, while other parts of it stretched further southward and westward, ringing the southern horizon of the Hippo plain.Footnote ¹⁶ The highlands are sometimes referred to as ‘the mountains of the Medjerda’. They are one of the few micro-zones in the Maghrib east of the Atlas in Morocco that boast a higher than average rainfall, indeed among the highest in all of North Africa. An intensive mixed arboriculture has traditionally been the backbone of the rural economy, distinguishing it from the preference for cereal culture in the plains lying below the highlands. The rural economy in Roman antiquity appears to have shared this same distinction between highland and plains regions in this part of Africa. It is not accidental, I think, that the one detailed epigraphical text suggesting agricultural development in the lands near Thullium concerns a Lucius Arrius Amabilianus, an arboriculturalist. Like our Gaetulus, he was a flamen perpetuus, probably in the same municipality of Thullium.Footnote ¹⁷ The octogenarian Amabilianus boasts of having established his domus and having improved its economic well-being. He laid out an orchard with apple trees and provided it with a well, and then he set out a second orchard of fruit trees that he furnished with a water reservoir and a well. Amabilianus was another hard-working bonus agricola of the time who rightly boasted of the improvements that he made to his patrimony.Footnote ¹⁸ Arboricultural crops appear to have been the ones of which he was especially proud.

Figure 4.2 The Hippo Regius Region: Hippo Regius and Thullium.

Based on IGN 1960 ‘Carte d’Algérie’ 1:200.000.

Since both Gaetulus and Amabilianus held municipal priesthoods, we might ask when and how the municipalization of the region, and therefore of Gaetulus’ home town of Thullium, took place. Far to the northwest an Augustan colony was established at Hippo Regius, and a little further away to the south the Flavian emperors founded a colony of veteran soldiers at Madauros. But these were exceptional Roman settlements made by the direct intervention of the Roman state. Otherwise, Thullium was right in the middle of a zone that was remote in terms of Roman municipal development. The closest municipal centers were located on the periphery of a fifty-mile radius extending outwards from Thullium: Hippo Regius to the northwest, Thuburnica to the southeast, and Thagaste to the southwest. There was no known move to formal Roman municipal status made by any of the towns in the Cheffia highlands throughout the whole period of the high empire. We must therefore suspect that the advancement of Thullium to formal municipal status took place – if it happened at all – in the later empire.Footnote ¹⁹ Just how far the forming of municipal institutions eventually proceeded and what the process meant in the highlands is difficult to say. A comparable village in a similar highland environment at Henchir Aïn Tella (ancient Castellum Ma […] rensium), in the far western Khoumirie to the north of Thullium, was still governed by seniores or a council of elders as late as the age of the Tetrarchs.Footnote ²⁰ Generally speaking, then, it seems that the communities in the mountainous highlands from which Gaius Julius Gaetulus came were not as intensely connected with the main patronal resources of the empire. They were not able to develop the costly apparatus of urban Romanity sufficiently to convince governors or emperors that they were worthy of elevation to colonial or municipal status. In this fashion, the political ecology of the region determined elements of the identity of its inhabitants.

If the important colonial harbour city of Hippo Regius was only about forty kilometers from Thullium as the crow flies, the experiential distance was considerable. The accidence of the terrain and the heavily forested environment contributed to a palpable sense of difference from the metropolitan world of a well-connected Mediterranean sea port. Even within the Cheffia, Gaetulus’ village of Thullium was a satellite outlier, being located towards the northwestern periphery of the region. As such, it was much closer to the outer eastern periphery of the Hippo Regius region than to the subzone of the Bagrada (the modern Medjerda) river valley to the south. If Thullium was most probably still a simple civitas at the time that Gaius Julius Gaetulus served in the army, we know that its inhabitants were gradually adopting Roman norms. Formal municipalization was slow, only coming in the later empire when similar small towns were achieving higher status in the flush of what can be called a late imperial rural ‘boom economy’ in Africa. In the early fifth century, the village was known to Augustine, the Catholic Christian bishop of Hippo. He referred to Thullium and to a man there with the African name of Kurma who was a curialis of the municipality.Footnote ²¹ Augustine’s words are rhetorically construed (for him, Kurma’s unusual life-and-death experience was being used as an example), but they strongly suggest that to be a member of the town council of Thullium and to be in the ranks of its duoviri did not require particularly great wealth.

The ethnic group of the Misiciri to which Gaetulus belonged is one of the better-attested ‘tribal’ entities in Roman-period North Africa.Footnote ²² By studying the distribution of inscriptions in both palaeo-Tamazight and Latin, or ones that were bilingual, using both languages simultaneously, it is possible to plot the region in which the people who identified themselves as Misiciri lived. Their distribution on a map (see Fig. 4.3) shows that their region was a zone between the Bagrada valley and hilly lands to the south and the coastal plain inland of Hippo Regius to the northwest. If there were long-term interactions between the inhabitants of the Cheffia and their environment, it is hardly surprising that they came to share common identities. The peoples inhabiting the region would have shared a common distinctive environment in which they lived and worked. The larger montane zone consists of distinctive subzones, and so it is speculatively possible to identify five major subgroups of which the Misiciri were formed and to map their locations in the highlands of the Cheffia.Footnote ²³ The concentrated location of inscriptions belonging to each subgroup argues in favor of an ecological component in its formation and identity. Each seems to be located within a fairly well-defined territory that was formed by a valley – that is, by distinctive mountain and riverine confines.Footnote ²⁴ What is more, if the ecology of these regions in the Roman past resembled that of the later nineteenth and early twentieth centuries (which I have no reason to doubt), then the dense habitation of the mountain highlands was matched by an intense fragmentation of ethnic identity. In addition to the five subgroups of the Misiciri, seven other ethnic groups have been identified in the highlands immediately adjacent to the Cheffia. Most probably, like the subsections of the Misiciri, they were local groups of the much larger peoples collectively named the Numidae.Footnote ²⁵

Figure 4.3 Proconsular Africa: The ‘Cheffia’ Region in Context.

Copyright: author.

Having been raised in this ecology, who was our Gaetulus? Was he was a high-ranking Roman citizen, a soldier in the Roman army named Gaius Iulius Gaetulus? Or was he Keti son of Maswalat from the tribe of the Misiciri, from the subtribe of the S’RMMi? Almost certainly he was both at the same time. He was like the man from Gaul who boasted on his tombstone found at Aquincum on the Danube: “I am a citizen of the Franks and a Roman soldier under arms.”’.Footnote ²⁶ Like many Gauls and Germans serving the empire, this man maintained a bifurcated identity. One was local and the other imperial, the second being determined by the existence of the empire and its army. Gaius Julius Gaetulus was like the Roman citizen from Tarsus who called himself Paulus. Several times Paul insisted on his possession of the Roman citizenship before high-ranking officials of the imperial state. At the same time, he was Saul, a man who self-identified before his fellow Jews as Jewish, a descendant of Abraham from the tribe of Benyamîn, and belonging to a family of strict Pharisaic upbringing.Footnote ²⁷ A special aspect of Gaius Julius/Keti’s split identity is that it was not new to him. It had been maintained over a number of generations. The original citizenship of Gaetulus’ remote male ancestor, and hence Keti’s own praenomen and nomen of Gaius Julius, almost certainly dated to the time of Julius Caesar. That ancestor had probably received land and citizenship from the great generalissimo in the mid-40s BCE as a reward for military service. A precise date for our inscription is difficult to specify with any certainty, but we have argued earlier that some point in the early to mid second century makes the most sense of all of the evidence. If so, our Gaetulus was part of a family that had connections with individuals and institutions that were Roman for about two centuries (possibly more if there were Marian antecedents in his line of armed service). But let us say, provisionally, that we are looking at approximately two centuries. Assuming an arbitrary calculation of about thirty years to a generation, our Gaetulus was part of a family whose service connections with Rome (or, at very least, Roman citizen identity) had continued through no less than five to six generations. We are fortunate to have evidence of another man from Thullium who did army service and who also bore the name of Maswalat. Having served in the Roman army, like our Keti, he is similarly called a ‘servant of the great chief’ (i.e. the Roman emperor). There is a considerable likelihood that his Roman name was also Gaius Julius. If not these names, however, he surely bore the Roman tria nomina. But this Maswalat, despite having had the same army service as Keti, had all of his identity recorded solely in his native language and in words taken from his own African tongue to describe his imperial service.Footnote ²⁸

As shown previously all by his army service, our Keti was also a Roman. Indeed, as has been acutely observed, in this respect he could hardly have been more Roman.Footnote ²⁹ Yet in his native language he chose to present himself as an African who belonged to an ethnic group, the Misiciri, and more specifically to a smaller subgroup of the Misiciri, the S’RMMi. Such men who performed imperial service, and persons related to them, added the cognomen Gaetulus, Gaetulicus, or variants to their Roman names and were proud of it.Footnote ³⁰ The problems with our Gaetulus, however, are not so easily solved. Without doubt, during all of the years that he served in the army, he would have forefronted his imperial identity. He would ordinarily have spoken Latin and he would have been committed to the military values that enabled him to win the honors that he did. If he enlisted at the usual age of 18 to 20 and received a normal honesta missio, Gaius Iulius Gaetulus would have returned to his home town in his mid forties. He died at the age of 80, so more than half of his adult life was lived not in the Roman army but back home in the highland society of the Cheffia. In this context, inherited aspects of his behavior, inculcated from infancy and early childhood, like the native language that he spoke, would have come back into play in his daily interactions with the local people with whom he now lived. Who our man was depends very much on the time when we are considering his personhood. Was army service a usual gateway to imperial membership for men in these highlands? We are fortunate to know of other cases precisely like Keti’s. There were, for example, several known men from the same region who bore the name Iasuchthan: from the highlands around Mactaris to the southwest of the Cheffia, but also in the Cheffia itself.³¹ A man most probably from our region, also bearing a ‘republican’ praenomen and nomen, Marcus Porcius Iasucthan was a centurion serving in the Roman army who left record of his service at the distant desert post of ad Golas (modern Bu Njem, Libya) in the 220s CE.³² The long metrical poem erected at Iasucthan’s behest is filled with indications that for him Latin was manifestly a second language.³³ At least in terms of language, but probably much else, Iasucthan shared the same kind of double identity as did Keti son of Maswalat, and this some three generations later.

That a culture and therefore a personal identity is confirmed and continued by the inculcation and adoption of a language goes without saying. Language is a verbal and written encoding of the canons of a culture taken on by humans from birth without their assent or permission. What is significant about the region that Keti came from is that it gives all of the appearances of being a particularly strong container of indigenous African cultures and languages. By contrast, the lowlands and plains areas below the Cheffia seem to have been caught up in a series of large-scale economic acculturations led by Punic city-states in Africa that led to the proletarianization or transition to peasant agriculture by the local farmers.Footnote ³¹ Through the years of the high empire and into late antiquity, the rural dwellers in these lowlands continued to speak Punic as their primary or first language. In the highlands, however, where this economic shift did not take place, the inhabitants apparently continued to speak various dialects of native African languages that we rather misleadingly call Libyan, but which were most probably distant ancestral forms of Tamazight. The reason that we know this is because of truly striking concentrations of inscriptions. Almost all of them are on funerary stones, written in a script developed to write the local language, as a form of writing that was distinctively different from the Punic and neo-Punic scripts used to write ‘Punic’ or the Roman script that was used to write Latin.Footnote ³²

Better to understand the physical and social context, we might reconsider the position of Thullium in comparison with the small village of Fussala. Fussala was equidistant from Hippo Regius, probably located about 15 to 20 km southwest of Thullium. We know that the first language spoken at Fussala, indeed practically the only one spoken by the majority of the peasant farmers in the region, was some form of what we (and the Romans) call Punic.Footnote ³³ The palaeo-Tamazight speakers of the highlands, who lived in places like Thullium, had a linguistic buffer zone of non-Latin speakers placed between them and the large imperial urban center of Hippo Regius and the plains region immediately adjacent to that city. It would have been a rather permeable buffer, however, since it is likely that there was a greater linguistic proximity between palaeo-Tamazight and Punic, and related spoken languages, than there was with Latin.Footnote ³⁴ Further nuances are evident, even within a small community like Thullium. Two separate cemeteries have been found in the village: one that contains gravestones with the palaeo-Tamazight and bilingual Latin/palaeo-Tamazight epitaphs, and a second one where the writing on the gravestones that do boast epitaphs (admittedly relatively few) is only in Latin.

A basic and simple sign of empire in the highlands around Thullium was the pursuit of what has been called the epigraphic habit. It seems likely that the heyday of the production of epigraphical texts in non-Latin languages tracked the chronological arc of the production of Latin inscriptions for the same purposes, mainly funerary memorialization and the recording of public honors.Footnote ³⁵ That writing in this particular script was used demonstrates a type of cultural continuity in which Gaius Julius Gaetulus must have shared. The palaeo-Tamazight script is found widespread across the entire face of Africa. Variants of it are found in distant Mauretania Tingitana far to the west and in the form of casual wall graffiti in lands much further to the east on the desert periphery of Tripolitania, perhaps significantly, in this latter instance, in connection with a Roman army base.Footnote ³⁶ Perhaps even more important than the simple use of this script for interpreting Keti’s social position is apparent from the distribution of writing in all of the highland regions east of the Guelma/Calama line in North Africa (Table 1). Of all these zones, the subregion of the Cheffia is the only one where Latin/palaeo-Tamazight bilinguals are found. There are few of them, so there is every reason to believe that those who chose to have their final memorials recorded in both languages were themselves rather special cases. They were not special, however, in that they were surrounded by an unusually high number of palaeo-Tamazight inscriptions.

Lowland beliefs and religious institutions inflected the nature of local culture in the highlands of the Cheffia and therefore customs of burial and commemoration. Many of the iconic themes on funerary stelae – crescent moons, rosettes, caducei, crowns, and palm branches – are the same as ones conventionally found on the Saturn stelae of contemporary African cult in the Roman-type transformation of the cult of Ba’al Hammon. However pervasive these signs were on the cultic imagery found in the highlands, it was not for any engagement with the cult of Ba’al Hammon or Saturn that Gaius Julius Gaetulus was noted. Most significant is the fact that he came back to his home municipality of Thullium to hold the position of flamen perpetuus, the local priest in charge of the cult of the emperor. He was not alone. The ‘good farmer’ from this same region, Lucius Arrius Amabilianus, also held the same position, probably also at Thullium. Without large numbers or ways of tracking and quantifying all such ritual adherences, it is still notable that a few of the wealthiest and highest-status men were careful to note their engagement with the imperial cult. The position of flamen, which would have engaged the holder of the title in the annual celebration of the emperor’s birthday and the administration of public oaths of loyalty, was surely selected to emphasize Gaetulus’ Roman identity. He lived before the times when Christian ideas and practices were beginning to have wide influence in Africa. The emergence of Christian institutions at Thullium was probably roughly analogous to the pattern found at Fussala, located in the southeastern borderlands of Hippo Regius. There are some signs of Christian building activities in the Cheffia, including a chapel built at Bar el-Ghoula by a patron of the church, but they are rather few in number and do not seem to betoken anything like the comparatively intense Christian presence at lowland sites in the regions around the Cheffia.Footnote ³⁷ From Augustine’s words about the ironworker Kurma, there appears to have been no Catholic bishop at Thullium in the first decades of the fifth century. Like the village of Fussala, Thullium appears to have been nested within the large Christian bishopric of Hippo Regius. Neither the acts of the general conference held at Carthage of 411 nor the Notitia of bishoprics of 484 indicate that there was any bishop, Catholic or ‘Donatist’, in the town. By the late Vandal period, however, Thullium had been able to assert its autonomy from the diocese of Hippo. There was a bishop of the Christian church from the town who was present at the conference at Carthage in 525.Footnote ³⁸ From the point of view of the formalities of the Christian church, Thullium, like Fussala, was a late developer. Fussala got its own bishop as early as the 420s, whereas the establishment of a bishopric at Thullium was delayed by as much as eighty or ninety years later.

In the provinces of the early to mid second century, optative identities like being a Roman or, later, being a Christian were largely matters of personal adhesion. But language and kinship were not. You were born and raised with given ones. The largest social group named in inscriptions from the Cheffia, both in Latin and in palaeo-Tamazight texts, was called the Misiciri. They were members of a kinship unit who were present on a geographic and demographic level that covered most of the region. It was the largest African group to which Keti claimed to belong. But he also recognized a subgroup of the Misiciri, called the S’RMMI or Saremmi. Manifestly, they were a smaller and more specific kinship group, to which he also belonged. Arguments have been proffered that there were at least four other similar subgroups of the Misiciri that are attested in the palaeo-Tamazight inscriptions in this same region: the NSFH, NNDRMH, NFZIH, and the NBIBH.Footnote ³⁹ Whether these were all the subunits of the Misiciri, and whether or not they confirm the existence in Roman period Africa of the ‘five-fifths’ segmentary systems found in some modern-day Amazigh groups in the Atlas and Rif far to the west, will need further investigation and discovery to confirm. It is sufficient to note here that the Roman Empire kinship identity in this region was nested in complementary segmentary units that had apparently existed for a fair period of time. These same nesting arrangements of kinship groups are attested for other similar ecologies in the Roman world. An inscription from Rawwâfa in northern Arabia records a temple built by a man, one Sa’dat, who identifies himself as from the Sisthioi, a subgroup of the Thamudenoi (Thamûd), and the Sisthoi themselves were from the ‘tribe’ (phylê) of Rhobathos.Footnote ⁴⁰ Here is found the same trifold nesting of one ‘ethnic’ unit within another.

Furthermore, a gendered aspect of the public notation of ethnic identity is regularly discernible in the epigraphy of the region. In all the funerary epitaphs from the Cheffia, all of the females are identified with names that look very Latin. Not one of them is memorialized in the palaeo-Tamazight script or in a palaeo-Tamazight/Latin bilingual. No woman in any type of epigraphical text identifies herself as a member of any of the kinship groups in the Cheffia. In the public sphere, it seems that men, but not women, deliberately marked elements of traditional culture and ethnic affiliation. And yet there must surely be a strong presumption that indigenous women, as in many comparable instances in the western provinces of the empire, were special bearers of local identity, such as being a Misiciri or a S’RMMi. Apparently these women, who were surely in the majority, simply did not present themselves in the field of public epigraphy.

There is one strong qualifier to all of these observations on kinship and ethnic identity. Our man boasted the Roman Latin cognomen of Gaetulus. In Roman terms, there is no doubt that he was seen and classified as a ‘Gaetulian’. The problem is that there was almost certainly no ethnic group defined in terms of kinship (like the Misiciri, for example) that identified itself as Gaetulian. Such a term never appears in the indigenous palaeo-Tamazight script or in any contemporary epigraphical texts as the name of a distinct ethnic group.Footnote ⁴¹ The name appears to be an external identifier, one of the generic categories of ‘Africans’ that were used by imperial administrators, and by the geographers and ethnographers who provided them with ‘ethnic information’. The name seems to have emerged as a convenient label for a generic class of indigenous persons who happened to engage in armed service for the Roman state. Gaetulians were seen as a grab bag of sometimes southern, sometimes highland, sometimes autonomist, occasionally violent peoples. Various peoples who were occasionally involved in resistance to programs of settlement and integration that were fronted by Mediterranean states with which they came into contact were categorized as ‘Gaetulian’ regardless of their own self-ascribed ethnic identity. In defeating any people who fell under this external rubric, Roman generals assumed the ethnic name ‘Gaetulicus’ as a victory cognomen.Footnote ⁴² Many of the specific ethnic groups who fell under this external rubric were subsequently absorbed into the armed forces of the Roman state. Evidence of such armed service dates early into the pre-Roman past of the Carthaginian hegemony in Africa. We are told that Hannibal recruited Gaetulians for service in his army.Footnote ⁴³ Men of this extraction formed a pool of potential recruits for armies whether they were Carthaginian, African, or Roman. Marius recruited important elements from ethnic groups called Gaetulian for his African campaigns against Jugurtha.Footnote ⁴⁴ In the civil wars of the 40s, these same Gaetulians, along with Numidae, were recruited by the Pompeiani and served them until Julius Caesar, the descendant of Marius, appeared on African shores. At that point they defected en masse to his side.Footnote ⁴⁵ In regions further to the west, around Cirta and Calama, other men of this same background went over to the side of Caesar’s self-appointed freelancing baronial ally, the Campanian freebooter Publius Sittius. It is very likely that many of the Africans who later bore the cognomen Sittius were among the Gaetulians who were enfranchised by Caesar’s man in the west.Footnote ⁴⁶ It is similarly probable that the Gaetuli who loyally served Julius Caesar in the battles in the old Republican province of Africa in 46 BCE account for considerable numbers of men who were enfranchised by him. They later bore the praenomen Gaius and the nomen Julius, as our man Keti did many generations later. This particular Roman connection with the so-called Gaetulians deserves closer inspection.

In taking over command of the war against Jugurtha, Gaius Marius recruited heavily not just from among newly eligible Roman citizens in Italy but also from, as is often not noted as part of this same process, among ‘ethnic’ peoples in North Africa. These latter men also provided important additional manpower for the war against Jugurtha. Being well acquainted with local languages and customs and thoroughly experienced with the climate and terrain, they were perhaps among the most useful of Marius’ new recruits. He drew many of them from indigenous peoples living along the frontiers of the Republican province. When the war was over, he arranged land rewards not only for his Roman citizen and Italian veterans but also for his African soldiers. They were settled in towns and in rural regions in the same interstitial zone along the western border of the Roman province.Footnote ⁴⁷ In Roman parlance, these Africans doubtless became his clients and were understood to be so, although surely no Roman technical term would have been necessary in the minds of the Gaetuli themselves to describe the social gratitude that linked them to their benefactor. They had served him, and now he had served them. As was traditional, they assumed close ties of loyalty by kinship and military service with Marius’ descendants. In the factionalism of the civil wars that rent Africa in 46 BCE, it was natural that the descendants of Marius’ Gaetulian recruits rallied to support Julius Caesar, who was Marius’ close familial relation, against his personal enemies, the Pompeiani. In response, Caesar had extended the citizenship to these men and had made grants of land to them.

In consequence we encounter many descendants of these Gaetulians in the high empire who bear the praenomen Gaius and the nomen Julius.Footnote ⁴⁸ Several generations after the age of Caesar, we find cohorts of Gaetulians in the service of the Roman army. One of them, the Cohors Prima Gaetulorum, is reasonably well documented.Footnote ⁴⁹ The geographic distribution of the cognomen ‘Gaetulicus’ reveals heavy concentrations just to the west of the old provincial boundary, the Fossa Regia, one of them in eastern Numidia where Keti’s home town of Thullium was located. Most of the other groups of men bearing the cognomen of ‘Gaetulicus’ were connected with various army bases in Africa, including Ammaedara, Theveste, and Lambaesis. Such men are also found concentrated in colonial settlements of veterans, like Madauros and Thubursicu Numidarum.Footnote ⁵⁰ There are good reasons to believe that Marius was responsible for the settlement of his African veterans who would otherwise have been labeled as Gaetulians but who, after their receipt of citizenship, bore the praenomen Gaius and the gentilicium Marius. Significant numbers of the descendants of such men are found in the borderlands of the old republican province in Africa: that is to say, in lands south and east of the Fossa Regia. The distribution manifestly points to the pattern that we would expect from the historical scenario just described.Footnote ⁵¹ As with the men and women who later have the cognomen Gaetulicus, or variants like Gaetulus, those bearing the praenomen-nomen Gaius Marius when they are found outside the core area to the west of the republic province in Africa, are also attested in the big army camps or in veteran colonies.

All these facts indicate that army service continued to define who these men were over several generations. Gaetulian was manifestly an external Roman label used to cover such armed servitors. The men themselves, however, had their own local identities: for some of them it was that of belonging to the Misiciri. To use the name of Misiciri, however, only raises further questions of identity and representation. The Misiciri were just one of a number of small ethnic groups lying south and west of the frontier of the old province of Africa from whom the Roman state continued to recruit in the empire. Men from smaller local groups like the Misiciri were usually recruited under larger headings, being considered Gaetuli, Afri, Musulamii, Numidae, or the like, for the administrative purposes of the Roman state. Men in all of these groups continued to contribute manpower to the auxiliary units of the imperial army. Many of the Gaetulians who were already Roman citizens from the days of Marius and Caesar, however, were eligible for direct entry into the legions of the imperial army. This historical background and the claims with the Roman state that could be based on it are significant. In the wider context of the empire, it is manifest that there were firm ethnic prejudices held by the big power holders, against Gauls and Germans and other northerners for example, that were effective barriers to advancement in the imperial system.Footnote ⁵² Even for Gauls and Germans, however, a gateway into the ranks of imperial power – and one that became dominant for excluded northerners from the early third century onward – was through service in the army.

This is the role in which we find our Gaius Julius Gaetulus at Thullium, and most likely not a few of the other men whose gravestones we find in the Cheffia – men like Lucius Postumius Crescens, also from Thullium. Crescens was remembered not just in Latin on his memorial stone, but also in a parallel inscription in palaeo-Tamazight where he self-identified as belonging to the Misiciri, probably assuming that everyone knew that his being one of the local S’RMMi was understood.Footnote ⁵³ There were others who, like Nabdhsen son of Cotuzan, noted that he was from the tribus Misiciri. Perhaps because he died at the age of twenty, Nabdhsen never made it into active army service. This is the only information that appears in Latin in Nabdhsen’s epitaph. But the words in Latin are accompanied by a parallel inscription in palaeo-Tamazight which shows that, like Gaetulus, Nabdhsen also identified himself as a member of the S’RMMi.Footnote ⁵⁴ Nabdhsen probably shared a status similar to Sactut son of Ihimir who was also memorialized in a parallel palaeo-Tamazight inscription on his tombstone.Footnote ⁵⁵ Such was probably also the case with Chinidial son of Wisicir from the tribus Misiciri from a site just to the southwest of Thullium, whose Latin epitaph is also accompanied by one in palaeo-Tamazight. In this case, interestingly, the text in the indigenous script does not say that he belonged to the Misiciri but rather to the NChPi, who, like the S’RMMi, were most probably a smaller subgroup of the Misiciri.Footnote ⁵⁶ Chinidial therefore preferred to note his membership in the small kinship group, assuming that everyone understood that the NChPi were a subgroup of the Misiciri. On the other hand, one Paternus son of Zaedo, like our Gaetulus, is named as a member of the Misiciri only in the palaeo-Tamazight text on his stone, as is another son of the same father in the same town.Footnote ⁵⁷ Similarly, one Aug[e?] son of Sadavo, Numidian from the tribus Misiciri, includes his affiliation with the larger ethnic group as a significant element of his identification.Footnote ⁵⁸ An important ancillary point revealed by the nomenclature of these men is that the default mode of ethnic identity in the highlands of the Cheffia shows no sign of any obvious Punic influences. These personal names are not cast in a formal Roman Latin mode; they are Latin transcriptions of African names. Nabdhsen, Cotuzan, Chinidial, Zaedo, Auge, Sadavo, and so on, are not Punic names but African ones – just like the Kurma from Thullium who was mentioned by Augustine. This much is evident from other regions in Africa, where some locals who also had the praenomen -nomen Gaius Iulius, like Gaius Iulius Arish and Gaius Iulius Manulus from Calama (modern Guelma), adhered rather to the use of the Punic language and to the worship of Lord Ba’al.Footnote ⁵⁹ They are two examples of Africans from non-highland areas whose culture had become Punicized before becoming Roman.

Through a series of ingenious and insightful parallels, it has been shown that the palaeo-Tamazight on Keti’s funerary stone reading MWSi MNKDi probably means something like ‘servitor of the supreme chieftain/king’. This was a local Misicirian way, so to speak, of describing Keti’s service in the Roman army. Several of the Latin bilingual speakers in the Cheffia, including our man Keti/Gaetulus, Postumius Crescens, and Sactut son of Ihimir, were army veterans.Footnote ⁶⁰ For them and, we must suspect, for many others like them, the army was one of the main instruments of imperial integration and identity. The social and disciplinary regimes in the hothouse of the legion and the auxiliary formations helped to shape an identity vitally linked with the empire.Footnote ⁶¹ The role of the uniform requirements of a type of a national military service in forming identity was surely as significant here as it has been in many modern instances.Footnote ⁶² While the majority of recruits of the Legio III Augusta in Africa seem to have come from the more densely populated urban centers of the old province of Africa, especially from their urban proletariats, recruiting also continued from ‘ethnic zones’ of the African provinces. Many of the non-citizen recruits from these social groups probably gained citizenship and the ability to enroll in the legionary forces of the empire through auxiliary service in one of the units of Afri, Mauri, Numidae, or Gaetuli that are well attested in the auxilia of the high empire. The recruiting of highland peoples, whether the Ituraeans in the Lebanon, Thracians from the Balkans, or Isaurians from southern Asia Minor, was as normal. It was as typical as it later was for the armies of early modern Europe for whom military service by Scots, Swiss, Auvergnians, Pyrenaeans, or other impoverished highland men was normal. We can therefore say that armed service for the Roman state was a choice that a young man like Keti might make. The rider is that the Roman army as specifically configured in the late Republic and Principate had to exist as a viable institution for men like him to be able to make such a decision. Even given the availability of army service, the degree of freedom of choice is still open to debate. Four or five generations of male ancestors of Keti’s had already been in Roman army service. Given the tendency for a behavior like this to be inherited, we might ask how much this element was an embedded element of our man’s ethnic identity. Did the peoples of the Misiciri who had performed armed service for the Roman state for generations, like the Nepalese Gorkhas, the Gurkhas of the British army, come to be defined by that service?

The empire consisted of many different types of social and political units. These included kingdoms, principalities, and baronies, followed by city-states of various types, conventionally labeled as ‘free and autonomous’ or wholly subject to the dictates of the states. There followed ethnic units variously known as gentes, nationes, tribus, or populi. Since the first of these political units tended gradually to be squeezed out by managers of empire who considered such quasi-autonomous entities to be incompatible with the fact of empire, it became conventional to view the empire, ideally, as an amalgam of ‘cities’ on the one hand and of ‘peoples’ on the other. Therefore one way of envisioning the imperial project is to see it as an entity composed of distinct modular units: ethnic peoples on the one side and cities or urban communities on the other. More of some were found in certain regions, and more of the others in others. Such a taxonomy was always complicated by the fact that urban groups were themselves sometimes construed as ethnic groups, as, for example, the Cirtenses, Madaurenses, Thuggenses, and the Capsitani in Africa. A different way of thinking about the same process would be to see this divided composite of identity as potentially running internally through individual subjects all the way down to the ground level of any given locale. There is every reason to believe that the empire was filled with persons of such divided identity. The one individual case of Keti son of Maswalat powerfully indicates how moveable and changeable some of the elements were that contributed to ethnic identity in this mix. Mommsen long ago made a fundamental observation that is worth repeating: the empire was a continuous revolution, a thing always in the process of remaking itself. The effects of this continual refashioning were felt at local level. In this hybridity, there are some elements that seem more stable or longer term than others, but change was ever present.

Anchoring one end of this polarity were long-term, almost inherited aspects of identity that surely had a large impact on how the person saw himself or herself. Of these, the inculcation and learning of a native language must be one of the most formative, even in a multi-lingual environment. There is no reasonable doubt that a variant of proto- or palaeo-Tamazight was spoken in the highlands to the east and south of Hippo Regius and that it was in all likelihood Keti’s language of birth. Since the region was surrounded and, to some extent, penetrated by native speakers of Punic, we might suspect that Keti might have acquired knowledge of this other language, especially given its closer relationship to his native tongue – that is, when compared to Latin. The odd thing, perhaps, is that the palaeo-Tamazight speakers in the region developed, adopted, and propagated an idiosyncratic script of their own in which to write their language in public. This cannot be accidental. The continued use of a distinctive and peculiar script that first appeared in these regions, broadly speaking, with the first African kings must have been a deliberate choice. Both of these facts distinguished this propensity from the speaking of Latin. Although the inscriptions in the Cheffia highlands belong to a relatively restricted Roman time frame, nonetheless the script has a known time span that had already covered about three centuries or so by the time that Keti’s relatives were using it for his funerary epitaph.Footnote ⁶³ On the other hand, from prolonged army service, if nothing else, Gaius Julius Gaetulus would surely have acquired a reasonably good command of Latin. Even if learned and even if a second or a third language, the language of empire was present in relatively remote villages and hamlets like those of the Cheffia.

Our Gaius Julius Gaetulus or Keti son of Maswalat belonged to the peoples of the Misiciri and the S’RMMi. And he might even have considered the Gaetuli to be some larger such notional kinship-like entity to which he also belonged by virtue of the fact that the Misiciri were labeled as Gaetulians by the Romans whom he served. But he was also a Roman citizen of a family who had been Roman citizens for many generations. He was a citizen of the great imperial metropolis of Rome as also of the town, perhaps municipality, of Thullium. The managers of empire must have been aware of how normal a circumstance this was in the formation of their state. As we have said, one ideal way of reading the standard claim that the empire was made up of ‘cities’ and ‘peoples’ was to stress the existence of different and exclusive categories of social groups out of which the empire was composed: cities on the one hand and peoples on the other. The empire is made of apples and oranges, chalk and cheese. On the other hand, the concept could be understood as indicating a different sort of composite of which the empire was made in which individuals were simultaneously members of cities and also of peoples, like our Gaius Iulius Gaetulus/Keti son of Maswalat.

Of course, it is not possible finally to sort out something as complex as ethnic identity on the basis of a few inscriptions, some scattered literary references, and a few comparative data. But at least the following seems reasonably certain: even after many generations of integration into the parallel apparatuses of the Roman state, an apparently Roman man who served a lifetime in its army and in its municipal institutions still maintained a separate African identity. And he was not alone. The nomenclature of other persons in the Cheffia, the widespread use of a palaeo-Tamazight script, and the presence of stereotypical units of common ethnicities indicate a general social system of which he was part. Demonstrably in his case, and probably in the others, this local African culture was a living fact over a significant number of generations. This vibrant cultural world and the language in which it functioned was not a choice in the formation of Keti’s identity. There were also other elements that were outside free choice, and one of the big ones, surely, was the sea change in shape and structure that the Roman Empire went through in the late third and early fourth centuries. In Africa, as elsewhere, the third-century crisis of the empire was a crisis for ethnic identity. It is an observable phenomenon on the southern frontier of the empire, as well as on and beyond its northern ones. A host of identifiable ethnic groups entered this crisis and then disappeared from view. In the mid third century and at the end of it, new groups and new identities emerge. Very few of the old ones made it through the crisis unscathed.Footnote ⁶⁴ The Gaetuli and Gaetulians, like our Gaius Julius, disappear from the record, as do the Misiciri and, needless to say, the S’RMMi and the other four subgroups of the Misiciri of whom we know from the high empire.Footnote ⁶⁵ In some sense, it seems, these groups had their identities confirmed and maintained by being part of an imperial system that recognized them as being a specific people and that continually underwrote that identity by an administrative computation and by a specific type of integration within the empire – in the case of the Misiciri by armed service. When that system profoundly shifted in structure, so did the ethnic identifiers that were part of it.

On the other hand, there is no reason to diminish the Roman parts of Keti’s identity: his knowledge of Latin, Roman citizenship, municipal service, and Roman name. All of these elements, and the values associated with them, had also been maintained by his family over several generations. These other Roman elements of his identity, however, required the continued presence of the empire, indeed a particular type of that empire, to sustain them. Without an imperial army that entailed specific kinds of recruitment, integration, training and service, there are serious questions whether much of the imperial identity would have taken hold among the men of the Cheffia in the way that it did. As long as those specific connections existed, however, the case of Keti points to the presence, in non-trivial numbers, and in considerable parts of Africa of persons with this type of split identity. And there are surely good reasons to suspect that Keti son of Maswalat/Gaius Julius Gaetulus was a typical figure of empire. Everywhere we look, from the Gaulish noblemen in the west to persons like Saul/Paul in the eastern provinces of the empire, we witness the same inside schism that ran along the internal fault line between local society and central state.Footnote ⁶⁶ The different strands in personal identities ran from the top to the bottom of the social orders of the empire, confirming in reality the ideological claim that the Roman Empire was an empire of cities and peoples. Frequently, we must suspect, it was so within each person. It was an ever-changing, ever-adapting social schizophrenia that was maintained over many generations in not a few of the provincial families in the empire. It was as essential a characteristic of empire as were the elements of its grander political unity.