1.1 The Traditional Model

The traditional demographic model of Roman slavery essentially states that, by the Late Republic (133–27), Roman Italy was unquestionably a society permeated by the use of slavery, with roughly one-third of its population being slaves. This model dominated for the better part of the late nineteenth and the twentieth centuries, with Beloch, Brunt, Patterson, Bradley, and Finley, despite significant disagreements on the actual number of slaves, all concluding that slaves comprised between c. 30–40 % of the population of late republican and early imperial Italy.^[4]

In addition to this high rate of slave-ownership, Romans were also thought to have been very liberal in freeing their slaves. Scholars have pointed to a noticeable increase in the number of inscriptions from Rome from the latter half of the second century BCE onwards in which the status of libertus/a is recorded. This epigraphical evidence reveals a high proportion of those who declared freed status, particularly in funerary epitaphs, compared to those who were unambiguously freeborn.^[5] Scholars have also argued that, based on onomastics, even many of those who did not explicitly state status – incerti – were freed slaves.^[6] Evidence from the early imperial period was also used to support such a view: a first-century CE album from Herculaneum seemingly suggests that over 60 % of the adult male citizenry there were freed slaves.^[7] As Alföldy further notes, many epitaphs of freed slaves from the early imperial period exhibit rather young ages. On these grounds, he contends that across much of Late Republican and Early Imperial Rome, it was almost certain that a slave would be manumitted in their lifetime, and rather young, too – often in their 20s or 30s.^[8]

On top of this, many earlier demographers also argued that this high level of enslavement drove the displacement of freeborn (ingenui/ae), particularly rural, Romans and, as a result, the Roman citizen population suffered a catastrophic collapse in the Late Republic.^[9] Yet, despite this apparent shrinkage of the freeborn population, demographic models suggest that the overall citizen body still grew healthily – therefore, the replacement population must have come from somewhere. In his influential work on Roman demography, Brunt argues that frequent manumission was the logical solution: “The enfranchisement of […] slaves certainly swelled the size of the citizen body, but this increase only concealed a considerable diminution in the old Italian stock”.^[10]

The conflux of a high level of enslavement, regular manumission, and declining freeborn population led to a model in which freed slaves were extremely numerous in Roman society by the Late Republic. In his highly detailed analysis, Brunt determines that the urban population of the city of Rome was around 500,000 in 70 and 750,000 by the end of the Republic, of which 100–200,000 were slaves and around 130,000 were freeborn citizens.^[11] This implies that there would have been between 170,000 to 520,000 freed slaves in the city of Rome alone at the end of the Republic. In other words, Brunt proposes that the freed population outnumbered the freeborn population. Other relatively contemporaneous scholars largely concur – Dumont estimates that over 50 % of Rome’s urban population might have been freed slaves even as early as the mid-second century BCE.^[12] Indeed, freed slaves were thought to have outnumbered free citizens to such an extent that the categories of urban plebeians and urban freedmen were often conflated. In the words of Purcell, “the history of the Roman libertinus is the history of the plebs urbana and the history of the city, socially, economically, and culturally”.^[13]

This model naturally impacted our understanding of how freed slaves interacted with the political sphere. The sheer number of mostly destitute urban freedmen with little to no patronal oversight implied that, if unregulated, they could dominate the legislative assemblies. Scholars were thus inclined to view the freedmen voting rights as a perennial struggle between populist politicians mass-redistributing freedmen in the comitia tributa/concilium plebis and conservative politicians curbing the political influence of freedmen by restricting their tribal distributions, a political seesaw starting from the censorial reforms of Ap. Claudius Caecus (cos. 307, 296; cens. 312–308) down to the turbulent career of P. Clodius Pulcher (trib. pleb. 58).^[14]

1.2 The New Model

The three fundamental assumptions upon which this high estimate of freedmen was based – high level of slaveholding, high frequency of manumission, a collapsing freeborn population – have come under increasing criticism in recent years. As previous research often approached the issue by examining the supply of slaves, Scheidel re-evaluated slaveholding numbers by examining demand for slave labour and carry capacity for slaveholding instead.^[15] Roman agricultural handbooks suggest that Roman estate owners during the Late Republic and Early Principate did not regularly rely on large slave workforces.^[16] Rather, they typically kept a small, enslaved skeleton crew to keep a farm or estate running throughout the year and hired free labour during busy periods such as harvesting or sowing seasons (Cato agr. 10–11; Colum. 1.7.6, cf. Plin. nat. 18.21).^[17] In urban areas, too, the use of slave labour was likely lower than previously assumed: for example, large-scale building projects often used free workers, not slaves.^[18] By examining demand and carrying capacity instead of supply, Scheidel argues that there were far fewer slaves in Roman Italy at the end of the Republic and the start of the Principate than traditionally assumed, both in raw numbers and as a percentage of the total population – around 10–20 % of the population of Italy.^[19] Scheidel later mildly revised that estimate to 15–25 % of the Italian population and about 10 % of the total population across the Roman Empire.^[20]

While Scheidel’s numbers are not universally accepted, they have been deeply influential since their proposal, with most current scholars concurring with Scheidel’s analysis.^[21] Morley concludes that slaves likely made up 5–20 % of the urban population of the city of Rome.^[22] Herrmann-Otto separately determines that slaves likely made up only around 5–10 % of the urban population in the City.^[23] Lavan allows a lower limit of slave numbers at 6 % of Roman society overall in his demographic reconstructions.^[24] Although both Scheidel’s ‘bottom-up’ and the previous ‘top-down’ methods are only educated guesses at slave numbers, the latter method has far less supporting evidence and the suggestion that slaveowners always maximised supply regardless of demand is economically unsound.

Furthermore, the assumption that Romans regularly and ‘excessively’ freed their slaves has also increasingly come under attack, and the overreliance on epigraphical evidence (and all its biases and interpretative problems) has been thoroughly criticised. Wiedemann convincingly argues against the idea of ‘automatic’ manumission, stating that regular manumission was merely an ideal that was not often followed in practice.^[25] In essence, freed slaves may have been over-represented in the epigraphical evidence due to a desire to commemorate themselves in a permanent fashion, possibly driven by their previously enslaved status.^[26] The often young ages shown in freed slaves’ funerary epitaphs may have been because deaths at the peak of youth were considered particularly tragic and thus more likely to have received some form of commemoration that would survive to the modern day.^[27] Finally, it is unrealistic to make arguments about the entire population based on epigraphical records – at best, we could perhaps claim that freedmen may have been disproportionate amongst those who were wealthy enough to leave permanent forms of remembrance, not that they were the dominant social group amongst the entire Roman population. This phenomenon, called the ‘freed epigraphic habit’, is now a well-studied and acknowledged part of the study of Roman slavery.^[28]

Finally, the model suggesting that the freeborn population was collapsing and required the manumission of slaves to maintain the citizenry, typically called the ‘low count’, is no longer held as defensible in its original form. Recent supporters of the ‘low count’, such as De Ligt, re-form the theory by allowing for a modest natural growth in the free population.^[29] The opposing ‘high count’^[30] and the alternative ‘middle count’^[31] both argue for a higher population and rate of population growth than the ‘low count’. Of course, all three counts suffer from different problems and none of them dominates our demographic understanding of Rome.^[32] None of the revised models, however, argues that the free population declined, but rather that it grew healthily during the Late Republic and was certainly not reinforced purely through manumission.

Therefore, our latest research in Roman demography is disinclined to support any of the three fundamental assumptions (very high levels of enslavement, high levels of manumission, a collapsing freeborn population) upon which the earlier estimates of freedmen in Rome was based. Consequently, most scholars have turned away from such estimates whereby freed citizens outnumbered freeborn citizens. Nicolet suggests that there might have been 200,000 freedmen in the city of Rome by the end of the Republic.^[33] Scheidel further revises the estimate down to 200,000 freedmen across all Italian urban centers in the first century BCE, though he does not provide an estimate for Rome itself.^[34] Mouritsen and Morley both offer 100,000 freedmen in the city of Rome as an estimate for the Early Principate.^[35] Tacoma suggests 60,000–135,000 freedmen in the City for the High Empire.^[36] Herrmann-Otto argues for an even lower number, 50–70,000 freedmen, in the city of Rome at the start of the Principate.^[37] Similarly, López Barja De Quiroga estimates that there were around 75,000 freedmen in the city of Rome by the Late Republic.^[38]

1.3 Rethinking Freedmen and Roman Politics

With this new demographic basis, we shall now briefly re-examine the politicisation of freedmen and manumission to construct the context necessary for our exploration of the lex Sulpicia de libertinorum suffragiis. Long has the censorship of aristocratic ‘enfant terrible’ Ap. Claudius Caecus in 312–308 been held as the beginning of the issue of freedmen voting rights since he allegedly restructured the distribution of the voting tribes and raised sons of libertini to the Senate (Liv. 9.46, cf. Diod. 20.36, and Plut. Publicola 7.5).^[39] His reforms, Livy complains, allowed the ‘forensis factio’ to rise to power. This reform of Ap. Claudius was rather short-lived as it was negated by the censors Q. Fabius Rullianus (cos. 322, 310, 308, 297, 295) and P. Decius Mus (cos. 312, 308, 297, 295) in 304 (Liv. 9.46.14 f.; Val. Max. 2.2.9b). According to Livy, these censors restricted all the humillimi, a category including the libertini (Plut. Publicola 7.5 and vir. ill. 32.2 f.), whom Appius Caecus had allowed “to register in the tribe of their domicile”,^[40] to the four urban tribes (Liv. 9.46.14). Because of these actions, Fabius was nicknamed ‘Maximus’ (‘the Greatest’) by his senatorial peers, undoubtedly a signal indicator of the senatorial aristocracy’s strong feelings in this matter.

It is, however, highly unlikely that the reform and counter-reform of 312–308 and 304 targeted freedmen exclusively or explicitly.^[41] Indeed, Cels-Saint-Hilaire strongly doubts that the humiles referred exclusively to freedmen, whose numbers she believed to have been negligible at the time.^[42] Cels-Saint-Hilaire and Humm further argue that the libertini here in question were not necessarily exclusively freedmen, but simply ‘new citizens’, viz. peregrines recently integrated into the Roman citizenry (perhaps following a procedure similar to the manumissio censu).^[43] These ‘new citizens’ were politically and socially devalued (at least in our sources of aristocratic inspiration) and possibly corresponded to Latins or Campanians whom the Roman expansion of the late 4^th century BCE had recently forcefully integrated into Roman citizenship.^[44] By integrating the sons of libertini into the Senate, Ap. Claudius was thus not acting out of demagogy (as later sources have often interpreted) but in the interest of the state (res publica) as he sought to integrate into the governing body representatives of the Latin or Campanian elites who had recently entered Roman citizenship (such as Cn. Flavius, son of Annus or Annius, curule Aedile in 304 and perhaps from Praeneste, or his colleague Q. Anicius Praenestinus).^[45] Quite possibly, these new citizens were simply lumped together indiscriminately with the actual freedmen in this era, which the former would no doubt have experienced as an additional sociopolitical grievance.^[46]

The beginning of genuine senatorial and censorial concern regarding freedmen specifically, then, should likely be traced to the late 3^rd century BCE. The combined evidence of Liv. per. 20 and Liv. 45.15.1–3 indicates that the libertini were restricted by a pair of censors between 230 and 220 to the four urban tribes, possibly with certain ad hoc and ongoing exceptions for certain categories (ad hoc: those who had a son over five years old; and indefinite: those who had property in the country valued at over HS 30,000 = 75,000 asses, then the minimum requirement for enrolment into the second property class in the comitia centuriata: Liv. 1.43.4 and Dion. Hal. ant. 4.16–21). After the censors of 179, M. Aemilius Lepidus (cos. 187, II 175) and M. Fulvius Nobilior (cos. 189) perhaps reverted to a more generous approach vis-à-vis the freedmen (comp. Liv. 40.51.9^[47]), the censor Ti. Sempronius Gracchus (cos. 177, II 163) in 169 prevailed over his hesitant colleague C. Claudius Pulcher (cos. 177) to terminate any exceptions granted during and after 230–220, enrolling all freedmen alike to one city tribe chosen by lot – tellingly, an arrangement much welcomed by the Senate (Liv. 45.15.3–7; Cic. de orat. 1.38, quoted infra; and vir. ill. 57.3, flagging a more serious rift between the censors on the matter).^[48] While the freedmen issue in the early to mid-Republic remains murky and scholars still debate the precise details of each event, it nevertheless seems that the issue of freedmen voting rights consistently remained the prerogative of the censors.^[49]

If censorial and senatorial concern about freedmen voting issues began in earnest around 230, and if freedmen were not nearly as numerous as we have previously assumed, then why did the censors and the senatorial aristocracy become so concerned at this juncture and onwards? According to Mouritsen, the senatorial aristocracy’s misgivings about the power and status of freedmen “were essentially ideological concerns”.^[50] The censorial activities, and our later examination of the lex Sulpicia, however, rather suggest that many of the senatorial elite and even non-elite genuinely feared the (re)distribution of the libertini. Using some case studies from the Second Punic War (218–201), we will corroborate, as a basis for our examination of 88, that (1) freedmen nonetheless formed a significant voting bloc, even if they did not number as greatly as previous thought, and (2) significant numbers of freedmen were attached to wealthy and politically active patrons, including those outside of the traditional senatorial class, such as equestrian and first census/property class businessmen and Italian domi nobiles.

In 217, after the horrific loss at Lake Trasimene, the dictator Q. Fabius Maximus (cos. 233, 228, 215, 214, 209) drafted freedmen of military age and with children with the other urban dwellers of Rome: magna vis hominum conscripta Romae erat; libertini etiam quibus liberi essent et aetas militaris in verba iuraverant (Liv. 22.11.8). Those under thirty-five were sent to the navy, while those above that age formed two legions to garrison Rome (Liv. 22.11.9). As freedmen were not exclusively targeted for recruitment, we have no idea what proportion of this magna vis hominum was comprised of libertini, though one imagines that they must have numbered in the hundreds or low thousands, at least, to make targeting them for recruitment worthwhile.

In the immediate aftermath of Cannae in 216, when Roman ranks had been seriously depleted following a raft of costly defeats and the city of Rome itself was vulnerable to direct Carthaginian assault, and after taking the desperate measure of enrolling four legions and a thousand horse of young men (iuniores) “from the age of seventeen” and even “some who still wore the purple-bordered dress of boyhood”, the Senate also had the treasury buy some 8,000 “young and stalwart slaves” and armed them, “first asking each if he were willing to serve” (Liv. 22.57.11). Valerius Maximus also records this but increases the number of slaves recruited to 24,000 (Val. Max. 7.6.1a). Feig Vishnia’s argument that these were predominately slaves of senators is shaky – if the Senate could have internally raised these slaves, then there was little reason why the tribunes had to announce it to the people and establish a commission (Val. Max. 7.6.1a).^[51] According to Livy and Macrobius, the Senate preferred this option over the less expensive alternative to ransom Roman prisoners-of-war, with Macrobius adding the valuable information that these volunteers fought on behalf of their Roman owners. Clearly the owners so excused from the fight would have been no ‘ordinary’ Romans but rather men of some consequence, sufficiently justified by their other responsibilities and activities, and sufficiently well-off to be able to part with their slaves in wartime.

Significantly, in 215, forming the bulk of the consular legions of Ti. Sempronius Gracchus (cos. 215, II 213), these so-called voloni/es (“volunteers”) first scored a significant victory over the rebellious Campanians at Hammae and next even repulsed Hannibal’s army as it besieged Cumae.^[52] In the spring of 214, authorised by the Senate and the consul M. Claudius Marcellus, the now proconsul Sempronius Gracchus promised freedom to every slave-volunteer legionary if they fought bravely in a pitched battle against the Carthaginian Hanno and his formidable confederate army at Beneventum. The Roman army prevailed over Hanno, with losses amounting to a mere c. 2,000 men, after which the consul duly made good on his promise.^[53] This was the first battlefield victory of the Second Punic War on Italian soil and the fact that the victorious army fielded thousands of voloni probably explains why this remarkable success did not result in any triumphal honours.

In 214, on account of a dire shortage of sailors for the Republic’s critically important naval forces^[54],

consules ex senatus consulto edixerunt ut qui L. Aemilio C. Flaminio censoribus milibus aeris quinquaginta ipse aut pater eius census fuisset usque ad centum milia aut cui postea tanta res esset facta, nautam unum cum sex mensum stipendio daret; qui supra centum milia usque ad trecenta milia, tres nautas cum stipendio annuo; qui supra trecenta milia usque ad deciens aeris, quinque nautas; qui supra deciens, septem; senatores octo nautas cum annuo stipendio darent. Ex hoc edicto dati nautae, armati instructique ab dominis, cum triginta dierum coctis cibariis naves conscenderunt. Tum primum est factum ut classis Romana sociis navalibus privata impensa paratis compleretur. (Liv. 24.11.7–9)

“the consuls, following a senatorial decree, issued an edict that if in the censorship of Lucius Aemilius and Gaius Flaminius any man had had his own property or his father’s assessed at between fifty thousand and one hundred thousand asses (or if it subsequently reached that level), he was to provide a single sailor along with six months’ pay. Anyone assessed above one hundred thousand and up to three hundred thousand was to provide three sailors along with a year’s pay. For assessment above three hundred thousand and up to a million, it was five sailors, and above a million, seven; senators were to supply eight sailors, with a year’s pay. The sailors provided in accordance with this edict boarded their ships, armed and equipped by their masters, with a thirty days’ supply of cooked rations. That was the first occasion on which a Roman fleet was manned with crews raised from private funds.”

A second request for sailors was made in 210. This request was originally refused until the senators offered to contribute funds as well, whereupon the rest of the populace (presumably the wealthy section) followed suit (Liv. 26.35 f.).

While the above passages reveal that slave ownership likely extended into the ‘middle reaches’ of Roman society, they also confirm that slave ownership was largely confined to the equestrian and first three census classes, as we can note that the number of slaves possessed declined disproportionately between the first and third census/property classes.^[55] These requests further differentiated the senatorial elite from other wealthy citizens. There thus existed unambiguously numerous slaveowners who were not a part of the senatorial elite by the latter half of the third century BCE. In so saying, even in a time of emergency, the number of freedmen enrolled in the army or slaves requested by the Senate did not reach numbers suggestive of a high-count of slaves and freedmen. It would not be before the depths of the crisis of the Social War in 90 that the Senate authorised the levy of twelve full cohorts of – probably Roman – libertini (i.e., c. 6,000 men, slightly over the equivalent of one full-strength legion) for service in the coastal region between Rome and Cumae, where, significantly, they “performed memorably and with courage”.^[56] As the Roman state was confronted with a severe shortage of soldiers as it sought to stave off disaster in Italy,^[57] the fact that no more than twelve cohorts of libertini were enrolled should probably not just be explained in terms of status-conscious reticence on the part of the Senate: rather, it corroborates the above analysis and invalidates the idea of a seemingly inexhaustible potential reserve of freedmen.^[58]

Of course, a lower count of freedmen does not suggest that they were politically insignificant. On the conservative assumption that freedmen made up around 5–10 % of Romans enrolled in the census and departing from the last recorded pre-Social War census figure of 394,336 civium capita in 115/114, there would have been c. 20,000–40,000 libertini in 90, signifying the number of 6,000 enrolled is eminently sensible. A politician wishing to mobilise voters hardly needed to have hundreds of thousands poor freedmen to vote to make a difference – even if only numbered in the lower tens of thousands, the libertini, if well motivated, would still have made a formidable voting bloc. In sum, our literary sources confirm that there existed notable slaveowners beyond the senatorial elite, and that the number of slaves and freed slaves mobilised during wartime are congruent with more recent lower counts of slaves and freed slaves.

1.4 Why manumit?

Once we remove the assumption that freedmen made up the majority of the plebs urbana, a result of Roman slaveowners wantonly and dissolutely freeing their slaves on a regular basis, and before we move on to a detailed analysis of relevant events in 88, we therefore must determine why a slaveowner would want to manumit: after all, regardless of reciprocal socio-legal obligations, a slaveowner undoubtedly had less control over a slave once the slave was freed: a freed slave was still a citizen sui iuris, with all the protections of that status.

Freeing a slave out of personal affection was likely one of the most significant motivators.^[59] Socio-political rationales – namely, to swell one’s ‘clientele’ – also played a part.^[60] We have clear evidence that freedmen acted as trusted go-betweens in the political sphere. These relationships were undoubtedly still based on personal affection and trust: their statuses as trusted confidantes naturally led them to adopt such roles. In so saying, manumission purely based on personal affection must have been relatively limited – few slaves would have had sufficient intimate contact with their owners to be manumitted for this reason alone. Such slaves were likely to have been favoured urban domestics.^[61]

Another important reason to manumit came with the expansion of Rome’s sprawling provincial territories, particularly after the First (264–241) and Second (218–201) Punic Wars, and with the increasing sophistication of Rome’s economic and financial structures, when freed dependents started to become invaluable in commercial affairs. A slave or a son under potestas could not conduct legal business independently, enter into contracts, or own property and therefore could not be legally held responsible if any issues arose.^[62] A series of praetorian edicts gradually introduced various actiones against slaveowners who acted through such dependent agents by making slaveowners potentially fully liable for any losses or problems incurred by their slave agents.^[63] Although we have clear evidence of the use of slave agents in business dealings, the practice remained legally unideal.^[64]

The utilisation of freed slaves in commercial dealings, however, solved many of these issues. A freedman was a Roman citizen sui iuris, in control of his own property, and held the ius commercii. Slaveowners could first ensure that the slave they were manumitting was trusted and had already displayed business acumen. The former owner could then much more safely invest in such freed agents knowing that, since they were separate legal entities, their own liabilities were limited. At the same time, they could expect a return on the profit by striking formal agreements of societas or loans with the freed slaves. Furthermore, the slaveowner might also expect long-term returns, as freed slaves were, in an increasing number of circumstances, legally obliged to leave behind parts of their estate to their former owner when they died.^[65] Freed slaves, too, would have also been incentivised to maintain a good relationship with their former owners post-manumission in order to access their patrons’ social and financial support networks.^[66] Thus, a slaveowner could use a freed slave in commercial enterprises, with a host of formal and informal incentives and coercive measures to enforce a continued relationship and to ensure returns while limiting liabilities in ways that utilising a slave agent could not.

This matches broader theoretical understandings as well. Both Scheidel and Hawkins, though acknowledging certain exceptions, demonstrate that highly skilled activities that demanded the ability to make rapid decisions without constant oversight, such as commerce and finance, required ceding greater autonomy to the agent and thus needed a more complex system of rewards, such as manumission.^[67] Broekaert argues that merchants needed agents like freed slaves to operate efficiently in pre-industrial commercialised economies like that of Rome.^[68] The fact that talented slaves in commercial endeavours were much more regularly freed than in other pursuits has been well-recognised.^[69]

Those who freed slaves for commercial reasons include certain senatorial families, who indirectly utilised freed slaves in commercial enterprises since senators were typically barred from direct engagement in commerce.^[70] Members of the urban plebs media also did so.^[71] Hawkins convincingly shows that urban artisans were highly likely to free slaves inter vivos and to utilise operae as a way to modulate labour depending on market demand.^[72] He further contends that urban artisans were far more likely to leave their workshops behind to freed heirs, even if they had natural heirs.^[73] Andreau reveals a similar phenomenon amongst bankers and financiers.^[74] Holleran also shows that shopkeepers were often freed slaves managing the shop on behalf of their former owners.^[75]

Businessmen belonging to the equestrian order or the upper census classes, however, were likely the largest subset of this group, as they undoubtedly had freed agents and utilised them liberally in a wide range of commercial enterprises.^[76] Cicero mentions that a highly reputable equestrian, Q. Minucius, had his freed agent, Timarchides, with him in Sicily (Cic. Verr. 2.2.69). Cicero, in a letter to P. Cornelius Lentulus Spinther (cos. 57), recommends to him a businessman, A. Trebonius, along with Trebonius’ own freedmen (Cic. Fam. 1.3). In Petronius’ Satyricon, Trimalchio is portrayed as having received a hefty inheritance from his former owner. Having become financially established, Trimalchio started investing in his own freedmen and used the profits to purchase land (Petron. Sat. 76). Well into the imperial period, Gaius reveals that it was still common to use freedmen as business agents abroad (Just. Dig. 40.9.10).

Significantly, many Italian domi nobiles also fell into this group. Italo-Roman merchants arrived in great numbers when Delos was made a free port in 167 and the ensuing period saw rapid commercial growth in the Aegean and Asia – areas that would later come under direct threat by Mithradates.^[77] Commercial interest in Asia Minor grew further when C. Gracchus (or perhaps some legislation passed shortly before his tribunate) allowed the publicani to farm taxes there (Diod. 34/35.25).^[78] Onomastic investigations of inscriptions in areas of the East controlled by Rome during this time also reveal a large presence of Italians and Romans with financial and commercial interests, though there are still debates as to the exact proportion of Roman citizens to non-Roman Italians there.^[79] The massacre of Mithradates shows the scale of Italo-Roman operations in Asia by the 80s. During the so-called Asiatic Vespers, some 80,000 Romans and Italians were reportedly killed, whom Valerius Maximus specifies as being there to do business (Val. Max. 9.2e.3, cf. Cic. Manil. 7, App. Mithr. 23, and Cass Dio. 31.101). Likewise, when Mithradates’ general Archelaus sacked Delos in 88, some 20,000 people, “of whom most were Italians” (ὧν οἱ πλέονες ἦσαν Ἰταλοί), were massacred (App. Mithr. 28).^[80]

This wealthy echelon of Italians would certainly have benefited from, or even agitated for, Roman citizenship in the run-up to the Social War, as they would likely have become equestrians or be assigned to the ranking census classes.^[81] Becoming Roman citizens would not only have lifted from them the burden of much of the taxes required to support their military contingents (the tributum had been abolished for Roman citizens in 167),^[82] but also allowed them to compete for lucrative state contracts, especially since tax farmers were now established in Asia.^[83] They would have had the wealth and time to travel and vote in Rome and thereby sway foreign policy.^[84] They could also potentially compete in elections themselves.^[85] With the conclusion of the Social War, this influx of new voters was dramatic – the census of 115/4 had 394,336 citizens recorded whilst the census of 70/69 had no fewer than 910,000, more than double the number of citizens from before the Social War. Even if we cautiously accept that only a minority of these new voters had the time and means to engage meaningfully in politics, this would certainly have upset traditional voting patterns.^[86]

Italo-Roman businessmen most certainly had freed clients, who themselves became wealthy and influential. Hatzfeld, in his investigation of Delian inscriptions, finds that out of 221 inscriptions of Rhomaioi with known statuses, 88 are freeborn, 95 are freedmen, and 48 are slaves.^[87] Three slaves and a freedman of likely an Oscan-speaking owner dedicated an inscription to Jupiter Liber on Delos (CIL I² 2203 = ILLRP 194). Three freed magistri of Mercury and Maia were also likely to have been Oscan speakers (CIL I² 2240 = ILLRP 749).^[88] Two freedmen gave the Italo-Romans at Delos a laconicum (CIL I² 2247 = ILLRP 289). On the Italian peninsula, inscriptions recording libertus status, prior to the first century BCE, centered in certain locations. Large Etruscan cities (Caere and Tarquinii, in particular), cities in Latium and Campania (Capua for example), cult centres such as Praeneste, and Rome all have far greater numbers of freed inscriptions surviving than other locations. While Rome surpassed the other cities in the number of freed inscriptions, prosperous and highly urbanised Italian cities, particularly those linked with maritime trade, also had high numbers of slaves and freed slaves before the Social War.^[89]

Significantly, Appian tells us that freedmen of Italians and Romans were explicitly targeted for slaughter in the opening stages of the Mithradatic War (App. Mithr. 22 f.). Of course, we do not know the exact numbers of freedmen with Italian owners, but there is nothing to suggest that the patterns of manumission in Italy might have been significantly different from rates in Rome – it might even be possible that Italians were more liberal with manumission, as they were not subject to Rome’s manumission tax. Importantly, Italo-Roman businessmen who had commercial concerns in Asia now also had a vested interest in the swift and successful pursuit of the Mithradatic War. These freed clients of the Italo-Roman sub-senatorial elite actively engaged in business alongside their patrons, and some became quite wealthy and influential themselves. As a result of the consecutive mass enfranchisements of 90/89 and 87, the Italian subset of this elite, with their freed clients, were to become Roman citizens and were now eligible to vote in Rome’s electoral and legislative popular assemblies.

In the light of all these considerations, we can establish a new contextual understanding of freedmen. Freedmen were not nearly as numerous as we have previously assumed, and a mass of ‘patronless’, destitute urban freedmen likely did not exist to any significant extent. Most politically notable and active freedmen were instead attached to, or themselves part of, the plebs media and above, with sub-senatorial equestrians (and later Italian domi mobiles) likely the group with the greatest need for and therefore the greatest number of freed clients. It is no surprise then, that the bulk of senatorial and censorial activity to restrict freedmen came during and after the Punic Wars, when these sub-senatorial equestrians rapidly increased in influence due to the expansion and complexification of Rome’s trade and financial systems.

Concerns about wealthy non-senators must have been at the forefront of the minds of the censors between 230–220, whereas in 218 a plebiscite carried by the tribune Q. Claudius, supported by C. Flaminius (cos. 223, 217; cens. 220), allegedly the only senator to support the bill, forbade senators from owning and operating seafaring vessels that could carry over three hundred amphorae (Liv. 21.63.2–5). Scholars have long debated whether this law was beneficial or detrimental to the senatorial class.^[90] Yavetz and D’Arms contend that this law forced senators to rely on urban merchants to conduct their commercial activities.^[91] Davenport suggests that the law may have been sought by equestrians who wanted to prevent senators from competing with them commercially.^[92] As Linke has further argued, C. Flaminius must have had wealthy supporters in the comitia centuriata who were dissatisfied with the Senate for him to have had so much electoral success despite senatorial opposition.^[93] The censors of 169, C. Claudius Pulcher and Ti. Sempronius Gracchus, too, clashed with this class: in addition to their notable move against freedmen (cf. supra), they treated the ordo equester very harshly, depriving many of them of their public horses (Liv. 43.16.1 f.). They furthermore quarrelled with the publicani, denying all of them of the contracts entered into with the previous censors (Liv. 43.16.2). The Senate supported the censors’ decision and the publicani were forced to turn to a friendly tribune of the plebs, P. Rutilius, to support their case. The situation quickly unravelled and led to an ingens tumultus (Liv. 43.16.9), resulting in the censors subsequently being charged with perduellio, a form of high treason tried before the People (iudicium populi) and representing a capital offense.^[94] Many principes civitatis, presumably senators or nobiles, supplicated on the censors’ behalf during their trial (Liv. 43.16.14). Despite this, a sizeable portion of the first few census classes voted to condemn. Eight out of the eighteen equestrian centuries and many of the prima classis found Claudius guilty. Only with difficulty did Gracchus the Elder eventually manage to persuade the others to acquit Claudius (Liv. 43.16.15 f.).

Furthermore, freedmen’s collective confinement to the urban tribes had the additional advantage of reinforcing the existing socio-political order as it played the freeborn plebs against the freed in that it endowed the former with a gratis if important marker of status differentiation and ditto sense of social superiority: a costless if effective way to divide the sub-elite electorate and reinforce the existing socio-political order – akin to setting poor whites against freed blacks in the southern United States after 1865 by subjecting the latter to continued electoral and socioeconomic discrimination and segregation. In this context, we should point to the evidence in Velleius (2.20.2) that the mass enfranchisement laws of 90/89 stipulated that the new citizens were to be enrolled in but eight tribes “for the sake of preserving the dignity of their benefactors”, viz. the numerically inferior body of ‘old’ citizens – a provision doubtlessly deeply offensive to the newly enfranchised freeborn Latin and Italian constituents as putting them on a par with the libertini in the all-important electoral and legislative popular assemblies.^[95] Freedmen, likely keen to assert their newly found liberty and social status, may have even been prone to vote in greater numbers than their freeborn compatriots.^[96]

It is certainly no coincidence that the first accounts of a politician relying on or rebuking freed voters came from around this time – freed voters had become politically notable. In 142, P. Cornelius Scipio Africanus Aemilianus (cos. 147, 134; cens. 142) allegedly had freedmen, who were not only “frequenters of the Forum” but were described as agitators and mobilisers, support his candidature for the censorship (Plut. Aemilius Paulus 38.3 f.).^[97] Soon afterwards, the consul M. Aemilius Scaurus in 115 passed a (presumably tribal) lex de libertinorum suffragiis (vir. ill. 72.5 f.: [Marcus Aemilius Scaurus] consul legem de sumptibus et libertinorum suffragiis tulit). Even if we have no further information about the scope of this law and given what we do know about the Sulpician Law on the freedman suffrage, it likely hardened the existing censorial confinement of the freedmen to the urban tribes into statute law.^[98] Probably intended both as a powerful public affirmation of senatorial policy and a precautionary measure – ‘firebrand’ Fulvius Flaccus had after all been consul in 125,^[99] raising the spectre of someone with similar political convictions making it all the way to the censorship – this measure thus terminated centuries of censorial discretion in the matter of the tribal enrolment of the freedmen. The nobility’s deep fears of the potential electoral clout of the libertini and the ensuing need to confine them to the urban tribes are, furthermore, plainly articulated in De orat. 1.38, where Cicero has Q. Mucius Scaevola (consul in 95 and together with his colleague L. Licinius Crassus author of a contentious law compelling large numbers of Italians who had illegally enrolled in the Roman census records to re-register as citizens of their home communities^[100]) say that if the censor Ti. Sempronius Gracchus (cos. 177, II 163) had not enforced (and tightened) this requirement in 169, “we should long ago have lost the Republic which, as it is, we preserve only with difficulty”.^[101]

All of the above, then, provides us with ample grounds for a re-evaluation of the events of 88: politically active freedmen were not overwhelmingly from the destitute part of the plebs urbana but could be relatively well-off and had strong ties to the Italo-Roman elite, especially those with commercial interests. This new subset of the elite had already repeatedly clashed with the senatorial elite, especially when its financial or political endeavours were under threat. As the senatorial consensus began to splinter in the Late Republic, exacerbated by the Social War and the question of the enfranchisement/suffrage of Latins and Italians, the tensions surrounding the issue of freed tribal distributions intensified, contributing to the spiralling and unprecedented political violence in the notorious year 88.

2.1 Annus horribilis 88 – Tentative Timeline and Scope of the Sulpician Laws

An approximate timeline of the events of 88 BCE must first be established before any analysis can be performed, since all the surviving narratives differ slightly. As this paper will argue, precisely when Sulpicius put forward his freedmen bill impacts on how we interpret why it was proposed and passed in the first place. As the first full-fledged Roman civil war is a crowded field,^[102] we will home in on matters relevant to the issue of freedmen.^[103]

Sometime in the autumn of 89, shortly before the consular elections (i.e., under the terms of the Sempronian Law^[104]) and while fighting still raged in parts of Italy, the Senate assigned Italy and Asia with the war against Mithradates VI Eupator of Pontus as consular provinces for 88, the latter command conveniently falling to Sulla ‘by lot’ at the outset of his tenure.^[105] Instructed by the Senate to stamp out the remnants of the Social War in and around Nola before taking the six legions encamped there to the East to fight Mithradates, Sulla left Rome early in his consulship, the administration of affairs there as well as oversight of events in northern Italy being the province of his – ideologically closely aligned – colleague, Q. Pompeius Rufus.^[106] After a relatively quiet start to the year, however, the situation in Rome significantly deteriorated when the tribune of the plebs P. Sulpicius broke with the Senate and the consuls and instead allied with the equites and the ageing but ever ambitious C. Marius.

An outstanding and popular speaker in his own right, Sulpicius had been a favourite and fervent admirer of L. Licinius Crassus (cos. 95, cens. 92/91) and M. Antonius (cos. 99 and among those killed in 87 by supporters of Cinna and Marius) as well as a close friend and supporter of Crassus’ foremost political protégé, M. Livius Drusus, tribune of the plebs in 91.^[107] In 90 and 89, he had distinguished himself as one of the many legati serving in the Social War.^[108] Cicero, who spoke with Sulpicius on several occasions during his tribunate, tells us he witnessed how he harangued the people almost daily.^[109] Probably sometime in the summer of 88, however, Sulpicius clashed violently with the equally eloquent and influential C. Iulius Caesar Strabo Vopiscus, younger brother of L. Iulius Caesar (cos. 90 and cens. 89).^[110] After serving as one of L. Appuleius Saturninus’ decemviri agris dandis attribuendis iudicandis in either 103 or 100 and being elected twice to the office of tribunus militum, Caesar Strabo held the curule aedileship in 90.^[111] In that capacity, he had been a vocal opponent of Q. Varius and his (equestrian-backed) extraordinary quaestio as well as a champion of his brother’s famous enfranchisement bill. In 89, he saw brief if meritorious service in the Social War and was probably prominent in supporting the lex Plautia iudiciaria (“with the aid of the nobles”, this law had divided the criminal court juries between senators, equites, and some first census class members of the plebs) as well as instrumental in Varius’ subsequent condemnation under his own law,^[112] actions which incurred him the hatred of the equites and C. Marius: after Rome’s surrender to Cinna and Marius late in 87, both L. and C. Caesar were among those killed in their murderous reprisals against real or alleged political opponents, their heads among those placed on the Rostra.^[113] Confident of his achievements and backed by hired street muscle, Caesar Strabo now aspired to run for the consulship (of 87) without holding the praetorship.^[114] Though once on friendly terms with Strabo and united with him in his admiration of L. Crassus, Sulpicius was determined to thwart Strabo’s illegal bid for office.^[115]

Initially, Sulpicius worked with his colleague P. Antistius to stop Strabo by legal means, “but when the dispute grew excessively fierce” and Strabo would not give up on his attempts to secure the required legal exemption from the lex Villia annalis, Sulpicius “resorted to weapons and armed force”. Still according to Asconius, the conflict between Caesar Strabo and Sulpicius thus “was a cause of civil war”.^[116] In order to stop the well-connected Caesar Strabo in his tracks and disappointed with the lack of senatorial support, Sulpicius took the radical decision to realign himself with Strabo’s foremost enemies, a choice that also put him on a collision course his former close friend and ally, the consul Q. Pompeius Rufus, as well as the latter’s colleague (and ally) in the field, L. Cornelius Sulla.^[117] First, he formed an “alternate Senate” (ἀντισύγκλητος: pro senatu, rather than “anti-senate”) composed of 600 elite equestrian iuniores (ἱππικῶν νεανίσκων: probably mostly current and prospective members of the sex suffragia) who formed a personal bodyguard as well as an alternative advisory body, further complemented by a private militia of some 3,000 swordsmen – probably hired at considerable expense since Plutarch claims that Sulpicius left behind a personal debt of no less than 3,000,000 denarii.^[118] Second, he also formed the fateful alliance with C. Marius (cos. 107, 104–100, 87), who likewise enjoyed strong equestrian support.^[119] After his disastrous alliance with the maverick tribune of the plebs L. Appuleius Saturninus and the praetor C. Servilius Glaucia in 100 losing him much political support from across the political spectrum and failing to regain his status in the Social War,^[120] Marius had lingered as a privatus and was, at least according to all the surviving sources, eager to regain his former prominence. Thanks to Diodorus Siculus, we know that both Strabo and Marius were covetous of the Mithradatic command, believing it would yield easy and rich prizes, and that the populace was divided in its support for both contenders.^[121] Had everything gone to plan, the anticipated proceeds of the Mithradatic War would, furthermore, have enabled Marius to bail out Sulpicius and provided the funding for future popular largesse. Even though the sources remain silent on the issue, Sulpicius was likely also sorely disappointed in the consuls: in return for his energetic blocking of Strabo’s ambitions and thus safeguarding of the s.c. de provinciis consularibus of 89, an outcome favourable to Sulla in particular, he may well have expected their support as he sought to complete the enfranchisement projects of his late friend Livius Drusus by legislating the enrolment of the new citizens as well as the libertini across the thirty-five old tribes. Let down in his expectations, he resolutely turned to other powerful individuals and interest groups to achieve this ambition.^[122] In other words, after pursuing a political strategy inspired by that of his late friend and political mentor Livius Drusus, viz. to achieve important reforms with the support of the Senate, Sulpicius, profoundly disillusioned in his dealings with consuls and Senate, embarked upon a radically different course of action.^[123]

Before progressing our reconstruction of relevant events in 88, it is at this stage important to restress that the enfranchisement laws of 90/89 provided for the enrolment of a very significant body of newly enfranchised Latins and Italians (excluding the so-called dediticii: those rebellious allies who had fought on until comprehensive defeat and unconditional surrender, as well as those diehards still resisting into 87, chiefly Samnites and Lucanians) in eight new tribes who were to vote only after the thirty-five ‘old’ tribes had voted first. Significantly, this envisaged enrolment of the newly enfranchised into just eight new tribes – reminiscent of the fairly similar treatment of new citizens in the late 4^th century, cf. supra – also put their wealthier segment – especially the domi nobiles – at a distinct disadvantage in the comitia centuriata since they were to content themselves with just sixteen new first class centuriae (one century of iuniores and one of seniores for every new tribe), as opposed to the seventy centuriae controlled by the ‘old’ citizens in the existing thirty-five tribes.^[124] Indeed, sometime between 241 and 218, voting procedures in the timocratic comitia centuriata had been reformed in that the centuries of the first class were reduced from 80 to 70 to achieve equal distribution across all thirty-five tribes, the geographical administrative units for the administration of conscription (dilectus) and war-taxes (tributum – abolished for Roman citizens in 167 but not so for the Latin and Italian allies: cf. App. civ. 1.7). Since every tribe now received one first class century for iuniores (those under 46) and one for seniores (those 46 and older), this reform especially favoured the first-class inhabitants of the eighteen ‘new’ tribes further removed from Rome since each area was now guaranteed equality regardless of voter turnout. The catalyst for this momentous reform, which gave greater voting weight to the periphery by literally mapping the first class centuriae onto the various regions of central Italy, had been the trials of the First Punic War (and perhaps the ensuing costly wars in Cisalpine Gaul), when wealthy voters from the first class across all tribes alike had made enormous sacrifices in terms of manpower and tributum.^[125] In the context of this digression, it is, moreover, also important to note that recent scholarship concerning voter behaviour in the period from the introduction of the secret ballot to the outbreak of civil war in 49 indicates that rural citizens of property down to approximately the fourth census class (estimated at roughly the equivalent of a farm of 20 iugera of land) constituted an ‘active centuriate electorate’ that would travel to Rome to vote in large numbers every year on the occasion of the consular and praetorian (and other, contemporary) elections in an electoral landscape that was – within ‘aristocratic’ bounds – relatively open and competitive, with a critically important role for the locally influential men in the municipia, coloniae and praefecturae across Italy.^[126] In sum: the stakes of any laws that sought to fully enfranchise large numbers of people, both freeborn and freed, and regulate their (re-)distribution across the tribes, were very high indeed and would significantly impact the political participation and clout of rural voters down to the third and even fourth census class.

At all events, after repressing Caesar Strabo and his supporters with his newfound allies, Sulpicius boldly moved to propose a range of sweeping bills, according to Livy all promulgated at the instigation of Marius. The most notorious provided for (1) the recall of those exiled under the Varian Law, the distribution among the ‘old’ thirty-five tribes (in practice the 31 existing rural tribes) of (2) all newly enfranchised citizens as well as (3) the freedmen, and (4) the appointment of C. Marius to the command against Mithradates in the capacity of proconsul (instead of the consul Sulla).^[127] Earlier in his tribunate and no doubt likewise motivated by his desire to thwart Caesar Strabo, Sulpicius had used his veto against the recall of those exiles who had not been permitted to plead their case before the quaestio Variana. Now, however, he supported their return, representing them as having been expelled by violence.^[128] By means of this manoeuvre, Marius and Sulpicius no doubt aimed at dividing the nobility and building their own support base: prominent among those who promptly returned to Rome was none other than Q. Varius himself, whose controversial tribunician law had been passed with strong equestrian support in the waning days of 91.^[129] First, the consecutive if closely aligned measures redistributing both the new citizens (i.e., the formerly Latin and Italian allies) and the libertini in all thirty-five (existing) tribes – the lex Sulpicia de libertinorum suffragiis – would realise some of the late Drusus’ ambitions.^[130] Second, and more significantly, these bills also had the additional advantages of outflanking the Caesars and offering Sulpicius and Marius tremendous future dividends in the popular assemblies (App. civ. 1.55; cf. also infra).^[131]

According to Plutarch (Sulla 8.2), Sulpicius furthermore proposed a bill providing for expulsion from the Senate of those with a debt of over 2,000 denarii. Tantamount to a public shaming of the senatorial order, this measure, too, should probably assigned to the ‘second phase’ of Sulpicius’ tribunate as it was bound to create a significant and immediate number of senatorial vacancies for Sulpicius’ and Marius’ equestrian associates to fill since the Social War had no doubt worsened the financial situation of many senators.^[132] Marius, for his part, elderly and corpulent, now made daily appearances in the Campus Martius to participate in equestrian military exercises, courting the favour of the young equites and keen to show he was still fit for purpose.^[133] That the equites strongly supported Marius’ designs on the command against Mithradates should come as no surprise: heavily invested in the affairs of Asia, they wanted a speedy resolution to the war in the East and would have been unhappy with the perceived delay by the Senate and Sulla and their ongoing inability to recover their financial losses owing to Mithradates’ occupation of the province.^[134]

The bill concerning the redistribution of the new citizens, however, immediately sparked spiralling violence between the new and old citizens, the former outnumbering the latter and the latter therefore fearing to be outvoted by the former.^[135] Advised of these alarming developments by his colleague, Sulla hurried back to Rome from Campania. Desperate to buy time to shore up support, the consuls issued an edict declaring a cessation of all public business (iustitium) for several days.^[136] As the consuls were addressing a contio near the temple of Castor and Pollux, the patron deities of the equestrian order,^[137] Sulpicius moved in with his armed militia and ordered the consuls to revoke what he argued was an illegal cessation. When they bluntly refused, the mood turned violent: whilst Pompeius and Sulla were able to make their escape (the latter fleeing into the house of Marius near the Forum), Q. Pompeius’ son (also Sulla’s son-in-law) was killed as he stood his ground and spoke his mind freely. Left with no other choice by Marius, Sulla eventually reappeared and annulled the iustitium before making off and galloping back to Nola.^[138] Sulpicius, now in complete control of the situation in Rome,^[139] lost no time in putting his aforementioned bills to the vote, including the suffrage bills and, immediately thereafter, the transfer of the province of Asia with the Mithradatic command from Sulla to Marius, duly invested with the first extraordinary proconsulship since the waning days of the Second Punic War.^[140] As this law probably also reassigned the legions encamped at Nola to Marius, the newly minted proconsul immediately sent his legatus Gratidius and two military tribunes with orders to lead this army to him. Adding insult to injury, Sulpicius also abrogated the consulship of Q. Pompeius, who ipso facto lost his membership of the Senate.^[141]

The sources are unequivocal that Sulpicius’ legislation, foremost the unprecedented law concerning the Mithradatic War, plunged Rome into its first-ever full-fledged civil war.^[142] Twice publicly humiliated by his opponents, a furious Sulla had little difficulty in convincing his army, fearful others might take their place in the lucrative Mithradatic War, to do the unthinkable and march on Rome “to deliver her from tyrants”. Joined by his colleague Q. Pompeius he took the city by force, after which they had the Senate first annul the Sulpician laws as passed illegally through violence (per vim) and during an official cessation of public business (iustitium) and next passed a series of consular laws declaring Marius, Sulpicius and some ten other ranking associates, public enemies and enforcing a reactionary new constitutional settlement.^[143]

Having set out this rough timeline, we must now establish that Sulpicius’ freedmen bill was passed very close in time to the Italian and Mithradatic bills (likely in between both measures), and not sometime earlier in Sulpicius’ tribunate, when he was more closely aligned with the consuls and the Senate and before he broke with them over their lack of support against Caesar Strabo’s illegal consular ambitions.^[144]

Livy’s Periochae records that Sulpicius introduced a whole series of laws all at once, including, at least according to the wording of the epitomiser, what seems to have been a single bill redistributing new citizens (i.e. newly enfranchised Italians) and freedmen into all the voting tribes.

“When P. Sulpicius was the tribune of the plebs, he, instigated by C. Marius, passed pernicious laws to recall exiles, to distribute the new citizens and freedmen across the tribes and to make C. Marius the leader [of the war] against Mithridates, King of Pontus” (per. 77).^[145]

Asconius, too, lists the freedmen bill as one of the reasons why Sulpicius consulum armis iure oppressus […] est (Ascon. 64C). Plutarch was also evidently aware of the connection between Sulpicius’ tribunate and issues relating to the voting issues of Italians and freedmen, though he presented it in a highly sensationalised and biased way, claiming that Sulpicius “sold the citizenship to freedmen and foreigners” (i.e. Italians) (Plut. Sulla 8). Although the issue of freedmen is entirely missing in Appian’s much more detailed account, which may have been due to a differing authorial agenda,^[146] we can safely conclude that sources that do mention the freedmen bill unanimously place it very close in time to the Italian and Mithradatic bills, and not earlier in Sulpicius’ tribunate.

Second, it is also important to note that subsequent events in 87, 84, 82/81, and 67 corroborate that the lex Sulpicia de libertinorum suffragiis was a distinct measure, and not simply a subclause of the Italian tribal (re-)distribution bill.^[147] Soon after Sulla’s departure from Italy in the early spring of 87, the renegade consul L. Cornelius Cinna (cos. 87–84) vigorously if fruitlessly tried to re-enact both Sulpicius’ tribal redistribution laws as separate measures against the violent opposition of his colleague C. Octavius, the Senate, and the old citizens, attempts that promptly triggered another full-scale civil war in Rome and parts of Italy. After political calculations had caused Cinna and his associates to delay both suffrage measures after regaining control of Rome late in 87, his death at the hands of his own troops in Ancona early in 84 and the urgent need to muster maximum support in the face of Sulla’s imminent invasion of Italy eventually had the sole consul Cn. Papirius Carbo (or a perhaps a friendly tribune of the plebs) carry both measures with the support of the Senate, albeit in tense circumstances, with his troops purposefully encamped near Rome.^[148] Perhaps in the spring of 82, engaged in a battle of life and death with his enemies in Italy and as the loyalties of many Italian communities and recruits were wavering, Sulla at long last made a compact (foedus) with the peoples of Italy, pledging that he would not deprive them of their Roman citizenship and that he would also uphold the measures of 84 concerning their enrolment in all thirty-five tribes. Conversely, as soon as he had acquired mastery of Rome and Italy and secured an unprecedented plenipotentiary dictatorship, the law of 84 ordering the redistribution of freedmen into all thirty-five tribes was cancelled along with a raft of other laws passed during the supremacy of Cinna and Carbo.^[149] Unsurprisingly, Sulla also especially targeted the equestrians as he conducted an unprecedented and comprehensive proscription in 82/81.^[150]

In December 67, finally, the newly minted tribune of the plebs C. Manilius again took up the freedmen issue as the very first action of his tribunate. Thanks to Asconius (64 f.C) we know that Manilius “just a few days after he entered his tribunate” (i.e. on 10 December 67) re-enacted one of Sulpicius’ nullified bills (64C: hanc eandem legem; 65C: legem eandem). Had Sulpicius really introduced the Italian and freedmen redistributions in a single bill, then Manilius would have been required to draft a new bill, as his bill only affected freedmen. Cassius Dio (36.42.2–43.2) further records that Manilius carried his lex de libertinorum suffragiis toward the evening of 29 December – according to Ascon. 65C the day of the Compitalia – “after bribing some of the populace” (παρασκευάσας τινὰς ἐκ τοῦ ὁμίλου). Significantly, the law provided for the enrolment of the freedmen in the tribes of their former owners as it “granted the class of freedmen the right to vote with those who had freed them” (τῷ γὰρ ἔθνει τῷ τῶν ἀπελευθέρων […] ψηφίσασθαι μετὰ τῶν ἐξελευθερωσάντων σφᾶς ἔδωκεν). We are also told that the Senate, after learning of it the following day (i.e., 1 January 66, in the consulship of M’. Aemilius Lepidus and L. Volcatius Tullus), promptly cancelled the law, doubtlessly on technical grounds,^[151] and that – other important detail – the urban plebs too were terribly angry about this measure (ἐπειδὴ τὸ πλῆθος δεινῶς ἠγανάκτει).^[152] Fearful of retribution, Manilius allegedly first vainly tried to pin the idea on M. Licinius Crassus (cos. 70 and, important detail, indefatigable champion of equestrian interests^[153]) and next secured the protection of Cn. Pompeius (cos. 70) first by allying with A. Gabinius, then one of Pompeius’ foremost protégés, and next going as far as carrying a law granting Pompeius command of the war against Tigranes and Mithradates as proconsul of Bithynia and Cilicia, very much against the will of the Senate. Thanks to Asconius 45C, where the lex Manilia is singled out as an example of “populist lunacies” (populares insanias), we furthermore know that Manilius was “supported by a gang of freedmen and slaves” as he passed “an utterly immoral law to allow freedmen the vote in all of the tribes”, and that he went as far as pursuing this aim through rioting and blockading the climb to the Capitol. The quaestor L. Domitius Ahenobarbus (cos. 54), a staunch conservative, eventually “scattered and broke through the gathering so violently that many of Manilius’ men were killed. By this action, he gave offence to the lowest ranks of the plebs and acquired great goodwill in the Senate”.^[154] Significantly, the plebs infima, the lowest echelon of the urban plebs, took offence at Domitius’ murderous repression of the humble folk ‘commissioned’ by Manilius for the legislative comitia rather than at his rabid hostility vis-à-vis the suffrage law itself. Seemingly puzzling, this can easily be explained in that the freeborn urban plebs overwhelmingly consisted of ‘old’ (i.e., pre-Social War) citizens or their immediate descendants. As such, many would have felt that Manilius would give freedmen suffrage rights superior to their own, notwithstanding their superior social status as ingenui.^[155]

After a conscious delay first by Sulla and next by ‘his’ Senate for over a decade, the all-important census of 70, conducted at the behest of the consuls Cn. Pompeius and M. Licinius Crassus, who notably for most of their tenure refused to disband their large armies encamped near Rome, returned a total of 910,000 capita civium (draftable male citizens), as opposed to the last attested pre-Social War total of 394,336 capita civium (115/114)^[156]. As the newly enfranchised consequently significantly outnumbered the old citizens in the popular assemblies, the latter constituency may therefore also have felt that Manilius’ measure would have further strengthened the hand of the former in enabling their presumably more numerous freedmen to vote in their respective rural tribes. Indeed, if we again cautiously assume freedmen made up of 5–10 % of the ‘old’ Roman citizenry, this would have amounted to c. 20,000–40,000 ‘old’ libertini. If we estimate a similar level of manumission in Italian cities, the enfranchisement of Italians would have added c. 25,000–50,000 ‘new’ libertini, making for a combined total of c. 45,000–90,000 libertini – certainly an electoral force to be reckoned with, even should we accept only a minority of them had the means or the inclination to vote. Conversely, many if not most ranking new citizens were, therefore, probably supportive of the idea of redistributing the freedmen across all thirty-five tribes. From all of this, it can be reasonably inferred that (1) already in 88, the ‘old’ citizens, rural and especially urban, would likewise have been very hostile vis-à-vis Sulpicius’ freedmen bill, too, and not just towards his bill fully enfranchising the former Latin and Italian allies, and that (2) in 88, too, the aspiring freedmen would have found more supporters among the new than among the old citizens.

2.2 Cui bono? Seeking out further Beneficiaries of the lex Sulpicia de libertinorum suffragiis

Upon determining the above timeline, several issues arising from the traditional view – that there were so many urban freedmen that Sulpicius wished to redistribute them into all thirty-five tribes to overwhelm the votes – become immediately obvious. Therefore, we need to look for other explanations, circumstantial as well as structural.

Immediately we must question why Sulpicius would have wanted to redistribute freedmen, if he was already planning on redistributing the numerous Latins and Italians. The simplest explanation would be that Sulpicius and his equally ambitious backer Marius wanted as much support as they could muster for the contentious Mithradatic bill, Marius’ most immediate objective. Indeed, the last time an attempt was made to transfer a major military command to a privatus happened in 131, when P. Cornelius Scipio Africanus Aemilianus (cos. 147, 134) tried to gain the command against Aristonicus of Pergamon. It went poorly and only two tribes voted in favour (Cic. Phil. 11.18).^[157] No doubt mindful of this precedent, Sulpicius would have wanted to secure sufficiently strong political support before attempting the Mithradatic bill.

This, then, begs the question as to how precisely Sulpicius might have gained immediate political support by legislating to spread the freedmen into all thirty-five tribes. Cassius Dio’s valuable clarification that freedmen were to be redistributed into their patrons’ tribes raises further questions – if Sulpicius simply wanted as many freedmen to vote in as many tribes as possible, then other methods would have sufficed, such as redistributing them at the tribune’s discretion or redistributing them by lot. The fact that they were being redistributed explicitly into their patrons’ tribes suggests that there was more in play than just a bid to win the freedman vote. Indeed, it is unclear why anyone, bar those radically opposed to the idea that freed slaves should have weightier votes on ideological grounds, would be against this: surely even members of the conservative senatorial elite would benefit from their freed clients having more weighty votes and being placed into their own tribes, where these votes could be more easily monitored and influenced?^[158]

Furthermore, prior to the Aemilian Law of 115, issues of freed tribal distributions had been the exclusive province of censors, who had the power to change a citizen’s tribe, whilst enrolment in the tribes by the censors was an absolute prerequisite for citizens being able to vote in both the comitia tributa and centuriata.^[159] It remains unclear whether Sulpicius’ legislation empowered him to change the tribes of the freedmen himself or whether this was to be done by the next pair of censors.^[160] If Sulpicius’ envisaged empowering himself to this undertaking, then even by a low count of the number of freedmen (cf. supra), one would imagine it would have taken some time to redistribute tens of thousands of freedmen into their patrons’ tribes. If the freed citizens needed to wait until the next census to complete their transfer of tribes, then they could hardly have been very useful in voting for the Mithradatic bill. After lifting the ban on public business and before the Italian bill was formally passed, Sulla hurriedly left for Nola. He had only just arrived at his army before the news broke that his command against Mithradates had been stripped from him by law. Plutarch’s narrative presents Sulla and Marius’ legates arriving one rapidly after another at Sulla’s camp (Plut. Sulla 8 f.).^[161] Even if the law empowered Sulpicius to conduct a redistribution immediately, the short interval between the freedmen bill and the Mithradatic bill would have meant it was neither feasible nor necessary for Sulpicius to mass-redistribute all freedmen – a small subset of freedmen, adequately motivated, would have sufficed.

Moreover, we cannot simply assume that politicians could continue to mobilise freedmen after they had been redistributed, and that freedmen would be so pleased with their redistribution that they would continue to support the politician responsible unquestioningly.^[162] As Lintott astutely concludes for Ti. Gracchus: “The political following he acquired during his tribunate […] could only be converted into a true and permanent clientela if it were bound by permanent and tangible benefits which could outweigh any other allegiance”.^[163] There is little reason why this would not hold in general.

In addition, a series of laws from the 130s to the 100s BCE, starting with the lex Gabinia of 139, had made voting secret (Cic. Leg. 3.34). While there is still debate on how, if at all, these laws changed voting outcomes,^[164] there was very little a politician could do to force their supporters to attend the vote or vote in their favour.^[165] If Sulpicius needed freedmen, or those who benefited from freedmen being redistributed, to continue to support him, he needed to continue to pass laws that would unequivocally benefit them. Moreover, the tribunate of the plebs often acted as an early stepping stone for a Roman politician’s career.^[166] It would be most short-sighted if a tribune burnt all his bridges with the most valuable voters in the comitia centuriata by the ill-considered courtship of the alleged ‘poor masses of urban freedmen’ useful only in the comitia tributa, especially if their tribal redistributions might not have taken effect before his year-long tenure as tribune was over.

In sum, the redistribution of freedmen would only provide Sulpicius with a continuously loyal novel constituency if his subsequent bills continued to benefit freedmen or those who would benefit from their redistribution across all the tribes. Sulpicius’ intent hardly amounted to an idealistic ideological statement that all citizens should have equal votes, regardless of their origins.^[167] Due to the exigency of the situation, Sulpicius certainly had more pressing and practical concerns. Since his subsequent bill was the transfer of the Mithradatic command from a sitting consul to a private citizen (political rivals at that), the freedmen bill and Mithradatic bill, despite being two separate bills, should therefore be examined as being closely interconnected: Sulpicius clearly expected that whoever would benefit from his freedmen bill must also benefit from his Mithradatic bill enough that pursuing such a contentious proposition was worth the trouble and expenditure of political capital. The fact that the parallel measure providing for the tribal redistribution of the much more numerous newly enfranchised Latin and Italian ingenui would already make for a formidable hurdle to take gives further salience to this question. In other words, who would have wanted to see Marius instead of Sulla secure the coveted Mithradatic command, and what was their connection with freed slaves?

2.3 Towards a Revised View of the lex Sulpicia de libertinorum suffragiis

Thus far, the lex Sulpicia has been largely interpreted as an attempt to mobilise the mass of urban freedmen. For example, Tatum contends that Sulpicius had already clashed with the Senate regarding the consulship bid of Caesar Strabo, though Sulpicius had no idea how seriously he had antagonised the Senate at the time. Subsequently, and unaware of any senatorial animosity, Sulpicius introduced the freedmen bill, separately though alongside the Italian bill, explicitly to appease the urban plebs. It was then that violence ensued and Sulpicius realised the extent of the senatorial hostility towards him. After this, Sulpicius turned to Marius who used his equestrian supporters to pass all of Sulpicius’ measures.^[168] If, however, Sulpicius did not realise that he had already broken with the nobiles until after he introduced his Italian bill, then it is unclear why the freedmen bill had to be introduced alongside the Italian bill, as he would not have expected his Italian bill to have faced such opposition from the nobiles that he needed significant comitial support from the urban plebs. Moreover, if Sulpicius was trying to pacify the urban plebs, then his freedmen bill would have only benefited a small subset of them. In addition, if the target really was the urban plebs, then Sulpicius could have only hoped to have gained the support of the four urban tribes, hardly enough to win him the vote. And if he already had enough support amongst the rural tribes, he need not have cared about how the urban tribes voted. Lastly, if this bill was purely a tactic to appease the urban plebs, it is then odd why Sulpicius did not drop it when it became clear that his attempt to forestall opposition from the urban plebs did not work and that the urban plebs rioted anyway. We must search for another more likely support group.

After the disastrous close to his sixth consulship in 100 BCE due to his ill-advised alliance with Saturninus and Glaucia, Marius departed for Asia in 98 BCE. According to Plutarch (Marius 31), he ostensibly went for religious reasons but, allegedly, it was so that he would not have to be present for the recall of his enemy Q. Caecilius Metellus (Numidicus) from exile and in order to stir up a major conflagration in Asia – though these supposed reasons likely rose from anti-Marian sources.^[169] Regardless, when Marius returned either in 98 or 97, the affairs of the Roman province of Asia and wider Asia Minor became very prominent in Roman politics. As praetorian proconsul of Asia in 97, Q. Mucius Scaevola (pr. 98?, cos. 95) and his consular legate P. Rutilius Rufus (cos. 105) moved to curtail and punish the abuses of the influential equestrian tax farmers. In 95, not coincidentally in Scaevola’s very consulship, these powerful equestrian interests struck back at Scaevola by securing the scandalous conviction and subsequent exile of the equally incorruptible Rutilius in the wholly equestrian quaestio repetundarum in a trial that “tore the Republic apart”. Sometime late in 92, Q. Servilius Caepio (pr. 93?, son of the homonymous consul of 106) furthermore indicted princeps Senatus M. Aemilius Scaurus (cos. 115) under the same lex Servilia (Glauciae) repetundarum for having allegedly taken bribes during an embassy to Asia (perhaps in 93/92).^[170] Following his tenure as praetor urbanus in 97, Sulla was sent to Cappadocia with the rank of pro praetore pro consule, where he reorganised the nearby kingdoms and restored Ariobarzanes to the throne of wealthy Cappadocia.^[171] After his return (in 94 or possibly as late as 93), he was charged with repetundae in an equestrian court, though he escaped the charges not because he was acquitted, but because the prosecutor C. Marcius Censorinus for some reason failed to show at the trial (Plut. Sulla 5.6).^[172] Subsequently, the outbreak of the First Mithradatic War (89–85) and the genocidal slaughter of Italo-Roman businessmen and their entire familiae by Mithradates’ agents in the province of Asia in 88 all but crashed the financial market in Rome (Cic. Manil. 19).

The 90s thus saw certain ranking conservative politicians interfere with the extensive financial interests of the Italo-Roman elite in Asia. Following the mass enfranchisements of 90/89, the Latin and Italian elite now potentially had a real say in Roman politics, provided they could be distributed in the existing tribal voting units.^[173] Since both Italian and Roman negotiatores and the like had suffered disproportionately from the Mithradatic War, they would naturally look to someone who would champion their cause to lead the war against Mithradates, both out of a desire for revenge and for the war to end quickly so that their financial activities in Asia could resume on terms dictated by a friendly Roman imperator (cf. supra p. 557).

Sulpicius likely wanted the Italo-Roman businessmen and the non-senatorial elite to support the transfer of the Mithradatic command from the consul Sulla – perceived as dallying in Campania as per the Senate’s wishes and less favourable to their interests and activities in Asia^[174] as well as a notorious and relentless commander against the Italians during the Social War^[175] – to Marius, a novus homo from Arpinum who had strong ties to these new interests. Marius was able to win multiple consular elections without support from conservative members of the senatorial aristocracy (suggesting significant support from wealthy voters in the comitia centuriata who were not beholden to the senatorial class)^[176] and had close enough bonds with commercial and financial groups that rumours circulated that he was once himself a publicanus (Diod. 34/35.38).^[177] In fact, during the Jugurthine War, it had been the Roman equestrians doing business in Africa that mobilised their networks back at Rome which voted in Marius as consul despite opposition from conservative politicians (Sall. Iug. 65.4).^[178] During the decisive battle at Vercellae in July 101, Marius on the spot (and possibly illegally) granted citizenship to no fewer than two full cohorts (c. 1,000 men) from the Latin colony of Camerinum, likely earning him the goodwill of those Latins and Italians pushing for citizenship. Furthermore, during the Social War, Marius took the field against the Italian insurgents rather reluctantly and eventually abandoned his command before the end of the conflict on the pretence of infirmity.^[179] In other words, both Italian domi nobiles and wealthy Roman businessmen had a vested interest in seeing Marius and not Sulla gain the command against Mithradates. Indeed, both Mitchell and Lintott suggest that Marius was still popular enough with the equestrians to mobilise them to vote for the lex Sulpicia de uno imperatore contra Mithridatem constituendo.^[180] Levick also adds Italians to those supporting Marius.^[181] Carney proposes, explicitly, the publicani as the support group ‘par excellence’.^[182]

In light of the above analysis, we argue that Sulpicius’ freedmen bill and, indeed, the bill concerning the newly enfranchised Latins and Italians, must have disproportionately benefited this group. When Sulpicius first promulgated his predictably contentious suffrage bills, he and Marius must have been confident that they enjoyed support from enough influential ‘old’ citizens, despite their demonstrable unpopularity with large swaths of the pre-Social War constituency. This inference is further corroborated by the fact that Sulla promptly decided to interrupt his military operations in Campania and hurried back to Rome, where, together with his colleague, he had to resort to declaring a cessation of all public business (iustitium) to buy time to shore up their own support base.^[183] Sulpicius and Marius’ foremost supporters were likely Roman equestrian and first census class businessmen^[184] who would have greatly benefited from both their newly enfranchised freeborn counterparts and their combined bodies of freedmen being redistributed into their respective tribes. These ranking ‘old’ citizens might have been relatives, friends, allies, and business partners of many of these Italian (esp. the newly enfranchised domi nobiles) commercial and agrarian elites, all of whom would have been keen to see their libertini, who would have shared similar financial and political goals, gain more political influence if these bills and the ensuing mass tribal redistributions succeeded.^[185] For Latins and Italians in particular, it would indeed have been their domi nobiles especially who would have had the means to travel to Rome, possibly with their extended retinues, to take part in the comitia, once the tribal redistributions had been affected.^[186] Now that most of the Latins and Italians had finally gained Roman citizenship (in 90/89, excepting the Italic dediticii, enfranchised only in 87), their freed clients and future freed slaves would also become Roman citizens. It would be natural for these new constituencies to wish for their freed clients to have weightier votes, especially since Sulpicius offered to transfer all freedmen into their patrons’ tribes, ‘old’ and ‘new’ alike. Wealthy freedmen would also have been motivated to attend to increase their own political standing.

Furthermore, a significant number of Roman libertini had distinguished themselves serving in the Roman army in the Social War (cf. supra p. 537). Even if the sources remain silent on any rewards or hopes dangled by the Roman authorities at the time of their enrolment into the ranks, it is not unreasonable to speculate full enfranchisement would have ranked highly on their wishlist, especially since issues of citizenship and suffrage were at the very heart of the Social and first Civil Wars in 91–88. Even if we assume these hopes had materialised before the tribunate of Sulpicius and the freedmen combatants levied in 90 had been enrolled in the tribes of their patrons, likely as a one-off measure, both these fully enfranchised freedmen veterans and the other (non-combatant) Roman freedmen would have been encouraged to militate together for a comprehensive and structural legislated equalization with the freeborn Roman citizenry.

Should the bill have allowed Sulpicius to redistribute tribes immediately, he could have reregistered as many of the Italians and freedmen who attended the vote (both at the voting of the redistribution laws itself, or perhaps immediately before the Mithradatic bill) that he could, which would have provided a powerful incentive for them to be present. While the sources state that the Italians were spread into all thirty-five tribes, Taylor’s detailed examination of these new citizens and communities shows that, when they were finally redistributed, they were placed into the thirty-one rural tribes.^[187] Therefore, Sulpicius pushed to have the Latins and Italians spread into the existing thirty-one rural tribes, where their first census class elites would consequently also be able to vote in all seventy first-class centuriae in the comitia centuriata (instead of in the sixteen newly created first-class centuriae representing the eight newly created tribes),^[188] and next put forward a freedmen bill that benefited all Roman patrons, ‘old’ and ‘new’ alike, further by placing their freed clients into their own tribes, rather than have their freed clients be restricted to the urban tribes with less useful votes. Now the mass enfranchisements of 90 and 89 were an irreversible fact, the ‘old’ Roman equites probably felt that lifting the discriminatory tribal arrangements built into the enfranchisement measures of 90/89, fully enfranchising the soon-to-be-enrolled new equestrian and first census class citizens (i.e., largely the Latin and Italian domi nobiles), would considerably strengthen their ranks/cause vis-à-vis the ‘unfriendly’ Senate, weakened as a result of the carnage of the Social War^[189] and which would not see its numbers significantly reinforced before the census of the 80s and especially Sulla’s dictatorship. Then Sulpicius introduced the Mithradatic bill in a short enough time frame that his supporters from this subset of the Italo-Roman elite, with their freed clients and retinues, would still have been in Rome to vote for another bill that ostensibly benefited and protected their interests since the revival of Marius fortunes at home and abroad was bound to yield handsome political dividends.

Even if Sulpicius probably needed to wait until the next census for the tribal redistribution of Italians and freedmen,^[190] he and the Italian/freedmen cause must thus have had support amongst many of the old citizenry (sub-)elites (i.e, equestrian and first census class), the ranking business partners or co-investors of the domi nobiles. Indeed, some of the domi nobiles, at least, were already present in Rome (as Appian records their clash with some of the old citizens) in order to persuade their citizen allies or sympathisers to support the Italian cause, even if they themselves could not yet meaningfully vote. Championing their equal enfranchisement would also have allowed Sulpicius and his most prominent associates to facilitate their continuing rise up the cursus honorum since the wealthy Italo-Roman businessmen and their well-off freed clients would have been useful especially in the all-important first census class segment of the comitia centuriata as Marius was ageing and none had expected Sulla to react the way he did after being granted refuge in Marius’ home^[191] – they obviously expected the redistribution of freedmen and Italians to have lasting benefits for their careers. Publicani, the other equestrians, domi nobiles, and freedmen were all singled out in the Commentariolum petitionis as useful voters to court in a consular election (Cic. comment. pet. 13, 29–31). Regardless, Sulpicius’ bills were short-lived. Sulla promptly marched on Rome, had Sulpicius outlawed and killed, and all his laws annulled.

In summary, a close analysis of the events of 88 BCE shows that the freedmen bill was not about creating an overwhelming new voting bloc in the popular assemblies, particularly the comitia tributa, allegedly to be filled with the indiscriminate mass of poor urban freedmen. It was not even about the sheer number of freedmen per se, but rather the type of freedmen and who would have benefited from their broader redistribution – fully enfranchised, even a small subset of the libertini with their patrons would have made for an electoral force to be reckoned with. Sulpicius was likely trying to court a certain segment of valuable voters: wealthy Italo-Romans, particularly those with business interests in Asia, who had been enfranchised only to be subject to the same sort of discrimination imposed upon the libertini in their tribal enrolment, rendering them consequently not all that different from cives sine suffragio.^[192] After falling out with the Senate and the consuls of 88 over their lacklustre support in his political struggle against Caesar Strabo, Sulpicius openly allied with Marius and championed the Latin/Italian and freedmen cause to raise maximum support for the transference of the Mithradatic command to Marius and eying their ongoing political support for future purposes. In other words, his bill should be understood not so much as a ‘class-conscious’ ideological manoeuvre, even if it fits the popularis political profile, but rather as both an ad hoc attempt to raise as much political support and mobilise as many voters as possible for a specific, hitherto wholly unprecedented, attempt at transferring military command from a consul to a private citizen, and as a structural measure to consolidate long term political support for himself and his political associates.