There's a lot to unpack here, and you're asking some seriously debated questions in rapid succession. You're also asking some questions for which the orthodox opinion has a pretty clear answer, but for which the orthodox opinion is, in my view and that of others, hopelessly out of date.

In the first place, what were the "gangs" to which Syme refers? I, and a few others, take some serious issue with the description of these people as "gangs," but the term has been current for well over a century, and there's really not a particularly suitable word in English, so you still see it in publication. The use of the term "gangs" to describe the violent political units employed by certain politicians rests on the single individual of P. Clodius, whose violent activities and organization are known almost exclusively through Cicero. Cicero describes Clodius as using operae to commit his physical violence. The meaning of operae is notoriously hard to translate. It means "labor" or "work," but it can also be used to describe the service of one to another (or even to a concept). It has other meanings--leisure, care, etc. After Cicero it can come to mean "day-laborer" or "workman," but it doesn't seem to mean that yet in Cicero's lifetime. This word is only applied by Cicero to Clodius. Milo and Pompey, who similarly fielded violent units, never get the label. And while this is not quite the only term used by Cicero to refer to Clodius' violent supporters (he refers at times to mercenarii, "day-laborers" and to slaves aiding Clodius in his violence, though never with the indication of as much centralized organization as operae suggests) it is by far the most common.

Why translate this as "gangs?" The standard view on Clodius' violence is to be found in Lintott's 1968 book, Violence in Republican Rome. A somewhat expanded and updated version can be found in Tatum's 1999 book, The Patrician Tribune. Lintott and Tatum both argue that Clodius created a new type of political conflict in the late 60s or early 50s. Prior to Clodius, Lintott argues, political violence was a more or less spontaneous response by people holding a stake in whatever the matter at hand was--land distribution and Italian citizenship appear to have been particularly contentious topics in the run-up to the Social War. Clodius, the argument goes, established a base of support in the individual vici and collegia of the city. He was thus able to co-opt an existing structure of organization and simultaneously divorce his support from any particular group limited by interest or location. The argument is that Clodius was able to do this in part by his wide popularity, but especially because he revived the ludi Compitalicii, a local series of festivals celebrated in the vici of the city, which had been outlawed by a senatus consultum, and by allowing the collegia to form as they willed, through a lex that superseded a senatus consultum of several years earlier that had outlawed most collegia. The idea is that Clodius' actions allowed unprecedented access to lower-level organized associations in the city, whose members were now freed from the limits of profession that had previously held them. The argument is that the new collegia, rather than being professional associations as before, were essentially political clubs (this very term was used by the German scholar who first started describing these violent groups as gangsters) without clear purpose, and created more or less solely to act as manpower for Clodius.

There are a lot of problems here, however. In the first place, the Lintott-Tatum model does not think there is any trace of organized, "professional" political violence before the 60s. Lintott says so directly. This is questionable at best, and in my view can be easily refuted by pointing to any number of earlier incidents. As early as Ti. Gracchus we already find some evidence for semi-professional fighting forces (Asellio claims he was accompanied by several thousand clients at all times, which is echoed by Plutarch, who adds that some three hundred of these died with him). Saturninus' use of veterans is dismissed by Lintott as being the mobilization of individuals who would have benefited from his laws. Sulpicius pretty clearly seems to have had some sort of core group of violent adherents separate from his electoral base, although the tradition is quite confused as to what this constituted. And of course the Catilinarians are said by Sallust to have tried to enlist the duces multitudinum in an effort to bust their arrested members out of jail, an attempt that failed. Equally importantly, there is no evidence for the relationship with the vici and collegia that the Lintott-Tatum model proposes. It rests on only a handful of lines in Cicero, which speak of Clodius ordering the people vicatim and into collegia. These lines are all in the context of Clodius' lex de collegiis, which enrolled new collegia, recorded existing ones, and established a list of collegiati (this last easily explains vicatim, which means "by vici"--Clodius included in his list of names the residences of the enrolled). The organization of the vici under magistri and so forth is an Augustan innovation, and there is little evidence for any formal or informal organization in the Republic. Attempts by Mommsen to construct a class of collegia compitalicia and collegia tenuiorum, both of which might be connected both to the vici and to lower-class violence, have been by now pretty soundly torpedoed by Flambard and recently Bendlin. Apart from religious colleges--a class of their own--our knowledge of Republican collegia consists entirely of professional collegia.

Moreover, these collegia are absent from the ancient discussion of Clodius' violence. Cicero never refers to collegia's involvement in violence. He refers a few times to tabernarii, shopkeepers, many of whom would have been members of collegia, but simply because one is a shopkeeper and a member of a collegia does not mean that collegia were turning up in organized units to fight for Clodius. Collegia, vici, or any derived words never appear in Cicero as synonyms for operae. No other author refers to Clodius' use of collegia at all. On the rare occasions when Clodius is mentioned before his death his violence is described as being initiated by slaves--famously, the incident in the forum in 57, when Clodius borrowed a troupe of gladiators from his brother and attacked Q. Cicero. Even Cicero starts to drift away from free people towards the end of Clodius' life, mentioning the use of slaves from Clodius' country estates. Similarly, Cicero repeatedly refers to Pompey's actions in bringing his own slaves and country clients into the city to do battle.

The argument could go on for some time longer, but suffice to say that I see no good reason to accept the Lintott-Tatum model, and there seem to be significant obstacles in supposing it is correct. Instead, all the evidence I know of suggests strongly that Clodius' violence did not greatly differ from that of his predecessors. He, and they, seems to have relied on really quite small groups of "professional" (if we may really use the term) fighters--slaves, gladiators, his own freedmen, the private armed retinues with which all senators were surrounded in public--which might initiate violence, into which the people at large might join in particular incidents. This model is still observable. We frequently still find that small groups of "professional" protesters are easily able to create or provoke group activities much larger than themselves as individuals--usually with some personal interest but strictly speaking unaffiliated with the originators of the action--join in on either side. In addition, this helps explain what the texts actually say about this type of violence. Treatments throughout most of the twentieth century following Syme often refer to open "gang warfare" as breaking out in the 50s, as the gangs of Clodius, Milo, Pompey, etc. constantly battled out in random streets. Cicero paints, to a certain degree, such a picture, in an effort to depict Clodius' lawlessness. Nobody else does, and Cicero's rhetorical purpose should be obvious. In fact the number of known incidents of violence in the 50s is fairly limited, definitely countable, and largely lines up from one account to the other, even if the precise details do not. We do not see "gangs" engaging in turf wars or roaming around and fighting with each other as they bump into each other. Instead, we see organized, pre-planned political actions that generally turn into larger conflagrations on the spot once they are initiated. Clodius used his muscle mainly to disrupt courts, elections, voting, and the public speeches (contiones) of his opponents. Precise actions with definable and realizable goals, particularly the prevention of the mobilization of opposition voters or the obstruction of individuals from addressing the people and making themselves heard. These last points fit in well with Morstein-Marx's model of routing opponents from the political discourse at contiones, to which I will refer again somewhat later. And these actions are consistent in earlier incidents as well. Particularly after Saturninus most organized political violence we hear about is directed at voting assemblies or public contiones, not just general slaughter in the city.

1/2