So this is a complicated question because of the nature of identity claims, even more than the massive time span during which people identified as "Roman" in various ways, and Romans were particularly bad at sorting out their political and ethnic ideas, for instance by designating a certain package of civil rights as "Latin" - why a package of legal rights would come designated in ethnic terms is hard to grasp for Americans, at least, that have a relatively strong sense of legal citizenship. People have different identities they deploy for different reasons in different circumstances. When asked who you are you might answer by naming your citizenship (if you were, perhaps being interrogated by ICE in certain parts of the USA recently), or you might answer with your familial relationship (visiting someone in the hospital) or your occupation in some other circumstance, etc. For Romans, where "Roman" was both an ethnicity and a citizenship, when we ask who a Roman is and he (since we don't have any female writers, most likely, left for us to ask) says "Roman" we then have to sort out in what sense he means he's Roman. But sometimes its perfectly clear. Accius, a contemporary of Cato the Elder, said "Homo Romanus natus in Latio" - "I am a Roman born in Latium." That's a pretty explicit rejection of the ethnic term "Latin." Accius is just one guy though, and other people had different ideas of citizenship and ethnicity. We'll come back to this in a moment.

As I said being “Roman” meant something for a very long period of time, and it didn’t always mean the same thing. I’m a Late Republican/Julio-Claudian guy and most of my comments are restricted to that period, but we’ll start with a brief sketch of the expansion of Roman citizenship.

I am very skeptical about any positive statements regarding Roman history and culture before Punic 2. I am pretty sure, whatever happened around Rome’s founding and the development of republican government, it wasn’t what the sources say happened. But it seems Roman citizenship was connected with the group of people in the city much like any Mediterranean citizenship. Rome was more accepting than some states, say, Athens: Latins for instance (ethnically similar) could become Roman by relocating to Rome; individuals or communities might become Roman by legislation. Rome also developed “levels” of citizenship such as the citizenship sine suffragio – citizenship without the vote. Through the wars with Carthage Rome extended its citizenship in these forms, but began prefer managing the Italian peninsula through treaties, and by the mid 2nd century the practice had ceased. Some scholars think this closing of the citizenship was one of the causes of the Social War of 91-88 BCE. I’m not sure about this, but the fact is in 90 and 89 Rome passed a lex Julia and a lex Plautia Papiria that extended (or imposed) Roman citizenship to (or on) first Rome’s loyal allies, and then Rome’s pacified former allies. There was another broad grant of citizenship sometime around 44 BCE, when probably Julius Caesar promoted Cisalpine Gaul from a province to Italy proper and this standing was probably confirmed by the ratification of Caesar’s acts after his assassination. At the end of the Civil War, Augustus established a number of veteran colonies in the provinces that extended citizenship outside the Italian peninsula, and in 48 CE the emperor Claudius successfully argued in the Senate that the elite of the Gallic provinces (some already citizens) should be allowed to participate in Roman government (well, he didn’t use the word Roman; he used Italian, which is a significant point for thinking about ethnicity) as members of the Senate. That’s the temporal extent of my own study, but it’s important to note the Antonine Constitution of 212, that extended citizenship (in a different form) to all free men in the empire, and of course those East Romans who insisted on calling themselves Romans (legitimately in my view) until 1453 CE. Various other people have asserted the claim to Roman, for instance Mehmet II is sometimes associated with the title Imperator, the Tzars, the Kaisars, etc.

The fact that Roman citizenship was legal does emphatically not mean it lacked an ethnic component! We saw earlier Accius’ claim to the name “Roman”; that claim was made in the context not of claiming any legal right or exception – it was made to excuse his poor grasp of the Greek language. He’s saying he doesn’t know Greek, he’s a Roman and a rustic one at that: that is a claim about ethnicity. In the first-century foundation narratives of Livy and Cicero we also see elements of ethnicity as components that justify participation in political life at Rome. Cicero describes the importance of Latinity – Tarquin is an acceptable king, despite being a foreigner, because among other things he Latinizes his name. Livy’s Sabines are acceptable because their children and grandchildren (by rape) are Romans through their fathers, and so there is the connection of blood for participation in government. Further, even though the Roman citizenship as some writers conceived of it in the first century BCE could accommodate diverse ethnicities if they sufficiently acculturated, there was still a tension between the legal and ethnic requirements for citizen status. The rebels of the Social War made a… well, if not quite unique, a rare thing in the history of coinage and struck a number of series of bilingual coins! They declared their state Italia in Latin and in Oscan. When these rebels were eventually subsumed into the Roman state, we find historians like Appian struggling somehow to distinguish this group and their descendents from “real” Romans (Appian calls them “Old” and “New” citizens). Well into the imperial period ethnic origin remained something to exploit positively or negatively; Farney has a very good book on Ethnic Identity (2007 I think) that shows how Romans deployed different ethnic stereotypes – Sabines (such as the Claudian clan) could be wise and stoic or ostentatious; Etruscans could be fat and soft, Marci unbeatable in battle. These ethnicities could be signaled by names among other things, such as Sabinus (The Sabine) or Vitellius (Mr. Italy in Oscan, I think), and Farney notes complaints that there were a lot of people claiming to be Sabine that apparently weren’t. Now, if you’re defining all these different kinds of “other” you’re defining them against a normal, which would be Roman, “real” Roman because all these other guys also have the legal standing of Roman.

Cicero seems to be the first guy who tried to tackle this problem directly. In probably 55 BCE, so about 40 years after the Social War (in which Cicero fought, on the Roman side) and the inclusion of the allies in the Roman Citizenship, he set out to write his dialogues de Re Publica and de Legibus; in this second work he proposes a model of citizenship in which each rural citizen (i.e. citizens not from the physical location Rome) have two fatherlands; one of nature, to which he owes his affection, and one of law, to which he owes his duty. As far as I know this is the first attempt to divorce citizenship entirely from ethnicity in the Classical tradition, though I’m not as up on Greek models of citizenship. Cicero knew this was a controversial position when he wrote it because in his dialogue his friend Atticus has conniptions: How can anyone have two homelands? Does that mean Cicero is not a loyal citizen of Rome? Are you saying Cato the Elder (who was born at Tusculum) not really Roman? WTF MATE? Cicero won the argument and talked Atticus around to his point of view (well, he would, he wrote the damn book).

Place of origin continued to matter throughout the Julio-Claudian period. Claudius was famously born not in Rome or even in Italy but in Gaul, and as you see from reading the Apocolocyntosis that was grounds for mockery and for attempts to de-legitimize his rule. Trajan I think had long-standing family connections in Spain, although his family was originally Italian. All the families that produced emperors I’m familiar with I think had Italian origins, even if they had relocated. The first purely provincial emperor I’m aware of is Severus, though that is well outside of my period, and maybe I don’t know his family background as well as I think. I’ve been caught making that kind of mistake before.

So, there you go, I guess. Roman was a legal citizenship, but it was also an ethnicity, at least for a time, and it’s hard to extricate the two ideas, and Cicero might have been the first one to try.