The question of the legionarii origo (legionaries origin) in the I century is very complex and connected with the linguistics romanization of the not Greek part of the Empire.
In the matter linguistic investigation has achieved more concrete than historical research: Devoto has recognized in the linguistics Romanity that “new waves shows when are no more the Capital or the city élites those who spread the Latin but the legions and the colonies of veterans; of them, starting from Augustus age and until the recruitment or the early barbarian contingents,” – historicians would say: until the Flavii – “those of Italian extraction was the main part”.
Apuleius, a writer of the II century, when wants show a Roman soldier speaking Latin in Greece, makes he saying a very blunder: ubi as adverb of motion to place, as then it become in Romance languages. Actually it is that vulgar Latin from which differentiated the various dialectal areas that, later, will originate the Romance languages. Now, how this vulgar Latin was considered?
Also in this case another writing (Lucianus), also from II century, can help us in the answer. The language of the soldier who, for example, uses ubi for motion to place, was designated as the language of Italians. It was, in fact, the language of italian country proletariate recruited in the legions of I century and main autor of the common spoken Latin: just that common designation, that kept until V century (the historician Priscus indicated the tongue of the Dacian Danuvius bank as ‘language of italians’), makes us sure that, in the common sense, the legionay language was characterized by those uses of sermo vulgaris of the italian proletariate, and that they spread in the Illyrican regions.
Further we can better ascertain as the liguistic base that was in charge of the Rumen language forming substantially corresponds to the language of the italian proletariate. Diffrently, for example, from Sardinia, Spain and France, romanized by the leadership of the ancient Roman stock, todays Rumania has a neo-latin language that dropped the final S, exactly as in Italian: as in the typical examaple of the Rumen DOI and Italian DUE, while Lugudorese (a Sardinian language) DUOS, Spain DOS and French DEUX. The Italian-Rumanian similitude cannot than link to the language spoken by the legionaries and traders that started the current Rumanian language forming.
Therefore the language spoken in the Illyrican regions, ancestor of the current Rumanian, was just the ‘language of Italian’ country people, that just since III century BC, shown – in certain areas – the tendence to drop the final S. While the western provinces romanization happened under the classic Latin, viceversa the Danubian provinces romanization happened under the italians Latin.
In other words: those who colonized the Danubian and Illyrican regions spoken a ‘language of italians’ that already differentiated from the vulgar language spoken in the western provinces of the Empire (Sardinia, Spain, Gallia); they spoken the ‘language of italians’, characterized already in II century by the drop of the final S from the declension.
This is enough to show how relevant was the italian (of origin) part in the legion and in the creation of the canabae (the tends, or the quarters): from both them comes the formimg of the Rumanian neolatin.
Of course not everywhere the ‘language of italians’ was the same: actually ‘language of italians’ designates the confused existing of a sermo vulgaris coming from italian country proletary people but that not always identify with it. Also the resistance of the S in northern Italy (and also in some way in ceratin cities of Southern Italy), in Gallia, Spain and probabely in Africa (where, for its Latin, is strictly connected with Spain), shows as the more ancient layer of romanization reacted with its ‘language of cities’ to the ‘language of italians’. The recruitment of italian peasants in the I century legions and their presence in any Imperial province of the Empire, in the everyday life replaced the Roman ruling class language with a vulgar language of italians. A consuetudo vulgaris that Plinius opposed versus the élites linguistic consuetudo that Plinius loved instead: the language of Italians versus the language of Latins (Latinitas).
This phenomenon is very important in the Empire history. It matured during the II century, when christianity was already spreading even in the western provinces and mainly in Africa. Christian preachers talked to the people in the people language, Irenaeus did it even in Celtic to Gallia peasants, generally they spoken in vulgar Latin language. Thus christianity used the language of italians.
At that time, in later II century, the Bible was translated; the translator was African maybe but the translation was named Itala (even some modern scholars has tried to deny this name existance). But now we can easely understand it: Itala means ‘translation in the language of italians’, while the Bible was in translation, itala was sinonimous of vulgaris.
For the same reason, the translation made by Girolamus in late IV century (initially just a a revistiting of Itala, was called Vulgata. Therefore the two terms are equivalent but in II century Itala shows as the vulgar language always is, more or less, the ‘language of italians’. Where, in late Empire, Vulgata shows the fact the language of italic legionaries of I century (and their descendants) could no more designate the general consuetudo vulgaris, because it was already becoming diffrent in the many various dialects.
But, even in V century, Priscus still was considering ‘language of italians’ the latin spoken in the Ripa Dacica, that’s the ancestor of todays Rumen.