Previous installment: Substantive adjectives "Latīna, Graeca" as language names
Related Qs: Deriving adjectives from city names, Someone of someplace
I'll start by answering the questions:
- no and no
- it doesn't indicate Italian descent or culture, but a looser connection
- it's a generic solecism, except pehaps in millipede Latin
First of all, the word doesn't seem to exist in dictionaries either Classical, Medieval (Corpus Corporum), (Logeion) or Modern Latin (Del Col), (Morgan-Owens). In Koebler one finds things like (note that the /ā/ is wrong and the /ī/ a poetic license):
- italiciēnsis: an inhabitant of the ancient city of Italica in Spain and another in Italy
- italiciānus: concerning Italy as a prefecture; a cognomen
- italius: Italian (only also found in Cassel, probably a corruption); a cognomen
- but no *italiānus
Next, using Perseus Hopper, let's check whether such a formation makes sense by seeing whether there are adjectives in -iānus formed from placenames in -ia: Appiānus and Asiānus are Greek, and so looks to be the hapax(?) Indianus; Sardiniānus occurs exclusively in L&S and must be a corruption - and any way, it would be Greek as all the names relating to Sardinia. And that's everything that I've spotted. The vast majority of these results are formed to gentilic names in -ius, referred to as adoptive cognomina in the literature, used to derive affiliated family names and those of libertines. Sure enough, Leumann 1977 p.324: "used to form personal names to indicate affiliation/belonging [..] The i comes from gentile names in -ius", also treating the adoptive phenomenon. Already I conclude that
1.A: No, the word italiānus is not correctly formed to Italia, but backformed from its look-alikes in modern European languages: in Latin's highly-neutral terminology, it's either macaronic or a barbarism - or at best, Greek.
I've found this word on Vicipaedia, where it's presented in all its unadulterated glory in the name of the article on the Italian language. I've skimmed through the Vicipaedia discussion. There it's claimed to be "a modern Latin word", with such excellent authorities as the Facebook group "italianus sum, omnes latini a me alienum puto" — and a millipede. The one source of some credibility is the news website Ephemeris, where a handful of examples can indeed be found - against ~180 hits for italicus.
An anonymous contributor to Vicipaedia at the end of the section opines:
Ut est Afer "homo in Africa natus", Africus "Africae nativus",
Africanus "qui non est origine Africus sed aliquid cum Africa commune
habet"; sic etiam Italus "homo in Italia natus", Italicus "Italiae
nativus", Italianus "qui non est origine Italicus, sed aliquid commune
cum Italia habet".
('As Afer means "a person born in Africa", Africus "native to Africa", and Africanus "someone who isn't native to Africa but has something to do with it", likewise Italus means "a person born in Italy", Italicus "native to Italy", and Italianus "someone who isn't native to Italy but has something to do with it."')
This prompted me to google a bit. An English translation of Zumpt concurs:
From some of these countries, adjectives are formed with the
terminations -ensis and -anus, as Graeciensis, Hispaniensis,
Siciliensis; Africanus, Gallicanus, Germanicianus, which must be
carefully distinguished from the adjectives derived from the names of
the respective nations. Thus exercitus Hispaniensis signifies an army
stationed in Spain, but not an army consisting of Spaniards; but
spartum Hispanicum is a plant indigenous in Spain.
A conjecture from Velleius Paterculus, if correct, would nicely illustrate this point (the suffix being functionally equivalent):
nōn Hispāniēnsis nātus, sed Hispānus ('not just born in Spain but an
actual Spaniard)
A precise parallel by Martial, a Spaniard himself:
nē Rōmam[..]nōn Hispāniēnsem librum mittāmus, sed Hispānum ('or the book I send to Rome migh not just be from Spain, but Spanish')
Leumann 1977 p.324 for unknown reasons says "Āfricānus nach Āfrica, aber bezogen auf Āfrī" - I don't know in what collocations it refers to the African people instead of Africa.
One of the linked Qs mentions Coriolānus, of whom Wikipedia says:
He received his toponymic cognomen "Coriolanus" because of his
exceptional valor in a Roman siege of the Volscian city of Corioli.
Searching Google Books produces many hits of Italianus as a medieval last name/cognomen, which also accords well with the usages outlined above. I think at this point it's safe to say that:
1.B: Even if it were correctly formed, it wouldn't have been capable of expressing the intended meaning, since
2: Unlike italus or italicus, it doesn't refer to someone of Italian descent or something inherently Italian.
Speakers and writers from all over the world have been writing in Latin about Italic peoples as well as the Italian people for as long as the language existed, and all of them appear to have agreed on the inappropriateness of the word italiānus to refer to conteporary Italians or their language. When one finds that their language competence directly conflicts with that of 2300+ years' worth of speakers, both native and otherwise, it seems as suitable a situation as any to accept their judgement on the matter and wait till one's own competence adjusts itself to reflect linguistic reality. Thus, to answer the last question:
3: The term originates in two cases: a) whenever a latinist forgets or doesn't know the Latin word for "Italian", and b) whenever they're writing in intentionally macaronic, or consciously barbaric Latin, or even - conlangtin.
The primary justification for using italiānus on Vicipaedia seems to be the need to disambiguate the meanings "Italic, pertaining to the pre-Roman peoples of Italy" and "Italian, pertaining to the post-Roman people of Italy". The term whose increased use introduces a need for disambiguation from the old one is the reference to ancient Italian peoples, and it's this term that is open to coining - luckily, this is just as easy to do as disambiguating Modern English from Old English.