I was reading my father's (Philip Davis) new book "Ancient Loons: Stories David Pingree Told Me" last night (parenthetically, I entirely recommend it if you're interested in this kind of thing, though on the whole not very relevant to mythsoc) and I stopped short at the following sentence, in a quoted letter from David Pingree:
"Do you know about Virgilius Maro Grammaticus, who lived in Merovingian Gaul? ... He invented twelve different types of Latin with a grammar for each and forged 'classical' verses to exemplify them."
Since Virgilius Maro Grammaticus (not, of course, the famous poet) is not mentioned in either Okrent's "In the Land of Invented Languages" or Eco's "In Search of the Perfect Language", and since he lived 3 centuries or so before Hildegard of Bingen, whose Lingua Ignota is presented in Okrent as the earliest known invented language, I thought I should check him out.
There is a short Wikipedia article about VMG
and there is a book about him, "Wisdom, Authority, and Grammar in the 7th Century" by Vivien Law. The best short account I've found online is a review of Law's book, at
which let me quote a bit:
Virgilius makes up words; the authorities he cites are either fictitious (at least, both unknown and improbable) or endowed with works or statements that could not possibly be theirs; instead of the normal austere and tedious lists of forms, there are reminiscences, conversations, and debates. Terrentius and Galbungus are found arguing for fourteen days and nights over the vocative of ego; there are two lists (which do not match one another) of the twelve kinds of Latin, of which only one is that familiar to the average student of the language; the history of the grammarians begins with a Donatus -- which would be at least some link to the normal grammatical tradition, if Virgilius' Donatus had not come from Troy to Rome, known Romulus, and lived for a thousand years.
So I'm not sure whether Virgilius counts as a bona fide inventor of
invented languages, but he is certainly somewhere in the fantasy/language nexus.