Hello all, I found something I can’t really understand.
I know that usually, initial g before n was lost around the time of pre-classical Latin, and so the n- form is recommended over the gn- form in the dictionaries, but it looks like it’s not the case with the word “(g)nārus”, where the gn- form is the main entry rather than the n- form. Is there a rational explanation?
It’s important for me because I have to choose the “best” spelling.
I doubt there’s what you would call a “rational” explanation, and I think the very idea of the “best” spelling is a will-o-the-wisp. How could you hope to define such a thing? Still, it’s possible that the negative, ignarus, influenced the lexicographers’ decision in favor of gnarus over narus—that and the fact that narus is scarcely attested.
And so long as they’re cross-referenced, what does it matter?
I agree that “best spelling” should be first defined, and may vary depending on who’s talking. Some words have various spellings, but usually only one of them is recommended (at least when we talk about Classical Latin), for example caelum is better than coelum ou cœlum, or etiam is much better that eciam, The “best” spelling should reflect the dominant classical use and not some bad interpretation of old philologists based on bad manuscripts or folk etymology.
As for the “rational” explanation, I’m expecting a linguistic fact that explains why the gn- version is favored while it opposes to the natural form n-. And if I ask an explanation, it’s just because I like to understand why things don’t match the pattern.
I’m no expert on this sort of thing, but my understanding is that whereas the n- forms largely replaced the original gn- forms, the g tended to hang on in compounds, hence ignarus, cognatus etc. As for gnarus itself, well, narus is scarcely attested, as I noted, so if it’s “the dominant classical use“ that you’re after, gnarus will probably be it.
But such orthographical questions are of purely historical interest. Varro and Quintilian are both indispensable sources of information, as I expect you know. I recommend Bob Kaster’s work on the ancient grammarians, and above all the incomparable Leofranc Holford-Strevens on Aulus Gellius.
Excellent answer, thank you. My goal is to have the best possible speling that would be recommanded in textbooks, for example, and which form would be better to memorize first.
Sometimes it’s impossible to say with 100% certainty. What’s “attested” is sometimes not even what the original authors wrote, and it took many years of text criticism to get the texts into the states they are today. So that’s why you see in some earlier texts coll- (like collegium) whereas a later editor would write conl- (conlegium).
Fortunately, with gnarus, we do have some information. Cicero indicates that it was actually narus, not gnarus:
Noti erant et navi et nari, quibus cum IN praeponi oporteret, dulcius visum est ignotos, ignavos, ignaros dicere quam ut veritas postulabat.
We had the forms noti, navi and nari; when the syllable in was prefixed, it sounded better to say ignoti, ignavi, ignari, than to use the true form.
As the note to the Loeb (whence the translation comes) points out, Cicero is making a mistake here, thinking that these were original, when in fact the originals were gnoti, gnavi, and gnari. This does suggest that Cicero might not have even known or realized that the initial consonant was a G. The true “Classical” form perhaps ought to be narus after all, and the results in our manuscripts due to later corrections by copyists.
There are other possibilities, too, but it at least shows the difficulty in speaking definitively about spelling during the period these sound changes were occurring.
Does that mean “conlegium” is the recommanded form and what was really said in Classical Latin, or is it just an etymological spelling?
But from the sources I have, gnārus is the main entry, and nārus the secondary one, as if nārus was the rarer-almost-incorrect variant. What must we understand from that? That it’s a matter of time before the dictionaries integrate this data?
My problem is that one one side if I use forms that are hardly in the dictionaries, I won’t be understood, and on the other side, I’d like to use the best forms possible at the current state of our knowledge (otherwise we’d still be writing cætera and so on).
Early on the N in words like “conlegium” were nasalized, and you see inscriptions with it, with it dropped out, and with it doubled. But that doesn’t mean it was always pronounced one way. Consider the English word “gimme” - certainly there are times when it’s clearly pronounced “give me”, and it would be written as such, but also times where it’s written “gimme” and written as such. Same is true for would have / would’ve / would of, etc.
But from the sources I have, gnārus is the main entry, and nārus the secondary one, as if nārus was the rarer-almost-incorrect variant. What must we understand from that? That it’s a matter of time before the dictionaries integrate this data?
My problem is that one one side if I use forms that are hardly in the dictionaries, I won’t be understood, and on the other side, I’d like to use the best forms possible at the current state of our knowledge (otherwise we’d still be writing cætera and so on).
These decisions are your call to make. It really depends on what you want to do. I think “consider your audience” is a good but by no means the only maxim to use when writing.