Hacker News Reader

Comment by jacquesm | Hacker News Reader

The article doesn't seem to mention this at all, but Latin definitely lives: it's called modern day Romanian. There are many differences but if Latin survived anywhere into modern times it would have to be Romania. I had some Latin in high school and that came in surprisingly handy, much more so than in other Latin derived languages.

View on HN

Sardinian is considered to be closer to Latin than Romanian, and arguably Italian and Spanish are closer to Latin too.

According to a book that I haven't read[1], but is cited by Wikipedia[2], "A 1949 study by Italian-American linguist Mario Pei, analyzing the degree of difference from a language's parent (Latin, in the case of Romance languages) by comparing phonology, inflection, syntax, vocabulary, and intonation, indicated the following percentages (the higher the percentage, the greater the distance from Latin): Sardinian 8%, Italian 12%, Spanish 20%, Romanian 23.5%, Occitan 25%, Portuguese 31%, and French 44%."

[1] Pei, Mario (1949). Story of Language. ISBN 03-9700-400-1 [2] http://en.wikipedia.org/wiki/Sardinian_language

Looking at an excerpt in another wikipedia entry, it looks like the metric is based on phonology only:

> Demonstrates a comparative statistical method for determining the extent of change from the Latin for the free and checked accented vowels of French, Spanish, Italian, Portuguese, Rumanian, Old Provençal, and Logudorese Sardinian. By assigning 3½ change points per vowel (with 2 points for diphthongization, 1 point for modification in vowel quantity, ½ point for changes due to nasalization, palatalization or umlaut, and −½ point for failure to effect a normal change), there is a maximum of 77 change points for free and checked stressed vowel sounds (11×2×3½=77). According to this system (illustrated by seven charts at the end of the article), the percentage of change is greatest in French (44%) and least in Italian (12%) and Sardinian (8%). Prof. Pei suggests that this statistical method be extended not only to all other phonological, but also to all morphological and syntactical, phenomena.

http://en.wikipedia.org/wiki/Classification_of_Romance_langu...

Sounds rather arbitrary. Linguistics has come a long way since the 1940s, so I would take that one with a big grain of salt without confirmation from more modern references.

That said, as another comment in this subthread says, Latin survives in all Romance languages. There's no objective basis in claiming that Romanian is the one true descendant.

I have heard similar anecdotal evidence on Italian, and it's linguistic similarity to Spanish would support that. I'm a little surprised that Portuguese strayed so far.

Portuguese has some influence from French in terms of character set as the first King was of French origin.

At the time of the country origin, Galician, Catalan, Spanish, Asturian, Arab were all spoken in the penisula, which might have contributed to such deviation.

At its roots Portuguese came from Galician mixed with some Arab.

Using your way of putting it I think all Romance languages have a legit claim to "being Latin". [I seem to recall reading that there were time periods in which all the local languages were still casually called "Latin" while being quite distinct.] Some are more conservative than others in terms of preserved phonetics or grammar etc. But none of them "are Latin", and yet all descended from it quite naturally. Is Romanian, which has had some changes over centuries just as any other, any "more Latin" or "less Latin" than Italian, Spanish, Portuguese -- each of these having unique and sometimes overlapping changes relative to Latin?

Take some trivial word like "bene". The Italians say "bene". Their spelling is the same, so OK, it superficially resembles Latin when it's written out. But the Spaniards say "bien". That diphthong "ie" looks like a radical change but IIRC it's actually capturing the fact that the length of the Latin word's first vowel is different from the 2nd one. Less Latin? Who is to say, really? Then the Portuguese say "bem", which looks ridiculous with that M in there ... until you realize that nasal consonants come and go, they change into each other in certain positions, is it really "less Latin"? Would it really be shocking if some Roman pronounced an M in some position we think of today as having Ns? We don't have recordings or time machines so we can't say if they did.

At some point in my own amateurish study of Romance languages I realized this sort of thing and gave up trying to classify as "less Latin" and "more Latin". If you want to make a statement about who's speaking Latin today, the best way to capture it is that they all are, really (and at the same time, they are not).

I think you're mostly right, though there is a notion of a "more conservative" or "less conservative" pronunciation.

For example, surely we could see Spanish "luna" as just closer to Latin "lūna" than the Portuguese "lua" is, and Spanish "consonante" as closer to "cōnsonāntem" than Portuguese "consoante" is. (But that doesn't make Spanish as a whole closer to Latin than Portuguese as a whole... but it's possible to imagine a language typically being more conservative in some respect, having fewer sound changes applied and so on, than another language.)

Yes, I am aware of this. Even that strikes me as the wrong metric. For instance, AFAIK standard Italian is mostly conservative due to standardization efforts. I had known that it was largely based on the Tuscan dialect, but when I saw what actual Tuscans do to their consonants I couldn't believe it. Though this conservatism is now present in the native language of a country, it seems to have come about through somewhat artificial means, rather than a long unbroken chain to Rome.

I recall there were other such examples of a romance culture re-importing Latin words from literature, and it being fairly common to have two words with the same root, where one is conservative and another drops syllables, changes consonants, or both. Spanish is the one I know best, just to throw one out there from my head, you have "llave" and "clave", or "llano" "plano" and "plan". (All of this from the consonant + L -> ll change, then I presume a re-import from Latin happening later.)

Secondarily, who is to say some of the sound changes we see in Romance languages couldn't have somehow originated with regional or informal pronunciation of Romans? I am sure many have studied this question more than I have, but I will bet the answer here is nonzero.

Those are good observations that I hadn't considered before -- thanks!

Romanian was NOT so "Latin" as the people like to put it. Yes, cherry-picking one could find instances of astonishingly well preserved Latin traces, but the fact is that the old Romanian had only under 40% Latin-descending words and was heavily infused with words from Slavic, Turkish, Magyar, Bulgarian, German, and what not! By the way, those layers were being ousted continuously in a political effort to "re-Latinize" the Romanian language. To give you an example, the word "slobod" was a common Romanian word of Slavic origin, meaning "free" and was used broadly in various formal contexts (like administration, where "slobozie" was officially a locality free of taxes), and now its default meaning is the content that was "freed" on ejaculation, thus avoided in all but vulgar instances. For a while now Romanian proffered Latin-originated borrowings and thus become more related, but I think you could force that pretty much on any language if you try hard enough.

Funny how Latin sounds so prestigious, while Romanian (even though the word clearly claims, well, Roman ancestry) has no such bling to it. Only because I was exposed to it by a Romanian student I got to know this, otherwise Romania's image made me think they used a slavic language.

It's a weird mixture actually, there are quite a few Slavic influences, and even a bit of Greek.

http://en.wikipedia.org/wiki/Vulgar_Latin

In fairness, Latin got a lot of Greek in it in Imperial times.

Reply on news.ycombinator.com