Development of French, by request

[lethargic_man]

Adrienne Ross wrote in response to my last blog post, "At some stage you could look at the Latin based/Romance languages to see how they have developed." My first thought was 'All I've got time for, especially when I'm not as familiar with the Romance languages as the Germanic ones', but then it occurred to me I could probably write something interesting just from what I already know. So here we go.

Archaic Latin

At school I studied Latin, and at one point we did a bit of Archaic Latin; it was interesting to see the differences from Classical Latin, including endings in "-os" and "-om" instead of "-us" and "-um", which already makes it look much more like Greek, with which Latin of course shares a Proto-Indo-European ancestry. I remember at the time going away and looking up how the number one evolved from PIE into Latin: PIE oinos > oenos > Archaic Latin unos > Classical Latin unus.

At a later stage, I looked a bit further into the oldest attestations of Latin, of which a nicely concise one was then (and is apparently again now, having been in between times denounced as a forgery) the Praeneste fibula ("fibula" meaning a toga clip, and not a leg bone!), which reads (in an alphabet that still shows evidence of its Phoenician (≈ Palaeo-Hebrew) origins) "MANIOS MED FHEFHAKED NVMASIOI", meaning "Manios made me for Numasios".

This shows a number of interesting features:

• The -os (and -oi) instead of -us (and -o as referred to above.
• med for "me" intead of Classical Latin me.
• Fhefhaked as past tense of facere instead of Classical Latin fecit brings to my mind the dictum that all irregular forms are remnants of once regular ones; in this case it looks like the first syllable was reduplicated, then the version with the nonstandard vowel dropped.
• Numasios for Classical Latin Numerius shows a shift S > R. Such a shift seems not to have been uncommon in at least Indo-European languages: there was a shift Z > R in the Germanic languages after the East Germanic ones had branches off (which is why Gothic has Zs where other Germanic language has Rs), and there's been a certain amount of back and forth between these sounds even since, which is why you have erkoren in German but chosen in English, and indeed internally to English, which is why you have both "was" and "were".

There were other sound shifts as Latin evolved, such as D > L, changing Archaic Latin dingua instead lingua, which is why it's not obvious that Germanic "tongue" and Latin-derived "linguistics" come from the same root. This same sounds shift also explains why the Romans called the Greek hero Odysseus Ulysses.

Classical Latin and Vulgar Latin

This blog post is primarily about languages, not writing systems, but I couldn't resist sharing this. Before we go any further, have a look at this inscription. (You don't need to be able to understand it!)

I was astonished to discover ten or fifteen years ago that the Romans used diacritics called apices, not dissimilar to acute accents, to mark vowel lenth. Have I simply been failing to see such diacritics every time I've looked at a Roman inscription, I wondered. The next time I encountered such an inscription, after I'd already attempted to read it, I remembered about this, and had a second look, and blow me if diacritics I'd completely failed to see didn't suddenly leap into view.

Now have a second look at the above inscription; you should be able to see the apices too now.

(I further discovered during the preparation of this blog post that the letter I did not take an apex, but instead was elongated; you can see this in the word filii in the inscription.)

Right, now back to the development of the Latin language. By the Golden Age two thousand years ago, Latin had separated into two languages, Classical Latin, that was spoken by the elite, and used for writing, and Vulgar Latin, which was spoken by the man on the street, but almost never recorded in writing. As an example of the difference between them, the Classical Latin word for "cat" was felis (which gave us "feline" in English); the Vulgar Latin word was cattus. As you can tell from looking at the forms of the words in modern day Romance languages, such as chat and gato, it's Vulgar Latin that evolved into the modern day Roman languages; Classical Latin eventually died out as a spoken language.

As I alluded to above, one of the characteristics that one thinks of when one thinks of Latin is a lot of words ending "-um". However, I understand that by the Golden Age the final M was already hardly pronounced.

At school I was taught to pronounce Latin in the classical manner, parodied in 1066 And All That when Julius Caesar declaims "veni, vidi, vici" (I came, I saw, I conquered) but the Ancient Britons thought he was calling them weany, weedy and weaky. My Ph.D. supervisor, however, who is Italian, was taught to pronounce Latin much like Italian; this is also the way Church Latin is pronounced. I remember being astonished to attend a Latin love poetry reading at a friend's wedding and hear it pronounced with the sounds I had been taught, but the singsong prosody of Italian.

Proto-Romance

After the breakup of the Roman Empire, there was no longer enough travelling around Europe to keep the language together, and the regional dialects began to diverge into separate languages. The earliest attestation of French was in the Oaths of Strasbourg, in which two of the grandchildren of Charlemagne made an alliance against the third. Each made his oath in the language of the other's army, so that they could understand him; that of Louis the German read:

Pro Deo amur et pro christian poblo et nostro commun saluament, d'ist di en auant, in quant Deus sauir et podir me dunat, si saluarai eo cist meon fradre Karlo, et in adiudha et in cadhuna cosa, si cum om per dreit son fradra saluar dift, in o quid il mi altresi fazet. Et ab Ludher nul plaid nunquam prindrai qui meon uol cist meon fradre Karle in damno sit.

I find this in some ways more reminiscent of modern Spanish than French; for example the B in poblo from Latin populus; compare Spanish pueblo with modern French peuple (and of course English "people"). Possibly it's like this in modern Occitan too; I don't know.

Old and Mediaeval French

Moving on into the Old French period, the major source of the language are the troubadours, particularly Chrétien de Troyes (late twelfth century). A century before Chrétien, however, nobody was writing in French (they were all writing in Latin), and a major source on the state of French at the time was the Biblical and Talmudic commentator Rashi, who sometimes explains obscure Hebrew terms by giving their translation into French.

During this period and beforehand (and afterwards), there were shifts in sounds, which can be traced by seeing what form words and names took when they were borrowed into other languages. For example, C in Latin became CH: Late Latin cappa "cape" spawned Old French chapele (originally the place where the cape of St Martin of Tours, a holy relic in Christianity was kept), which ultimately became chapelle in Modern French. But the pronunciation of the word as brought into English by the Normans tells us that the pronunciation of CH in French at the time was like English CH, and not English SH as today. There are lots of other examples (the name Charles, to give just one example, as pronounced in English and French, compared to its German form Karl).

I once heard the following delightful (but completely erroneous) characterisation of French: The Latin-speaking Romans conquered the Celtic-speaking Gauls, who proceeded to treat Latin like a Celtic language and not pronounce a third of the letters in each word. I was surprised to learn that, though this was partly due to scribes, who were paid by how much paper their works filled up, reintroducing no longer pronounced letters like the H in homme, the terminal letters of French works were still largely pronounced well into the last millennium (as Rashi's transcriptions of French in the Hebrew alphabet^* evidence). (Which is as much detail as I can give on this.)

* Incidentally, if anyone has seen these and is wondering why he uses ש to render S rather than ס, when one might expect ש to represent /ʃ/ (English SH), apparently it was not uncommon for transcriptions to use ש for this purpose in languages that had no /ʃ/.

I was also suprised to learn that it was as late as the sixteenth century that the French first started experimenting with accents. But once again, that's as much as I can say on the subject.

Comments

[ewx]

You (or any other fans of ancient languages) might like https://www.youtube.com/watch?v=MvAEMz64O9c 'House Of The Rising Sun 1270 A.D (Cover in Old French 800-1400 A.D)'.

[lethargic_man]

Nice find!

[ewx]

Also: “Looking for a faster way to learn a language? Try historical linguistics”, which chimes with my own limited experience. For instance knowing some of the Romance language sound changes meant that Italian orthography seemed logical instead of arbitrary (having a bit of Latin vocabulary helped as well).

[lethargic_man]

Interesting. Of course, that was my approach all along with German, including going to great lengths to find^* an etymological dictionary of the German language in English at a time when my German wouldn't have been good enough to read one in German.

* I ended up using a print-on-demand service to print one dating from the turn of the last century, which was all I could find.

[lethargic_man]

Actually, I've just spotted a couple of possible mistakes in it; it renders CH as /ʃ/ and J as /ʒ/ like in modern English; but as I've pointed out above, the former was /tʃ/ at the time, and the latter as /dʒ/, as you can see from the English pronunciation of loanwords from Norman French such as "judge".

I said "possible" here, because I'm not sure off the top of my head when the shift to the softer sounds happened, possibly it had already happened by then.

your footnote about Rashi

[michael_gilbertkoplow]

Interesting and excellent, as always. In Early Yiddish, ס was almost never used for /s/ (except of course in Hebrew- or Aramaic-based words containing the letter). ש was used for both /s/ and /∫/. (I’m making assumptions about EY pronunciation that may be incorrect, such as that דש [for example] was pronounced /dVs/ and not with a /∫/.) So I'm not sure that the use of ש indicates the lack of /∫/.

Re: your footnote about Rashi

[lethargic_man]

I'm pretty sure Early Yiddish would have had a /∫/ sound, because Middle High German, from which it diverged, did. My contention (or rather, my repeating something interesting I heard from elsewhere) was not that the use of ש indicates the lack of /∫/ but rather that languages without /∫/, such as Old French, use(d) ש to represent /s/. I have no informed view on the use of ש in languages that had both sounds.