Két Sheng has a post at Poemas del río Wang that discusses the propensity of some Renaissance writers to proclaim the stylistic similarity of their own languages to Hebrew (which of course was considered the original language of the Garden of Eden), ending as follows:
Pei Di thinks that the discovery that the translator’s own language is the closest one in its style and metaphors to the Hebrew, may have been general in this period. “If you regularly read 16th-century vernacular literature in more than one language, you will clearly see that their archaic way of speaking still used many metaphors in every vernacular. The contemporaries, however, usually knew only their own vernacular version to the extent to realize this, while the common school Latin was in fact poor in metaphors. This is why, when they discovered the richness of Hebrew figurative language, they may have felt that they found the closest relative of their own mother tongue, certainly in style, and perhaps also regarding its its origins.”
Was this idea, the stylistic closeness of the various vernaculars to Hebrew, really so widespread in the Renaissance? We want to ask for the help of the polyglot readers of Río Wang in this question. If you have ever encountered any contemporary declaration on the similarity between the Hebrew language and any Renaissance vernacular, please share it with us.
I thought it was a striking hypothesis and a good question, so I’m passing it along.
Not Renaissance, but the 1970s: “Y’a eune affiche sus l’haut d’la Crouaix, I.N.R.I., en trais langues – lé Latîn, lé Grec et l’Hébreu, la langue dé nout’ Seigneur. Lé Latîn est coumme lé Français – pas tous l’pâlent; lé Grec coumme l’Angliais – tous l’savent; et l’Hébreu coumme lé Jèrriais.” (report of a sermon preached in Saint Helier in 1971 – http://members.societe-jersiaise.org/geraint/jerriais/sermon71.html)
Of course it was Aramaic that Jesus spoke. He probably spoke Greek, too; he and Pontius Pilate could scarcely have had any other common language.
In Renaissance times, many people in Western Europe were eager to associate their lowly vernaculars with prestigious languages. Hebrew and Latin were both prestigious, for different reasons. Romance speakers could easily be satisfied that their languages came from Latin, the language of the Church, but non-Romance speakers had no such obvious linguistic ancestor to appeal to, so Hebrew, the language of the Bible and probably of the Creation, was the natural choice. Latin texts studied in school (history, oratory) were dry and unimaginative by comparison with both everyday speech and the poetic sections of the Hebrew Bible.
GJ, doesn’t the comment on the trilingual inscription on the cross mean that Latin was like French, moderately well-known, Greek, like English, universally known, and Hebrew as little known as Jèrriais?
A little later, but this is Byron (Don Juan, first canto, describing Juan’s mother):
‘She liked the English and the Hebrew tongue,
And said there was analogy between ’em;
She proved it somehow out of sacred song,
But I must leave the proofs to those who’ve seen ’em;
But this I heard her say, and can’t be wrong,
And all may think which way their judgments lean ’em,
” ‘T is strange — the Hebrew noun which means ‘I am,’
The English always used to govern d—n.”‘
Quite what the quoted sermon was saying is open to interpretation – and all we have is a newspaper report of the sermon as preached, not the actual text of the sermon. One interpretation could be that Latin was an administrative language like French, Greek was a language of trade and commerce like English, and Hebrew was a language of identity, but lesser-used, like Jèrriais. I just thought it pertinent to bring it up since there’s a chance that the “vernacular like Hebrew” image has been trundling along since the Renaissance. In fact I’d be surprised if 19th century language revivalists hadn’t trotted it out in various places.
Marie-Lucie: during the Renaissance things were a little murkier, alas: non-Catholic Romance speakers were sometimes quite hostile to the idea that their language derived from Latin: many French Protestants and Jews claimed that French derived from some language other than Latin (Hebrew, Greek, Gaulish were some of the more popular choices), for example.
Conversely, many non-Romance speaking Catholics tried to prove that their language derived from Latin: some Lithuanians, for obvious religious reasons, tried to show that their language (but *not* the East Slavic vernacular spoken by their Orthodox subjects/neighbors…) derived from, or was related to, Latin.
Anyway, back to the main question: John Davies, who revised the translation of the Bible into Welsh, wrote a Welsh grammar in 1621, in which he claimed that Welsh bore similarities to Hebrew due to the former’s great “purity” (since all languages come from Hebrew, of course). What’s interesting is that he actually quoted some genuine linguistic similarities between Welsh and Hebrew that separate them both from Latin and Greek: lack of nominal case-marking, for example.
He also showed that Hebrew and Welsh poetry are quite similar to one another.
Obviously, as a Protestant he had every motivation to find some non-Latin language to derive Welsh from, but it’s interesting that he does quote some genuine similarities between the two languages in support of his thesis.
I’ve read that Welsh Hebrew similarities handbook but I can’t read Hebrew although I did realise that the diacritical markers to indicate the Hebrew vowels weren’t include.
@Etienne: Much more striking, in my opinion, than Welsh and Hebrew’s common lack of nominal case-marking, is that they both inflect prepositions to indicate pronominal objects. Obviously Davies knew English, which I think would make the latter similarity less impressive (and if he knew any of the Romance languages, then the same would be true there); but my impression is that inflected prepositions are pretty rare outside of Celtic and Afro-Asiatic languages.
Uh, for “latter” read “former”. :-/
@Ran: actually, there is a linguist, Theo Vennemann, who has argued that the pre-Celtic language(s) of the British Isles was/were Afro-asiatic, and that the (Insular) Celtic languages (and some English dialects) owe their Afro-asiatic-like traits to this substratum.
Inflected prepositions are indeed rare in the languages of the world: however, I was surprised to learn that another language family where inflected prepositions are found is Oto-Manguean, whose member languages are typically verb-initial, like Hebrew and Modern Celtic languages: perhaps verb-initial syntax and inflected prepositions are related features of a certain language type.
Etienne: it is good that you are around to add the right details to my sometimes overgeneral statements.
Hebrew-Welsh typological resemblances:
– lack of nominal case-marking: I understand that negative similarities are not considered very decisive markers of relationship, since they could arise independently and for very different reasons;
– verb-initial word-order is quite common in the languages of the world, whether as the basic order or a “marked” order (it is quite common in Spanish, for instance, although it is not the most basic order);
– inflected prepositions: in some languages, many equivalents of locational prepositions are nouns referring to body parts, for instance “in” = literally “belly”. It is not impossible that “prepositions” arising from such nouns should have preserved noun-like inflections. This seems to be me more plausible than linking inflected prepositions to basic word order (word order is more likely to influence whether there are pre- or post-positions in a language). Howeer, I am not a typologist.
– Vennemann: I have read some of his work and attended two of his conference presentations. His work is intriguing, but in my opinion not very convincing. But I am not a specialist in either Afro-Asiatic or Celtic, and I am willing to keep an open mind.
Vennemann isn’t alone in his African-Celt theory; I have a book printed in 1995 theorizing actual the same notion. The book’s title is “Pre-Celtic Languages, The African Substratum Theory” by Ahmed & Ibrahim Ali.
I have heard the theory that the language spoken in Western Europe before the arrival of the Indo-European Celts was something akin to Berber.
Unfortunately I don’t know anything about Berber.
Prepositional pronouns are characteristic of Celtic languages. It would be interesting to determine if this or other Celtic language constructs derived from some previous language stratum.
Very informative commentary; go raibh maith agaibh!
[agaibh: inflected preposition; ‘thank you (all)’ = ‘may-there be good to-you(pl.)’]
Marie-Lucie: considering the many detailed comments you leave here, that was *quite* a compliment you paid me.
As for Vennemann, I share your impression: however, I have been assured that his work on early Germanic is much more solid.
Theories positing a strong substratum influence on Celtic have always been popular because of their highly non-Indo-European appearance: indeed it took a while before they were recognized as Indo-European at all. We know more about Gaulish, Celtiberian and other early Celtic languages today, however, and they certainly look quite “un-exotic” when compared to Latin or Greek.
I’m very strongly inclined to believe that the “exotic” appearance of Celtic languages today is primarily due to rather extreme sound changes, which left the Modern Celtic languages looking (superficially) very “un-Indo-European”, just in the same way that French looks far less “Romance”, at first glance, than Italian does. The system of initial consonant mutation, for instance, clearly derives from an earlier system of suffixes which must have been quite banal for an Indo-European language. Indeed, typologically, its genesis is very similar to that of French liaison.
My guess is that Modern Celtic inflected prepositions are due to such sound changes too: the prepositions proper have cognates in other Indo-European languages, and I think that because of various phonological reductions of preposition + pronoun combinations, some of the reduced pronouns came to look like the person-marking morphemes of the verb system, thereby triggering the creation of entire paradigms of “conjugated prepositions”.
First, thanks for everyone for a fertile discussion and the good contributions! I appreciate it very much.
Marie-Lucie and Etienne: sometimes even Catholic Romance language speakers did not like the idea that their vernacular should be a corrupted, thus less valuable version of Latin. Regarding Spanish, in late 16th Century there was even a short lived but spectacular attempt by Gregorio López Madera to prove that Spanish was one of the 72 languages of Babel, thus as old and noble as Latin – and he tried to popularize this idea in a time when the “corrupted Latin” theory was already accepted by the majority. For turning back to my original question, even the Spanish etymologist Sebastián Covarrubias prefers Hebrew etymons over Latin ones with the not so hidden agenda of proving the antiquity and thus the noblesse and dignity of his vernacular, as we also can read it in the preface of his Tesoro, cited in my original post on Río Wang. Interestingly, this same aspiration of proving the dignity of Spanish gives the deathblow to the above non-Latin theories in the turn of the 16th and 17th Century, but now dictated by the needs of proving the uninterrupted continuity between Imperial Spain and the Roman Empire. Supporting this aim, the Latin theory was just perfect.
Geraint Jennings: I agree with your interpretation of the sermon. It gives a quite precise mapping of the sociolinguistic situation of Palestine in the time of Jesus: Latin, the language of administration of the occupying powers, spoken by the few, Greek, the lingua franca of the time, and Hebrew, the only true mother tongue, the vernacular. (For truth’s sake, it is a somewhat simplified mapping of the situation, but let us not open that can of worms. 🙂 )
GJ and KS, I agree with you about the relative functions of Latin, Greek and Hebrew (or Aramaic) at the time of Jesus, but the sermon (if reported correctly) compared Latin to French, Greek to English, and Hebrew to Jerriais, in terms of relative numbers of speakers. At the time of the sermon (only a few decades ago), it was English that was the administrative language, known by the whole population, Jerriais the local vernacular, not spoken outside the island, and I am not sure of the position of French, but I don’t think that Jersey has ever been under French (as opposed to Norman) political or administrative control.
English became an official language in Jersey in 1900 – before that French was the only official language, although Jèrriais was the language of the majority of the population. The first Jersey legislation to be drafted in English was the 1929 income tax law. It was only a couple of years ago that property contracts stopped being drafted in French. My oath of office as an elected politician is in French (as is the law under which I am elected and which I am sworn to uphold). Sittings of our (officially bilingual) legislature always start in French. French political control and Francophonie are not the same thing. The report of the sermon compares Hebrew and Jèrriais without explicitly mentioning numbers of speakers, but the Jèrriais speaking congregation presumably took away the point that the indigenous languages were the closest to God (and the context was, after all, an act of worship in the vernacular).
Gerraint has a blog about Jèrriais that you can read if you click on his name. I’m guessing he also wrote some of the interesting Wikipedia article on Jèrriais (and possibly those on the other languages of the Channel Islands: Auregnais, Guernésiais and Sercquiais, too).
Thanks for the plug for the blog – and I might just squeeze in a tangential mention for my co-edited and about-to-be-published anthology of the vernacular literatures of the Channel Islands. George Métivier, the “Guernsey Burns”, states in one of the poems included in the anthology that Rabbinical writings are among his favourites. He notes further in his trilingual collection “Fantaisies Guernesiaises” (1866): “Notwithstanding a residence of thirty years in England, French and Hebrew are the languages that the author of ‘Les Fantaisies Guernesiaises’ prefers.” Incidentally, the full title of Métivier’s book is “Fantaisies Guernesiaises, dans le langage du pays, la langue de la civilisation, et celle du commerce” (to be understood as – Dgèrnésiais being the language of the homeland, French being the language of civilisation, and English being the language of commerce).
Etienne and M-L, you comments about verb-initial word order and intial consonsnat mutations accord with what I have heard of Spanish and know of Irish, but it seems remarkable to me that Celtic would deviate from IE in the matter of verb position exactly in the direction of Semitic and possibly also Berber, or that Celtic would by its own devices develop a system of intial mutations so similar, if only superficially, to Atlantic (especially Senegambian). Remarkable as in weird coincidence.
Apparently Nivkh has something similar, so I’m not saying there’s some mystical areal effect.
Supposedly when King James got his translation crew toghether and they started in on the OT, they were all amazed at how much less work it was to go from Hebrew to English than from Greek to English, and given the sentiments of the time, that must have been very momentous for them.
But it sort of reminds me of a comment that a Japanese kid made to me about learning Chinese – all you had to do was use Japanese (learned) vocabulary and English word order. The similarities are about on the same level.
Jim: it seems remarkable to me … that Celtic would by its own devices develop a system of intial mutations so similar, if only superficially, to Atlantic (especially Senegambian)
It seems well-established that the Celtic consonant mutations are not terribly ancient (apparently they did not exist in Old Irish), and they are triggered by the shapes of the preceding grammatical morphemes, such as possessives, at least some of which can be reconstructed to PIE. Etienne mentions that the phenomenon is comparable to that of French liaison (where the last consonant of a word, lost at the end, shows up if it precedes a vowel in the next word): the regular plural suffix -s, which only shows up (as [z]) before a vowel, shows up at the beginning of the cognate nouns in French-based creoles. Another example concerns the “mutating” prefixes of some Uto-Aztecan and Philippine languages, which cause a modification of the initial of the base word (eg nasalization, spirantization), just as in Celtic. In all those cases, the basic shape of the triggering morpheme (attested in the same language or recoverable through morphophonemic analysis or comparative study) is the factor responsible for the “mutation”.
Jim: Supposedly when King James got his translation crew toghether and they started in on the OT, they were all amazed at how much less work it was to go from Hebrew to English than from Greek to English
I wonder where that comes from.
it seems remarkable to me that Celtic would deviate from IE in the matter of verb position exactly in the direction of Semitic
Not so remarkable. There are, after all, only a limited number of possibilities for verb position. There are much more remarkable coincidences in language (English bad = Persian bad is one of my favorites).
it sort of reminds me of a comment that a Japanese kid made to me about learning Chinese – all you had to do was use Japanese (learned) vocabulary and English word order
That poor kid. In Mandarin word order is often radically different from English. Something as simple as “I put my watch on the table” is literally “I grab (ba3) my watch put the table on.”
‘There are, after all, only a limited number of possibilities for verb position.”
Yes. But statistically that is not a very likely way to go.
“That poor kid. In Mandarin word order is often radically different from English. ‘
No shit. But I thnk the kid saw SVO and stopped about there.
“It seems well-established that the Celtic consonant mutations are not terribly ancient (apparently they did not exist in Old Irish), ”
I thought Thurneyson formulated rules for them. In fact he formulated rules for one that didn’t relly exist, gemination, but that showed up in the orthography. So I guess that’s not much proof then that they existed. As for the age of the mutations, I think the jury is probably our. Old itish went through some major ructions in the century the country Christianized, with Old Irish only gradually supplanting other varieties of spech, however close or unlike those might have been (Galeoin etc.)
“Another example concerns the “mutating” prefixes of some Uto-Aztecan and Philippine languages, which cause a modification of the initial of the base word (eg nasalization, spirantization), just as in Celtic.”
Yeah. I am a litle familiar with that in Philippine languages. So it all comes down to a normal but rare process that just coincidentally gives this areally similar result. Sounds plausible to me.
“I wonder where that comes from. ”
I think it was pretty superficial. The word order was more similar in Hebrew and English maybe, and then there’s the preference for nominalization they supposedly share.
Jim, Marie-Lucie: you’re both right. “Classical” Old Irish, of the ninth century, certainly did have Mutations, only some of which could be represented in the orthography. “Early Old Irish” (fifth/sixth centuries) lacked them: the nominal and verbal endings whose fall caused the birth of mutation were still present at that stage of the history of the language. The continental Celtic languages, likewise, have preserved inherited Indo-European morphology as well as Latin or Greek and thus do not exhibit Mutation.
A language which is creating a system of very Celtic-like mutations today is, in fact, Spanish: in phonologically conservative varieties of the language the opposition between MI DEDO/ MIS DEDOS “My finger/My fingers” is an opposition between singular zero and plural /s/. Now, Spanish also has a rule whereby intervocalic voiced stops are realized as fricatives, and as stops in other positions: crucially, this rule is not limited by word boundaries: the initial phoneme of DEDO “finger”, in isolation, is realized as a stop, but as a fricative in MI DEDO.
In more innovative forms of Spanish, final /s/ is weakened to /h/ and sometimes lost altogether: in which case the opposition between singular and plural MI DEDO/MI(S) DEDO(S) is carried by the initial phoneme of DEDO, which is a fricative in the singular and a stop in the plural.
This is much closer to Celtic Mutation than to French liaison, in terms of diachrony: the reason that the Welsh (invariable) definite article Y(R) causes a weakening (mutation) of the initial consonant of a following feminine noun, but *not* of a following masculine one, is because this weakening originally took place when a consonant was intervocalic, and when the feminine form of the article ended in /a/ (hence caused the mutation) and the masculine in /os/ (hence did not): after /a/, /os/ and other endings were lost, what had been a purely allophonic matter became phonemic, with the newly-born mutations maintaining the distinctions once expressed by the endings.
Oh, and I may have solved my own problem (about conjugated prepositions and VSO): VSO and SVO languages would indeed be likelier to create Mutations and conjugated prepositions than SOV ones. SOV languages typically use postpositions and suffixation instead of prefixation: hence a pronoun + postposition would be unlikely to be re-analyzed as an instance of inflection, since prefixed inflection is incompatible with the ‘typical’ SOV template. In a language with both prefixation and suffixation, as well as prepositions, however, such as VSO and SVO ones, a preposition + pronoun combination would be liable to be perceived as an inflected preposition, especially if (as in Indo-European languages) person-marking morphology elsewhere (such as on the verb) is suffixed. Furthermore, Mutation being rather prefix-like, the re-analysis of the (word-initial)allophony caused by (lost) endings as morphology would be very unlikely to occur in a purely suffixing language.
Etienne, very interesting and pertinent example about colloquial Spanish. This is the type of thing that has happened also in some Uto-Aztecan.
“Inflected prepositions”: short, unstressed morphemes can easily merge with other such morphemes, as in English don’t from do not, or French au, from al, ultimately from à le (where the preposition has merged with the following article). French uses a stressed pronoun after a preposition (à moi, à toi), but if the earlier norm had been to use an unstressed pronoun (as in English to me), “à me, à te” could have evolved into the “inflected prepositions” ame, ate.
Latin had the mildly inflected prepositions mecum, tecum, nobiscum, vobiscum for the expected cum me, cum te, etc. The singular ones got another cum placed in front in Spanish, leading after the usual sound-changes to comigo and contigo respectively; comigo was mildly rationalized to conmigo later.
Oops, didn’t mean to save.
There is also consigo and the archaic connosco [sic], convosco.
JC: Latin had the mildly inflected prepositions mecum, tecum, nobiscum, vobiscum for the expected cum me, cum te, etc.
These forms could be said to be additional case forms of the pronouns rather than inflected propositions. In any case, they appear to be survivals from Old Latin, with -cum ‘with’ comparable to -que ‘and’ (as in “Senatus Populusque Romanus) (and these two postposed elements may have the same origin). Cum must have gradually become independent, able to be used before a noun and thereby becoming a preposition (as in cum grano salis), except in those old complex pronominal forms, so that at a later date speakers felt the need to use cum before those now irregular forms in order to make them conform to the new prepositional structure. In turn, the new preposition became prefixed to the complex pronoun, the “suffix” deriving from the original -cum having lost its identify.