Sunday, May 21, 2017

Latin-speaking Muslims in medieval Africa

In the Middle Ages as today, Christians and Jews regularly called God "Allah" when speaking Arabic, just as Muslims did . It is perhaps not as well known that the converse was often also true: from a very early period, North African Muslims called God "Deus" when speaking Latin. This can clearly be seen on the 8th century Umayyad coins of Tunisia and Spain, which include statements such as:
  • Non deus nisi Deus solus - There is no god but God alone (لا إله إلا الله)
  • Deus magnus omnium creator - God is great, the creator of all things (الله أكبر خالق كل شيء)

I had always assumed it more or less stopped there, as Latin-speaking Muslims shifted to Arabic. But in the towns of southern Tunisia, the former Bilad ul-Jarid, Latin was still being spoken well into the 12th century. In his recent book La langue berbère au Maghreb médiéval (p. 313), Mohamed Meouak uncovers a short recorded example of spoken African Latin from between these two periods, which otherwise seems to have escaped notice so far.

The 11th-century Ibadi history of Abu Zakariyya al-Warjlani, he gives a brief biography of the Rustamid governor Abu Ubayda Abd al-Hamid al-Jannawni (d. 826), who lived in the Nafusa Mountains of northwestern Libya. Before assuming his position, this future governor swore an oath:

Bi-llaahi (by God) in Arabic, and bar diyuu in town-language (بالحضرية), and abiikyush in Berber, I shall entrust the Muslims' affairs only to a person who says: "I am only a weak being, I am only a weak being."
In al-Shammakhi's later retelling, the languages are named as Arabic, Ajami, and Berber (بلغة العرب وبلغة العجم وبلغة البربر). As Mohamed Meouak correctly though hesitantly notes, diyuu must be Deo; he leaves bar uninterpreted, but it is equally clearly Latin per, making the expression an exact translation of Arabic bi-llaahi. The Berber form is probably somewhat miscopied, but seems to include the medieval Berber word for God, Yuc / Yakuc.

The earliest Romance text is the Old French part of the Oaths of Strasbourg, made in 842 and opening Pro Deo amur... "for the love of God". The Ibadi phrase recorded above curiously echoes this, although it predates it by several decades.


Imed Adel said...

I really wonder if Latin has any influence on the Gafṣi dialect today.

And, if African Latin survived until today, how would it sound? Would it sound like Romance languages?

Anís del moro said...

You can seemingly find another version of this oath in Mohammed Umadi's article on Berber religious terminology in Ibadite manuscripts (2003?, p. 13), where "by God" in Berber is recorded as سـ أيكوش.

David Marjanović said...

Would it sound like Romance languages?

Perhaps as much as Romanian does.

Matthew Smith said...

It would be one of the Romance languages if it still existed today. Bear in mind, they all sound very different; Portuguese sounds very different to Spanish, despite looking very similar.

Did these Berbers actually speak Latin, though, or use a few Latin phrases like the ones quoted, in their normal Berber speech as a remnant of when they were Christians?

Tom Dawkes said...

Considering how different French and Italian are from each other, particularly in sound, and how much closer Italian is to Latin compared with French, it's only a matter of speculation what a North African Romance would have become, though on the basis of Maghrebi Arabic, with its reduced forms such as /ktæb/ for كِتاب/ kitāb/ and /klæm/ for كَلام /kalām/, it might have echoed French developments, such as droit from DIRECTUM ; moitié from MEDIETATEM; or vergogen form VERECUMDIAM

PhoeniX said...

@Abu Ilyas:

Well that makes more sense: The preposition s- is 'with (instrumental)' in Berber.

And there the transcription makes more sense ʔykwš is a pretty good transcription for *uyakuš

Anís del moro said...

Yeap. Umadi's version makes more sense... in many senses. Unfortunately, it is not clear which manuscript he is referring to, or even whether he is quoting or just paraphrasing from it.

In any case, it seems to be Abderrahman Ayoub, editor of Al-Warjalani's work (Tunis, 1985), who should be credited for his rendition of the text as "par Dieu" (cf., p. 129, n. 488).

Anís del moro said...

Or rather Ismail Al-Arabi, sorry, responsible for the Algerian (1979) and Lebanese subsequent editions.

Anís del moro said...

وكلمة «بريديو» قد تكون تحريفا من النساخ بزيادة الراء [كذا] لكلمة «ديو» اللاتينية فتكون‫ ‏"Par Dieu".
ونحن نتذكر أن اللغة اللاتينية كانت لا تزال مستعملة في بعض الأوساط بشيوع في المدن في هذا العهد.

(كتاب سير الأئمة وأخبارهم، المعروف بتاريخ ابي زكرياء، ت. إسماعيل العربي، المكتبة الوطنية، 1979، ص. *)

Lameen Souag الأمين سواق said...

What would African Romance sound like? No way to know, but I'd try applying the same sound changes as happened in Tunisian Arabic, or Djerbi, to Vulgar Latin. Maybe something like "Bŭṛṛay bin safər kumḏu syăn falabăn li ʕžăm di Tunəs lukan nun dunirən ʕṛăb."

Abu Ilyás: Thanks for tracking that down. U Madi is a smart guy, but as far as I know he's not trained in text edition. I suspect he just substituted his own reconstruction of what it said (certainly there's no way the original text had سـ أيكوش written with a non-final sin separated from the next word). Time to dig up Abderrahman Ayoub's edition, I guess...

Etienne said...

As a fictional scientist might have put it (had he existed and been interested in historical linguistics): fascinating.

A Romance scholar once wrote that the story of the interaction between Romance languages and Arabic was yet to be written: this is just as true today as it was when he wrote that observation. This sample of African Latin definitely should be part of the story.

Imed Adel: Considering how dissimilar various Romance languages are from one another today, phonologically, I suspect that, if a Romance language had emerged and survived to the present day in North Africa, it would be quite distinctive today, phonologically. We do know that North African Latin had preserved final /s/, so that if we could jump aboard a Wellsian time machine and go back a thousand years (or so...) and hear what this Spoken Tunisian Latin was like, I suspect it would sound more like Spanish or Sardinian than Italian.

As for its influence upon Arabic (and Berber?): We should not make the mistake of thinking that such influence will be confined to wherever Latin lasted the longest in North Africa. Within Romance itself, many words of Gaulish origin have spread far beyond where Gaulish was ever spoken.

Mathew Smith: since the oath was in Arabic and Berber as well as in this third language, it seems more reasonable to assume that this was indeed a separate language (i.e. Latin/Romance) and not a Latin-flavored Christian lect of Arabic or Berber. Doubly so considering that we have good evidence that Latin remained in existence as a spoken language in North Africa long after the date when this oath was sworn...

Lameen: this sample indeed predates the Strasburg Oaths, but there exists one possible sample of Romance which predates both:

(Have a look under "first sample", and references given).

Tom Dawkes said...

For a possible influence of Latin on Berber, with Latin IPSE/IPSA influencing the development of Berber ta-/tha-/tsa- see
Eduardo Blasco Ferrer "Tipología e ricostruzione: una nota su latino IPSA e berbero ta" (1989)
[text in Italian]

Unknown said...

Martin Posthumus, a conlanger (language inventor) imagined such a descendant of North African Romance which he named "Tunisian" :
The point of divergence from our timeline seems to be a Norman invasion in 1135 CE, reviving the indigenous Christian community.