Wednesday, March 11, 2009

Arabic (and Berber?) loanwords in southern Italy

Just came across a little monograph on Arabic and Berber loanwords in the dialects of the Basilicata (southern Italy): Sopravvivenze lessicali arabe e berbere in un'area dell'Italia meridionale, la Basilicata by Luigi Serra. Most of the loans listed are from Arabic, some quite obvious (eg taūt "coffin" < تابوت, źir "a copper or terracotta container for liquids" < زير, zammîl "big pannier with which various goods are transported on a beast of burden's back" < زنبيل), others rather less clear-cut.

Only three loans (and one placename) are claimed as from Berber. Two of them look acceptable, but all of them seem questionable, and they all refer to objects that there would have been no obvious reason to borrow terms for. It's possible that Berber influence can be found in southern Italian dialects, but this doesn't present a terribly convincing argument. Still, here they are:
  • źembr / źimbr / zimr / źimmr "billy-goat" (caprone, becco) < pan-Berber izimmər "ram", p. 39. (Looks good, but why the shift in species? - Also, see comments for an alternative Greek etymology.)
  • aččáta "big meal" (scorpacciata, mangiata, spanciata) < pan-Berber əčč "eat", p. 11. (The semantic and phonetic match are great, but the word is so short that coincidence seems hard to rule out.)
  • šéḍḍa "wing" (ala) < Zenati Berber "bird", eg Siwi ašṭiṭ, p. 26. The author mentions an alternative possibility - deriving it from Italian ascella "armpit" - that seems much more plausible.
  • Zaza (placename) < Berber azəzzu "thorny broom (plant sp.)" - not discussed in any detail (author cites Renisio), p. 41.


Anonymous said...

The first of the etymologies ("billy-goat") seems especially plausible: I am no specialist in Southern Italian dialectology, but I believe that in general, in these dialects, no native word can have an initial voiced sibilant. The shift in species doesn't bother me too greatly: borrowed terms for plants and animals frequently undergo such shifts, and are especially likely to be borrowed.

What I wonder is whether the term entered Southern Italian directly from Berber, or whether it entered via Arabic (possibly a now-extinct Berber-influenced variety, unless the word is found in North African varieties of Arabic).

bulbul said...

That's funny, according to Aquilina, Maltese zanbil is a loan from Berber "azenbil". But then again, Aquilina also cites bilħaqq as a loan from Berber :)
As for 'zembr' etc., Cortelazzo's/Marcato's Dizionario etimologico dei dialetti italiani (p. 476 "zìmbaru") gives an alternative source - Greek 'χίμαρος' = "bocco giovane", Liddell and Scott's "he-goat".

Lameen Souag said...

Nice find Bulbul - that looks like a good explanation, especially given the location. Aquilina's claims on Berber loans are certainly unreliable; Borg's dictionary of Cypriot Maronite Arabic gives relevant cognates for زنبيل both within Arabic and in other Semitic languages, like Syriac zabbīlā "basket".

Rhaeticus said...

This is the first time I commment on this blog, so hello everybody!

I am actually from the opposite end of Italy and I am only an amateur linguist, so this is neither a native-speaker comment nor an academic opinion, but here it goes:

(1) zìmaru from Greek χίμαρος fits with everything I know of Greek borrowings in Southern Italy. This and Cortelazzos/Marcato's opinion are enough to settle the case for me.

(2) In the claim about aččáta, -ata is interpreted as the common participial suffix meaning a one-time action; an initial a- in feminine names can never yield conclusive evidence because it merges with the final -a of the articles; this leaves us with one common sound. "Big meal"? No big deal. There is simply no ground for the claim. Usually, -ata doesn't attach itself to foreign borrowings coming out of nowhere, but to recognizable elements already functioning in the language (see for instance one of the given translations: s-corp-acci-ata, "an action which causes your fat body to explode"). If indeed aččáta is to be analyzed as ačč-áta (first impression might be deceiving!), then its first part has been shortened beyond recognition, and nothing more can be said.

(3) Latin axilla < *ag-sl-ill-a may be the basis of the Romance words for "armpit", but it is thought to be a diminutive of āla < *ag-sl-a from the same root as agō (in this case, "to push"). Its original meaning (unattested, as far as I know) would then be "little wing" and šẹ́ḍda, certainly not from Italian but possibly directly from Latin, would be a case of conservation of an archaic meaning. This has to be proven, of course, but sounds hundreds of times more probable then the Berber derivation. In addition, while the change ll > ̣ḍd is commonplace in the region (dot under ̣d means retroflection), I will believe that an Arabic/Berber emphatic can change to a Southern Italian retroflex when I see real evidence.

(4) All I can say is that Zaza is also a surname, based around Bari, not far away from Basilicata. Look at this map and don't be deceived by the many immigrants in Rome, Turin and Milan. One might think that it is related to the Greek surname
Τζατζάς, but I see some difficulties (e.g. stress position) and I don't really know. In any case, I am always extremely suspicious of claims based on obscure placenames whose origin is really unknown.

Serra's is surely a valuable work, but of his four Berber claims, not even one might be correct.

shaden said...

Well, they could have been Arabic loans borrowed through Berber, though a Berber-influenced Arabic variety seems more likely given that Arabic was actually spoken in Sicily (the supposed ancestor of Maltese).

Offtopic: I'm looking for a (re)source that includes modern linguistic terminology in English and their Arabic equivalent; eg. Morpheme = the-Arabic-term. Preferably something along the line of Trask's Dictionary. I'm trying to read texts like this with more ease.

Lameen Souag said...

Rhaeticus: thanks for the comments. I agree, none of these look like good Berber etymologies.

Shaden: I have a couple of such dictionaries, but I'm not sure a list that includes modern linguistic terminology in English and their Arabic equivalent is what you need for that site; it uses quite traditional Arabic grammatical terminology and theory, for which a reasonably old English grammar or textbook of Classical Arabic may be more helpful.

shaden said...

Yeah, I was considering Ryding's grammar for that, but I'd also like to have something basic on hand. What are the dictionaries that you recommend? If they are published in the ME, I could probably pick one up on my next trip back in Cairo.

Lameen Souag الأمين سواق said...

This book is now online: