Monday, September 21, 2015

Berber substratum nouns in Beni-Tamer (Adrar, Algeria)

In most of the Arabic dialects of the Algerian Sahara that I've encountered, Berber influence is rather inconspicuous. Loanwords exist, of course, but they are so well adapted to Arabic structure that they tend to be difficult to spot; who would guess, for instance, that šəṛšmala "skink" came from Zenati Berber *a-sərm-šal (cf. attested Tashelhiyt asrmkal?)

In a few areas, however, Berber loans retain the Berber nominal prefixes a- and ta-, and hence stick out like a sore thumb. In such cases, they often keep Berber-style plurals as well, reproducing a Berber subsystem within the otherwise Arabic domain of the dialect's nominal morphology. The only major Saharan dialect that consistently does this, as far as I know, is Hassaniya in Mauritania and the Western Sahara. However, during fieldwork some years ago, I came across another case well outside of Hassaniya. The area around Adrar (medieval Touat), in southwestern Algeria, seems to have shifted from Berber to Arabic relatively recently, and the process is not complete even today. At least one village, Beni-Tamer just outside Adrar, accordingly borrows many Berber nouns with Berber nominal prefixes, including ones unfamiliar to other speakers from near Adrar that I met. I only spent a short time with the one speaker from Beni-Tamer that I met, but he gave me quite a few examples from his Arabic (he did not speak Berber):

With masculine a-:

  • aždəl "garden near town"
  • ažəlžim “hoe” (Taznatit ažəlžim)
  • afdam “palm fibre” (cf. Hassaniyya fdām)
  • afrag “palm-leaf fence” (Taznatit afrag, cf. Hassaniyya efəṛṛāg)
  • aqənnin / qənnin "palm stump" (Taznatit taqənniħt)
  • agžəm “cellar” (Taznatit ikzəm)
  • amazzər “sloped spot in an irrigation channel”
  • anfif “drainage hole” (Taznatit anfif)
With feminine ta-:
  • tadmayt "garden outside town"
  • tasgat “large basket” (Taznatit tasgawt)
  • taṣəṛbiṭ “skink”
  • tagəmmi “stable”
Most of these are not attested in Hassaniyya, and closely reflect the Taznatit Berber still spoken at Timimoun, confirming that they represent a substratum of Berber words retained by this town's people after they shifted to Arabic. This also fits with their semantic distribution, including a lot of agricultural terminology. At least one of them takes a metathesised Berber plural, originally with the Berber masculine plural suffix -awən: agžəm “cellar”, pl. agəžwamən. Unfortunately I didn't elicit plurals for the rest. I don't think I'll be able to go to Adrar in the near future, but it would be interesting to look at this dialect more...

Have you seen anything similar in a dialect you're familiar with?

References: Hassaniyya from Taine-Cheikh, Dictionnaire hassaniyya-français; Taznatit from Boudot-Lamotte 1964, "Notes ethnographiques et linguistiques sur le parler berbère de Timimoun".


Etienne said...

I am certain I have seen a reference to an L1 variety of Arabic spoken somewhere in Morocco (by monolingual speakers) with a substratum of words of Berber origin which maintain etymologically Berber plurals. In Northern Morocco there are some instances/remnants of a Romance (Medieval Spanish?)-derived /s/, /es/ plural ending productively used by monolingual speakers of either Arabic or Berber, possibly both.

In principle I should be able to find the references to both phenomena, should you or anyone be interested.

John Cowan said...

English, of course, shows similar effects with words borrowed from its superstrata.

David Marjanović said...

^ I see what you did there. :-)

Lameen Souag الأمين سواق said...

Etienne: That would probably be the Jebala varieties of northern Morocco, which Heath, among others, discusses in some detail. Those I do have references for, but thanks anyhow!

John/David: Indeed; it seems that almost any sufficiently intensive contact leaves such stigmata, and for centuries Europe's literati prided themselves on their Latin and Greek. No doubt much more direct data could be gathered on the origins of such phenomena if the great writers of the Renaissance were still alive; unfortunately, they are now with the cherubim and seraphim.

David Marjanović said...


Moubarik Belkasim said...

In Rif-Berber (northern Morocco), "aḥaremcař / aḥaremcal" means: "lizard" (small type). "Afrag" also excists in Rif-Berber and means "fence".

Lameen Souag الأمين سواق said...

Funny that change of s/z to ḥ - I wish I understood why that happens sometimes...

David Marjanović said...

Could be the same as the change from s to h that has happened so many times elsewhere in the world in languages that didn't have ḥ to begin with. In some kinds of Spanish [s] is becoming [x] in front of consonants and word-finally right now.

petre said...


I've never (or rarely) heard [x] for /s/, but right here in the Canary Islands, and (so I'm told) throughout America, preconsonantal and word-final [h] regularly represents /s/. This is not happening "right now", but is a long-standing feature of those varieties of Spanish. If that pronunciation is extending itself on the Peninsula (beyond the far south), that would be interesting. Given the number of Mexican and Argentinian soap-operas on TV, also entirely plausible.

Since we have no /θ/, "La Paz", for example, is also rendered as [la'pah]. Listening to locals disambiguate Las Palmas (capital town of Gran Canaria) from La Palma (small island to the west of Tenerife) is a rare treat for linguists, or at least phontecists/phonologists, which you must come here yourself to experience. In such a context, you may indeed here some [x].

petre said...

"hear" not "here". 8/10.

Unknown said...

Hallo i am from North caucasus and intersting in amazigh culture and history