Berber Ghosts in Arabic

In Federico Corriente’s 2013 Descriptive and Comparative Grammar of Andalusi Arabic, he dedicates a small section to the Berber contribution to the Andalusi Arabic lexicon. He says: “The B[erbe]r. contribution to the A[ndalusi] A[rabic] lexicon, reputedly scarce, has been traditionally treated with supercilious neglect by scholars who, not having deigned to acquire a smattering of that language, dispatched the issue by claiming, as an anachronistic article of faith, that the Berber invaders of Al- Andalus, numerically several times superior to their Arab co-religionists, were already linguistically Arabicised.” (135).

After such harsh critique on this undefined group of scholars, one would expect that the data of the  Berber vocabulary would be well-researched, cited and presented by Corriente. Irritatingly, the opposite is true.  The majority of the words he cites are ghost words. To avoid that such ghost words enter the academic discourse, I decided to discuss them.

All of the words that are presented are simply marked as “Br.”  While for ideological reasons, and general closeness of the Berber varieties one might choose to call it a single language, but from a scientific point of view it is unacceptable to not distinguish between them. The differences between Tarifiyt and the rest of Moroccan Berber are enormous. Tashelhiyt and Tamazight are admittedly closer, but should still not be treated as a single variety by a serious researcher.

arġís ‘barberry’: While this word looks Berber in structure, I was unable to find it in the Berber lexicons. Considering the semantics, it's probably that it would be found in a book more specifically about botanical items. This website suggests the word does actually exist. If I find it, I will provide a reference.

Tafúrma ‘serpent eagle’: While this word looks Berber, tafərma would be a fairly normal verbal noun formation of a verbal root frm, it is not actually attested in Berber. Citing it as a Berber word is therefore misleading (Corriente 2007: s.v. altaforma, is explicit on it being unattested).

arkas ‘hemp sandal’:  Not a ghost word. Tarifiyt arkas ‘leather sandal’, Figuig tarkass ‘leather sandal’. Somewhat differently in Tamazight arəks. The word in Andalusi Arabic is hirkása which corresponds quite well to another Riffian expressive formation ahərkus.

səgnəs: Berber almost always have a prefix. Only before full vowels in the next syllable is the prefix sometimes lost in Zenatic languages like Tarifiyt, cf.: Tamazight afus ‘hand’, Tarifiyt fus ‘id.’ This already gives us a hint that this word cannot be correct. The word that is meant is probably Figuig tisəɣnəss ‘knitting needle’, Tarifiyt isəɣnəs ‘needle of a brooch’, Tamazight tisɣənst  ‘pin’ (with ɣ instead of g!) which appears to be the word Corriente means, as he cites the meaning ‘needle, brooch’ for **səgnəs in Corriente (1997: s.v. ZḠNZ). This ghost word is probably the result of Corriente misreading Renisio (1932: s.v. ĠNS) who gives the verb səġnəs ‘mettre une broche’ (N.B. Corriente read ġ as g, not as ɣ), a causative derivation from the root ɣns.

Despite the wrong form cited by Corriente, he is probably correct that the Andalusi Arabic word záġnaz ‘clasp of a necklace’ comes from the this Berber word iṣəɣnəs.

təgra ‘bowl’: While technically Tamazight tagra ‘wooden bucket’ in the annexed state, Berber words are normally cited in their free state.

agərgit ‘spear’: I cannot find this word, despite Corriente (2007: s.v. gorgoto) saying it is ‘documented’ (no reference given). Final t on masculine nouns is highly unlikely. A feminine form tagərgitt looks more Berber, but is still unattested.

agzal ‘spiked stick’: This appears to be a ghost word. The root gzl usually means ‘to be short’ (e.g. Siwa agzal ‘short’). The Spanish word that was supposedly derived from this word, was a short spear, which makes the semantics likelier, but the meaning ‘spiked stick’ is certainly unattested. See also Corriente (2007: s.v. tragacete).

zaġáya ‘spear’: Unattested, as Corriente himself admits and morphologically clearly not Berber. He suggests it is a nomen instrumenti from Br. aġ ‘to throw’.

First, instrument nouns are made with an s prefix, e.g. the above-mentioned tisɣənst ‘pin’ from the verb ɣəns ‘to affix with a pin’. This s only becomes z due to consonant harmony with a z in the stem.

Secondly, aɣ does not mean throw, but ‘to hold, to take’ in all Berber dialects, as far as I know. So not only is the word artificial, it is also deverbal from a verb which does not have the meaning Corriente says it has.

afrag ‘sultan’s tent’: This word exists, but the meaning presented by Corriente certainly is not the primary one. Tamazight afrag ‘enclosure, fence, sheephold; dry thornbush hedge’; Tashelhiyt afrag ‘dry thornbush hedge’, Tarifiyt afray ‘hedge, enclosure; Royal tent’. The connection with the already not so great phonetic match Spanish alfaneque ‘sultan’s tent’ should be approached with some caution.

tfaya ‘a certain sauce’: Corriente (2007: s.v. atafea) is explicit about where this word come from, it is based on the attestation in Andalusi Arabic, Moroccan Arabic and a hapax legemenon (in an older Spanish document?). Corriente suggests the word comes from a phrase aman ntfiyi/a ‘water of meat’ or a/isswi ntfiyi/a ‘broth of meat’. tifiyyi (var. tifiyya) indeed mean ‘meat’ in Tashelhiyt Berber (Also Tamazight tifiyi). The word is not attested outside of these two varieties. The word tfaya as cited, is a ghost word, whether one accepts the etymology or not.

References

Corriente, F. 2007. Dictionary of Arabic and Allied Loanwords. Spanish, Portuguese, Catalan, Galician and Kindred Dialects. Brill: Leiden & Boston.
Corriente, F. 2013. A Descriptive and Comparative Grammar of Andalusi Arabic. Brill: Leiden & Boston.
Tamazight:          Taïfi, M. 1991. Dictionnaire Tamazight-français parlers du Maroc central. Paris: L’Harmattan-Awal.
Figuig:                   Kossmannm M. 1997. Grammaire du parler berbère de Figuig (Maroc oriental). Paris & Louvain: Peeters.
Tashelhiyt:          Sabir, A. 2010. Taknarit. Diccionario. Español – Amasigh; Amasight –Español.
Tarifiyt:                 Renisio, A. 1931. Étude sur les Dialectes Berbères des Beni Iznassen, du Rif et des Senhaja de Sraïr. Grammaire, Textes et Lexique. Paris: Ernest Leroux.
Siwa:                     Naumann, C. unpublished. Siwi English Arabic Dictionary.

Comments

[Idir]

Hi,
Regarding to > agzal ‘spiked stick’
It is probably based on γ+zl (zl: qqc droit...)

Dic. Taifi.M
aγezzal : bâton, baguette, trique
ica yas aγezzal : il l’a bastonné
taγezzalt: petit bâton, petit baguette

Ghat (Tamachek)
ẓel : tige, branche

Tachelhit:
taẓelyit: branche

[PhoeniX]

That's a good suggestion, thank you!

[Maarten Kossmann]

I am not sure the /h/ in Tarifiyt is expressive here. It rather looks like a loan from the Moroccan Arabic form /herkus/.

[PhoeniX]

That seems like a better explanation!

Reading back your comments on this in your book (Kossmann, M. 2013. The Arabic Influence on Northern Berber. Leiden & Boston: Brill, pp. 197-198, for those interested), the reflex of /h/ might point to a lost consonant *a-βərkus.

And then the Moroccan (and also Andalusi) form would be early loans from Berber.

[Anís del moro]

Concerning 'agərgit', see Dozy, Glossaire des mots espagnols et portugais dérivés de l'arabe, 1869, p. 279:
https://archive.org/stream/glossairedesmots00dozyuoft#page/279/mode/1up

[Anís del moro]

Concerning 'agzál', see Dozy, Documents arabes inédits sur la vie sociale et économique en occident musulman au moyen âge. Trois traités hispaniques de Hisba, Cairo, 1955, p. 28 (infra):
https://archive.org/stream/rasael_hisba#page/n29/mode/1up

On Touareg 'agda/el' or 'adjedel', maybe related, see: http://encyclopedieberbere.revues.org/1488

[Anís del moro]

To sum up, you can see Ignacio Ferrando, "G.S. Colin y los berberismos del árabe andalusí ", Estudios de dialectología norteafricana y andalusí, 2, 1997, p. 105-145:
http://ieiop.unizar.es/publicaciones/detalle.php?idnoticia=78