Judeo-Amazigh / Tamazight (Berber)
Description by Joseph Yossi Chetrit

Asfalou, an artistic musical video created by Miléna Kartowski-Aïach, based on an old Judeo-Amazigh song from Tinghir, Morocco

Historical Background

Berber, also known as Amazigh or Tamazight, was the natural language of the North African populations for thousands of years before the Arab conquest of the 7th and 8th centuries. After the installation of Arab kingdoms during the 8th and 9th centuries and the establishment of Arabic as their official language, Berber declined in urban centers but continued to flourish as the first language in the countryside and mountainous areas. Even now, more than 30% of the Moroccan population and some 20% of the Algerian population speak Berber/Tamazight as a second language or as a first language alongside Arabic. Until recently, numerous people of all ages spoke Berber as their only language, but the expansion of national education in Arabic slowly reduced the number of such monolingual Berber speakers. In Morocco, three major distinct varieties of Berber dialects evolved: Tashelhit (Arabic Shilha) in the South and the Anti-Atlas range, Tamazight in the Center and High and Middle Atlas, and Tarifit in the North. In Algeria, Berber continues to be used in the northeastern Kabylia in its variety Taqbaylit, as well as in other varieties, such as Mzab and Chaouia.


What about the language of Jews who lived in North Africa from ancient times, long before Islam and Arabic culture (Hirshberg 1974, first chapter; Chetrit & Schroeter 2003)? Did they also speak Berber before Arabic? We have not been able to completely clarify the linguistic habits of the early North African Jews. Apart from Latin and ancient Greek in the 2nd to 5th centuries, they certainly used Hebrew as an educational and liturgical language and Aramean as a vernacular language until the 7th century, like all the Jewish communities of the Mediterranean. From the mid-9th century, Arabic became their principal language. They adapted it to their Jewish culture and identity in the form of Middle Judeo-Arabic, and more recently as communal Judeo-Arabic dialects (Chetrit 2022).

Quick facts

Names of language:

Judeo-Berber, Jewish Berber, Judeo-Amazigh, Jewish Amazigh, Judeo-Tamazight (some consider Berber derogatory)

Territories where it was/is spoken:

Southwestern Morocco, especially the High Atlas, the Anti-Atlas and its valleys, and the
large Sous valley, plus some evidence of use in the Rif Mountains of Northern Morocco and
in Algeria and Libya


Estimated # speakers:

1900: 8,000
2022: A few dozen



Severely endangered

Writing systems:

Hebrew letters


Haggadah translations, songs


Language family/branch:

Afroasiatic > Berber / Amazigh

As for Berber or Judeo-Berber, to the best of our knowledge and despite this long and close association with Berber tribes, it seems that Jews generally did not use Berber as their principal language at any point in their history apart from in isolated and small communities, as we shall see below. Nonetheless, in rural communities, they usually used it as a second language in Berber-speaking environments, jointly with Judeo-Arabic as their first language. The first evidence of the use of Berber by Jews goes back to only the 19th century (Chaker 2004; Chetrit & Schroeter 2003; Chetrit 2007:230–32), and concerns principally rural Jews in peripheral regions, in Libya (in the Jebel Nefusa), and in Algeria (in Saharan Ghardaia). In Morocco, Judeo-Berber was spoken until recently by thousands of Jews, and for some time after their emigration. An estimate from 1912 suggests that among the Jews of Morocco, 77,000 spoke (Judeo-)Arabic, 16,000 spoke Haketia, and 8,000 spoke (Judeo-)Berber (Lévy 2009; Beider 2017).

Judeo-Berber was used as a second language by many Jewish men and women in hundreds of bilingual rural and semi-rural communities scattered in the High Atlas and Anti-Atlas ranges and their valleys, as well as in the large Sous Valley in southwestern Morocco (Flamand 1959). In the villages of those areas, Jews lived  either in small Jewish quarters or in isolated streets in the immediate proximity of the Berber population, or in mixed streets, where some Jewish families inhabited houses or rooms that Muslim Berbers rented to them in exchange for certain services. Jews used Berber and Judeo-Berber (in addition to Judeo-Arabic) for petty commerce and other professional activities conducted among their Berber clientele (Chetrit & Schroeter 2003; Chetrit 2018).

In Morocco, Judeo-Berber was also used by monolingual speakers in small and isolated communities of the Tifnout region in the Anti-Atlas range, in the territories of the Ait Wawzgit (Ouaouzguite) tribes (Zafrani 1970; Chetrit 2007:230–32). It would appear that the small Jewish rural communities of Ait Bu Ulli were comprised of monolingual speakers too; these Jews settled in the impenetrable mountains of the High Atlas range near Demnate and to the north of Marrakech. These monolingual communities gradually became bilingual during the first half of the 20th century, due to the roads and paths constructed by the French Protectorate between 1920 and 1940 in order to extend its domination of the Berber tribes residing in the High, Middle, and Anti-Atlas ranges. The new infrastructure allowed isolated rural Jews to strengthen their contacts with the urban Jews who spoke Judeo-Arabic; as a result, they became bilingual too.

Judeo-Berber Oral and Written Literature
As a rule, Jews who settled in Berber environments and adopted Berber and Judeo-Berber as their second language developed essentially an oral culture like their Imazighen (Berber) neighbors. They used Hebrew books and Hebrew scrolls for their liturgical Jewish life, but when they created texts in Judeo-Berber, they were transmitted only orally either in storytelling or in poetry, and in small pieces. It is only recently that I discovered the first Hebrew manuscript from Morocco containing a bilingual poem, written in Hebrew characters at the beginning of the 19th century. The poem uses Judeo-Arabic mixed with Judeo-Berber to tell a romantic communal mini-scandal (Chetrit, forthcoming). I discovered
another small Judeo-Berber poem written at the end of the 19th century in another Hebrew manuscript.


If oral tradition was so prevalent among Berber speaking Jews, what about the written translation in Judeo-Berber of the Haggada of Pessah published by P. Galand-Pernet and H. Zafrani from a Hebrew manuscript (Galand-Pernet & Zafrani 1970)? Such a question raises the problematic presentation of this translation by its renowned editors. In their introduction, the editors stipulate that the translation they uncovered and studied represents the common tradition of Judeo-Berber speakers to translate the whole Haggada into Judeo-Berber calque, just as other communities translated the Haggada into Judeo-Arabic, Judeo-Spanish, or Judeo-Persian. After extensive field work in Israel, Morocco, and France among Jews who came from rural communities where they spoke Berber and Judeo-Berber, I can certify that those rural communities, be they bilingual or monolingual, never translated the whole Haggada into Judeo-Berber. The Haggada of Tinghir edited and published by Galand-Pernet & Zafrani was as a matter of fact a commissioned – and not a spontaneous or traditional – translation into Berber. It was ordered in the 1950s by an influential Jewish figure in Casablanca, Raphael Benazeraf, who found a young Jewish scholar of Tinghir, Yossef Malka, through a Jewish merchant from Tinghir who knew his skills. Benazeraf wanted to save some cultural texts in Judeo-Berber for future generations. Later on, I was able to meet the translator of the Haggada in Kiriat Ata near Haifa and learned from the source the origin of this translation (Chetrit 2007: 292–321).

Moreover, at the same time, another commissioned translation of the Haggada was requested from the rabbi Masˤud Ben Shabbat, who served as rabbi in the rural communities of the Sous Valley, wrote poems in Hebrew and Judeo-Arabic, and spoke Berber perfectly. On the eve of the Second World War, he moved to Casablanca, where he opened a traditional school, a heder, and published his poems in separate leaves. His editors, the Hadida Brothers of Casablanca, recognized his skills and asked him to translate the Haggada of Pessah into Berber for them because they planned to publish it in a double version, Hebrew and Judeo-Berber, just as they did in Hebrew and Judeo-Arabic and Hebrew
and French. R' Masˤud prepared the translation, but the publishers did not publish it because the natural users of this new version emigrated to Israel. Fortunately, they kept the manuscript of the Haggada in their archive, from where I was able to get a copy, which I plan to edit and publish (Chetrit, forthcoming). I attempted to order a third commissioned version of the Haggada from a Judeo-Berber speaker, Yehuda Derˤi in Ashdod, who immigrated from Iligh in the Sous Valley to Israel, but he translated only a dozen sections (Chetrit 2007: 292–321).

So the Judeo-Berber speaking Jews never translated the whole Haggada for the Pessah seder. However, the author was able to record informants from rural communities where they spoke Berber as a second language chanting some sections of the Haggada, in particular the second section Ma Nishtanna [=What differs?] and the third section ˤAvadim
Hayinu [=we were slaves], as well as describing the Pessah service and its ritual symbols in Judeo-Berber. Some versions of these texts are parodic and it seems that Jews from urban communities wrote them to mock the rural Jews, whom they regarded condescendingly if not contemptuously.
These brief texts (in Berber and in English) illustrate such oral literature:

Despite the absence of intellectual works, whether oral or written, in Judeo-Berber,Jewish speakers nonetheless had rich repertoires of Berber songs and tales, which they shared with their Muslim neighbors and which they performed separately or jointly with them in their family festivals. Many families of rural origin continued until recently to perform their poetic and musical repertoire in their family festivals in Israel, and particularly upon the occasion of weddings, and the songs of some knowledgeable informants, both men and women, were recorded during the fieldwork (Elmedlaoui & Azaryahu 2014; Levin 2017, 2021). According to the information provided by many informants, there were even some Jewish experts named rrais (pl. rrwais), who composed original texts for the melodies of rural Moroccan Berber Aħwaʃ and Aħidus ceremonies, where poetic jousts on love and other lyrical issues were performed in the form of collective dances and individual songs, gathering men and women dancers in separate rows (Elmedlaoui & Azaryahu 2014; Levin
2017, 2021). Several sets of such songs were recorded in the course of our fieldwork, and other recordings of Berber songs performed by Jewish informants can be found at the Sound Archives of the National Library of Israel in Jerusalem.

Texts in Judeo-Berber

Linguistic Features
As a rule, the phonological and morphological structures of Moroccan Judeo-Berber dialects were very close to those of their neighboring Muslim dialects, because of the intimate proximity between Jews and Muslims in rural settings. However, the Jewish pronunciation of Berber and its Judeo-Berber varieties has several distinctive features, originating in a Judeo-Arabic first language that often interfered with its phonetic articulations.

Despite the lack of distinct phonological and morphological features, Judeo-Berber should be classified as a Jewish language, at least for its sociolinguistic and cultural features, which determine Jewish discourse and semiotics. In specific texts, Judeo-Berber also has distinctive syntactic and semantic features. Like many other Jewish languages, Judeo-Berber uses calque translations when translating the Bible. These texts mirror the syntax of the Hebrew original, resulting in an artificial grammar that differs from Berber syntax. On the semantic level, the distinctive Jewish vision of the world filters through the calque translations of biblical and traditional Jewish expressions, as well as through the implicit references to cultural and spiritual Jewish entities.

Apart from the tendency to centralize the articulations of |i| and |u|, the low central vowel |a| is pronounced with a lesser degree of lowness in ordinary phonetic contexts, i.e., without emphatic consonants, under the influence of the Imala pronunciation of Judeo-Arabic. This feature is represented by [ä]. In direct emphatic contexts, |a| is velarized, particularly directly after or before an emphatic consonant as the root of the tongue moves back and high; it is transcribed here as [ɒ]. In non-direct contact with an emphatic consonant, the Imala is canceled. Examples from an oral text: täddärt (the name of a village near Ifran/Oufran), nǝnnä i̯äs ‘we told him’; i̯ät lmɣɒṛɒ ‘a cemetery’ — loaned from Arabic; är i̯aqqṛɒ ‘he recited some prayer.’ Another particularity of Judeo-Berber is the tendency to expand, as in Judeo-Arabic, the pharyngealization of emphatic consonants, not only to neighboring vowels, but also to other syllables of the lexical unit. In such emphatic environments, the realization of the vowel |i| is lowered as for |e|; it is represented here by [ỉ]. The realization of |u| is also affected: it is rounded and lowered to [o].

Another feature is the frequent canceling of the labio-velarization of palatal and velar consonants: k   > k, kk  > kk, g  > g, gg  > gg, x  > x, ɣ  > ɣ: Muslim Berber nkk  ni/nukkninǝkkni ‘we’ (also found in some non-Jewish varieties). This tendency also distinguishes Judeo-Arabic dialects from Muslim Arabic dialects in Morocco, where labial, palatal, and velar consonants are also frequently velarized, but not by Judeo-Arabic speakers.

We also find the frequent realization of the post-alveolar consonant (ʃ, sh) as a dental-alveolar (s) with a weak stridency, particularly by speakers living in separate Jewish quarters, in big villages, or in urban and semi-urban communities. As in the Judeo-Arabic dialects, the stridency of this consonant is not the same from one community of Judeo-Berber speakers to the next and can range from [ʃ] to [s]. The unvoiced labial stop [p] of Hebrew and European languages often appears in Judeo-Berber, while Muslim speakers articulate it as [b] or [bb], e.g., palistin (Palestine) or pisäħ (Passover). Another feature is the occasional realization of the liquid consonant /l/ as [n] by Jewish speakers from southwestern Moroccan communities, such as Iligh, Tahala, or Oufran/Ifran. Here is an illustrative brief excerpt from a famous Jewish joke told by Jaïs Bensabbat, who was born in Ifran but lived for dozens of years in Marrakech, where he was recorded:

innä-k i̯än ḍ-ḍoṛ i̯uʃkäd i̯än nħǝzzän s i̯ät ţmäzirţ. tämäzirt-än id bäb-nns gän kunnu iɣ w jän, ur ssǝn wänu

‘It is said that a Rabbi once arrived in a [rural Jewish community]. The locals of the place, all of them were donkeys, they did know nothing’


Here nħǝzzän ‘a/the Rabbi’ is used in place of lħǝzzän; kunnu ‘all of them’ is used in place of kullu; and wänu instead of wälu ‘nothing’.

In Judeo-Berber translations of the first sections of the Pessah Haggada, of which there are many communal oral versions, as well as in other liturgical texts, we find a mix of calque syntax that imitates the original Hebrew and natural Berber syntactic structures (Chetrit 2007: 272–292). But in the individual written translations of Yossef Malka and Masʕud Ben Shabbat, as well as in the other two short oral texts, the calque schemes of the Hebrew — and, in fact, of the Judeo-Arabic — are respected and truly illustrated.

Lexicon and Semantics

In ordinary, specific, or professional discourse, we find Hebrew words, compounds, and formulas that refer to specific Jewish cultural entities from all domains of Jewish life, as in other Jewish languages (Chetrit 2009). The loans are perceived as a kind of technical terminology and are mostly borrowed from Judeo-Arabic and not directly from the Hebrew texts. In oral texts we have, for example: lħoṛban חורבן ‘the destruction [of the temple of Jerusalem]’; lqiboṛɒ קבורה, here ‘tombs,’ used along with lqiḅoṛ and lqḅoṛ; lmiʕɒṛɒ מערה ‘cemetery’ used with its Arabic/Berber equivalent lmɣ̣ɒṛɒ; the argumentative expression illä wǝddäi̯ אלא ודאי ‘that is sure’; and references to rabbinical works, like ʃulħan ʕarux שולחן ערוך, the canon of Halakhic precepts and behaviors, or to Biblical figures of ill repute such as Datan and Abira[m] דתן ואבירם (Numbers 16: 1–38) (Chetrit 2015: 124–125). On the other hand, there are no Hebrew words in the Haggada translation, but that is a usual and even foundational principle in calque translations of biblical and liturgical texts.

On the semantic and lexical level, and apart from the Hebrew and Judeo-Arabic elements that distinguish the Judeo-Berber dialects, the entire lexicon is shared with Muslim speakers. However, in the parodic texts, presented as supposedly used in monolingual Jewish communities, there are some lexical forms that indicate cultural Jewish entities in a comical manner. For example, the ħaroset חרוסת, which symbolizes the mortar that the Hebrews were forced to produce during their servitude in Egypt and appears on the Pessah Seder plate, is called ṣṣɒḅon ixṣǝṛn, literally ‘spoiled soap,’ instead of the simple Judeo-Arabic Hebrew loan ǝl-ħarusit (or ǝl-hilk in southeastern communities). Another example is the compound ämän wɒḍỉḷ, literally ‘water of grapes’ for ‘wine,’ instead of the Arabic loans ʃʃṛɒb (or ṣṣṛɒḅ in the usual Jewish pronunciation) or lxmǝṛ (lxṃǝṛ in Jewish use) (Chetrit 2007: 270–272).

Judeo-Berber as a Jewish Language
All of these distinctive features clearly show that modern Judeo-Berber is a hybridized language like all other Jewish languages (Chetrit 2007: 3–38, 407–543, 2013). It combines a Judeo-Arabic substrate, a Berber matrix and a Hebrew component borrowed through Judeo-Arabic uses. But unlike other Jewish languages, Judeo-Berber lacks an intensive manipulation of Hebrew and Aramaic texts through exegesis, homilies, teaching, and so on, as the long-lasting field research teaches us. Yet, as shown in other studies (Chetrit 2007: 8–14, 2013: 182–184), diglossia consists not only of differential social uses of two languages by the same speakers, but also of the formation of many intertwined and
hybridized forms and structures in the two languages, due to their continuous interference. Likewise, we have no evidence that a secret language, based on a combination of Hebrew terms and Berber syntax as in Judeo-Arabic (Chetrit 2007: 545–564, 2009: 241–249), was used in Judeo-Berber communities.

Contemporary Use
Remnants of Judeo-Berber have been maintained in the form of loanwords in the Judeo-Arabic of communities that historically had contact with Berber communities. An example is the term azmumg and its variants, which continued to indicate the Ħinna ceremony preceding Jewish weddings in numerous communities, but which had totally
disappeared from the Muslim Berber dialects. Another example is the term abǝṭṭɒl, which refers to a kohen “descendant of the tribe of Aharon the priest,” whose title of kohen was 
denied by his community on account of his transgression of his religious duties as a kohen (Chetrit 2018). In addition, in Morocco and Algeria, one Jewish personal name, Zemmour
(‘olive tree’), and several dozen surnames derive from Berber words. Examples of surnames include Amray (‘head, director’), Aouat (a Berber tribe), and Ouyoussef (‘son of Joseph’) (Beider 2017).

As for spoken language, today only a small number of elderly immigrants from Morocco to Israel and France maintain Judeo-Berber. There is little or no postvernacular engagement among younger generations. Documentation work is limited to select linguists and ethnomusicologists. If we do not research this language in the coming years, our
opportunity for additional documentation will be lost.








To cite: Chetrit, Joseph. n.d. Judeo-Berber. Jewish Language Website, Sarah Bunin Benor (ed.). Los Angeles: Jewish Language Project. https://www.jewishlanguages.org/judeo-berber. Attribution: Creative Commons Share-Alike 4.0 International.