Skip to navigation – Site map

HomeNuméros3Part IJews of Berber Origin: Myth or Re...

Part I

Jews of Berber Origin: Myth or Reality?

Alexander Beider

Abstracts

The article addresses the ideas of the theory of JUDEO-BERBERS that imply mainly using onomastic arguments that Jews who lived in Maghreb during the last centuries partly descend from the Berber proselytes. Actually, the Berber origin is valid only for one given name and several dozens of Jewish surnames from Morocco, as well as a few surnames in eastern Algeria. These names appeared in the Jewish communities that used a Berber idiom as their vernacular language. Nothing indicates that they existed already in the Middle Ages. All onomastic arguments suggested by proponents of the theory of JUDEO-BERBERS attempting to link these names to the Berber proselytes to Judaism are untenable.

Top of page

Full text

1. Introduction

  • 1 About some of these aspects see Daniel J. Schroeter, "La découverte des juifs berbères", in Relatio (...)

1It was only at the beginning of the 20th century that authors writing about the history and the culture of Jews from North Africa started to pay particular attention to the Berber population of the same region. During the century, abundant literature described links of various kinds (cultural, linguistic, and even genetic) between the two groups of population of Maghreb. Often, the way these links, factual or conjectural, were described was closely related to ideology of the author or, at least, was due to the general frame of analysis that in turn was deeply influenced by various ideological aspects1. One of the developed new concepts was that of JUDEO-BERBERS. It mainly consists in two hypotheses. The first one corresponds to the idea that before the Arabic invasion to North Africa in the seventh century, Judaism played a crucial role in the life of Berbers of the region, with a significant number of Berber tribes converted to that religion. The second hypothesis implies that numerous Jews who lived in Maghreb in modern times are descendants of these Berbers.

2The aim of this paper consists in shedding light only to one aspect of the Berber-Jewish connection: the surnames and the given names borne by Jews in Maghreb. This aspect is of particular interest because in the context of extreme paucity of historical documents, a number of authors use inadequately onomastic materials to support their theories. No discussion will be done concerning the ideology of these authors. For a study focused on objective knowledge, these psychological subjective factors can be ignored. Only the arguments themselves will be taken into consideration.

2. Inception of the theory about JUDEO-BERBERS

  • 2 The most detailed scientific discussion of the historical aspects of the issue of JUDEO-BERBERS app (...)
  • 3 The classical French translation of this passage by de Slane (Ibn Khaldoun [Ibn Khaldûn], Histoire (...)
  • 4 See Ibn Khaldoun, Histoire..., vol. 1, p. 198, 208, 213-214, 340-342 about Al-Kâhina, and p. 207-20 (...)

3No available Jewish text written before the 20th century makes any reference even to a possibility of the mass conversion of Berbers to Judaism. On the other hand, a number of medieval Arabic authors provide information that can be relevant, directly or indirectly, for the issue in question2. The earliest reference comes from the pen of al-Idrîsî (12th century) who mentions the existence in ancient times of Jewish tribes in North Africa. Another author, Ibn Abî Zar’ (first quarter of the 14th century) says that in days of the foundation of the city of Fez (that is, at the end of the 8th century) two Berber tribes Banû Zanâta lived in the area of Fez, of which one was composed of Muslims, and another of Christians, Jews, and pagans. Note that none of these two authors speaks about proselytes. Their texts are relevant in this context only because they make reference not to individual Jews, but to (members of) tribes professing the Judaism. The third author, Ibn Khaldûn (1332-1406) writes that it is possible3 that in the past some of Berber tribes adhered to the Jewish religion, which they had adopted from their powerful neighbors, the Children of Israel, in Syria. This may have been the case with the Jarâwa, the people of the Aurès Mountains (now in eastern Algeria). This may also have been the case with the Nafûsa (now in western Libya), Fandalâwa, Madyûna, Bahlûla, Ghayâta, and Banû Fazâz (all from western Maghreb that is, now in northwestern Algeria and Morocco). Immediately before the Arabic conquest, Berbers from both Eastern and Western Maghreb professed the Christian religion. Ibn Khaldûn also narrates the story of *Dihya/Dahya, nicknamed Al-Kâhina, the queen of the Jarâwa (one of the aforementioned tribes) who was the leader of the Berber struggle against the Arabic invasion to North Africa. It is only after her death in a battle that Islam became really widespread among Berbers, while before the events in question various Berber tribes were Christian or pagan4.

  • 5 Nahum Slouschz, Hébræo-phéniciens et Judéo-Berbères. Introduction à l’histoire des juifs et du juda (...)

4The writings by Ibn Khaldûn were crucial for the proponents of the theory of JUDEOBERBERS. The earliest formulation is due to Nahum Slouschz. He extrapolated the information provided by the Arabic historian in a very special way. The story of Al-Kâhina became the basis for his general theory. Slouschz states that her name means ‘female priest’ in Phoenician and ‘daughter of a Jewish priest’ in Hebrew5. With such etymological connection and taking into account the role of Al-Kâhina as the Berber leader in one particular period of history, he comes to the conclusion that before the invasion of Arabs, the dynasties of Jewish priests (Cohanim) dominated over other North African Jews and Berbers. On numerous other pages of the same book, Slouschz provides multiple additional details concerning the tribes mentioned by Ibn Khaldûn, often without giving any reference about his source of information, and corresponding to various periods. Moreover, he regularly discusses not the history of the tribes themselves, but that of communities, Jewish and non-Jewish, in the geographic regions where the tribes in question lived at some period. This way under his pen any reference to the areas of the Nafusa Mountains in Libya, the Aurès Mountains in eastern Algeria, the region of Tlemcen in northwestern Algeria, and those around the Moroccan towns of Taza and Azrou acquires JUDEOBERBER colors: for him, all these places are associated for centuries to powerful tribes of Berber warriors professing Judaism.

  • 6 Richard Ayoun (“Les Judéo-berbères entre mythe et réalité,” About the Berbers – history, language, (...)
  • 7 See M. Talbi, "Un nouveau fragment...”, p. 42.
  • 8 One can also note that Ibn Khaldûn (Histoire..., vol. 1, p. 177) quotes the Arabic geographer Al-Ba (...)

5The approach by Slouschz does not fit any standard appropriate for scientific studies. He ignores the fact that Ibn Khaldûn just says that Al-Kâhina was the chief of a tribe possibly professed Judaism at some period, without saying explicitly that this woman was Jewish herself6. Slouschz writes about her Judaism as if it were factual. Even based on the text by Ibn Khaldûn only, this conclusion is more than doubtful. Indeed, Ibn Khaldûn says explicitly about Berbers being Christian immediately before the Arabic invasion7. Consequently, for him the Jewish connection of certain tribes should be posited to previous centuries. Moreover, the quote from Ibn Abî Zar’ given in the first paragraph of this section clearly indicates the possibility of different members of the same tribe professing different religions. One can also question the reliability of the information presented by Ibn Khaldûn who lived many centuries after the events he describes: note that his text is the unique source in which the seven tribes in question are said to have a (possible) Jewish past8.

  • 9 Compare, for example, Hans Wehr, A Dictionary of modern written Arabic, Ithaca, Spoken Language Ser (...)

6The link between the name Al-Kâhina and Cohanim postulated by Slouschz has no basis. This name is neither of Phoenician, nor of Hebrew origin. Indeed, kâhina كَاهِنَة just means ‘female soothsayer, fortuneteller’ in Arabic9. This meaning perfectly fits the information provided by Ibn Khaldûn. He indicates that the actual given name of this woman was Dihya, while Al-Kâhina was her nickname. He also precedes the word kâhina by the Arabic definite article al- and not a Phoenician or Hebrew linguistic element. In other words, he describes the story of Dihya ‘the Soothsayer’ and certainly not that of Dihya ‘the (daughter of the Jewish) priest’.

  • 10 See N. Slouschz, Hébræo-phéniciens..., pp. 449-451 and Nahum Slouschz, Un voyage d’études juives en (...)
  • 11 Ibn Khaldûn Histoire..., vol.1, p. 209, 215.
  • 12 N. Slouschz, Hébræo-phéniciens..., p. 412-447

7Slouschz was also the author who introduced another major hypothesis related to his concept of JUDEO-BERBERS. He states that, if we exclude the descendants of Jewish migrants who came from Spain (1391, 1492) and centuries later from Italy, other, “autochthonous,” Jews who lived in North Africa in modern times are descendants of the JUDEO-BERBER tribes of warriors known since the Early Middle Ages. He insists particularly on this local origin for Jews from Tripolitania (northwestern Libya), southern Tunisia (including the Djerba Island), eastern Algeria, and southern Morocco10. For this thesis, no direct historical evidence can be found. Quite on the contrary, according to Ibn Khaldûn, by 720 Islam was already a dominant religion among Berbers, and at the end of the eighth century Idrîs the Great (the founder of the city of Fez) wiped out all the remnants of other religions that were present in his kingdom situated in the territory of modern Morocco11. Moreover, even if certain Berber nomadic tribes professing the Judaism survived at that period, their traces would disappear after the religious persecutions of the times of Almohads (12th century). Globally speaking, the exposal of the Jewish history in North Africa after the conquest of this region by Arabs suggested by Slouschz himself12 looks much closer to the genre of erudite literary fantasy than to a scientific study. Maybe, the only substantive argument that this author proposed to support his theory concern the surnames used by Jews in Maghreb. It will be addressed in section 3.

3. Inception of the idea of “JUDEO-BERBER surnames”

  • 13 Davis Cazès, Essai sur l’histoire des israélites de Tunisie, Paris, A. Durlacher, 1888, p. 169-179.

8The first discussion of Jewish surnames in Maghreb appears in the history of Jews of Tunisia written by Cazès13. This author divides the corpus of surnames into several lists classifying them by type and by language (Hebrew, Aramaic, Arabic, Spanish, Portuguese, Italian, Greek etc.). His first list (p. 175) corresponds to those whose etymology is unclear. Cazès notes (p. 174) that: (1) these surnames are older than other surnames, and (2) either their roots are related to Berber or Coptic languages, or their current obscure forms are due to distortions that took place during their long history.

  • 14 N. Slouschz, Hébræo-phéniciens...

9The second discussion appears in the appendix to the study on JUDEO-BERBERS, by Slouschz14, who provides a list of 74 surnames known in Tripolitania (western Libya) and Tunisia that according to him are “ethnic names of doubtless Berber origin.” Among them, he indicates nineteen surnames derived from Berber place names in the area of the Nafusa Mountains (western Libya): Fitoussi, Ghaloula, Jami, Jarmon, Jouari, Jouili, Meghdis, Messari, Nefoussi, Sagroun, Seroussi, Sfez, Setbon, Sitrouq, Temsit, Toubiani, Zagron, Zemagi, and Zerousi. According to the author, this series displays “almost the whole map of ancient Jewish settlements” of the area. To that, he adds four surnames drawn from the “names of the nomadic tribes of Jewish religion: Abrahami, Alouch, Lellouch, and Mimoun. In his statement, the focus on the Nafusa Mountains is of particular importance because this geographic area is related to the Nafûsa, one of the seven tribes that, according to Ibn Khaldûn, were Jewish before the Arabic invasion to North Africa.

  • 15 Among them are Hamet (Les Juifs...), Eisenbeth (Les Juifs...), and Tolédano (Une histoire...) (see (...)
  • 16 Paul Sebag, Les noms des juifs de Tunisie. Origine et significations, Paris, l'Harmattan, 2002.
  • 17 D. Cazès, Essai..., p. 175.
  • 18 The comparison of the two sources yields the following results: (1) 58 names appear in both lists, (...)
  • 19 Isaac D. Abbou, Musulmans andalous et judéo-espagnols, Casablanca, Antar, 1953, p. 391

10The information provided by Slouschz about Jewish surnames derived from Berber place names in the Nafusa Mountains was used, without any critical analysis, by various authors who studied Jewish surnames from Maghreb15. It was Paul Sebag16, who quoting Eisenbeth rather than Slouschz, first expressed doubts about the veracity of the “facts” in question. He indicated that he was unable to find the villages suggested as sources for Jewish surnames in one detailed geographic study of the Nafusa Mountains. An attentive look into the list suggested by Slouschz strengthens Sebag’s skepticism. We can observe that a large majority of “Berber ethnic names” given by Slouschz actually originated from one source unrelated to Libya: the list of Tunisian Jewish surnames that appear in D. Cazès17, in the category of names whose etymology is unclear18. Apparently, Slouschz simply concluded that obscure names should be of local Berber origin: this idea, representing a dramatic simplification of the concepts by Cazès, “corroborated” his global theory. In his history of Jews of Morocco, Abbou proceeds to a similar simplification19. According to him, surnames that “have sense neither in Spanish, nor Arabic can be of the Berber origin only.” This idea is false: a surname that cannot be easily explained from Spanish and Arabic (or Hebrew) lexicons can still be unrelated to the vocabularies of Berber idioms. Indeed, several categories of other sources are possible. First, it can be derived from an archaic and/or rare word of one of these non-Berber languages, a word that does not appear in standard dictionaries. Note also that numerous forms morphologically derived from basic forms (for example, diminutives and participles) are usually absent from dictionaries. Second, the surname can be drawn from a given name or some toponym. Thirdly, as noted by Cazès, it can be due to some distortion. Finally, the surname can be a ready-made form brought to the area under analysis by migrants from other countries.

  • 20 Compare Adolphe de Callasanti-Motylinski, Le Djebel Nefousa. Transcription, traduction française et (...)
  • 21 Mordecai Cohen, Higgid Mordecaï (ed. Harvey E. Goldberg), Jerusalem, Ben Zvi Institute, 1981, p. 22 (...)

11The link to the toponyms from the area of the Nafusa Mountains postulated by Slouschz for nineteen surnames is doubtless only for two of them: Nefoussi (based on the name of the area in question) and Seroussi (derived from the name of an important locality that existed in that area in older times). For other seventeen surnames, we do not find any fit in detailed descriptions of local villages, inhabited or in ruins20. Of course, formally speaking such negative correlation does not exclude a theoretical possibility that some additional surnames from the list are also derived from the toponyms of the area: the geographic books in question are certainly not exhaustive. However, this possibility looks rather unlikely. On the one hand, Slouschz himself never corroborated his etymological statements by references to geographic sources. On the other hand, the etymology of several among these seventeen surnames is clear and it is related neither to the Nafusa Mountains, nor to Berbers. For example, Jarmon (ג׳רמון) and Sitrouk (סטרוק) are both known as Jewish male given names in Maghreb. The first of these given names is of Arabic origin. The second given name represents a colloquial form of Astruc, the given name of Old Occitan origin, widespread in medieval southern France and Spain. For it, a phonetic coincidence with a putative toponym in Libya is particularly implausible. Ghaloula (גלולה) is derived from the identical female given name (of Arabic origin) used in Libya at the end of the 19th century21. Jami (spelled ג׳אמיע and ג׳אמע in Jewish sources) is related to the Arabic word jâmiجَامِع compiler, collector’. The Arabic origin of Sagroun and its phonetic variant Zagron is also out of doubt. These names are based on the Arabic ṣaḡîr صَغِير small’ and end on the Arabic suffix –ûn common in Jewish surnames and given names in Maghreb.

12The absence of any Berber link is also characteristic to numerous other surnames among those shared by the lists present in the books by Cazès (as names of “unclear origin”) and Slouschz (already as “ethnic names of doubtless Berber origin”). Some of them correspond to surnames of Romance origin borne by Spanish Jewish refugees to North Africa (Gabizon and Ketorza). A large number of other surnames are of Arabic origin. Among the most obvious examples are: Younès < the vernacular pronunciation of yûnus يونس’ Jonah’ (a common male given name used by both Jews and Muslims), Fillouz (a phonetic variant of a much more common surname Fellous) < /fəllûs/ فلّوس chicken’, and Douib < */ḏwî(y)əb/, a diminutive form of ḏîb ِذيب ‘dieb’ (North African jackal). Any link to “nomadic tribes of Jewish religion” cannot be discerned for Abrahami and Mimoun (both derived from male given names, Abraham and Mimoun, commonly used by Jews) and Alouch (from Arabic ‘allûš عَلُّوش lamb’, also used as a male given name by both Jews and Muslims).

13Globally speaking, in Slouschz’s list of 74 “ethnic names of doubtless Berber origin surnames” only one, Zemour, is really derived from a Berber word, while Melloul and Temsit are likely to be derived from the Berber toponyms in Morocco. We will return to their exact etymologies in section 6. All other surnames have non-Berber etymons and almost all of them are related to the Arabic language.

  • 22 Eduard Pérez i Pons (ed.), Fonts pera l'estudi de la comunitat jueva de Mallorca: regesta i bibliog (...)
  • 23 Simon ben Ṣemaḥ Duran, Tashbaṣ I-III, Amsterdam, 1738/1739, n° I.21
  • 24 Henri Bresc, Arabes de langue, juifs de religion: l'évolution du judaïsme sicilien dans l'environne (...)
  • 25 Paul Wexler (The Non-Jewish origins of the Sephardic Jews, Albany, State University of New York, 19 (...)

14The only two surnames related to the toponyms of the Nafusa Mountains, Nefoussi and Seroussi, both end in the Arabic suffix –î used to designate residents or people who originated from the place in question. In other words, these two surnames are typically Arabic. For example, the first of them simply means ‘(one) from (the area of) the Nafusa Mountains’ in Arabic. Even if we know that Berbers lived and still live in the area in question, nothing indicates that during the centuries that followed the Arabic conquest the local population consisted in Berbers only. Variants of both names are known in the Mediterranean region since the Middle Ages. Jews called Nafuci are known in the Majorca Island already during the second half of the 14th century22. During the 15th century, Jewish sources indicate the presence in the town of Béjaïa (now in northeastern Algeria) of a rabbi called נפוסי who originated from Majorca23. Naffusi and Nifusi, as well as Xirusi and Chirusi (both related to modern Seroussi) also appear as Jewish surnames in 15th -century Sicily24. It would be a very bold hypothesis to draw from these references any conclusion corroborating the ideas of Slouschz. We know only the geographic origins of the progenitors of these surnames. Yet, nothing suggests that these Jews were descending from Berber proselytes to Judaism, or were in any way related to the Nafûsa Berber tribe mentioned by Ibn Khaldûn for the period seven centuries before the earliest references to the surnames. We cannot even assert that for these Jews or their ancestors their native language was Berber25.

  • 26 N. Slouschz, Un voyage..., p. 12, see also Nahum Slouschz, Travels in North Africa, Philadelphia, T (...)

15After presenting the list of “ethnic names of doubtless Berber origin” in his book of 1908, Slouschz made regular references to it in other studies published in different languages, with no additional details but with further extrapolations. For example, he insists that in the city of Tunis the majority of local Jews bear the ethnic names of Berber tribes and cities of the interior, while in the city of Tripoli the proportion of names of Berber or Saharan origin is even larger than in Tunis26. Since for many centuries the Berber presence in all of Tunisia and the area of Tripoli is marginal, such claims, if they were founded, would be indeed crucial for the history of the Jewish settlement of Eastern Maghreb. Yet, as shown in this section the onomastic arguments proposed by that author are untenable and, moreover, the information he provides seems to be partly based on the intentional falsification on his part. The results of the above detailed analysis are primarily important because the ideas by Slouschz influenced, directly or indirectly, other authors who wrote on the historical and/or onomastic aspects of the Jewish culture in North Africa.

4. Developments of the idea of “Berber” surnames

4.1 Hamet (1928)

  • 27 Ismaël Hamet, Les Juifs du nord de l'Afrique (noms et surnoms), Paris, Société d'Éditions Géographi (...)
  • 28 All other authors whose works are discussed below in this section actually took almost all the corr (...)
  • 29 I. Hamet, Les Juifs du nord de l'Afrique…, p. 14.
  • 30 The surname and its spelling variants appear in Jewish sources as ברהאנץ M. Cohen, Higgid..., p. 22 (...)

16The first serious study of etymologies of Jewish surnames in Maghreb is due to Hamet27. He suggests etymologies for dozens of surnames whose origin is not self-evident for native speakers of Arabic. Among them are a series of surnames of Berber origin for which Hamet often provides exact quotes from various French-Berber or Berber-French dictionaries28. This pioneering study also introduces basic notions for the domain of Jewish onomastics in North Africa. One of them corresponds to his idea that Jews bear surnames of several Berber groups such as Dray, Soussi, Touati, Zenaty, and Branes29. Inside of his dictionary, Hamet proposes the derivation from the tribe names for a number of other surnames. Actually, the first three surnames from his list all correspond to words designating inhabitants of territories (Draa, Souss, Touat) that during the 19th century were parts of southern Morocco. Muslim Berbers constituted a majority in these regions. However, it would be improper considering that the surnames in question designate Berber groups. Note that all three surnames end in the Arabic suffix -î, that is, they are not even derived from words used by Berbers internally. Only the last two surnames of the list suggested by Hamet are indeed related (at least, for their Muslim bearers) to Berber tribes, Zenata (again, with the final “y” in the surname Zenaty coming from the Arabic suffix –î) and Branes, respectively. However, the historical and linguistic analysis of the Jewish surname B(a)ranes shows that is unlikely to be directly related to the Berber tribe Branes30. On the other hand, most likely, the ancestors of modern bearers of the Jewish surname Zenaty lived in the territories populated by the tribe in question. In other words, when borne by Jews, this surname can simply indicate the geographic origin of its first bearers, exactly as Dray, Soussi, and Touati discussed above. Moreover, the tribe names often gave rise to place names. As a result, some of surnames that seem to be derived from tribe names can actually be drawn from the place names in question. For example, the town of Zenata (Arabic znâta زنَاتَة) exists in the northwestern Algeria. Finally, we cannot exclude a possibility that some of these names were borrowed by Jews from local Muslim notables: note that Muslim bearers are known for all of the surnames in question.

  • 31 I. Hamet, Les Juifs du nord de l'Afrique…, p. 4-6

17It is clear that the general idea by Hamet about tribes was, at least partly, influenced by writings by Slouschz. Indeed, Hamet makes regular uncritical quotes from the list of (pseudo)- toponyms from the Nafusa Mountains suggested by Slouschz. He also speaks about “numerous indigenous tribes converted to Judaism before the arrival of Arabs” again making quotes from Slouschz31. Hamet also provides his own (speculative) explanation for that “fact:” the Judaism professed in the past by Berber tribes would be due to their collaboration with local Jews in their wars against common Greek and Roman enemies.

4.2 Eisenbeth (1936)

  • 32 Maurice Eisenbeth, Les Juifs de l'Afrique du Nord – Démographie et onomastique, Algiers, 1936.
  • 33 M. Eisenbeth, Les Juifs de l'Afrique du Nord…, p. 72.

18The book by Eisenbeth32 represents the first systematic study of Sephardic onomastics. The dictionary portion lists 4,063 names classified into 1,146 entries called “roots” (French souches). Eisenbeth generalized the idea about the importance of the names of tribes for the corpus of Jewish surnames in Maghreb already present in the book by Hamet. However, one the one hand, he extends that idea to include not only Berber, but also Arabic tribes, and, on the other hand, totally dissociates it from the concept of JUDEO-BERBERS. In the introduction to his book, when speaking globally about surnames of Arabic or Berber origin, Eisenbeth postulates: “Jews seem to simply adopt the names of tribes among which they were living.”33 In other terms, he considers all these surnames to be of toponymic origin. Eisenbeth applies this principle as if it were a mathematical axiom. As a result, hundreds of surnames from North Africa are explained by him as derived from tribe names. A detailed analysis of the surnames in question shows that the principle introduced by Eisenbeth is methodologically untenable. But for a very few exceptions, one can suggest for all of these names other, much more plausible, etymons. To illustrate this statement we will consider only a sample of surnames starting with A-: Abdoun, Abensour, Addi, Afergane, Aiach, Aidan, Alcheikh, Allouch, Amouche, Amsetat, Assal, Athon, Azeroual, and Aziza.

  • 34 About the directions of migrations of Jews between various countries of Maghreb see: Alexander Beid (...)

19Firstly, a number of “tribal” connections suggested by Eisenbeth are more than doubtful from the point of view of geography. Among them are the explanations of the Algerian surnames Abdoun and Athon from the Moroccan tribes Oulad Abdoun and Oulad Hattoun, respectively; Abensour and Aziza, both widespread in various parts of Maghreb including Morocco, from the tribes Aziza and Sour, respectively, both in eastern Algeria; Moroccan Afergane and Amsetat from the tribes Beni Fergane and Amsettas, from the territories situated east to Algiers; Algerian Alcheikh from the Moroccan tribe Aït Cheikh. Of course, we cannot exclude a possibility of migrations that could occur between the corresponding areas. However, such a scenario represents already by itself an independent hypothesis and the logical probability of the etymologies suggested by Eisenbeth diminishes significantly. Note that, for certain etymologies proposed by Eisenbeth to be valid, we also need to conjecture that all branches of the corresponding families moved to another, distant, region. Moreover, some directions of migrations needed to validate the “tribal” etymologies are rather opposite to those known from historical sources. This is particularly the case for the putative displacement of Jewish families from eastern Algeria to Morocco34.

  • 35 See exact references to these given names in the corresponding entries of: A. Beider, A dictionary

20Secondly, in the sample of names in question, Abdoun, Aiach, Aidan, Addi, Athon, and Aziza are known not only as surnames, but also as given names used by Jews, female in the last case and male in all others35. If a priori a derivation from a tribe name cannot be ruled out for a Jewish surname, it is highly implausible for a given name. Moreover, a majority (or, maybe, even the totality) of the given names in question were borrowed by Jews from Muslims. The etymologies of these Arabic given names are well-known. They have nothing to do with names of tribes: Aziza means ‘dear, beloved’, Abdoun comes from the root ‘abd عَبْد servant’ and the suffix –ûn, and Aiach from the expression meaning ‘full of life’. A possibility that the origins of a surname and an identical given name can be independent can be dismissed. We can be confident that this surname is derived from the given name in question.

  • 36 It is not necessarily derived from the word meaning ‘lamb’ (see A. Beider, A dictionary…, p. 308).

21Thirdly, a number of surnames from our sample correspond to words from the general vocabulary, Arabic (Alcheikh ‘the elder’, Allouch ‘lamb’, Assal ‘beekeeper’) or Berber (Afergane from ifergan ‘fences’, Azeroual ‘blue (eyes)’. The meaning of some of these words implies the possibility of a direct construction of a nickname based on them. For certain of them, the intermediary role of a given name is also plausible: for example, Allouch is known as Muslim and Jewish male given name too36. For the meaning ‘fences’ the intermediary role of a toponym appears more than plausible and indeed a place name called Ifergan is known in southern Morocco. From the semantic point of view, the derivation of the corresponding Jewish surnames from the common nouns, adjectives, given names, or toponyms is much more plausible than their putative derivation from the names of tribes.

22For some surnames from the same sample one can also propose additional arguments showing the inadequacy of the approach by Eisenbeth. Note that the change of the final consonant in Amsetat and the exact form of Allouch both remain unexplained by etymons suggested by him, the tribe names Amsettas and Alalcha, respectively. Forms related to Abensour are already known in medieval Spain and, as a result, any link to the tribe Sour in eastern Algeria sounds implausible. In Algerian civil records, bearers of the surname Amouche sign נחמוש in Hebrew. This form corresponds to the Jewish male given name *Nahmouch, a diminutive of Nahman.

23The sample of surnames starting in A- was taken just as an example. For any other sample of surnames for which Eisenbeth suggests a derivation from a name of an Arabic or Berber tribe an analysis similar to the one proposed above would yield similar results invalidating the particular cases of application of the general idea by Eisenbeth. Yet, it would be inappropriate saying that here we deal with a totally fortuitous similarity between Jewish surnames and names of Muslim tribes from North Africa. The number of tribal clans that lived during the last centuries (and often still live) in Maghreb is large. Their names are often derived from the personal name of an ancestor common to all members of the tribe in question. This personal name in turn usually either corresponds to a Muslim given name, or comes from a nickname itself drawn from some Arabic or Berber word. For example, the tribe name Aït Aïache are certainly derived from the Muslim male given name Aïache (Arabic ‘ayyâ عَيَاش), the first word Aït coming from the Berber word meaning ‘children, sons’. In other terms, both Jewish surnames and names of Muslim tribes often have the same sources and it is precisely for that reason that we find numerous similarities between both groups of names. Eisenbeth observed this similarity and made a global methodological error misinterpreting the exact basis for that phenomenon.

  • 37 M. Eisenbeth (Les Juifs..., p. 124) seems to be the first to suggest the Berber origin for the surn (...)
  • 38 André Chouraqui, Marche vers l’occident: Les Juifs d’Afrique du Nord, Paris, PUF, 1952, p. 132.
  • 39 David Cohen, Le parler arabe des juifs de Tunis. Vol. 1: Textes et documents linguistiques et ethno (...)

24For long years following his publication, the book by Eisenbeth became the standard reference source on Jewish surnames from Maghreb. A large number of authors who wrote on the surnames, history, languages, and culture of North African Jews considered the etymologies suggested as if they were “factual”37. The connection of numerous surnames to Berber tribes made a particularly strong impression. For example, in his popular history of Jews of the area in question, Chouraqui extracts from the book by Eisenbeth a long list of surnames of the “Berber” origin and concludes about the existence of the “deep roots that plunge the Jewish population into the Berber lands of North Africa”38. Cohen, in the introduction to his detailed scientific linguistic analysis of the Judeo-Arabic idiom in Tunis, says about the Jews of Tunisia that “one cannot fail to be struck by the importance of the Berber element in their onomastics”39. He bases this (false) impression on the information found in works by Cazès and Eisenbeth.

4.3 Laredo (1978)

  • 40 Abraham I. Laredo. Les noms des juifs du Maroc – Essai d'onomastique judéo-marocaine, Madrid, Insti (...)
  • 41 About the age of Jewish surnames in Maghreb see A. Beider, A dictionary..., p. 57-84.
  • 42 Jean-Frédéric Schaub, Les Juifs du roi d'Espagne (Oran 1509-1669), Paris, Hachette Littératures, 19 (...)
  • 43 Joseph Ben Naim, Malkhei Rabbanan, Jerusalem, Abecassis, 1931, p. 29.

25In his monumental study of Moroccan Jewish names, Abraham Laredo adheres to the theory of JUDEO-BERBERS40. He gives an exact quote from the translation by de Slane of the text by Ibn Khaldûn about the seven Berber tribes converted to Judaism before the arrival of Arabs. However, instead of presenting the exact quotes from the Arabic historian about the struggle of Berbers against Arabs and the role played by Al-Kâhina, he adds details absent from the book by Ibn Khaldûn, namely, about the same seven tribes being particularly distinguished in the Berber resistance. Laredo also states that those Judaized Berbers who were not converted to Islam by Idrîs the Great merged with ethnic Jews to compose local Jewish communities. Among the examples of Moroccan Jewish surnames related to the historical events in question he mentions Al Mediuni and Bahlul derived, according to Laredo, from the names of two among the seven tribes mentioned by Ibn Khaldûn: Madyûna and Bahlûla, respectively. This explanation is hardly tenable. On the one hand, the surname Bahlul can be explained as a nickname based on the Arabic word bahlûl بَهْلُول silly’. On the other hand, Al Mediuni can be derived from the toponym Mediouna: towns with such name (derived in turn from the name of the Berber tribe in question) exist in Morocco and Algeria. This surname is typically Arabic: it starts with the Arabic definite article al- and ends in the Arabic suffix –î. In other words, Al Mediuni just means ‘the one from Mediouna’ in Arabic. The above explanations are direct, much simpler, and, therefore, much more plausible than the etymologies proposed by Laredo. Yet, his theory encompasses several independent hypotheses each of them being far from self-evident. Indeed, it implies that: (1) the information provided by Ibn Khaldûn about the tribes in question is accurate; (2) some of Berbers converted to Judaism kept this religion after the reigns of Idrîs the Great and the Almohads; (3) some of them took hereditary surnames reminding about the Berber tribe to which their ancestors belonged; and (4) these surnames survived in Jewish communities of Morocco until our days. The problematic character of the first among the above hypotheses was already discussed in section 2. The second hypothesis is purely speculative: no historical document, Muslim or Jewish, corroborates it. The third hypothesis goes against the common sense. It is difficult to imagine a convert to Judaism who wants to preserve in his surname the direct indication to the non-Jewish origin of his ancestors. The fourth hypothesis is badly correlated with the history of surnames borne by Jews in that region. Indeed, for the period before the arrival of numerous Jewish refugees from Spain at the end of the 15th century, we know a very few Moroccan Jews who used hereditary surnames. The earliest references to family names date from the 12th century only, that is, more than three hundred years after the Arabic conquest of western Maghreb41. The surname spelled El Medioni in Spanish and למדיוני in Hebrew is mentioned in the 17th century only42, while the earliest reference to Bahlul (Hebrew בהלול ) corresponds to the start of the 18th century43. Globally speaking, if we apply the Principle of Simplicity (also called Occam’s Razor), standard in sciences, the theory by Laredo appears untenable.

  • 44 A. I. Laredo, Les noms des juifs du Maroc…, p. 171-172.

26Laredo indicates the Berber origin for numerous other surnames said to be derived from to Berber tribes, toponyms, words, and given names. If the surnames in question include Arabic morphological elements (such as the suffix –î and/or the definite article Al-), he calls them “Arabo-Berber.” If a surname is based, according to Laredo, on a Berber diminutive form of a given name (mainly biblical) having a Hebrew root, he calls it “Hebrew-Berber.” Globally speaking, he comes to the total of 186 (or more than thirteen percent) of Berber-based surnames from the global total of 1,409 of Moroccan Jewish surnames for which he identified their source language44. Of these 186 surnames, 75 are “Berber”, 68 “Arabo-Berber,” 39 “Hebrew-Berber,” and four “Berbero-Phoenician.”

  • 45 Hugues Jean de Dianoux,"Remarques à propos de l’origine apparemment berbère des noms de familles ju (...)

27Dianoux45 reuses a number of Berber etymologies proposed by Laredo. He also adds a series of Berber derivations for surnames that were not considered to be of Berber origin by his predecessors. All his original etymologies are scientifically unreliable: they are based on fortuitous partial phonetic coincidences.

  • 46 Joseph Tolédano, Une histoire des familles – Les noms de famille des juifs d'Afrique du Nord: Maroc (...)

28Hundreds of entries in the popular dictionary of Jewish surnames from Maghreb written by Tolédano46 indicate the Berber connection of the names in question. Actually, all his Berber or pseudo-Berber explanations are taken from Eisenbeth, Laredo, or Hamet. In some cases, he indicates, without providing any argument that the name is of “uncertain, but surely Berber” origin.

4.4 Taïeb (2004)

  • 47 Jacques Taïeb, "Juifs du Maghreb: onomastique et langue, une composante berbère?", in Encyclopédie (...)

29The statistical calculations made by Taïeb47 for Morocco, Algeria, Tunisia, and Libya yield a close, though slightly smaller, percentage in comparison to that by Laredo. In that text, the most detailed study of the Berber origins of Jewish surnames in Maghreb published until now, its author says that among the 1,450 groups of related surnames (French souches), more than eighty are “almost surely of Berber origin” and more than sixty are either of “partial Berber origin” (that is, including one Berber and one non-Berber morphologic elements) or of “possible Berber origin.” It we take all the 150 surnames into account, they cover about ten percent of the total.

  • 48 The Jewish surname Taous is known in the city of Algiers only (A. Beider, A dictionary…, p. 598). M (...)
  • 49 N. Slouschz, Hébræo-phéniciens..., p. 469.
  • 50 D. Cazès, Essai..., p. 175.
  • 51 N. Slouschz, Hébræo-phéniciens..., p. 469.

30In his list, Taïeb distinguishes several categories of surnames. The first one corresponds to given names. Among the four examples provided by him, one, Taous, is actually unrelated to Berber idioms being derived from the Arabic word ṭâwûs طَاوُوس peacock, peahen’48. The second, by far the largest, category covers surnames derived from toponyms. Taïeb includes in it, among others, almost the whole series of “Berber villages” in the region of the Nafusa Mountains including Gallula, Magaïdès, Sfez, Sitruk, and Tubiana. As pointed above in this article, the list of these (mainly fictive) toponyms is ultimately due to Slouschz49. Among the members of the category of toponymic surnames one also finds: Amsili, Datchi, Demri, Ed-Dadsi, El Anfaoui, Es-Skuri, Ghamrasni, Mlili, Tedgui, and Zenati. Actually, according to their morphology, namely the Arabic suffix –î and—for some of them—the Arabic definite article El- (with variants Ed- and Es- before the corresponding Arabic “solar” letters), all of them should not be counted among “Berber” surnames: they are typically Arabic. The Berber origin of some of the place names in question is relevant for the study of North African toponyms. Yet, for the study of Jewish surnames it is totally irrelevant. Taïeb also cites a number of surnames related to occupations or personal characteristics of their first bearers. For them, he mainly provides a correct derivation that is ultimately due to Hamet. Finally, Taïeb includes in his group of surnames of “possible Berber origin” a number of names whose etymology is unclear or, at least, uncertain for him. Many of them (including Guedj/Guez and Lellouche) appear already among names unclear to Cazès50 that are also present among the list of “ethnic names of doubtless Berber origin” proposed by Slouschz51.

31If a large number of errors and misinterpretations of his predecessors were uncritically retaken by Taïeb, this author explicitly distances himself from the proponents of various aspects of the theory of JUDEO-BERBERS. The main conclusions he draws from his onomastic analysis are totally accurate. He insists on the hypothetical character of the data provided by Ibn Khaldûn. He observes that the surnames of Berber origins are mainly peculiar to Jews of Morocco and points that Jews living in modern times in the Berber-speaking areas were often having Arabic as their first language. Finally, he states that the existence of surnames of the Berber origin provides no corroboration for the idea about one part of Jews from North Africa descending from the Berber proselytes to Judaism.

5. Given names of Berber origin and surnames based on them

  • 52 For example, Encyclopedia Judaica, vol. 6, p. 270 and P. Wexler, The Non-Jewish origins..., p. 127- (...)
  • 53 Ibn Khaldoun, Histoire..., vol. 3, p. 252.

32*Dûnaš (דונש (is the oldest given name used by Jews that various authors consider to be of Berber origin52. Two of its bearers became famous during the 10th century: (1) Dunash Ibn Tamîm, a physician and scholarly erudite in Kairouan (now in Tunisia) and (2) Dunash ben Labrâṭ, a Hebrew poet and grammarian who before coming to Spain lived in Fez (now in Morocco) and Baghdad. Actually, nothing implies the existence of any link between them and the Berber culture. For both these medieval scholars, Arabic was their native language. If the origin of Labrâṭ is uncertain, Tamîm is a typical Arabic name meaning ‘perfect’. The given name *Dûnaš (or *Dûnas?) was clearly borrowed by Jews from Muslims: note that Abu-’l-Attaf Dounas, the son of Hammama, was the Emir of Fez in the 11th century53. No reference to this given name was found in various books dealing with given names used by Muslim Berbers.

  • 54 In this section, Jewish surnames derived from the discussed given names appear in brackets [].
  • 55 The only reference known outside of Morocco corresponds to the birth record of Idir Ben David Ben Y (...)
  • 56 Miloud Taïfi, Dictionnaire tamazight-français (parlers du Maroc central), Paris, L’Harmattan-Awal, (...)
  • 57 J.-M. Dallet, Dictionnaire..., p. 152.

33In modern times, Jewish given names of Berber origin were marginal in North Africa. Iddir (also spelled Yedder and Yeddir, pronounced yəddîr) [Iddir, Idder]54, used only in Morocco55, was the unique basic name for which such origin is doubtless. This participle of the verb idir ‘to live’56 is also used by Muslim Berbers57. Jews could borrow it from their Muslim neighbors because of its semantics. Note that a series of other given names used by Jews in Maghreb also have the root meaning ‘to live’ or ‘life’ (‘Ayyâš, Al-‘Ayš, Ya‘îš, Ḥayy and Ḥayyûn, all of Arabic origin, and the Hebrew Ḥayyim).

  • 58 David Corcos, Studies in the history of the Jews of Morocco, Jerusalem, R. Mass, 1976, p. 184-195.
  • 59 About these suffixes see, Annemarie Schimmel, Islamic Names. Edinburgh, Edinburgh University Press, (...)
  • 60 In section 5 of this article, the Judeo-Arabic pronunciation of various given names discussed by va (...)

34In his study of Jewish male given names used in Morocco, Corcos postulates the Berber origin not only for Iddir, but also for more than forty other names from the total of about 300 forms he considers to be basic58. However, a large majority of names in his list are not basic, but diminutive forms (Ašu, Bəddûḥ, Dâdi, Dâni, Məšîš, Məsîsu, Yakhlûf , etc.), some of them even include in their structure typical Arabic suffixes –ûš, -ûn, or -ân (‘Aṭṭûš, ‘Ayšûš, ‘Azrân, Ḥarbûn, Naḥmûš)59. Corcos himself indicates in a few cases that the given name originated as a diminutive form but later (when the link with its base form was lost) became considered to be basic. Often, he does not provide any explanation or explicitly indicates that the etymology of the name is unknown to him. In all these cases, it is unclear on what basis he made a decision that the name should be Berber. Actually, he provides Berber etymons only for two names in his list: Yəddîr (Iddir) ‘living’ and Məšîš ‘cat’60.

  • 61 I. Hamet, Les Juifs du nord de l'Afrique…, p. 16.

35Corcos is the only author who (incorrectly) postulates the Berber origin for numerous basic forms of Jewish given names. Other authors usually insist only on the Berber origin of numerous diminutive forms. Hamet was the first to pay attention to such Jewish given names as Ḥaqu [Haco] < Isaac, ‘Aqqu [Acco]< Jacob, Mûmu [Moumou] < Moses, Dûdu [Doudou, Dodo] < David etc.61 He also notes that the same Berber pattern is common for Muslims too. Laredo (Les noms...) and Tolédano (Une histoire...) generalized this approach: but for a very few exceptions, if a surname is derived from a diminutive form of some given name (often a biblical one), they invariably write about the “Berber” origin of the form in question.

  • 62 M. Taïfi, Dictionnaire..., p. 879.

36The pattern noted by Hamet, corresponds to the addition of the suffix –u (that can also be pronounced [o]) concomitant to the truncation of the basic form. The resulting diminutive most often has only two syllables. Among the male examples additional to those of Hamet are: Ḥammu [Hammou] < Hebrew Ḥayyim, Išu [Ichou] < biblical Yehoshua, Iṣu [Issou] < Arabic Isḥâq ‘Isaac’, Ṭûbu [Toubou] from biblical Tobias, and ‘innu [Innou], perhaps < ‘Amrân ‘Amram’. The surname Scando implies a possibility of the existence of a given name *Skandu derived from Arabic (I)skandar ‘Alexander’. The above diminutives of given names could all be formed by Jews. Among the rare female given names from the same group that contributed to the corpus of Jewish surnames are Tamu [Tamo] < biblical Tamar and Inu [Innou] < any name having /n/. The female name Zennu [Zennou] is used by Moroccan Berbers62 from whom Jews seem to borrow it.

  • 63 In the lists of Berber given names used in southern Morocco (M. Taïfi, Dictionnaire..., p. 878-879) (...)

37We also find diminutives following a pattern similar to the one discussed in the previous paragraph, but with the final vowel being either –i or, rarer, -a. Examples: Bîhi [Bihi] and Bâha [Baha] < Arabic 'Ibrâhîm ‘Abraham’, Dâdi [Dadi] < Dâwud ‘David’, Ḥaqi [Haki] < Isḥâq ‘Isaac’. All these patterns are typical for Berbers63.

  • 64 I. D. Abbou (Musulmans andalous.., p. 400) speaks about “ridiculous transformations” of biblical na (...)
  • 65 G.S. Colin, Arabe..., p. 74, 76.
  • 66 See Victor Hayoun, "Jewish names, surnames, and nickname of Nabeul, Tunisia", in Pleasant are their (...)
  • 67 G. S. Colin (Arabe..., p. 73) mentions the existence of the following Arabic diminutive pattern: th (...)

38For certain Moroccan Jews, forms like those cited in the two previous paragraphs were associated with the Berber-speaking south of the country64. However, the influence of the Berber patterns on Jews could also be indirect: through the intermediary of Arabic. Indeed, even if these patterns are of ultimate Berber origin, they became an integral part of dialectal Arabic, at least, in Morocco. For example, Colin mentions the existence of two Moroccan Arabic patterns that produce bisyllabic forms ending in –u: (1) with two consonants of which the second one is doubled and preceded by a short reduced vowel [ə] (‘əbbu < ‘Abd-allâh, Ḥəmmu < Muḥammad, Hənnu < Mhâni); (2) with two consonants of which the second one is preceded by the long vowel [â] (Ṭâmu < Fâṭma, Zânu)65. Moreover, a few examples of Jewish given names following patterns similar to those discussed in the two previous paragraphs are also found in Nabeul (northern Tunisia) where any Berber influence is unlikely: Gbayla < Gabriel, Khaylu < Makhlûf, Maylu < Samuel, Miru < Meir, Tiri < Esther66. For diminutive forms with the addition of the suffixes –u, -i, or –a, but without any truncation (that is, with all consonants kept), their Arabic rather than Berber origin is even more plausible67. The female name Məs‘ûdi [Messodi] < Messaouda represents an example.

  • 68 D. Corcos, Studies..., pp. 183-197.
  • 69 Numerous forms using one of the two patterns discussed in this paragraph appear in the list of Berb (...)
  • 70 V. Hayoun, "Jewish names...", pp. 176-179.
  • 71 M. Cohen, Higgid..., pp. 224-228.

39Another group of diminutives encompasses those in which we observe the reduplication of one of the consonants present in the basic form. Its first subgroup covers bisyllabic forms in which both syllables have the same consonant. These forms end in –u, –i and, for women only, -a. Among the examples are: female Ṭâṭa [Tata] and Tîtu [Tito], male Bîbi [Bibi], Kâku [Kakou], Qîqi [Kiki], Lûlu [Loulou/Lolo], Mûmu [Moumou], Nûnu [Nounou/Nono], Sîsu [Sissou], Tîtu [Tito], Yûyu [Youyou], and Zûzu [Zouzou/Zozo]. The consonant in question appears in the basic form only once and it can, in theory, be present there at any position. For example, traditionally Nûnu was considered a diminutive of Hârûn ‘Aron’, Qîqi of Jacob, Lûlu of Elijah, Mûmu of Moses, Sîsu and Zûzu of Joseph, and Yûyu of Jonah. In Dîdi [Didi] and Dûdu [Doudou], both from Dâwud ‘David’, we are not necessarily dealing with the reduplication: note that the basic form has two /d/. In the second subgroup, given names keep two different consonants of the basic form and one of them is reduplicated. The resulting form can end in a vowel or a consonant. Among the examples from Morocco (from which no surnames seem to be created) are: Azîzər from Eleazar, Brîri from Abraham, Šîšir from Asher, Šlîlu from Shalom or Solomon, Mîmin from Benjamin, Mrîru from Meir, and Zazak from Isaac68. The same pattern can allow reconstructing the following given names being based on known surnames: *‘Aqûqa [Acoca] and *Qaqûb [Kakoub] < ‘Yaqûb ‘Jacob’, *Ḥaqûq [Hakouk] < Isḥâq ‘Isaac’, *Brîru [Beriro] < Abraham, *Dnînu [Danino] < Daniel, *Mšîšu [Ben Mshisho] < Mûši ‘Moses’ (another diminutive form, M(ə)šîš [Mechich] is well attested, see above in this section), *Ḥbîbu [Habibou] and *<Ḥbâbu [Hababou] < Ḥabîb, *Ḥmîmu [Benhamimo] < Hebrew Ḥayyim, *‘Nânu [Ananou] and *‘Nûnu < [Anounou] < Emmanuel. Forms in which one of consonants was reduplicated are known in both Berber and Arabic onomastics69. However, numerous forms cited above in this paragraph sound typically Arabic, namely those that start with two consonants followed by /î/, that is, exactly the same beginning as in the most common pattern of constructing Arabic diminutives in Maghreb. Moreover, a large number of diminutives with one consonant reduplicated and ending for men in –u or, rarely, -i and for women in –a or, rarely, -u, was used in eastern Maghreb where any Berber influence is implausible; compare Baybu from Abraham, B(ə)nîni < Benjamin, Daydu < David, Gâgu < Isaac, Qîqi < Jacob, Mûmu < Solomon, Sûsu and Sîsi < Joseph, as well as Lâla < Lea, Tîta and Tîtu < Esther, Kûka, and Jayja, all in Nabeul70. The presence in some of the above forms of the diphthong ay or /î/ in the first syllable, a feature typical for diminutives in the Tunisian dialect of Judeo-Arabic, also links these forms to the Arabic rather than the Berber morphology. Note also that Jews from Tripoli used among others the following male names: Dâdi, Da‘di, Dûdu, Kûku, Sâsi, Tâtu, Tîtu, as well as Bîši and Dâna71.

  • 72 D. Corcos, Studies..., pp. 183-191.
  • 73 Examples taken from René de Segonzac, Voyages au Maroc (1899-1901), Paris, A. Colin, 1903, p. 97.
  • 74 M. Cohen, Higgid..., p. 224.

40In a series of Jewish diminutive forms in Morocco, the initial B(ə) was added to the truncated forms of basic given names, sometimes, the final vowel was added as well: Bâli < Elijah, Bakha(y), Bəddûkh [Beddoukh], Bkhâš, and Bkhas < Mordecai, Bâfu < Joseph, and Bakhlöf < Makhlûf72. This pattern was borrowed from Berbers: compare such Muslim Berber forms found in Morocco as male “Baadi” (Ba‘di) < Arabic Sa‘îd and “Bali” (Bâli) < Arabic ‘Alî, female “Berri” (Bərri) < Məryəm73. Several names from the same category are also known in Tripoli: Bâbâham < Abraham and Bâli74.

  • 75 See A. Beider, A dictionary..., p. 183-188.

41Numerous other patterns of creation of diminutives used by Jews in Maghreb are purely Arabic75.

6. Surnames with Berber morphologic elements

6.1 Surnames derived from Berber words or toponyms

42In Libya, the largest Berber Muslim group lives in the northwestern part of the country in the Nafusa Mountains. Its dialect is called Tanfusit in Berber and Nafusi in Arabic. No surname seems to be related to that idiom. A few Jewish surnames are connected to that region being derived from local toponyms such as Nefusi, Ghariani, and Seroussi. Here, the whole derivation surname is as follows: (Berber toponym) > (Arabic toponym) > (Arabic adjective based on the toponym) > (Jewish surname). As a result, these surnames should be counted among those of Arabic origin. The ultimate Berber source of the toponym is a fact of the science that studies Berber and Arabic toponyms. For the study of Jewish surnames in question this Berber source is pre-historical.

43In Tunisia, Berbers are marginal and a few surnames of this linguistic origin known in that country are due to migrants from Algeria (Atelan, Zemmour) or Morocco (Azoulay, Bouzeldou, Melloul).

  • 76 This section gives the Hebrew spellings of surnames if they are available. These spellings are usef (...)
  • 77 J.-M. Dallet, Dictionnaire..., p. 948.
  • 78 Solomon ben Ṣemaḥ Duran, Tashbaṣ IV: Ḥut ha-Meshullash, Amsterdam, 1738/1739, n° 6.
  • 79 J.-M. Dallet, Dictionnaire..., p. 839.
  • 80 J.-M. Dallet, Dictionnaire..., p. 518.
  • 81 M. Taïfi, Dictionnaire..., p. 878.
  • 82 Lloyd Cabot Briggs and Norina Lami Guède, No more forever: a Saharan Jewish town, Cambridge MA, Pea (...)
  • 83 M. Taïfi, Dictionnaire..., p. 165.

44Three major Berber idioms are spoken by Algerian Muslims: (1) Kabyle (also called Taqbaylit) in the Northeast, (2) Shawi (also called Tacawit), in the mid-East, (3) Tuareg in the South. The influence made by these languages on the corpus of Jewish surnames from Algeria is small: Jews primarily lived in areas where the Arabic-speaking population was predominant. Only a few names from these territories are based on Berber words. One of them is Zemmour (זמור)76; compare Kabyle azemmur ‘olive tree’77. The earliest reference to this name comes from the town of Biskra, in the area populated by Shawi speakers (16th century)78, (אטלאן) Atelan. from Kabyle aṭelan ‘Italian’79, seems to originate in the Constantine area. In the same area we also find references to the surname Aouat (עווט ,עוואט), the Berber tribe Aouat, and the village of Ouled Aouat. The surname is related either to this Berber toponym or directly to the tribe name. Amray, known only in the city of Algiers, is derived from the Kabyle word amṛay meaning ‘head, director’80. The etymology of Ainouz (עינוז) is uncertain. Its earliest references are from Algiers and the region of Constantine. It is unclear whether the coincidence between this surname and the male given name ‘Aynuz used by Muslim Berbers in Morocco81 is fortuitous or not. Guerbaz/Guerbas (Algiers) is derived from gerbas meaning ‘hardened goatskin water bag’ in the vernacular idiom used by the population of Mzab, central Algeria82. This word is surely related to Berber Tamazight agerbuz ‘goatskin’83. However, its exact form, without the initial aand with a different vowel in the last syllable both indicate that, perhaps, we are dealing with a dialectal Arabic word borrowed from Berbers. As a result, the assignment of the surname Guerbaz to the category of surnames based on Berber idioms is doubtful. Amahoua, also from Algiers, is of unclear origin. Its Berber origin is plausible because its first letters coincide with the Berber prefix am-. No Tuareg-based Jewish surname is known.

  • 84 Simon Lévy, Parlers arabes des Juifs du Maroc. Histoire, sociolinguistique et géographie générale, (...)
  • 85 See section 7.
  • 86 E. Ibáñez, Diccionario..., p. 35.
  • 87 Edmond Destaing, Dictionnaire français-berbère (dialecte des Beni-Snous), Paris, E. Leroux, 1914, p (...)
  • 88 In this list, only one source for Tamazight is used: M. Taïfi, Dictionnaire.... It is by far more d (...)
  • 89 M. Taïfi, Dictionnaire..., p. 38.
  • 90 M. Taïfi, Dictionnaire..., p. 252.
  • 91 M. Taïfi, Dictionnaire..., p. 351.
  • 92 M. Taïfi, Dictionnaire..., p. 354.
  • 93 M. Taïfi, Dictionnaire..., p. 408, A. Jordan, Dictionnaire..., p. 31.
  • 94 M. Taïfi, Dictionnaire..., p. 417, M. Dray, Dictionnaire..., p. 238
  • 95 M. Taïfi, Dictionnaire..., p. 447, M. Dray, Dictionnaire..., p. 60.
  • 96 M. Taïfi, Dictionnaire..., p. 639, A. Jordan, Dictionnaire..., p. 108.
  • 97 A. Jordan, Dictionnaire..., p. 155.
  • 98 M. Taïfi, Dictionnaire..., p. 82, M. Dray, Dictionnaire..., p. 277.
  • 99 A. Jordan, Dictionnaire..., p. 46, M. Dray, Dictionnaire..., p. 430.
  • 100 A. Jordan, Dictionnaire..., p. 51, M. Dray, Dictionnaire..., p. 238.
  • 101 A. Jordan, Dictionnaire..., p. 52.
  • 102 M. Taïfi, Dictionnaire..., p. 805
  • 103 M. Taïfi, Dictionnaire..., p. 449.
  • 104 M. Taïfi, Dictionnaire..., p. 122.
  • 105 E. Destaing, Etude..., p. 291.
  • 106 M. Taïfi, Dictionnaire..., p. 684
  • 107 M. Taïfi, Dictionnaire..., p. 684
  • 108 M. Taïfi, Dictionnaire..., p. 803.
  • 109 M. Taïfi, Dictionnaire..., p. 390.
  • 110 M. Taïfi, Dictionnaire..., p. 117.
  • 111 M. Taïfi, Dictionnaire..., p. 122.
  • 112 M. Taïfi, Dictionnaire..., p. 416
  • 113 M. Taïfi, Dictionnaire..., p. 83.
  • 114 A. Jordan, Dictionnaire..., p. 107.
  • 115 M. Dray, Dictionnaire..., p. 337.

45Morocco represents the only country in which Berber languages made an important contribution to the corpus of Jewish surnames. This fact is closely related to the widespread use of these idioms by Muslim population. Even today, after centuries of Arabization, Berbers account for about one half of the population of the country. References to Berber-speaking Jews are known since the Middle Ages. For thousands of Moroccan Jews, Berber was their vernacular language even at the beginning of the 20th century. According to one evaluation, in 1912 its Jewish population of Morocco was composed of about 77,000 Arabic-speakers, 16,000 speakers of Judeo-Spanish (also called Haketia), and 8,000 Berber-speakers84. However, Jews who spoke Berber languages were also usually fluent in Arabic and, moreover, for some of them, it was their first language85. In Morocco, one usually distinguishes three principal Berber idioms: (1) Tarifit (also called Riffian) in the Northeast (Rif Mountains); (2) Tamazight in the mid-eastern provinces (Central Atlas Mountains), and (3) Tashelhit (also called Shilha or Chleuh) in the South. Among surrnames based on Tarifit are: Afenjar / Ifenzar (איפנזאר) > afenjar ‘cup’86 and Izerzer (יזרזר ,אזרזר) > izerzer ‘gazelle ‘87. The following surnames are based either on Tamazight or Tashelhit88: Abettan (אבטאן) > abeṭṭan ‘item of clothing’, ‘skin’89; Aferiat (אפרייאט) > aferiaḍ ‘hare-lip’; Aharfi (אחרפי)> aḥerfî ‘eaten neat (alone and/or without condiments)’90; Aksas (אכסאס) > aksas ‘sheep with short curly wool’91; Aksoul (אכסול) > singular of akšuln ‘arm muscles’92; Amghar (אמגאר) > amḡar ‘old man’93; Amlal (אמלאל) > amlal ‘gazelle’94; Amouzig (אמוזיג) > amaziḡ ‘Berber’95; Amsallem (אמסלם ), perhaps < agent noun based on the verbal root sellem ‘to resign’, ‘to greet’96; Amzallag (אמזלג) > agent noun constructed from the verbal root zleg ‘to make tow ropes’97; Aoudai (אודאי) > uday ‘Jew’98; Assouline (אסולין) > asulil ‘rock’99, with a dissimilation of consonants from lil to lin; Azencot (אזנקוט) > azenkoḍ, azenkuḍ ‘gazelle’100; Azeroual (אזרוואל ,אזרואל) > azerual ‘blue’ (about eyes)101; Azoulay (אזולאי) > azulay ‘frizzy, kinky’102 Ben Mezour (בן מזור) > amezzur ‘broom’103 + Arabic prefix ben ‘son of’; Benakdar/Ben Akdar (בן אקדר) > aqeddar ‘potter’104 + Arabic prefix ben ‘son of’; Bohbot (בוחבוט) bu ‘man of’ + aḥebbuḍ ‘belly’105; Boucidan (בוסידאן) > bu šiḍan ‘alluvium’106; Boughanim (בוגאנים) > bu(w)ḡanim ‘flutist’107; Bouzeldou < bu ‘man of’ + zzelḍ ‘poverty, indigence’108; Ellouz (ילוז ,אילוז) < (e)lluz‘almonds’ (collective)109; Ifenzi (יפנזי) > ifenzy ‘toes’, ‘foot tip’110; Iferghan (אפרגאן ,יפרגאן) < ifergan ‘fences, hedges’111; Melloul (מלול) > word derived from the verbal root mellul ‘to be white’112; Ouddiz < uddiz ‘fist’113; Siksou (סיכסו ,סקסו) > seksu ‘couscous’114; Tamsout (תמשות) > tameššut ‘food’115; Timeztit and its variant Timsit, perhaps < the village of Timstiggit (southwestern Morocco). In the above list, certain surnames are derived from Berber words not directly, but via toponyms that in turn are based on these Berber words. It is almost surely the case of Iferghan and Melloul and can be the case for Assouline and Boucidan. The surname Aflalou (אפלאלו ) can be derived from the Berber village Afelilou in Central Atlas or the name of the region of Tafilalet in southeastern Morocco (Berber tafilalt).

46Surnames like Amsili, Drai, Filali, Roudani, Tedghi, and Zagouri all have toponyms of Berber origin as their roots. However, all of them end in the Arabic suffix –î designating inhabitants of the corresponding places. As a result, these surnames should be counted among those having Arabic (rather than Berber) etymons.

47Another series of Jewish surnames of Berber origin encompasses those starting with the patronymic prefix u- ‘son of’ (spelled ou or o n French). All of them originated in Morocco. For one part of these surnames this prefix is placed before the male given name: Ouanounou (ואענונו) > *’(A)nûnu (diminutive of Emmanuel), Ohamou < Ḥammu, Ouhayoun (וחיון) > Ḥayyûn, Ouahnoun (ווחנון) > Ḥ(a)nûn, Ousday < Ḥasday, Oussadon < Sa‘dûn, and Ouyoussef (ויוסף) > Yûsəf. The surname Ohnouna (וחנונא) is based on the female given name Ḥ(a)nûna. Ouhanna/Ohana (אוחנא) and Ouhnich (וחניש) can be derived from either male or female given names starting in Hebrew with חנ.

  • 116 P. Wexler (The Non-Jewish origins..., p. 128) points to the existence of the surnames Ohayon and Be (...)
  • 117 It is usually considered that Aknin represents a Berber diminutive of Jacob (A. Laredo, Les noms... (...)

48The majority of these surnames include Ou- from their inception. However, we cannot exclude the scenario according to which Ou- was added to an already existing surname. For example, Hayo(u)n (חיון) was commonly used as a given name by Jews in Maghreb. It is composed of the Arabic root ḥayy حَيّliving’ and the Arabic suffix –ûn. Yet, it was also a common surname (derived from the given name). Moreover, the forms Hayon and Aben Hayon (Arabic for ‘son of Hayon’) are both attested as Jewish surnames in medieval Spain. As a result, Ouhayoun/Ohayon does not necessarily mean ‘son of someone whose given name is Hayo(u)n’. It can also represent a secondary surname meaning ‘son of someone surnamed Hayo(u)n’116. For Ouakrich, its secondary character is even more plausible than for Ouhayoun: compare the medieval Iberian Jewish surname Acrix. The morphologic variation could also take place in the opposite direction. According to the Algerian civil records, during the 19th century in the city of Bône at least in one family the original form Ouaknine (ואעקנין), with the root of unclear origin, lost its Berber prefix giving rise to the form Aknine117.

49In Berber morphology, the prefix u- (singular) / ayt- (plural) is also used to construct words designating the names of inhabitants of various places. A group of surnames with u-derived from toponyms includes those coming from the following regions of Morocco: (1) southwestern: Outguergoust (וותגרגושת) > the village of Taguergoust, Outmezguine (וותמזגין) > the village of Timzghouine, and Outmezguit (וותמזגית) > the village of Timzguida, (2) southeastern: Ouizman (ויזמאן), perhaps < the tribal area of Aït Ismen, and (3) central: Ouizgane (וויזגאן) > the tribal area of Aït Ouizgane.

  • 118 M. Taïfi, Dictionnaire..., p. 539.

50The surname Ouakrat (ואקראט) also seems to start with the same Berber prefix. It can be an occupational name derived from the Berber word aqraḍ ‘slicing, chopping, carving’118.

51All the above surnames starting in O(u)- appeared in the communities whose vernacular idiom was Berber. In these cases, the linguistic origin of the given names, toponyms, and primary surnames in question is irrelevant. For example, Jews and Muslims called Youssef lived in the Berber-speaking communities. As a result, one can consider that this given name was a Berber borrowing from Arabic. In Arabic, /yûsəf/ (the phonetic transcription of Youssef) represents the vernacular form of Classical Arabic yûsuf that in turn is derived from the biblical Joseph, of Hebrew origin. Yet, the surname Ouyoussef is just Berber. The Arabic and the ultimate Hebrew origins of the given name on which this surname is based are pre-historical for it.

  • 119 The dates are mainly taken from J. M. Toledano, Ner... and J. Ben Naim, Malkhei... (see exact refer (...)

52The age of numerous surnames having Berber morphologic elements is uncertain. The oldest known references are: (1) circa 1500: Abouzaglou, Azoulay, and Ben Akdar; (2) 16th century: Assouline and Zemmour; (3) 16th -17th centuries: Ben Ifenzi, Ohana, and Zemmour; (4) 17th century: Aflalo, Ainouz, Amzallag, Azeroual, Ben Amouzig, and Boucidan, and (5) 17th -18th centuries: Azencot, Bohbot, Melloul, Ouyoussef (as a part of the double surname Ben David Ouyoussef), and Tamsout119.

53A number of surnames of Berber origin including Aflalo, Amzallag, Assouline, Outguergoust, Outmezguine, Tamsout, Timsit, and Timeztit seem to be exclusively Jewish. Moreover, these names can, in theory, be monogenetic.

6.2 Linguistic characteristics of the Berber-based surnames

54Several morphologic peculiarities of Berber idioms help identifying names of Berber origin. A large majority of Berber masculine nouns starts with a vowel: most often /a/. Those starting with /i/ are less common, while those in /u/ are unusual. Examples: Abettan, Aferiat, Aherfi, Amghar, Assouline, Azeroual, Azoulay, and, possibly, Aflalo. Forms starting in a consonant—like seksu [Siksou] are rare. However, in certain cases, the initial vowel was lost giving rise to other forms starting with a consonant as the aforementioned Zemmour, from azemmur ‘olive tree’. The feminine nouns always start in /t/ and most often also end in /t/ too. The diminutive forms of masculine nouns are also constructed following the same pattern: the addition of /t/ at the beginning and the end of the noun. Names like Tamsot and Timeztit are related to one of these two patterns: feminine or diminutive. The addition of the prefix am represents one of the main methods of constructing Berber nouns based on verbs. Amzallag, Amsellem, and Amahoua can be related to this pattern. If the singular masculine form starts in /a/, the plural usually changes the initial vowel to /i/ and includes an additional suffix –en or – an. The name Iferghan is a typical illustration of that pattern. It coincides with the Berber word meaning ‘fences’ (singular afrag).

  • 120 Compare Tamazight azzaglu (M. Taïfi, Dictionnaire..., p. 797) and Kabyle azaglu (J.-M. Dallet, Dict (...)
  • 121 Alfred Louis de Premare et al., Langue et culture marocaines - Dictionnaire arabe-français, 12 vols (...)

55A series of names such as Boughanem ‘flutist’ (literally: ‘man of flute’, from aḡanim ‘flute’), Bou Habout ‘big eater’ (literally: ‘man of belly’), and Bouzeldou ‘man of poverty’ illustrate the morphologic borrowing by Berber of the Arabic vernacular prefix ‘man of’ (literally: ‘father of’). On the surface, the surname Bouzaglou (בוזאגלו) seems to belong to the same series. Indeed, az(z)aglu means ‘yoke’ in several Berber idioms120. However, this word was borrowed from Berber to the Moroccan dialect of Arabic where it is pronounced zâglu121. Moreover, we find the surname variant Abouzaglou (אבוזאגלו ,known already in the 15th century), while nothing indicates that the full form abû was at some moment a loanword in Berber. The Arabic vernacular form (a)bu-zâglu also fits all known forms of the surname (including the above Hebrew spellings) much better than the Berber bu(az)zaglu. As a result, most likely, the direct etymon for this surname was Arabic rather than Berber.

  • 122 A.L. de Premare, Langue..., vol. 6, p. 139.

56Another example of borrowing from Berber to the Arabic dialect in Morocco of examples is /səkso/ كسكس’ couscous’122. As a result, in theory, this word could be the direct etymon for the surname Siksou, though nothing precludes the direct Berber origin of this surname. The borrowing in the opposite direction existed as well. For example, the Berber ellouz‘almond’ (the etymon of the surname Ellouz) comes from Arabic al-lûz الّوز ‘the almond’.

  • 123 A.L. de Premare, Langue..., vol. 6, p. 139.

57In Aferiat, Azencot, Bohbot, and Ouakrat we find the final ‘t’ in French spelling and tet (ט) in Hebrew in place where the Berber etymon has /ḍ/. This devoicing is related to a distinctive feature of the Judeo-Arabic pronunciation in Morocco: the /ṭ/-pronunciation (as for the Arabic letter ṭâ’) instead of /ḍ/ (valid in other Judeo-Arabic dialects for both ḍâd and ẓâ’)123.

7. Berber-based names and the history of Jews in Maghreb

7.1 Names and the theory of JUDEO-BERBERS

  • 124 According to this characteristic, the inception of Jewish surnames from North Africa is similar to (...)

58The Jewish given names and surnames of Berber origin discussed in sections 5 and 6 provide no corroboration for the theory of JUDEO-BERBERS. On the one hand, they are simply irrelevant for the question of the possible existence of Judaized Berbers before the arrival of Arabs. On the other hand, the influence of the Berber idioms to the corpus of Jewish names in Maghreb can be easily explained by factors other than the putative genetic link between Jews who lived in North Africa in modern times and the Jewish Berber tribes that could exist in the Early Middle Ages. The surnames of Jews in North Africa evolved on a natural basis, with the gradual transformation of family nicknames or sobriquets into hereditary family names124. In this context, the language on which a surname is based just indicates what was the vernacular (mainly) or cultural (rarely) idiom of either its first bearer or the person who assigned the nickname to the first bearer. According to the information available to us, Jews in North Africa used as their first language the idiom based on the vernacular of the local Gentile population, that is, depending on the region, Arabic, Berber, or Judeo-Spanish (Haketia). As a result, it is quite normal that the Jewish surnames created in Maghreb are either of Arabic or of Berber origin. Haketia, the vernacular language of Jews in northernmost Morocco, seems to make no contribution to the corpus of Jewish surnames in North Africa. The ancestors of these Jews came to Morocco from the medieval Iberia already with surnames.

  • 125 Harvey E. Goldberg, "The social context of North African Jewish patronyms", Folklore Research Cente (...)
  • 126 The last two possibilities are purely theoretical: no historical evidence exists to corroborate the (...)

59Goldberg provides several other arguments against the inappropriate use of names by proponents of the theory of JUDEO-BERBERS125. Firstly, from the point of view of social psychology, it is clear that Judaized individuals and groups would attempt to alter their original names, or take on new “Jewish” names in order to “forget” their earlier identity and to claim links relevant to their new social status. Secondly, some Jews could acquire the name of their Berber lord or the Berber tribe that protected/owned them. Thirdly, surnames derived from Berber toponyms just attest about the Jewish migrations between various areas. Moreover, these toponymic surnames can indicate not only the places of origin, but also those of some (business) connections of their first bearers126.

60A consideration of surnames and given names used by Jews in areas where the Muslim majority is Berber-speaking can also be useful in the discussion of the theory of JUDEO-BERBERS.

  • 127 See N. Slouschz, Un voyage..., pp. 41, 50, 59. The same source (p. 37, 43, 59) also mentions the fo (...)
  • 128 Harvey E. Goldberg, "Patronymic groups in a Tripolitanian Jewish village: reconstruction and interp (...)

61The medieval Jewish inscriptions from the area of the Nafusa Mountains include no surnames. The given names mentioned in them are mainly Hebrew though several names of Arabic origin are known as well: male ‘Ayâd (עיאד),Tamâm (תמאם), and Marzûq (מרזוק), and female Ḥasâna (חסנה)127. In the town of Gharian (area of Nafusa Mountains, Libya), in mid-20th century three surnames covered 86 percent of all local Jews: Ḡwîta‘, Ḥajjâj, and Ḥasân128. All three surnames are based on Arabic words. Doubtless Berber-based names are known for neither medieval, nor modern periods.

  • 129 S. Duran, Tashbaṣ IV..., n° 15, 1, 24, 33.
  • 130 Eliahou Marciano, Les sages d'Algérie: dictionnaire encyclopédique des sages et rabbins d'Algérie, (...)
  • 131 See corresponding entries in A. Beider, A dictionary

62Rabbinical responsa from the 16th century make reference to the following surnames: *Messalati (מסאלתי) in the region of Mzab, al-Ḥâmî (אלחאמי’), Adda (עדא), and Abî Sdîd (אבי סדיד), all three in Touggourt, a Saharan town situated not far from Mzab129. All these surnames are of Arabic and not Berber origin. In Ghardaïa, the largest town of the region of Mzab in central Algeria, before the end of the 19th century (when the Saharan area was annexed by France), local Jews lived in a relative isolation from other Jewish communities and were characterized by a strong endogamy. As a result, during the first half of the 20th century, a large majority of the community members were bearing one of the following surnames: Sellam, Partouche, Attia, Chekroun, Sebban, Elbaz, Lahyani, Meslati, Cohana, and Zenou. But for the obscure name Cohana, all others clearly originated in other areas. Meslati (a variant of the above *Messalati) clearly came from the town of Msallata (northwestern Libya). However, the majority of families seem to originate in Morocco. Indeed: (1) this origin is documented for rabbis Attia, Elbaz, and Lahyani130; (2) the Moroccan origin of Zenou (the only name in the list that can be of ultimate Berber origin being related to the given name used by Berbers) and Chekroun follows from the comparative analysis of references to the variants of these surnames in the Maghreb; (3) the oldest known references to Sebban are also from Morocco; (4) the ancestors of Partouche came from Spain to Morocco bearing the name Sasportas131.

  • 132 Vincent Monteil, "Les Juifs d'Ifrane (Anti-Atlas Marocain)", Hespéris - Archives berbères et buille (...)
  • 133 Paul Pascon and Daniel Schroeter, "Le cimetière juif d’Iligh (1751-1955). Etude des épitaphes comme (...)
  • 134 V. Monteil, "Les Juifs...", p. 154; P. Pascon and D. Schroeter, "Le cimetière...", , p. 47.
  • 135 The list of oldest inscriptions from Ifrane appears in J.M. Toledano, Ner..., p. 4-5 and V. Monteil (...)

63For southernmost Morocco, we have lists of several dozens of surnames used by local Jews during the 19th century and the first half of the 20th century in Ifrane Anti-Atlas132 and Iligh133. In the first community, among the 22 surnames, Aferiat, Boughanim, Iferghan, Ohayon, and, possibly, Knafou are of Berber origin. In Iligh, the list of 32 surnames includes four for which the Berber origin is certain (Amzallag, Azoulay, Iferghan, and Ohayon), and two for which it is possible (Aflalo and Knafou). All other surnames are either Arabic (mainly) or Hebrew (few). Moreover, in both places Jews spoke between them (Judeo-)Arabic using the Berber Tashelhit idiom only in their contacts with local Muslims134. In oldest known tombstone inscriptions from Ifrane we find references to the following non-Hebrew given names (some of which are present in patronymics): male ‘Azzûz, Makhlûf, Mas‘ûd, Maymûn, Mîmûn, Mûsâ, Wâ‘îsh and Ya‘îsh, female Saʕîda, all of Arabic origin; (2) male Isu, ‘Abbu and Mûmu, representing diminutives that follow the Berber patterns known among Arabs too (the last two appear as second names and can actually be not patronymics, but last elements of surnames Bən ‘Abbu and Bən Mûmu, both starting with the Arabic prefix meaning ‘son of’). The list also includes a number of surnames. In addition to aforementioned Aferiat, Ohayon, and Knafou, as well as several Arabic surnames (Chriki, Sebag), all still in use in the same town in the 20th century, we also find Ben Amgar, in which the Arabic patronymic prefix is placed before the Berber Amḡar, and Arabic Siboni135.

64The above data indicate that a large portion of Jews who lived in modern times in Berber-speaking areas actually descended from migrants from the Arabic-speaking regions. In other words, they show a trend that is opposite to that taken for granted by the proponents of the theory of JUDEO-BERBERS.

65A very small percentage of surnames of Berber origin used by Jews in Maghreb in comparison to that covered by names of Arabic origin cannot be considered a valid argument against the concept of JUDEO-BERBERS. Indeed, the notion of hereditary family names is relatively young: generally speaking, very few surnames created in Maghreb are known before the 16th century. As indicated in section 6.1, oldest references to Berber-based surnames also date from the 16th -17th centuries only. In other words, surnames are many centuries younger than the historical period directly concerned by the theory of JUDEO-BERBERS.

  • 136 See details in Alexander Beider, Origins of Yiddish Dialects, Oxford, Oxford University Press, 2015 (...)
  • 137 In the cities of Fez (Louis Brunot and Elie Malka, Textes judéo-arabes de Fès, Rabat, Typo-litho éc (...)

66The absence of the Berber substratum in the domain of given names is much more important from the historical point of view. In contrast to surnames, individual names assigned to babies at their birth have always existed. As a result, for various Jewish groups their corpus is usually cumulative. The oldest layer is biblical. The second layer is post-biblical Hebrew and Aramaic. The following layers correspond to various other vernacular languages used by the ancestors. For example, for Ashkenazi Jews in Western Europe the German/Yiddish layer is preceded in time by the group of Romance given names, substratal for Western Yiddish, based on the vernacular language(s) the ancestors of these Jews spoke before shifting to the Germanbased idiom. In the corpus of Yiddish given names in Eastern Europe, we find a significant Old Czech and a tiny East Slavic substrata, both based on the everyday languages one part of the ancestors of local Jews spoke before shifting to Yiddish136. Yet, as discussed in section 5 no Berber substratum can be discerned for the corpus of Jewish given names in Maghreb. This result makes implausible not only the second position of the theory of JUDEO-BERBERS, but even a more general scenario according to which the ancestors of modern North African Jews (proselytes or not) were Berber-speaking. The absence of any Berber linguistic substratum in Judeo-Arabic dialects of Maghreb137 points to the same direction.

7.2 Linguistic origin of names and their age

  • 138 J. Tolédano, Une histoire..., p. VII.
  • 139 A similar false idea was extrapolated by Chouraqui (Marche..., p. 132) to an absurd conclusion. Una (...)

67Certain authors who wrote about Jewish onomastics in Maghreb were characterized by the general linear vision regarding links between the age of surnames and the languages on which they are based. The surnames derived from Hebrew and Aramaic would be the oldest ones. Some of them would appear in biblical times, other developed in Palestine during the following centuries. To this specifically Jewish substratum, the names related to the Berber language are said to be added when Jewish migrants came to North Africa before the Arabic invasion of this area. The names drawn from Arabic would be older than those derived from Spanish and Portuguese that, in turn, mainly appeared in the Maghreb during the 14th -15th centuries. Such a concept dominates Part 2 of Laredo’s book. Its general aspects are also shared by Tolédano138 who describes the chronology of various linguistic strata of surnames in the following order: Hebrew-Aramaic, Punic, Berber, Arabic, Spanish and Portuguese139. Actually, no global correlation between the age of a surname and the language of its etymon can be postulated. A majority of surnames created in Maghreb are based on a single language: Arabic.

68A much smaller number of surnames are due to Berber languages. All other languages are really marginal. The difference between Arabic and Berber names is primarily geographic: as discussed above, but for rare exceptions, the latter are limited to Morocco. No information is available to us that would suggest that Berber names are older than the Arabic ones. We have no evidence that before the Arabic invasion of North Africa the local population, Jewish or non-Jewish (Berber), was using hereditary family names. Since Berber was spoken in large areas of Morocco and Algeria well after the invasion and is still used as a vernacular language today, Jews could have acquired their surnames of Berber origin much more recently.

8. Conclusion

69During the 20th century, a number of authors who wrote about the history of Jews in Maghreb adhered to the theory of JUDEO-BERBERS. The two major postulates of this theory are: (1) numerous Berber tribes adhered to Judaism before the Arabic invasion to Maghreb during the 8th century; (2) Jews who lived in Maghreb during the last centuries partly descend from these Berber proselytes.

70The information present in this article shows that the Berber origin of one given name and several dozens of Jewish surnames from Morocco, as well as a few surnames in eastern Algeria is doubtless. Yet, the geography of these Berber-based names used by Jews perfectly fits the geography of the settlement of the Berber-speaking Muslim population groups in North Africa. We also know that in modern period Jews lived in the territories in question. As a result, these names appeared in the Jewish communities that used a Berber idiom as their vernacular language. Nothing indicates that these surnames existed already in the Middle Ages. All onomastic arguments suggested by proponents of the theory of JUDEO-BERBERS attempting to link these names—and numerous others, actually unrelated to Berbers—to the Berber proselytes to Judaism are untenable.

71This article does not state that both major positions of the theory of JUDEO-BERBERS are false. Actually any of them or even both can be accurate. This text refutes the onomastic arguments considered by various authors as “corroborations” of this theory. Today our knowledge concerning the first position of this theory is not significantly richer than the information known to Ibn Khaldûn during the 14th century. Indeed, no historical, archeological, linguistic, or onomastic elements have been discovered since his time that would corroborate or refute the existence in the seventh century of, at least, seven tribes of Judaized Berbers in North Africa as well as the exact names of the tribes in question. The second position of the theory of JUDEO-BERBERS stays purely speculative. No information in our possession allows asserting that some Berber proselytes to Judaism should be or not be counted among the ancestors of Jews who lived in Maghreb during the last centuries. Yet, the consideration of the corpus of given names and the linguistic features of Judeo-Arabic dialects shows the absence of the Berber substrata in this corpus and the idioms in question. This factor makes highly implausible the idea about the genetic contribution of the Berber-speaking ancestors, proselytes or not, being significant for the Judeo-Arabic speaking Jews who lived in modern times in North Africa.

Top of page

Bibliography

ABBOU, Isaac D., Musulmans andalous et judéo-espagnols, Casablanca, Antar, 1953.

AYOUN, Richard, “Les Judéo-berbères entre mythe et réalité”, in About the Berbers – history, language, culture and socio-economic conditions, ed. Bo Isaksson and Marianna Laanatza, Upssala, Upssala Universitet, 2004, pp. 74-88.

BEIDER, Alexander, A Dictionary of Jewish Surnames from the Mediterranean Region. Vol. 1: Maghreb, Gibraltar, and Malta, New Haven, Avotaynu Inc., 2017.

BEIDER, Alexander, Origins of Yiddish Dialects, Oxford, Oxford University Press, 2015.

BRESC, Henri, Arabes de langue, juifs de religion: l'évolution du judaïsme sicilien dans l'environnement latin, XIIe-XVe siècles, Paris, Bouchène, 2001.

BRIGGS, Lloyd Cabot, GUÈDE, Norina Lami, No more forever: a Saharan Jewish town, Cambridge, Peabody Museum, 1964.

BRUNOT, Louis, MALKA, Elie, Textes judéo-arabes de Fès, Rabat, Typo-litho École du Livre, 1939.

CALLASANTI-MOTYLINSKI, Adolphe de, Le Djebel Nefousa. Transcription, traduction française et notes avec une étude grammaticale, Paris, Ernest Leroux, 1898.

CAZÈS, Davis, Essai sur l’histoire des israélites de Tunisie, Paris, A. Durlacher, 1888.

CHOURAQUI, André, Marche vers l’occident: Les Juifs d’Afrique du Nord, Paris, PUF, 1952.

COHEN, David, Le parler arabe des juifs de Tunis. Vol. 1: Textes et documents linguistiques et ethnographiques, The Hague-Paris, Mouton, 1964.

COHEN, Mordecai, Higgid Mordecaï (ed. Harvey E. Goldberg), Jerusalem, Ben Zvi Institute, 1981.

COLIN, Georges Séraphin, Arabe marocain, Aix-en-Provence, Edisud, 1999.

CORCOS, David, Studies in the history of the Jews of Morocco, Jerusalem, R. Mass, 1976.

DALLET, Jean-Marie, Dictionnaire kabyle-français, Paris, Selaf, 1982.

DESPOIS, Jean, Le Djebel Nefousa (Tripolitaine), étude géographique, Paris, Larose, 1935.

DESTAING, Edmond, Dictionnaire français-berbère (dialecte des Beni-Snous), Paris, E. Leroux, 1914.

DESTAING, Edmond, Etude sur la Tachelḥît de Soûs -Vocabulaire français-berbère, Paris, E. Leroux, 1920.

DIANOUX, Hugues Jean de, "Remarques à propos de l’origine apparemment berbère des noms de familles juives d’Afrique du Nord", in Juifs et judaïsme en Afrique du Nord dans l’antiquité et le haut Moyen-Âge, ed. Carol Lancu and Jean-Marie Lassère, Montpellier, Centre de Recherches et d’Études Juives et Hébraïques, 1985, pp. 103-111.

DRAY, Maurice, Dictionnaire français-berbère – Dialecte de Ntifa, Paris, L’Harmattan, 1998.

DURAN, Simon ben Ṣemaḥ, Tashbaṣ I-III, Amsterdam, 1738-1739.

EISENBETH, Maurice, Les Juifs de l'Afrique du Nord – Démographie et onomastique, Algiers, 1936.

GAUTIER, Emile-Félix, Le passé de l’Afrique du Nord: Les siècles obscurs, Paris, Payot, 1964.

GOLDBERG, Harvey E., "Patronymic groups in a Tripolitanian Jewish village: reconstruction and interpretation", Jewish Journal of Sociology 9 (1967), pp. 209-226.

GOLDBERG, Harvey E., "The social context of North African Jewish patronyms", Folklore Research Center Studies 3 (1972), pp. 245-258.

HAMET, Ismaël, Les Juifs du nord de l'Afrique (noms et surnoms), Paris, Société d'Éditions Géographiques, Maritimes et Coloniales, 1928.

HAYOUN, Victor, "Jewish names, surnames, and nickname of Nabeul, Tunisia", in Pleasant are their names. Jewish names in the Sephardi diaspora, ed. Aaron Demsky, Bethesda, University Press of Maryland, 2010, pp. 176-178.

HIRSCHBERG, Haim Zeev, "The problem of the Judaized Berbers", Journal of African History 4 (1963), pp. 313-339.

IBÁÑEZ, Esteban, Diccionario Rifeño-Español, Madrid, Instituto de Estudios Africanos, 1949.

IBN KHALDOUN, Histoire des Berbères et des dynasties musulmanes de l’Afrique Septentrionale, Algiers, Imprimerie du Gouvernement, 1852.

JORDAN, Antoine, Dictionnaire berbère-français (dialectes techelhit), Rabat, Omnia, 1934.

LAREDO, Abraham I., Les noms des juifs du Maroc – Essai d'onomastique judéo-marocaine, Madrid, Instituto Benito Arias Montano, 1978.

LÉVY, Simon, Parlers arabes des Juifs du Maroc. Histoire, sociolinguistique et géographie générale, Zaragoza: Instituto de Estudios Islámicos y del Oriente Próximo, 2009.

MARCIANO, Eliahou, Les sages d'Algérie: dictionnaire encyclopédique des sages et rabbins d'Algérie, du haut moyen-âge à nos jours, Marseille, IMMAJ, 2002.

MARTY, Paul, "Folklore tunisien. L’onomastique des noms propres de personnes", Revue des études islamiques 4 (1936), pp. 363-432.

MONTEIL, Vincent, "Les Juifs d'Ifrane (Anti-Atlas Marocain)", Hespéris - Archives berbères et builletin de l’Institut des Hautes-Etudes Marocaines 35 (1948), pp. 151-162.

NAIM, Joseph ben, Malkhei Rabbanan, Jerusalem, Abecassis, 1931.

PASCON, Paul, SCHROETER, Daniel, "Le cimetière juif d’Iligh (1751-1955). Etude des épitaphes comme documents d’histoire sociale", Revue de l’Occident musulman et de la Méditeraanée 34 (1982), pp. 39-62.

PÉREZ I PONS, Eduard (ed.), Fonts pera l'estudi de la comunitat jueva de Mallorca: regesta i bibliografia, Barcelona, PPU, 2005.

PREMARE, Alfred Louis de et al., Langue et culture marocaines - Dictionnaire arabe-français, 12 vols, Paris, L’Harmattan, 1993-1999.

SCHIMMEL, Annemarie, Islamic Names. Edinburgh, Edinburgh University Press, 1995.

SCHROETER, Daniel J., "La découverte des juifs berbères", in Relations Judéo-Musulmanes au Maroc: perceptions et réalités, ed. Michel Abitbol, Paris, Stavit, 1997, pp. 169-187.

SEBAG, Paul, Les noms des juifs de Tunisie. Origine et significations, Paris, l'Harmattan, 2002.

SEGONZAC, René de, Voyages au Maroc (1899-1901), Paris, A. Colin, 1903.

SLOUSCHZ, Nahum, Hébræo-phéniciens et Judéo-Berbères. Introduction à l’histoire des juifs et du judaïsme en Afrique, Paris, Ernest Leroux, 1908.

SLOUSCHZ, Nahum, Travels in North Africa, Philadelphia, The Jewish Publication Society of America, 1927.

SLOUSCHZ, Nahum, Un voyage d’études juives en en Afrique, Paris, C. Klincksieck, 1909.

TAÏEB, Jacques, "Juifs du Maghreb: onomastique et langue, une composante berbère?", in Encyclopédie Berbère, 26, Aix-en-Provence, Edisud, 2004, pp. 3969-3975.

TAÏFI, Miloud, Dictionnaire tamazight-français (parlers du Maroc central), Paris, L’Harmattan-Awal, 1991.

TALBI, Mohamed, "Un nouveau fragment de l'histoire de l'Occident musulman, 62-196 / 682-812, L'épopée d'Al Kahina", Cahiers de Tunisie 19.73-74 (1971), pp. 19-52.

TAVARES, Maria J. P. Ferro, Os judeus em Portugal no século XV, vol. 2, Lisboa, Universidade Nova de Lisboa, 1984.

TOLÉDANO, Joseph, Une histoire des familles – Les noms de famille des juifs d'Afrique du Nord: Maroc, Algérie, Tunisie; des origines à nos jours, Jerusalem, Ramtol, 1998.

TOLEDANO, Jacob Moses, Ner ha-Ma‘arav, Jerusalem, Lunz, 1911.

WEHR, Hans, A Dictionary of modern written Arabic, Ithaca, Spoken Language Services, 1976.

WEXLER, Paul, The Non-Jewish origins of the Sephardic Jews, Albany, State University of New York, 1996.

Top of page

Notes

1 About some of these aspects see Daniel J. Schroeter, "La découverte des juifs berbères", in Relations Judéo-Musulmanes au Maroc: perceptions et réalités, ed. Michel Abitbol, Paris, Stavit, 1997, pp. 169-187.

2 The most detailed scientific discussion of the historical aspects of the issue of JUDEO-BERBERS appears in Haim Zeev Hirschberg, "The problem of the Judaized Berbers", Journal of African History 4 (1963), p. 313- 339. The following quotes from medieval Arabic authors are taken from his text (p. 314-315).

3 The classical French translation of this passage by de Slane (Ibn Khaldoun [Ibn Khaldûn], Histoire des Berbères et des dynasties musulmanes de l’Afrique Septentrionale, Algiers, Imprimerie du Gouvernement, 1852, vol.1, pp. 208-209) is inexact. It omits the expression “it is possible” (see Mohamed Talbi, "Un nouveau fragment de l'histoire de l'Occident musulman, 62-196 / 682-812, L'épopée d'Al Kahina", Cahiers de Tunisie 19 (1971), 73-74, p. 42).

4 See Ibn Khaldoun, Histoire..., vol. 1, p. 198, 208, 213-214, 340-342 about Al-Kâhina, and p. 207-209, 211, 215 about Christian and pagan Berbers.

5 Nahum Slouschz, Hébræo-phéniciens et Judéo-Berbères. Introduction à l’histoire des juifs et du judaïsme en Afrique, Paris, Ernest Leroux, 1908, p. 339, 340, 354.

6 Richard Ayoun (“Les Judéo-berbères entre mythe et réalité,” About the Berbers – history, language, culture and socio-economic conditions (ed. Bo Isaksson and Marianna Laanatza), Upssala, Upssala Universitet, 2004, pp. 74-88) accurately indicates the absence of any reliable source confirming the historicity of Al-Kâhina. Yet, he makes the same misinterpretation as Slouschz stating that Ibn Khaldûn indicates the Jewish religion professed by Al-Kâhina

7 See M. Talbi, "Un nouveau fragment...”, p. 42.

8 One can also note that Ibn Khaldûn (Histoire..., vol. 1, p. 177) quotes the Arabic geographer Al-Bakri (11th century) who asserted that Berber people were originally chased from Syria by Jews. As pointed above, links to Syria and “powerful Jews” also appear in the aforementioned sentence where Ibn Khaldûn discusses the origins of Judaism among Berbers. If to this observation we add the fact that this author describes the genealogy of Berbers in a style similar to biblical genealogies and making regular references to the biblical text (compare Ibn Khaldûn, Histoire..., vol. 1, pp. 169-179), one cannot exclude the possibility that the whole Jewish connection of Berbers was introduced by this author for symbolic reasons.

9 Compare, for example, Hans Wehr, A Dictionary of modern written Arabic, Ithaca, Spoken Language Services, 1976, p. 844. The same translation also appears in the footnote to the text by Ibn Khaldûn made by the French translator de Slane (Ibn Khaldoun, Histoire..., vol. 1, p. 340) where the Hebrew word Cohen is also mentioned. According to the quotes he provides, Slouschz used the same edition of the book by Ibn Khaldûn and consequently he could not overlook the footnote in question. Clearly, he ignored the direct Arabic explanation that was not fitting his own theory. The approach by Emile-Félix Gautier (Le passé de l’Afrique du Nord: Les siècles obscurs, Paris, Payot, 1964, pp. 210-211, 256) is similar enough. He (incorrectly) states that “Kahena” means ‘female priest’ in Arabic, Hebrew, and Punic, favors the Hebrew derivation and proclaims that the link between the Zenata Berber group (to which belonged Jarâwa) and the Judaism is undeniable. Actually, the Arabic ‘soothsayer’ and the Hebrew כהן (ּCohen) ‘priest’ are two independent words. Yet, the coincidence between their three consonants is not totally fortuitous: Arabic and Hebrew are related Semitic languages and the words in question are based on cognate roots.

10 See N. Slouschz, Hébræo-phéniciens..., pp. 449-451 and Nahum Slouschz, Un voyage d’études juives en en Afrique, Paris, C. Klincksieck, 1909, p. 13.

11 Ibn Khaldûn Histoire..., vol.1, p. 209, 215.

12 N. Slouschz, Hébræo-phéniciens..., p. 412-447

13 Davis Cazès, Essai sur l’histoire des israélites de Tunisie, Paris, A. Durlacher, 1888, p. 169-179.

14 N. Slouschz, Hébræo-phéniciens...

15 Among them are Hamet (Les Juifs...), Eisenbeth (Les Juifs...), and Tolédano (Une histoire...) (see section 4).

16 Paul Sebag, Les noms des juifs de Tunisie. Origine et significations, Paris, l'Harmattan, 2002.

17 D. Cazès, Essai..., p. 175.

18 The comparison of the two sources yields the following results: (1) 58 names appear in both lists, (2) 10 names listed by Cazès are not present in the list by Slouschz, (3) 13 names listed by Slouschz appear in other lists given by Cazès (Essai..., pp. 175-178), that is, among surnames whose etymology was clear to the author, and (4) only three surnames in the list by Slouschz (Arbib, Lalo, and Nefoussi) seem to be absent from the book by Cazès. Arbib and Lalou (equivalent to Lalo) appear in other sources from Tunisia. The “rigor” of the scholarship of the list provided by Slouschz can also be gauged by the fact that it includes an important number of spelling errors in comparison to the text by Cazès including Dania < the original form Dana, Fregoui < Fregoua, Messari < Messas, Metoch < Mettodi, Ouzani < Ouzan, Saied < Souid, Serous < Serour, Schoui < Stioui, Sinourf < Sinouf, Sir < Sis, Sitrouq < Sitrouk, Toubiani < Toubiana, and Zerousi < Zerouk.

19 Isaac D. Abbou, Musulmans andalous et judéo-espagnols, Casablanca, Antar, 1953, p. 391

20 Compare Adolphe de Callasanti-Motylinski, Le Djebel Nefousa. Transcription, traduction française et notes avec une étude grammaticale, Paris, Ernest Leroux, 1898, pp. 117-120 and Jean Despois, Le Djebel Nefousa (Tripolitaine), étude géographique, Paris, Larose, 1935.

21 Mordecai Cohen, Higgid Mordecaï (ed. Harvey E. Goldberg), Jerusalem, Ben Zvi Institute, 1981, p. 224.

22 Eduard Pérez i Pons (ed.), Fonts pera l'estudi de la comunitat jueva de Mallorca: regesta i bibliografia, Barcelona, PPU, 2005, p. 196.

23 Simon ben Ṣemaḥ Duran, Tashbaṣ I-III, Amsterdam, 1738/1739, n° I.21

24 Henri Bresc, Arabes de langue, juifs de religion: l'évolution du judaïsme sicilien dans l'environnement latin, XIIe-XVe siècles, Paris, Bouchène, 2001, p. 87, 289, 44, 151

25 Paul Wexler (The Non-Jewish origins of the Sephardic Jews, Albany, State University of New York, 1996, p. 127) misuses the medieval Iberian reference to a Jew called Nafuci as “evidence” about the Berber origin of certain Spanish Jews.

26 N. Slouschz, Un voyage..., p. 12, see also Nahum Slouschz, Travels in North Africa, Philadelphia, The Jewish Publication Society of America, 1927, p. 224

27 Ismaël Hamet, Les Juifs du nord de l'Afrique (noms et surnoms), Paris, Société d'Éditions Géographiques, Maritimes et Coloniales, 1928.

28 All other authors whose works are discussed below in this section actually took almost all the correct Berber connections and explanations from his book.

29 I. Hamet, Les Juifs du nord de l'Afrique…, p. 14.

30 The surname and its spelling variants appear in Jewish sources as ברהאנץ M. Cohen, Higgid..., p. 229) and בארהנץ (tombstone inscription from the Borgel cemetery of Tunis), both with internal he (ה) and the final tsadi (ץ). The same two letters are also present in the earliest reference in Maghreb to a variant of this surname in question: אלבראהניץ (* Albaranes) appears among the leaders of the community of expellees from Spain to Fez (1494) (Jacob Moses Toledano, Ner ha-Ma‘arav, Jerusalem, Lunz, 1911, p. 78). We find no equivalent to these two sounds in branes, the Berber Tamazight name for the tribe in question.

31 I. Hamet, Les Juifs du nord de l'Afrique…, p. 4-6

32 Maurice Eisenbeth, Les Juifs de l'Afrique du Nord – Démographie et onomastique, Algiers, 1936.

33 M. Eisenbeth, Les Juifs de l'Afrique du Nord…, p. 72.

34 About the directions of migrations of Jews between various countries of Maghreb see: Alexander Beider, A dictionary of Jewish surnames from the Mediterranean region, vol.1: Maghreb, Gibraltar, and Malta, New Haven, Avotaynu Inc., 2017, p. 84-98

35 See exact references to these given names in the corresponding entries of: A. Beider, A dictionary….

36 It is not necessarily derived from the word meaning ‘lamb’ (see A. Beider, A dictionary…, p. 308).

37 M. Eisenbeth (Les Juifs..., p. 124) seems to be the first to suggest the Berber origin for the surname Ergas; compare Berber Kabyle arḡaz ‘man’ (Jean-Marie Dallet, Dictionnaire kabyle-français, Paris, Selaf, 1982, p. 715). This etymology is regularly cited by other authors. Yet, the Berber connection is highly implausible: we deal with a fortuitous phonetic coincidence. In Maghreb, Ergas is known as a Jewish surname only in the city of Tunis and, more precisely, in the community formed by migrants from Italy. Earliest references date from the end of the 17th century. Their text indicates explicitly that bearers of the surname Ergas originate from the city of Livorno. Outside of Livorno, Ergas is commonly found in various other communities formed by former Crypto-Jewish migrants from Portugal and Spain including Amsterdam, Hamburg, and London, sometimes in combination with typical Iberian Christian surnames (Gomes Ergas, Ergas Henriquez). These references exclude the possibility of the Berber origin: the surname is clearly of ultimate Iberian origin. Note also references to the Jewish bearers of the surname Ergas in 15th -century Portugal (see Maria J. P. Ferro Tavares, Os judeus em Portugal no século XV, vol. 2, Lisbon, Universidade Nova de Lisboa, 1984, p. 57, 145, 242, 288).

38 André Chouraqui, Marche vers l’occident: Les Juifs d’Afrique du Nord, Paris, PUF, 1952, p. 132.

39 David Cohen, Le parler arabe des juifs de Tunis. Vol. 1: Textes et documents linguistiques et ethnographiques, The Hague-Paris, Mouton, 1964, p. 1.

40 Abraham I. Laredo. Les noms des juifs du Maroc – Essai d'onomastique judéo-marocaine, Madrid, Instituto Benito Arias Montano, 1978, p. 84-86.

41 About the age of Jewish surnames in Maghreb see A. Beider, A dictionary..., p. 57-84.

42 Jean-Frédéric Schaub, Les Juifs du roi d'Espagne (Oran 1509-1669), Paris, Hachette Littératures, 1999, p. 222; J. M. Toledano, Ner..., p. 124.

43 Joseph Ben Naim, Malkhei Rabbanan, Jerusalem, Abecassis, 1931, p. 29.

44 A. I. Laredo, Les noms des juifs du Maroc…, p. 171-172.

45 Hugues Jean de Dianoux,"Remarques à propos de l’origine apparemment berbère des noms de familles juives d’Afrique du Nord", in Juifs et judaïsme en Afrique du Nord dans l’antiquité et le haut Moyen-Âge, ed. Carol Lancu and Jean-Marie Lassère, Montpellier, Centre de Recherches et d’Études Juives et Hébraïques, 1985, p. 103-111.

46 Joseph Tolédano, Une histoire des familles – Les noms de famille des juifs d'Afrique du Nord: Maroc, Algérie, Tunisie; des origines à nos jours, Jerusalem, Ramtol, 1998.

47 Jacques Taïeb, "Juifs du Maghreb: onomastique et langue, une composante berbère?", in Encyclopédie Berbère, 26, Aix-en-Provence, Edisud, 2004, p. 3969-3975.

48 The Jewish surname Taous is known in the city of Algiers only (A. Beider, A dictionary…, p. 598). Most likely, it is derived from the identical female given name (‘peahen’) used in Algeria by both Jews and Muslims. As a result, any link between this surname and the toponym Et Taous in the Berber-populated area of southeastern Morocco is highly unlikely.

49 N. Slouschz, Hébræo-phéniciens..., p. 469.

50 D. Cazès, Essai..., p. 175.

51 N. Slouschz, Hébræo-phéniciens..., p. 469.

52 For example, Encyclopedia Judaica, vol. 6, p. 270 and P. Wexler, The Non-Jewish origins..., p. 127-128.

53 Ibn Khaldoun, Histoire..., vol. 3, p. 252.

54 In this section, Jewish surnames derived from the discussed given names appear in brackets [].

55 The only reference known outside of Morocco corresponds to the birth record of Idir Ben David Ben Youssef (1891, northwestern Algeria). The surname reveals the Moroccan origin of the family: it represents an Arabicized form of Ben David Ouyoussef known in Morocco only. In no other civil record for all of Algeria (19th century) we find any other reference to this given name or its variants.

56 Miloud Taïfi, Dictionnaire tamazight-français (parlers du Maroc central), Paris, L’Harmattan-Awal, 1991, p. 70.

57 J.-M. Dallet, Dictionnaire..., p. 152.

58 David Corcos, Studies in the history of the Jews of Morocco, Jerusalem, R. Mass, 1976, p. 184-195.

59 About these suffixes see, Annemarie Schimmel, Islamic Names. Edinburgh, Edinburgh University Press, 1995, p. 69, 75; Georges Séraphin Colin, Arabe marocain, Aix-en-Provence, Edisud, 1999, p. 75-77.

60 In section 5 of this article, the Judeo-Arabic pronunciation of various given names discussed by various authors is reconstructed. Corcos, Hamet, Hayoun and others do not provide the exact forms: they make no distinction between full and reduced vowels, as well as between the long and the short ones. A more plausible etymology for the given name Məšîš (Mechich) is given below in this section, note that this surname has the initial vowel that does not appear in Berber Tarifit amšiš ‘cat’ (Esteban Ibáñez, Diccionario Rifeño-Español, Madrid, Instituto de Estudios Africanos, 1949, p. 35). In other Berber idioms the cognate word does not sound close to this given name at all, compare Tamazight mušš (M. Taïfi, Dictionnaire..., p. 440).

61 I. Hamet, Les Juifs du nord de l'Afrique…, p. 16.

62 M. Taïfi, Dictionnaire..., p. 879.

63 In the lists of Berber given names used in southern Morocco (M. Taïfi, Dictionnaire..., p. 878-879) and eastern Algeria (J.-M. Dallet, Dictionnaire..., p. 1027-1035) bisyllabic forms ending in –u, -i, or (mainly for women) –a are commonplace.

64 I. D. Abbou (Musulmans andalous.., p. 400) speaks about “ridiculous transformations” of biblical names in the South giving such examples as Bagga, Beggo, Bihi, Dokho, Doudou, Haki, Hako, Heddo, Hekko, and Izzo.

65 G.S. Colin, Arabe..., p. 74, 76.

66 See Victor Hayoun, "Jewish names, surnames, and nickname of Nabeul, Tunisia", in Pleasant are their names. Jewish names in the Sephardi diaspora, ed. Aaron Demsky, Bethesda, University Press of Maryland, 2010, p. 176-178. Most likely, in the last two examples we deal with the long /î/ in the first syllable.

67 G. S. Colin (Arabe..., p. 73) mentions the existence of the following Arabic diminutive pattern: three consonants of which the second one is following by the long vowel /î/. Among his examples one finds: Jewish Brîhmu from (I)brâhîm ‘Abraham’, Muslim Ḥmîdu from Ḥməd ‘Ahmed’ and Mbîrku from Mbârək. This pattern, typical for northwestern Morocco, takes its origin from the Standard Arabic diminutive pattern. It is unclear whether its final vowel – u was influenced by Berbers or not. In Judeo-Arabic of Tunis, we find diminutive forms of given names following the same pattern, but with ay in place of /î/ (D. Cohen, Le parler..., p. 168). This pattern is structurally equivalent to the above pattern from northern Morocco because it is precisely in Tunis that the regular Arabic diminutive forms have a variant with the diphthong ay in place of the long vowel î peculiar to Morocco and Algeria. Note also that in all the above Jewish forms from northern Tunisia, the vowel of the first syllable is precisely /ay/ or /î/. In other words, these Tunisian given names are clearly of Arabic rather than Berber origin.

68 D. Corcos, Studies..., pp. 183-197.

69 Numerous forms using one of the two patterns discussed in this paragraph appear in the list of Berber given names from eastern Algeria compiled by J.-M. Dallet, Dictionnaire..., pp. 1027-1035): (a) first subgroup: Caca, Fafa, Jijji, Juju, Kuku, Nunu, Ṭiṭi, and Tutu; (b) second subgroup: ‘Jaja, Hbubu, Ḥmimi, and Ṭiṭem. Yet, we also find several forms—such as Aichoucha (from Aicha), Aloulou and Fafani (from Afouna)—used by Arabic-speaking Muslims in Tunisia (Paul Marty, "Folklore tunisien. L’onomastique des noms propres de personnes", Revue des études islamiques 4 (1936), p. 390).

70 V. Hayoun, "Jewish names...", pp. 176-179.

71 M. Cohen, Higgid..., pp. 224-228.

72 D. Corcos, Studies..., pp. 183-191.

73 Examples taken from René de Segonzac, Voyages au Maroc (1899-1901), Paris, A. Colin, 1903, p. 97.

74 M. Cohen, Higgid..., p. 224.

75 See A. Beider, A dictionary..., p. 183-188.

76 This section gives the Hebrew spellings of surnames if they are available. These spellings are useful to corroborate etymological hypotheses because they provide additional information in comparison to the spellings in Latin characters. For example, the distinction between /t/ and its emphatic equivalent /ṭ/ via Latin characters is impossible. Yet, in the Judeo-Arabic spelling these sounds correspond to different Hebrew characters, tav (ת ) and tet (ט ), respectively.

77 J.-M. Dallet, Dictionnaire..., p. 948.

78 Solomon ben Ṣemaḥ Duran, Tashbaṣ IV: Ḥut ha-Meshullash, Amsterdam, 1738/1739, n° 6.

79 J.-M. Dallet, Dictionnaire..., p. 839.

80 J.-M. Dallet, Dictionnaire..., p. 518.

81 M. Taïfi, Dictionnaire..., p. 878.

82 Lloyd Cabot Briggs and Norina Lami Guède, No more forever: a Saharan Jewish town, Cambridge MA, Peabody Museum, 1964, p. 44.

83 M. Taïfi, Dictionnaire..., p. 165.

84 Simon Lévy, Parlers arabes des Juifs du Maroc. Histoire, sociolinguistique et géographie générale, Zaragoza: Instituto de Estudios Islámicos y del Oriente Próximo, 2009, p. 146.

85 See section 7.

86 E. Ibáñez, Diccionario..., p. 35.

87 Edmond Destaing, Dictionnaire français-berbère (dialecte des Beni-Snous), Paris, E. Leroux, 1914, p  148.

88 In this list, only one source for Tamazight is used: M. Taïfi, Dictionnaire.... It is by far more detailed than any published dictionary of Tashelhit, for which the following sources have been used: (1) Antoine Jordan, Dictionnaire berbère-français (dialectes techelhit), Rabat, Omnia, 1934; (2) Maurice Dray, Dictionnaire français-berbère – Dialecte de Ntifa, Paris, L’Harmattan, 1998; and (3) Edmond Destaing, Etude sur la Tachelḥît de Soûs -Vocabulaire français-berbère, Paris, E. Leroux, 1920. A few names in this list can be based on Tarifit too since similar words are also found in Tarifit: Amghar, Amouzig, Aoudai, and Siksou.

89 M. Taïfi, Dictionnaire..., p. 38.

90 M. Taïfi, Dictionnaire..., p. 252.

91 M. Taïfi, Dictionnaire..., p. 351.

92 M. Taïfi, Dictionnaire..., p. 354.

93 M. Taïfi, Dictionnaire..., p. 408, A. Jordan, Dictionnaire..., p. 31.

94 M. Taïfi, Dictionnaire..., p. 417, M. Dray, Dictionnaire..., p. 238

95 M. Taïfi, Dictionnaire..., p. 447, M. Dray, Dictionnaire..., p. 60.

96 M. Taïfi, Dictionnaire..., p. 639, A. Jordan, Dictionnaire..., p. 108.

97 A. Jordan, Dictionnaire..., p. 155.

98 M. Taïfi, Dictionnaire..., p. 82, M. Dray, Dictionnaire..., p. 277.

99 A. Jordan, Dictionnaire..., p. 46, M. Dray, Dictionnaire..., p. 430.

100 A. Jordan, Dictionnaire..., p. 51, M. Dray, Dictionnaire..., p. 238.

101 A. Jordan, Dictionnaire..., p. 52.

102 M. Taïfi, Dictionnaire..., p. 805

103 M. Taïfi, Dictionnaire..., p. 449.

104 M. Taïfi, Dictionnaire..., p. 122.

105 E. Destaing, Etude..., p. 291.

106 M. Taïfi, Dictionnaire..., p. 684

107 M. Taïfi, Dictionnaire..., p. 684

108 M. Taïfi, Dictionnaire..., p. 803.

109 M. Taïfi, Dictionnaire..., p. 390.

110 M. Taïfi, Dictionnaire..., p. 117.

111 M. Taïfi, Dictionnaire..., p. 122.

112 M. Taïfi, Dictionnaire..., p. 416

113 M. Taïfi, Dictionnaire..., p. 83.

114 A. Jordan, Dictionnaire..., p. 107.

115 M. Dray, Dictionnaire..., p. 337.

116 P. Wexler (The Non-Jewish origins..., p. 128) points to the existence of the surnames Ohayon and Benhayoun in Maghreb and Hayon in the communities formed by expellees from Spain in the Ottoman Empire. From that, he concludes that the original Berber surname Ohayon was “de-Berberized” in Spain prior to the expulsion, while in Maghreb “de-Berberizing” was followed by a partial Hebraization yielding the form Benhayoun. As a result, according to Wexler, we have an argument to support his theory about the Berber proselyte origin of Jews in Maghreb and medieval Spain. Such logic is untenable. On the one hand, the Arabic form Hayon (given name, surname) is older than Berber Ohayon and the latter is unknown in Spain. On the other hand, in hundreds of Jewish surnames in Maghreb, the initial Ben ‘son’ comes not from Hebrew, but from the vernacular Arabic. Benhayoun is not an exception from this rule. It is a surname independent of Ohayon: both of them are just derived from the same Jewish given name Ḥayyûn, of Arabic origin.

117 It is usually considered that Aknin represents a Berber diminutive of Jacob (A. Laredo, Les noms..., p. 356; D. Corcos, Studies..., p. 134). However, the final –nin does not correspond to any known Berber pattern. Note also that the surname with the patronymic Arabic prefix, Aben Aknin (אעקנין אבן), is known in the city of Fez already in the 12th century (J. M. Toledano, Ner..., p. 38). It is older than Ouaknin, the form with the patronymic Berber prefix, whose earliest known reference, from the same city, dates from the 18th century only (J. Ben Naim, Malkhei... p. 84).

118 M. Taïfi, Dictionnaire..., p. 539.

119 The dates are mainly taken from J. M. Toledano, Ner... and J. Ben Naim, Malkhei... (see exact references in A. Beider, A dictionary…). Many of these references correspond to the city of Fez. Their presence in this Arabic-speaking area explains the use of the Arabic prefix ben ‘son of’ in some of these surnames.

120 Compare Tamazight azzaglu (M. Taïfi, Dictionnaire..., p. 797) and Kabyle azaglu (J.-M. Dallet, Dictionnaire..., p. 935).

121 Alfred Louis de Premare et al., Langue et culture marocaines - Dictionnaire arabe-français, 12 vols, Paris, L’Harmattan, 1993-1999, vol. 5, p. 276.

122 A.L. de Premare, Langue..., vol. 6, p. 139.

123 A.L. de Premare, Langue..., vol. 6, p. 139.

124 According to this characteristic, the inception of Jewish surnames from North Africa is similar to that of their coreligionists from the medieval Iberia and that of the Christian population of various European countries (with a notable exception of the Scandinavian ones). Yet, the initial development of surnames from Maghreb is dramatically different from that of the surnames of Ashkenazic Jews acquired only during the 18th -19th centuries as a result of the promulgation of laws

125 Harvey E. Goldberg, "The social context of North African Jewish patronyms", Folklore Research Center Studies 3 (1972), pp. 249-254.

126 The last two possibilities are purely theoretical: no historical evidence exists to corroborate them.

127 See N. Slouschz, Un voyage..., pp. 41, 50, 59. The same source (p. 37, 43, 59) also mentions the following given names of uncertain etymology: male בלחייא and חברון (both also known in Tripoli, M. Cohen, Higgid..., pp. 224-225) and female נספה.

128 Harvey E. Goldberg, "Patronymic groups in a Tripolitanian Jewish village: reconstruction and interpretation", Jewish Journal of Sociology 9 (1967), p. 211.

129 S. Duran, Tashbaṣ IV..., n° 15, 1, 24, 33.

130 Eliahou Marciano, Les sages d'Algérie: dictionnaire encyclopédique des sages et rabbins d'Algérie, du haut moyen-âge à nos jours, Marseille, IMMAJ, 2002, p. 96, 193, 229.

131 See corresponding entries in A. Beider, A dictionary

132 Vincent Monteil, "Les Juifs d'Ifrane (Anti-Atlas Marocain)", Hespéris - Archives berbères et builletin de l’Institut des Hautes-Etudes Marocaines 35 (1948), pp. 151-162.

133 Paul Pascon and Daniel Schroeter, "Le cimetière juif d’Iligh (1751-1955). Etude des épitaphes comme documents d’histoire sociale", Revue de l’Occident musulman et de la Méditeraanée 34 (1982), p. 39-62.

134 V. Monteil, "Les Juifs...", p. 154; P. Pascon and D. Schroeter, "Le cimetière...", , p. 47.

135 The list of oldest inscriptions from Ifrane appears in J.M. Toledano, Ner..., p. 4-5 and V. Monteil, "Les juifs...", p. 158-159. Dates present in it are unreliable (see V. Monteil, "Les juifs...", p. 155-158). The readable parts of certain inscriptions actually do not include any date. Moreover, for example, the inscription including the Arabic name Maymûn is “dated” 4 BCE, that is centuries before the Arabs came to Maghreb.

136 See details in Alexander Beider, Origins of Yiddish Dialects, Oxford, Oxford University Press, 2015, p. 382-386, 419-428, 433-436.

137 In the cities of Fez (Louis Brunot and Elie Malka, Textes judéo-arabes de Fès, Rabat, Typo-litho école du livre, pp. III-IV) and Tunis (D. Cohen, Le parler..., pp. 10-11) Berber influence is unknown not only in the system-level parts of the language, but even in the lexicon.

138 J. Tolédano, Une histoire..., p. VII.

139 A similar false idea was extrapolated by Chouraqui (Marche..., p. 132) to an absurd conclusion. Unaware about the actual age of surnames, he states that the small percentage of names of North African Jews based on Hebrew and Aramaic implies that the large majority of these Jews descend from (Berber) proselytes to Judaism. If one follows the same logic, one should also assert that Ashkenazi Jews who bear surnames with Slavic roots and/or suffixes (such as Abramovich, Bialik, Dubnov, or Slouschz) are descendants of Slavs converted to Judaism, while those bearing surnames of German origin (such as Einstein, Freud, or Landau) had German Gentile proselytes as their ancestors.

Top of page

References

Electronic reference

Alexander Beider, “Jews of Berber Origin: Myth or Reality?”Hamsa [Online], 3 | 2017, Online since 05 May 2021, connection on 18 March 2024. URL: http://journals.openedition.org/hamsa/693; DOI: https://doi.org/10.4000/hamsa.693

Top of page

Copyright

CC-BY-NC-ND-4.0

The text only may be used under licence CC BY-NC-ND 4.0. All other elements (illustrations, imported files) are “All rights reserved”, unless otherwise stated.

Top of page
Search OpenEdition Search

You will be redirected to OpenEdition Search