Introduction

1In 1912, Marcel Cohen, a scholar in the field of North African linguistics, carried out a study of the Arabic dialect spoken by Jews in Algiers. In the 1930s and 1940s, Louis Brunot and Elie Malka investigated the Jewish dialect of Fez, while 50 years after his namesake, David Cohen continued this work by studying the dialect of Tunis along with many other varieties of Arabic. It seemed to me important to finish this work while there were still Jews in the Maghreb, before linguistic assimilation and emigration to France silences their voices. [2] The task is urgent: from 270,000 in 1945, their numbers fell to around 20,000–30,000, and there are only a few thousand left today as the more educated among them have become French-speaking.

2Although they have been there for a very long time, official history as promulgated by schools and political authorities has already drawn a veil of forgetfulness over them. Yet, having overseen the expatriation of 100,000 Jews between 1960 and 1965, the political authorities were forced to reintroduce some into its political groupings, even in jails, to demonstrate that there were different kinds of Moroccans in order to gain the support of their American allies in the battle [against secessionists] in the Sahara. It is partly for this reason that Morocco today is still a pluralistic country.

3The language I grew up with had few distinctive features: a lisping pronunciation, some borrowings from Hebrew, and many Hispanic features, an aspect Isaac Abbou emphasized in his [1953] book Musulmans andalous et judéo-espagnols. [3] Among Moroccan Jewish intellectuals, a certain snobbish insistence on their Sephardic origins led them to see only the prized Spanish influence in the Jewish-Moroccan past. [4] Spanish researchers and Northern European Jews naturally emphasized Judeo-Spanish, or Ladino/ḥakitiya, [5] as well as popular Sephardi literature. It is also true that Moroccan dialectology was treated with disdain in universities during the 1960s because it had been promoted during the French protectorate. [6]

1 – The Heritage

4The linguistic landscape was undergoing rapid change. Among the Muslim population, the process of linguistic unification was ironing out local dialects. As regards the dialects of the Jewish communities, emigration, beginning in 1948 and gathering speed in 1962 and 1967, [7] led to the disappearance of some communities, while for increasingly larger sections of these groups, French was supplanting Judeo-Arabic dialects. It was a vanishing heritage: our ways of speaking, the language spoken within the family and the neighborhood—what was in fact the language of our identity—was disappearing and was being replaced by French, introduced after 1862 by the Alliance Israélite Universelle (AIU), which barely taught Arabic. While Brunot and Malka in the case of Fez and Pellat for Debdou published valuable work, [8] other dialects ran the risk of being lost, erased from our memory as Moroccan Jews, from the whole of Morocco, and as pieces of the linguistic mosaic of North African dialects. [9]

2 – Methodology

5Recent monographs on dialects often start with a theory or model, which is then applied to a specific dialect, its phonology, or a particular syntactic feature. While these studies are clearly valuable, few of them present a corpus, a body of language on which the studies are based and which can then be used profitably by other scholars for their own purposes.

6I myself chose to return to the proven methods of the masters of Arabic and North African dialectology such as Marçais, Cohen, Levy Provençal, Colin, and Brunot, who produced texts, translations, notes, and phonetic, morphological, and lexical studies that have stood the test of time. Reading these texts reminds us that dialectology is above all interdisciplinary in that it uses information drawn from history, human and physical geography, sociology, ethnography, and religious studies, among others.

7This is all the more important when we are studying the dialects of minority groups living among much larger populations. In such cases, we cannot be satisfied with a synchronic, internal description any more than it is possible to study Jewish dialects in isolation. The amassing of specific details is relevant only in comparison with other local and surrounding dialects (under the general heading of local varieties of Arabic) but also those that are less local (that is, other dialects, whether Jewish or not). Here, we abandon the synchronic approach so rigorously developed by modern methods and models in order to pursue comparisons with other contemporary ways of speaking, whether archaic or more evolved. In this sense, comparative dialectology is a living diachrony.

3 – The Historic Context

8In the case under study, historical and sociological factors are crucial. Why do some Jewish communities such as those of Midelt, Demnat, and Tahala remain Arabic-speaking in a Berber-speaking environment? Why has Judeo-Spanish (ḥakitiya) survived in Tetouan and Tangiers but not in Fez or Meknes? How did one community influence another? And above all, how did they arise in particular areas, at what time, and from which center? Why does the Jewish dialect of Marrakesh contain so many characteristics of urban Arabic while the speech of their Muslim neighbors has so many rural features? To answer these questions, we need to start by sketching out the historical and social context of this study and by undertaking a history of Jewish settlement in the host country while avoiding all Manichaean dualism or ghettoization.

9This history might have been shorter had the extensive existing literature [10] been more detailed for any period before the fifteenth century and if modern Jewish historiography had been able to overcome its contradictory tendencies. Thus Nahum Slouschz sees the Jewish presence in North Africa as consisting largely of Berber converts to Judaism, while sixty years later, Hirshberg [11] downplays or even denies such an origin. Meanwhile, Eisenbeth [12] and Toledano (in Ner hamaεarav, 1911) were loud in bewailing (qinot) the misfortunes or supposedly experienced by the Jews under Idris Ist in the eighth century and then their supposed annihilation under the Almohad dynasty in the twelfth century.

10Other writers such as Ortega [13]  and Laredo [14] followed in their footsteps, failing to qualify their judgments by simply consulting the Muslim author Al Marrakuchi or the Jewish Maïmonide (Iggeret Hašmad), [15] who explain clearly how the Jews remained Jews and were accepted as such even when they adopted the outward appearances of Islam. Corcos, [16] who was better read, was seduced by the Manichaean theory, which holds that the Marinid dynasty (thirteenth–fifteenth century) was systematically favorable to the Jews. This unfortunately too widely held view that history can be divided into periods and rulers between being either good or bad for the Jews makes it impossible to understand how Moulay Sliman, the “good” ruler who revoked the anti-Jewish edicts of his predecessor the “bad” Moulay al-Yazid could be the same man who ordered that the Jews of Rabat Salé and Tetouan be confined in the mellahs [17] (ghettos) when the Napoleonic Wars reached the Straits of Gibraltar (1808–1810). Instead, they should have made use of all possible available sources, Moroccan or foreign, Jewish or Muslim, and brought them together for critical comparison. These speak to us of lesser-known episodes—engravers of Jewish coins in Todgha and Ikem at the time of Idris, or the enlightening writings of Maimonides in his Iggeret hašmad. [18] They speak to us if we are prepared to abandon the Manichaean concept of Jewish history and to write history as such, that is, not by starting from what is good or bad for the Jews but from the analysis of social phenomena, known facts, texts, and other materials set in their global context.

11Thus, to return to the Marinids, we may see them either as benefactors—they legalized Judaism, unlike the Almohads, who had seen the yahoud as Muslims of a special kind, had Jewish ministers, and reestablished the dhimma, [19] or alternatively as the creators of the first mellah. In practice, in the thirteenth–fourteenth centuries, they were above all politicians, and effective ones at that. They replenished the state coffers through economic activity and trans-Saharan trade (not to mention the collection of the jizya tax) and in short, performed economic roles the Jews had not yet abandoned under the Almohads.

12Later administrations protected the Jews against riots, either militarily, as in 1273, or by conceding to them an autonomous and fortified part of Fez Jdid, near the Palace, namely the mellah. In a way, the mellah gave spatial form to the provisions of the dhimma, the legal status of Jewish or Christian minorities in Muslim societies, allowing them to be self-governing. Although this was not in the egalitarian spirit of the Universal Declaration of Human Rights (how could it be, in the Middle Ages?), it granted them a place in society as well as certain rights: fewer than those of Muslims, but genuine rights nonetheless. In Fez, during the Marinid era [1250–1474], the Jews needed military protection. We should try to understand from the perspective and in the light of the religious values of the time the shock and scandal caused by the official abjuration of Islam on the part of the neo-Muslims, their relapse into Judaism, and the resulting frictions in the Medina between those who were newly professing their Judaism, those who had definitively adopted Islam (the beldiyin), and the original Muslim inhabitants. We find a sense of this in writings of the time, especially in anonymous pamphlets and chronicles such as the Rawd al-Qirtas and Zahrat-al-As. [20]

4 – Isolation and Integration

13The dhimma consisted of an inward-looking community life, a specific religious culture, and an economic role largely open to the wider society. The conjunction of these isolating or integrating factors were in fact opposing forces expressed in certain features of the dialects spoken by the Jewish minority. Although their communities bathed in a society that surrounded them for part of their lives, they also had their own reserved, private, and specific domain. Within this melting pot were found dialects and their associated popular dialect and literature, which retained features that had died out elsewhere, forming useful milestones for the study of the history of Moroccan Arabic and refining the classification of dialects into pre-Hilalian and Bedouin. However, in addition to such preserved features, we also find certain lexical or morphemic innovations in these dialects, such as plurals ending in /ot/ in Fez and a phonetic evolution not present in classical forms, such as the sliding of (voiced) šin/žim toward the (unvoiced) sibilants /s/ and /z/. We find these uncorrected variants in the Muslim speech of Meknes, Tadla, and elsewhere as well as in the feminine Muslim speech of Fez, Rabat, and Tangiers.

14This similarity suggests a tendency to a corrected lisping pronunciation that was present in a number of early (pre-Hilalian) dialects and registers, notably the feminine one. Schools, the language of the Qur’an, and religious practice all played a role in affecting pronunciation, with the standard of linguistic purity set by male city-dwelling Muslims, less so by women.

5 – Some Phonetic Features

15The analysis of the Jewish dialects studied shows their close relationship to pre-Hilalian varieties of Moroccan Arabic, the so-called urban variants, as well as the mountain dialects (jbala).

16On the basis of comparison, they can be classified into three groups, with the chief feature they have in common being the pronunciation of the classical Arabic phoneme /ق/ (qaaf) (voiced post-velar occlusive). In the first group /ق/>/’/, qaaf becomes hamza, a strongly-voiced laryngeal glottal occlusive, as in the dialects of Fez, Sefrou, Meknes, and Rabat-Salé, in the second /ق/ > /q/ (voiced post-velar occlusive) in the dialects of Marrakesh, Essaouira, Safi, El Jadida, and Azemmour, and in the third /ق/>/k/ (a voiced velar occlusive), as in the Debdou, Tafilalt, and Oued Dra dialects. This shift is also found among a minority in certain communities in Marrakesh and on the Atlantic coast, which some refer to as hәdra-ssġera, as opposed to hәdra bәl-qāla (/ق/ = /’/), which reduces the /q/ to a simple strong vocal attack.

17A feature of almost all of these dialects is the lisping pronunciation in which š>s and ž>z, a feature found also in the peasant speech of the Tadla region, among elderly Muslim inhabitants of Fez, and in female speech in Muslim Fez, Tetouan, and Sale, with these Jewish dialects being uncorrected by education or religious norms (particularly the reading of the Qur’an) and having thus evolved freely. This is the hypothesis that could be put forward for Muslim female variants, for example. Yet this lisping is not found in the Jewish speech of Debdou (see below), such as the Jewish dialect of neighboring Oran, with which it shares other distinctive features.

18Does the prolonged contact between Judeo-Arabic and Judeo-Spanish in Jewish Meknes explain the atypical vocal features of this dialect? The comparison with older forms of ḥakitiya in Tetouan described by Benoliel in his work Dialecto judeo-hispano-marroqui o hakitia, republished in 1977 in Madrid, is enlightening in this respect.

6 – Examples: The Dialects of Debdou: Hədra Sġera and Others

19This study reveals that the lisping is not found in all Jewish dialects, contrary to what is often believed. Although some speakers always pronounce /s/ with a lisp, this is limited to certain individuals and families. It should be noted that the dbadba [21] dialects have a particular way of pronouncing consonants, as Debdou as well as its colonies, which took root in from Midelt to Taourirt, Oujda, Missour and the entire Eastern region, including Melilla, differentiate clearly between šin and žim and between sin and zay, as do the Jewish dialects of the Oran region, at least when they were spoken there prior to 1962.

20These dialects also have a particular way of pronouncing /ق/ as /k/. This feature is common to all Jewish dialects in Tafilalet and the Draa valley. Some speakers differentiate /k/ by depalatizing it to /tch/, a process that led to its pronunciation as /ţ/ in Tafilalet (as in sәţţar for sәkkar, sugar).

21Others retain the /k/ pronunciation for both ق [qaaf] and ك [kaaf]. This feature is also found in the mountain dialects of the Trara and Msirda regions of Algeria. In Safi, Marrakesh, and Azemmour, this pronunciation is optional among part of the Jewish population and is referred to as hәdra sġera, as opposed to hәdra bәl qạla (ق ›/ ′/), as found in Fez, Meknes, and elsewhere.

22Although this element of the study, namely the issue of qaaf and its various pronunciations, has to be relegated to the “unsolved” file, I have put forward some hypotheses for the Moroccan case by analyzing the influence of the Berber substratum.

23Similarly, the study has revealed another markers on the evolutionary road that could complete the chain that leads from the relative pronoun әldi and culminates in the use of the preposition diâl to express relationships. In the Jewish speech of Tafilalet and Draa, we find the analytic expression bnu li miyәr with the relative particle li (to), as in “his son to Meyer,” as well as de-li (which is mine), in other words de+l+affixed pronoun, which urban speech renders as diäli. There is probably where the missing link in the chain is to be found.

24I would also like to draw attention to the consonants added to the root in the continuous present tense formed using /a/. In the Jewish Sefrou dialect (Jjbala), we find anәrfed = “I carry” and adәrfed = “you carry.” Meanwhile, /da/ in Debdou becomes da-i-ţöno = “they are,” and in Tafilalet, there is the paradigm ti/te and ţa, equivalent to /ka/, hence ina ti-nεәrf = “I know,” nti ti t-εәrf = “you know,” and huwa ta iεәrf = “he knows.”

7 – The Berber Substratum

25I believe that these elements can contribute to the debate over the influence of the Berber substratum on the formation of North African Arabic dialects. This debate, which has been frequently curtailed as a result of ideological or cultural conditioning (whether conscious or unconscious), must continue if we wish to one day write a principled history of North African Arabic. Although the lexical facts are easy to discern, the similarities in the structures and constructions of the two languages are not always enlightening. That is, who is calquing whom? Furthermore, both belong to the same extended family of Afro-Asiatic languages. A study of dialects can identify concrete elements that may enable us to carry out a theoretically grounded reconstruction of the development from the Berber structure noun /n/ noun, as in afus-n-urgaz = “the man’s hand,” which produces l’yid dial әrražel. We can gain a better sense of the phenomenon of calquing when we examine ba (ma) in-flan, where the Berber particle -n- is adopted in Taza and the East before being replaced by /de/ or /li/ di-li—diali > dial flän.

26Linguistics has everything to gain by putting aside any a priori bias and studying the Maghreb, which is a kind of privileged linguistic laboratory due to the diversity of its dialects and its phenomena of dynamic multilingualism that are changing rapidly before our eyes.

8 – The Influence of Spanish

27Some (though not all) Jewish communities were bilingual in Arabic and Spanish, including in Tetouan, Tangiers, Asilah, and El Ksar until today and Fez, Meknes, and Rabat-Salé from the sixteenth to the eighteenth century.

28The relationship of Jewish dialects to Spanish needs to be studied more closely, combining a historical approach, for example by estimating the size of these communities during the fifteenth and sixteenth centuries on the basis of contemporary accounts, with sociological, juridical, and cultural approaches, for example by establishing the extent of the Sephardic elite (εazmiyin) compared to the bәldiyin, the number of rabbi judges (dayanim) and teachers in the sla, [22] among whom the use of the etymologically Spanish verbs bezar and asakkar [23] was widespread. The Jews from Spain had successfully imposed the Jewish law codified by Rabbi Yosef Caro [24] in the sixteenth century based on the work of a number of rabbis and scholars. They even set up a short-lived printing press in Fez in the sixteenth century and introduced many innovations in the area of crafts, especially jewelry, in the vocabulary of furnishings, such as körtina, “curtain” and sianto “seat,” and the female garment zәlteta “skirt” as well as terms related to pastry making such as masappan and fazuelos. [25]

29However, their influence remained limited to major cities and ports. Meanwhile, Draa and Tafilalet were influenced only indirectly, and the number of Hispanicisms found there is below the average number featuring in Muslim dialects in Morocco as a whole. [26] On the other hand, the original Jewish inhabitants, the beldiyin, experienced permanent influx from the south and had their own traditions and forms of solidarity, particularly in Fez, [27] and a well-organized and productive rabbinate, including the Kabbalists of Draa, studied by Haïm Zafrani. Finally, a united body of law was formed incorporating large swathes of local customs (or minhag) covering (among other things) marriage and inheritance in Fez, where bigamy had been reintroduced, even among descendants of Jews from the Iberian peninsula, as well as a host of familial practices (or εada), which spread widely by means of marriage.

30In terms of linguistics, ḥakitiya remained in use in towns such as Tetouan, Asilah, and Tangiers, whose original residents had been driven out by Spanish attacks in the fifteenth and sixteenth centuries and which were repopulated by remaining groups of Sephardis, who were able to retain a dominant position. Later, in the seventeenth and eighteenth centuries, Larache, Asilah, and especially Tangiers, which was liberated from the English in 1684, were resettled from Tetouan. We should note that in these cities, there was clearly bilingualism between Arabic and the Judeo-Arabic of the original Jewish residents, the forasteros (literally “strangers to the town”). Hakitiya naturally remained highly open to Arabic borrowings, of which some 1,500 have been recorded, more than the maximum number of Spanish borrowings, of which there are around 400 in Fez, Meknes, and Rabat.

31However, we should emphasize once more that not all the Spanish borrowings found in Jewish dialects are exclusively Jewish. Moroccan Arabic maintained strong relationships with Spanish during the historical formation of both languages: with the Mozarabic of Spain via Andalusian Arabic and with Castilian through trade, though often through the intermediary of a Jewish tordjman (interpreter). For Morocco, Spanish was the language through which it communicated with Europe as well as a diplomatic language even for relations with England, despite the latter’s influence at the time. Finally, with the arrival of the Spanish-speaking Moriscos (Muslims forcibly converted to Christianity) in 1610, Muslim cities such as Rabat and Tetouan experienced a new era of bilingualism, as did Jewish communities with the arrival of the Sephardi refugees. However, the record for the number of Hispanicisms in any Arabic dialect is held not by a Jewish dialect but by the speech of Muslims in Tetouan and Tangier, even before the period of the Spanish Protectorate (in the early twentieth century). [28] The sea, seafaring, and fishing also gave coastal cities a corpus of Spanish terms, or kәlma mәrsawiya, in both Arabic and Berber.

9 – Hebrew

32Jewish dialects are characterized chiefly by their borrowings from Hebrew, as one might imagine. However, it is frequently forgotten that Hebrew was not a spoken language for Jewish Moroccans. Rather, its status was that of the language of high culture, of study, of the synagogue school (ṣla), of the yeshiva: it was the language of texts and prayer. Even in teaching, discussions were conducted in Arabic, using translations and calques (šarḥ). Among Spanish Jews, these gave rise to Ladino. [29]

33Today, French has largely replaced Hebrew as the language of culture due to the efforts of the AIU, [30] while Hebrew remains the almost exclusive language of prayer. The traditional study of sacred texts committed to memory gave rise to a whole religious and legal vocabulary and a lexicon of philosophy and abstract ideas as well as common expressions such as baddaï “of course,” kebar “already,” afillo “at least,” and in Fez, the Aramaic bine-bine “meanwhile.” When studying this lexicon, it is especially important to distinguish between different registers, namely educated language or the normal speech of men or of women. Finally, there is lasöniya (ţalasunţ or tadoberţ), a secret argot used by Jewish merchants in the souk, supposedly unintelligible for Muslims, though nothing of the kind was the case since this argot was limited and repetitive, and some terms must surely have been picked up by Muslim customers and colleagues.

34I have attempted here to pick out the Hebrew borrowings that have passed into Arabic (excluding those from the sacred domain that were assimilated into classical Arabic). These include mazzal, ḥamẹs, and kašer, which were adopted, sometimes with a change of meaning. Most of these terms belong to the common language: duwwәε-ne (look at me secretively) is in fact vulgar, while kuwwәn (play dead) is not. These terms most probably came from lasöniya. Such words survived in spoken Moroccan for a long time. Finally, kašer (mortadella sausage) took on an autonomous existence and, through its use in fast food, gained widespread favor, not because the rabbinate had imposed Jewish dietary laws on Muslim grocers but simply because mortadella, formerly kašer [now kosher] or simply ḥalal, filled a certain commercial niche as a cheap sandwich filling.

35Several thousand of us in Morocco still speak these Judeo-Arabic dialects. Elsewhere in Israel, there are tens of millions who know them, use them, and transform them. Within the French diaspora, in Spain or in Canada, these dialects are more of a folk memory, tinged with nostalgia.

36French and Judeo-Arabic bilingualism is revealed more in the form of linguistic interference than by borrowings and mainly by the substitution of one language by the other. The setting up of the first AIU school in Tetouan in 1864 began a conflict of civilizations and cultures in which Judeo-Arabic was unable to take the offensive. Yet it managed to survive into the twenty-first century thanks to the power of inertia through succeeding generations, kept alive by the dynamism of Moroccan Judaism. In replacing one language by another, we have gained modernity and efficiency—at the price of part of ourselves.