Dialects: Classification

1. General remarks

Generalizing labels such as ‘ Egyptian Arabic’, ‘Syrian Arabic’, or ‘Moroccan Arabic’ are commonly used to refer to dialect types spoken in the respective countries. In textbook titles and names of courses in Spoken Arabic, they are used for the sake of convenience, although in fact they often refer to the dialects of the capital cities. This is not merely a simplification but, in a sense, it is also justified because of the ongoing trend toward regional standard dialects with the dialects of the urban centers as the models. A striking example of this development is the influence of Cairene not only on the dialects of the neighborhood, but on the dialects of urban centers such as Alexandria and Port Said as well, in which the structure of Cairene already predominates (Behnstedt and Woidich 1983).

When systematic classifications based on well-defined criteria are aimed at, there is more than one choice for the approach. If the interest is purely synchronic, the classifications can be made on the basis of an adequate selection of synchronically well-documented linguistic variables for each dialect or group of dialects, contrasted with their counterparts in other dialects, without consideration of diachronic and extra-linguistic criteria. If the interest is focused on cultural and historical points of view, diachronic and comparative data play a crucial role. Linguistic features mirroring the movements and interrelationships of various groups, as well as their ecological environments, stand out as relevant criteria for classification.

Dialects are identified and their boundaries defined by means of isoglosses. Drawing a number of isoglosses on a map normally exhibits border areas in which a number of isoglosses lie close enough together to constitute bundles of isoglosses marking boundaries between different dialect areas. The bundles normally reveal the focal area of a dialect, and between the focal areas there are transitional areas in which the isoglosses do not tally with the bundles and in which contrasting items may be used interchangeably. In a parallel way, groups of dialects can be identified. A prerequisite for exact synchronic classifications is thorough dialect-geographical study in which a sufficient number of relevant isoglosses have been drawn on a map. Early dialect maps of Arabic are Bergsträsser's map of Syria and Palestine (1915) and Cantineau's atlas of the Ḥōrān region in southern Syria (1940); more recent works are Behnstedt's dialect atlas of North Yemen (1985), Behnstedt's and Woidich's atlas of Egypt (1985), Behnstedt's atlas of Syria (1997), and the dialect-geographical appendix of de Jong's grammar of the Bedouin dialects of the northern Sinai littoral (2000). An exceptionally illuminating single example of the results of a systematic survey is the colored overview map of the sedentary Arabic dialects of Egypt (Behnstedt and Woidich 1983), which, in addition to twelve distinct dialect areas, specifies two areas in which different dialects coexist, six transitional areas, and a number of locally limited dialects, all defined by means of 50 selected distinctive features. This map, based on purely synchronic material, contains details – among them ‘town dialect with Cairene elements’ and ‘dialect in regression’ – which suggest that it can also be read diachronically.

2. Eastern vs. Western dialects

Dialect-geographically, the Arabic-speaking area can be divided into an Eastern (mašriqī) and a Western (maġribī) dialect group. Until the 1970s, the boundary between the two groups was commonly drawn from the western border of Egypt on the Mediterranean coast in the north to Lake Chad in the south. As the most distinctive individual isogloss, the conservative inflection of the 1st persons sg. and pl. in the imperfect (aktib, niktib) in the Eastern group, and the paradigmatically leveled innovative inflection (niktib, niktibu/níkitbu) in the Western dialects was used. However, more detailed investigations have shown that this isogloss runs through the western Delta and follows the Nile Valley between Asyūṭ and Luxor (Behnstedt and Woidich 1985, maps 210–213). Because the dialects of these areas share a significant number of distinctive features with the Egyptian dialects, this isogloss cannot be used as an absolute classificatory criterion, but other typologically prominent differences must be considered as well. Important distinctive features of the Western dialect group include the following: (a) loss of inherited short vowels in medial position; (b) non-phonemic vowel quantity; (c) aspiration of t [t^s] < *t and *ṯ; (d) the syllable patterns CVCC > CCVC (*rijl > ržəl) and CVCV- > CCV- (*katab > ktəb); (e) the use of an innovative indefinite article *wāḥid al-; (f) the use of -āš (*ʾayy šī < ʾayy šayʾ) to form adverbs and conjunctions (kīfāš ‘how?’, bāš ‘in order to’, etc.); and (g) high ratio of analytical genitives (Marçais 1977:iv–vii). Among these, (a), (b), (c), and (d) apply especially to the Western branch of the maġribī dialects. In addition, there are noticeable differences in the vocabulary, both in lexical items and their semantic sense.

3. Bedouin vs. sedentary dialects

Arabic dialects cannot be properly classified without attention to the stratification of society. One relevant point of departure is a sociologically-based grouping of them into sedentary (ḥaḍarī) and Bedouin (badawī) dialects. The sedentary dialects can further be divided into urban (madanī) and rural (qarawī ‘village’ or fallāḥī ‘peasant’) dialects. These divisions reflect the history of settlement and are applicable to the classification of the dialects in virtually the entire Arabic-speaking world, but with a wide range of variation as to the degree of mutual divergences. It has to be emphasized that these designations in this context refer to different dialect types, irrespective of the present-day division between urban, rural, and Bedouin populations. As a result of radical changes in the course of history, in a number of cities the majority of the population speak dialects of a Bedouin, or sedentarized Bedouin type, many villagers speak Arabic of an urban type, and in several old urban centers the inhabitants speak Bedouinized dialects.

In a classification exclusively based on linguistic contrasts, scarcely any single criterion besides the reflexes of *q distinguishing between the Bedouin-type and sedentary-type dialects can be found. However, there are a number of prominent typological features, some of which are shared by all sedentary dialects, yet without constituting a contrast with all dialects of Bedouin type, and vice versa, and there are features constituting significant partial contrasts between the two groups. In the following list, the kind of contrast is indicated by adding (A) to the criteria which are shared by virtually all the dialects belonging to the group, and (P) to the criteria which are shared by a substantial part of the respective group (Table 1).

Table 1. Features of Bedouin and sedentary dialects

Bedouin dialects	Sedentary dialects
retained interdental fricatives (A)	interdental fricatives > postdental stops (P)
partially retained and generalized indefinite marker -in (tanwīn) (P)	no indefinite marker -in, except in formulaic expressions (A)
retained gender distinction in plural (P)	no gender distinction in finite verbs and personal pronouns (P)
no verb modifiers in the imperfect (P)	different verb modifiers in the imperfect (A)
internal passives productively used (P)	absence of internal passives (A)
retained productivity of Form IV (P)	absence of Form IV (P)
very low frequency of analytical genitive	common use of analytical genitive structures (A)

Among Bedouin dialects, a division can be made between those which use phonetically-conditioned affrication of g and k (peninsular Bedouin dialects) and those which do not have affricated allophones (northwest Arabian dialects, Egyptian and North African Bedouin dialects). In the classification of a more restricted group of Bedouin dialects, the contrasting isoglosses are the different variants of the affricated allophones: g/j and k/č (Syro-Mesopotamian = ‘pre-ʿAnazī’) vs. g/ǵ and k/ć (northern Najdī = ‘Šammarī’ or central Najdī = ‘ʿAnazī’). Another criterion is the division between the inflectional suffixes in the imperfect: -īn, -ūn (Najdī, Ḥijāzī Bedouin, the Gulf dialects, sedentary and Bedouin Mesopotamian) vs. -ī, -ū (urban Ḥijāzī, northwest Arabian, Bedouin of Egypt and North Africa; common sedentary type).

As is obvious from these differences, the Bedouin dialects have retained more morphophonemic categories than the sedentary dialects. However, they also exhibit innovations. Thus, the partial retention of the tanwīn in the Najdī dialects is a striking conservative feature, but this tanwīn is different from the Old Arabic indefinite marker. It occurs in new morphological categories (e.g. suffixed to sound pl. masc.: muslimīnin) and has partly new functions. Another highly conservative feature is the retention of the internal (apophonic) passive. This category also exhibits innovations: in the northern Najdī dialects the passive vocalism in Form I (act. kitab, pass. ktib < *kitib) has been applied to derived forms as well: libbis ‘he was dressed’, ʿīlij ‘he was treated’, irsil ‘he was sent’, tgīsim ‘it was shared’, iʿtibir ‘it was considered’, stigbil ‘he was welcomed’. Another innovation in verbal morphology is the n-passive of the t-reflexives: yintalabbas ‘he can/should be dressed’, tinti-gāsimōn ‘you [pl. masc.] are apt to be shared with’ (Abboud 1979:474, 476–477; cf. the nt-/tn- forms in Algeria, Marçais 1958:195–196).

It is a general tendency in Arabic dialects to develop toward more analytical structures. The sedentary dialects have as a rule proceeded farther than the Bedouin dialects in this direction. The domains in morphosyntax most clearly displaying the sedentary vs. Bedouin dichotomy are the aspect/tense system in the imperfect and the genitive structures. In Bedouin dialects of the Najdī type, the old aspect-centered system is preserved, with only incipient development in the macro-structure toward a new tense-based system (Ingham 1994:87), whereas most sedentary dialects have a well-developed relative tense system implying the use of different verb modifiers. The very low frequency of analytical genitive structures is another synthetic trait in Bedouin dialects.

4. The sedentary dialects of the Eastern group

5. The Western dialects

The Western dialects can be divided into two major groups: the so-called pre-Hilālī sedentary dialects and the Bedouin dialects. The former hark back to the first phase of Arab immigration (7th–10th centuries C.E.). The rural dialects of the Jbāla in northern Morocco as well as those spoken around Nedroma in the northwestern corner of Algeria and in the neighborhood of Djidjelli and Collo in northeastern Algeria also belong to this phase. These dialects display considerable substrate influence from Berber languages.

In the 11th century the originally Najdī tribes of Banū Sulaym and Banū Hilāl and the southern Arabian tribe of the Maʿqil moved westward and occupied the North African plains and steppes. At present, Sulaymī Bedouin dialects are spoken in Libya, southern Tunisia, and northeastern Algeria; eastern Hilālī dialects in central Tunisia and eastern Algeria; central Hilālī in central and southern Algeria; northern Hilālī in the northern part of central Algeria; and Maʿqilī dialects in northwestern Algeria and Morocco. The differences between the Bedouin dialects in the whole Western dialect area are relatively slight. In the Maʿqil and northern Hilālī dialects *j > ž, ġ is retained, and the 3rd pers. sg. masc. suffix pronoun is -ah, whereas the counterparts in the central Hilālī dialects are *j > ž, *ġ > q, and -u (Grand'Henry 1976:4–5; Fischer and Jastrow 1980:31–38).

As a result of the Bedouin migrations, clear-cut distinctions developed between urban, rural, and Bedouin dialects. The long belt of urban pre-Hilālī dialects begins with the old Tunisian cities of al-Qayrawān, Mahdiya, Sousse, and Tunis. In Algeria it continues with the littoral cities of Skikda, Djidjelli, Dellys, Cherchell, and Ténès, and the interior cities of Constantine, Médéa, Blida, and Miliana. In the westernmost part of Algeria the pre-Hilālī dialects include the dialect of Tlemcen, the old urban center of Orania, surrounded by a wide area of Bedouin dialects, and to the northwest of it, the dialect of Nedroma. In Morocco, old urban dialects are spoken in Old Fes, Rabat, Salé, Taza, Tangier, and Tétouan; these constitute the northern group of urban Moroccan, with the present tense marker kā- as a salient feature, distinguishing the group from the southern urban Moroccan spoken in Marrakesh and New Fes, which have tā-. The new cities of Casablanca and Mogador represent Bedouin-type dialects (Fischer and Jastrow 1980:33–35).

The pre-Hilālī dialects of the Maġrib can be divided into an Eastern and a Western branch. The Eastern branch, comprising Libya, Tunisia, and easternmost Algeria, has a more conservative structure, as is apparent from the following phonological traits. The interdental fricatives are retained in all Tunisian dialects except Mahdiya and the Jewish dialects. Inherited short vowels – e.g., in Tunis a, i, u – are better retained, whereas in the Western group they have been reduced into a vs. ə or only one phoneme, as in Djidjelli. The reflexes of the diphthongs *aw and *ay are ū and ī, as in the Western branch, but in Mahdiya they are ū and ī, in Sfax aw and ay. Also, the women in Tunis and Sousse, as well as the Jews in these towns, have retained the diphthongs (Cohen 1975:65–67; Singer 1980:249–251)

One prominent feature shared by the oldest urban dialects of the area is the glottal stop reflex of *q, typical of the urban dialect of Fes, but occurring in Rabat, Tétouan, Tlemcen, and, significantly, also in Maltese. When classified according to the Eastern vs. Western division, Maltese undoubtedly represents the latter. In a contrastive analysis on the basis of 37 isoglosses, Maltese shared 25 with the urban pre-Hilālī maġribī dialects (Vanhove 1998). Its phonological innovations, interesting from the comparative Semitic point of view, include the merger of x and ḥ > ḥ, realized as [h], [ħ], or [χ], and of ġ and ʿ > ʿ, and further > Ø, still discernible as pharyngealization of the adjacent vowels (Schabert 1976:45–50).

In Algeria, the old urban dialects of the interior, except the prestigious dialect of Tlemcen, have been influenced by neighboring Bedouin dialects; in Morocco, this is the case with Marrakesh and Meknès. The rural dialects spoken in wide areas adjacent to Djidjelli and Nedroma have exerted a considerable influence on the dialects of these towns; in Morocco, the same development has taken place in Tangier (Fischer and Jastrow 1980:34; Iraqui-Sinaceur 1998:138–139). The population of Algiers, one of the pre-Hilālī urban centers, has during the last few generations grown too heterogeneous to render it meaningful to speak about its dialect any more (Boucherit 2002:24–25). In Libya, the most closely sedentary-type dialect is that spoken in Tripoli, which can be characterized as a Bedouinized former urban dialect. The few extant features of the pre-Hilālī urban type include the postdental stop reflexes of the interdentals and the reflexes of the verb *raʾā ‘to see’ (Fischer and Jastrow 1980:36).

6. Classification according to religious affiliation

In many Arab cities, religion correlates more or less closely with dialect. One of the most noticeable cases is the situation in the Mesopotamian dialects. The dialect of the Christians of Baghdad differs from that of the Muslims in several points, among them the following: the interdental fricatives have become postdental stops vs. retained interdentals; *r > ġ vs. retained r; retained q vs. g; retained k vs. k + phonetically-conditioned č; use of the present markers qad-, qa- vs. gāʿed, da-; use of the optative marker da- + 1st pers., e.g., daqūl ‘let me say’, danqūl ‘let us say’ vs. xaldangūl (Blanc 1964:20, 25–26, 115–118; Abu-Haidar 1991:7–9, 88–89). The Jewish dialect, which until the beginning of the 1950s was spoken by a significant number of the population of Baghdad, was to a high degree identical with that spoken by the Christians. Salient Jewish Arabic features were, e.g., retained interdentals, a strong ʾimāla, and the future marker (has)sa- (Blanc 1964:42–43, 117–118; Abu-Haidar 1991:29).

In Mosul, where the whole population irrespective of religious affiliation speaks qəltu dialects, the differences are minimal when compared with Baghdad (Jastrow 2004:141–142). In Aleppo differences are also found between the dialects of the Muslims and the Christians. At the beginning of the 20th century there still were different Christian dialects in different quarters, but since then these divisions have blurred (Behnstedt 1989:43–44). Salient Aleppine Christian traits include the retention of the diphthongs aw and ay; the use of the glottal stop reflex of *q without the back allophone of a and ā in juxtaposition to it; and ʾimāla in many cases in which the dialect of the Muslims does not have it (Behnstedt 1989:45–63).

The division between two different dialect types in Bahrain is parallel with the earlier development in Lower Iraq. Although representing the Shiʿi–Sunni split, it is in fact a result of two phases of settlement: the Shiʿi population speak the old rural Baḥārna dialect, which displays typical sedentary devices, whereas the Sunni newcomers speak a dialect of the ʿAnazī Bedouin type (Holes 1995:272–273).

In North Africa the Jewish Arabic dialects are of urban type and represent the first phase of Arab settlement. Their phonology is markedly urban: the reflex of *q is the glottal stop (Algiers); the old interdental fricatives have become postdental stops; h has disappeared; š and s, as well as ž and z have merged; short vowels have been highly reduced (Marçais 1977: 9–11). In Oran and some towns in the region of Algiers, the Jewish dialects represent the sedentary, and the Muslim dialects the Bedouin type. As pointed out by Blanc (1964: 16), the parallel with the distribution of the qəltu vs. gilit dialects in Lower Iraq is striking. Here, as in all other cases of dialect differences along the lines of religious affiliation, the differences – besides religious-cultural technical terms – can mainly be attributed to settlement history.

7. Classification of dialects on the Eastern–Western boundary

8. De- Bedouinization, sedentarization, and Bedouinization developments

The dialects spoken in the Arabian Peninsula, except its southwestern parts, are Bedouin or former Bedouin dialects. In sedentary environments the Bedouin dialects tend to adopt reductional and innovative traits, plausibly as results of increased dialect contact. The Meccan dialect, for instance, while displaying several Bedouin-type features such as the g reflex of *q and the productive use of Form IV, at the same time exhibits many traits typical of sedentary dialects, among them the following: absence of interdental fricatives; absence of gender distinction in plural forms of finite verbs and personal pronouns; optional use of the present continuous marker bi-; and frequent use of the genitive exponent ḥagg (Schreiber 1970:6 and passim; Ingham 1971:273 and passim).

In a corresponding way, the Arabic dialects of the Gulf area, which are a relatively recent offshoot of the ʿAnazī dialect group, during only two centuries of sedentarization have adopted considerable reductional changes, such as elimination of the internal passive, the indefinite marker -in, and, mostly, gender distinction in plural forms of finite verbs and personal pronouns. An innovation typical of sedentary dialects is the future and volitive marker bi-. These dialects have thus drawn away from their original central Arabian ʿAnazī dialects (Ingham 1982:33; Holes 2001:xviii). In maritime environments, the vocabulary naturally differs noticeably from the one mirroring the traditional nomadic culture. Even the rhythm and intonation patterns are at the present time quite different from the dialects of the ʿAnazī type (Johnstone 1967:18).

The Mesopotamian gilit dialects exhibit similar developments in the reductional and innovatory direction. The rural gilit dialects have still preserved several prominent features of Bedouin type, contrasting with the urban gilit dialects. These include the phonetically-conditioned affrication of the reflex of *q; the gahawa syndrome; retention of gender distinction in plural forms of finite verbs and personal pronouns; and infrequent use of the verb modifier gāʿid/jāʿid. In the Muslim dialect of Baghdad the sedentary-type development has advanced much more, obviously under influence from the old sedentary qəltu dialects. Typical sedentary traits, such as the use of the verb modifiers gāʿed, da-, and rāḥ/raḥ with the imperfect, d(e)- with the imperative, have been adopted. An important Aramaic substrate feature has also found its way to the Muslim dialect of Baghdad through the qəltu dialects, viz. marking the definite direct object by affixing to the verb a suffixed pronoun referring to the object and introducing the object epexegetically by the preposition l-, e.g., bāʿa lilbēt ‘he sold the house’ (Blanc 1964:128–130).

During the Ottoman period, in particular, new Bedouin tribes settled down in the neighborhood of towns and villages lying near the fringes of the Syrian Desert. One of the results was a progressive Bedouinization of the old sedentary dialects in these areas. Examples of this development are the qəltu dialects of the Euphrates group and the few sedentary dialects spoken to the east of the Jordan (Blanc 1964:26–27; Jastrow 1978:25–26; Palva 1994:468–469).

9. Isolated dialects

Cypriot Arabic, the dialect spoken by a few thousand Maronite Christians in Cyprus, synchronically to be classified as an isolated dialect displaying considerable superstrate influence from Cypriot Greek, can historically be classified as representing the old sedentary dialects of the Fertile Crescent. According to Borg's definition (1985:157), Cypriot Arabic “represents a now superstratally modified variety of a dialectal prototype antedating the present areal configuration obtaining among Arabic-speaking sedentaries in this region”. A prominent typological feature which Cypriot Arabic shares with northern Syrian dialects is the vocalically conditioned ʾimāla (Borg 1985:156–157; Behnstedt 1997, maps 43–62). Cypriot Arabic also shares a number of salient traits with the southeastern branch of the Anatolian qəltu dialects, among them, -n in the suffixed personal pronoun of the 2nd and 3rd pers. pl. com.; use of copulas derived from independent personal pronouns; use of reflexes of *ḥattā as a verb modifier marking the future tense; use of genitive exponents derived from *ḏayl; and dropping of h in personal and demonstrative pronouns. Important affinities to the present-day sedentary dialects of Greater Syria include the b(i)- non-contingency marker in the inflection of the imperfect; retention of reflexes of tāʾ marbūṭa in numerical constructs; reflexes of *hunnā, *-kun, *-hun; and a genitive marker harking back to šayʾat- (Borg 1985:154–155). An interesting trait is the partial retention of *raʾā, which attests its use in Syrian Arabic during the first Islamic centuries.

Bibliography

• Abboud, Peter F. 1979. “The verb in northern Najdi Arabic”. Bulletin of the School of Oriental and African Studies 42.467–499.
• Abu-Haidar, Farida. 1991. Christian Arabic of Baghdad. Wiesbaden: O. Harrassowitz.
• Aguadé, Jordi, Patrice Cressier, and Ángeles Vicente (eds.). 1998. Peuplement et arabisation au Maghreb occidental: Dialectologie et histoire. Madrid and Saragossa: Casa de Velázquez.
• Behnstedt, Peter. 1985. Die nordjemenitischen Dialekte. I. Atlas. Wiesbaden: L. Reichert.
• ——. 1989. “Christlich-aleppinische Texte”. Zeitschrift für Arabische Linguistik 20.43–96.
• ——. 1994. Der arabische Dialekt von Soukhne (Syrien). II. Phonologie, Morphologie, Syntax. III. Glossar. Wiesbaden: O. Harrassowitz.
• ——. 1997. Sprachatlas von Syrien: Kartenband: Beiheft. Wiesbaden: O. Harrassowitz.
• ——. 1998. “La frontière orientale des parlers maghrébins en Egypte”. Aguadé a.o. (1998:85–93).
• —— and Manfred Woidich. 1983. “Ägypten: Arabische Dialekte”. Tübinger Atlas des Vorderen Orients, AVIII-12. Wiesbaden: L. Reichert.
• ——. 1985. Die ägyptisch-arabischen Dialekte. II. Dialektatlas von Ägypten. Wiesbaden: L. Reichert.
• Blanc, Haim. 1964. Communal dialects in Baghdad. Cambridge, Mass.: Harvard University.
• Borg, Alexander. 1985. Cypriot Arabic. Wiesbaden: F. Steiner.
• Boucherit, Aziza. 2002. L'arabe parlé à Alger: Aspects sociolinguistiques et énonciatifs. Paris and Louvain: Peeters.
• Cantineau, Jean. 1938. “Remarques sur les parlers des sédentaires syro-libano-palestiniens”. Bulletin de la Société de Linguistique de Paris 40.80–88.
• ——. 1940–1946. Les parlers arabes du Ḥōrān. Atlas. Paris: Klincksieck.
• Cohen, David. 1970. Études de linguistique sémitique et arabe. Paris: Mouton.
• ——. 1975. Le parler arabe des juifs de Tunis. II. Étude linguistique. The Hague and Paris: Mouton.
• Fischer, Wolfdietrich and Otto Jastrow (eds.). 1980. Handbuch der arabischen Dialekte. Wiesbaden: O. Harrassowitz.
• Grand'Henry, Jacques. 1976. Les parlers arabes de la région du Mzāb (Sahara algérien). Leiden: E.J. Brill.
• Holes, Clive. 1995. “Community, dialect and urbanization in the Arabic-speaking Middle East”. Bulletin of the School of Oriental and African Studies 58. 270–287.
• ——. 2001. Dialect, culture, and society in Eastern Arabia. I. Glossary. Leiden: E.J. Brill.
• Ingham, Bruce. 1971. “Some characteristics of Meccan speech”. Bulletin of the School of Oriental and African Studies 34. 273–297.
• ——. 1982. North east Arabian dialects. London and Boston: Kegan Paul International.
• ——. 1994. Najdi Arabic: Central Arabian. Amsterdam and Philadelphia: J. Benjamins.
• Iraqui-Sinaceur, Zakia. 1998. “Le dialecte de Tanger”. Aguadé a.o. (1998:131–140).
• Jastrow, Otto. 1978. Die mesopotamisch-arabischen qəltu-Dialekte. I. Phonologie und Morphologie. Wiesbaden: F. Steiner.
• ——. 2004. “Jüdisches, christliches und muslimisches Arabisch in Mossul”. Approaches to Arabic dialects: A collection of articles presented to Manfred Woidich on the occasion of his sixtieth birthday, ed. Martine Haak, Rudolf de Jong, and Kees Versteegh, 135–150. Leiden: E.J. Brill.
• Johnstone, T.M. 1967. Eastern Arabian dialect studies. London: Oxford University Press.
• Jong, Rudolf de. 2000. A grammar of the Bedouin dialects of the northern Sinai littoral, bridging the linguistic gap between the eastern and western Arab world. Leiden: E.J. Brill.
• Marçais, Philippe. 1958. Le parler arabe de Djidjelli (Nord constantinois, Algérie). Paris: Adrian Maisonneuve.
• ——. 1977. Esquisse grammaticale de l'arabe maghrébin. Paris: Adrian Maisonneuve.
• Owens, Jonathan. 1995. “Minority languages and urban norms: A case study.” Linguistics 33.305–358.
• Palva, Heikki. 1994. “Bedouin and sedentary elements in the dialect of es-Salṭ: Diachronic notes on the sociolinguistic development”. Actes des premières Journées internationales de dialectologie arabe de Paris, ed. Dominique Caubet and Martine Vanhove, 459–469. Paris: INALCO.
• Schabert, Peter. 1976. Laut- und Formenlehre des Maltesischen anhand zweier Mundarten. Erlangen: Palm und Enke.
• Schreiber, Giselher. 1970. Der arabische Dialekt von Mekka: Abriß der Grammatik mit Texten und Glossar. Ph.D. diss., University of Münster.
• Singer, Hans-Rudolf. 1980. “Das Westarabische oder Maghribinische”. Fischer and Jastrow (1980: 249–276).
• ——. 1984. Grammatik der arabischen Mundart der Medina von Tunis. Berlin and New York: W. de Gruyter.
• Vanhove, Martine. 1998. “De quelques traits préhilaliens en maltais”. Aguadé a.o. (1998:97–108).
• Woidich, Manfred. 1980. “Das Ägyptisch-Arabische”. Fischer and Jastrow (1980:207–248).
• ——. 1993. “Die Dialekte der ägyptischen Oasen: westliches oder östliches Arabisch?” Zeitschrift für Arabische Linguistik 25:340–359.
• ——. 1996. “Rural dialects of Egyptian Arabic: An overview”. Centre d'études et de documentation économique, juridique et sociale: Égypte/Monde arabe 27–28.325–354.