Diglossia in North Africa
Summary and Keywords
Diglossia refers to a situation where two linguistic varieties coexist within a given speech community. One variety, labeled the ‘high variety’, is used in formal domains including education, while the other variety, labeled the ‘low variety’, is used principally in instances of informal extemporaneous communication. The domains of use, however, are not strictly separate and especially so with the increase in electronic modes of communication. This results in what has been described as diglossic code-switching, and the gradual encroaching of, in the case under consideration here, vernacular Arabic upon the domains of use of Standard Arabic.
While the genetic relationship between the two varieties is central in the definition of a classical diglossic situation as in the case of Arabic, the concept of diglossia has often been extended in the literature to cover situations of a functional distribution between languages that are genetically distant, such as with the situation of Spanish and Guaraní in Paraguay.
In North Africa, vernacular Arabic is in a classical diglossic distribution with Standard Arabic, while the Berber languages are often described as existing in a situation of extended diglossia with Arabic. However, distinguishing between diglossia as it exists between the Arabic dialects and Standard Arabic and the situation of bilingualism that involves Arabic, Berber, and European languages provides the best framework for describing the linguistic situation in North Africa. Diglossia is a key element in understanding the mechanisms of the region’s language contact and change as it plays a central role in shaping language attitude, language policy, and language planning.
1. Defining Diglossia
In its most accepted definition, the term diglossia refers to the existence of two or more clearly distinguishable varieties of a given historical language. The term was used early on in the literature to describe the case of Greek, considering not only lexical variation but also structural divergence between the vernacular Demotic variety and the standardized Katharevousa variety (Krumbacher, 1902; Psichari, 1928). As early as 1930, in his article “La diglossie arabe,” William Marçais explicitly addressed the diglossic situation in French North Africa. As a high-level administrator in the colonial education system, Marçais drew attention to the distance between written Arabic and the different Maghrebi dialects. He stated quite forcefully that the major challenge for education in the region was the existence of this situation, which he compared to a two-headed monster.1 In 1959, Charles Ferguson offered a theoretically grounded formulation of the concept of diglossia, referencing four cases that included both Greek and Arabic (Ferguson, 1959a). Most importantly, Ferguson’s definition included a list of traits that delineated the type of relationship between any two-language systems that are in a diglossic situation and the features that relate them and set them apart.
To summarize briefly, Ferguson postulated that, while the high variety (H) is learned formally and is held in high esteem by the majority of the speakers, the low variety (L) is acquired naturally and is often perceived as a corrupt version of the former. In addition, the H variety is standardized and has a long history of literary production, while the L variety is an oral vernacular that is generally not recorded in written form. As such, the H variety is the code for formal communication, including education, religion, and government, while use of the L variety is limited to informal interactions. Structurally, the difference between the two varieties is significant, but they continue to share major morphosyntactic features.
Despite some shortcomings, Ferguson’s definition seemed a suitable model at the time to describe the Arabic situation. This is especially true given his own familiarity with Arabic and emphasis on the relevance of the genetic relationship between the two varieties that are in a diglossic situation. The central role of this genetic relatedness and the speakers’ awareness of it were reiterated by Ferguson in his 1991 article:
My intention was that the users would always view the two as the same language: I excluded cases where superposed on an ordinary conversational language is a totally unrelated language used for formal purposes, as in the often-cited case of Spanish and Guaraní in Paraguay
(Ferguson, 1991, p. 223).
Ferguson’s remarks are a direct reference to the concept of extended diglossia that was introduced by Joshua Fishman (1967). Fishman proposed that any two languages that are in a functional distribution could be in a diglossic situation, regardless of the genetic distance that separates them. Since then, the term diglossia has become equated with social bilingualism with an ever-increasing range of scenarios of functional distribution all described as diglossic. At the same time, studies of cases similar to that of Arabic, i.e. classical diglossia, remain rare. A major exception is a significant body of work analyzing the development of the Romance languages through the theoretical framework of classical diglossia (Wright, 1982, 1994; Lloyd, 1987; Posner, 1996), of value to those interested in cases such as that of Arabic and Greek. Findings from studies on classical diglossic situations and their resolution, especially in the case of Maltese for Arabic, have also been relevant to those interested in dialect drift and the genesis of new varieties.
In North Africa, diglossia refers to the existence of different varieties of Arabic that are used in complementary domains. However, the strict functional separation proposed earlier on does not successfully capture the complexity of the situation. Be it in school, mass media, or even mosques, the local varieties of Arabic tend to find their way into domains that were previously off-limits or assumed to be, albeit with varying degrees. The exclusive use of the H variety is confined to contexts where extemporaneous discourse is not an option.
On the other hand, bilingualism describes the use of the European languages for a variety of educational and professional domains and the use of the Berber languages for intracommunal communication in areas where it is still maintained. Rather than describing the linguistic situation in North Africa in terms of both classical diglossia and extended diglossia, I will use diglossia to refer to the coexistence of different varieties of Arabic and bilingualism to describe the existence of Berber and the European languages alongside varieties of Arabic. More discussion on the problematic notion of extended diglossia is offered in the section on the critical analysis of scholarship.
2. Arabic in North Africa
2.1 Introduction and Spread of Arabic
Despite the swift progress of Muslim troops through western North Africa, the spread of Arabic was not as fast-paced.2 In fact, Arabic initially was more limited to coastal areas and newly established sedentary centers. But with the arrival from Egypt of numerous Bedouin tribes of Banu Hilal in the 11th century, estimated at 200,000 people (Laroui, 1977), the language spread to the interior areas of the region. The Berbers, who had for centuries confronted successive attempts of colonization, finally started to adopt Arabic as they converted to Islam. As they shifted to Arabic, several features from Berber were transferred and resulted in substantial lexical and structural differences that help set Maghrebi Arabic apart from other Arabic dialect groups. The prominent 14th century chronicler Ibn Khaldun noted the importance of contact with Berber in the formation of Maghrebi Arabic varieties when he stated that: “[…] there originated another, mixed language in which the non-Arab element was preponderant, for the reasons mentioned. The language spoken there [in North Africa] is more remote from the ancient language than other dialects” (Ibn Khaldun, 1958, p. 352).
Today, although Arabic is without a question the dominant language, the extent of Arabization varies from one country of the Maghreb Union to another. Overall rates oscillate between a fourth and a fifth part of the population in Morocco and Algeria who speak Berber (Boukous, 1997; Ennaji, 2005; Chaker, 2001; Benrabah, 2007), while in Tunisia Berber speakers are often estimated at a mere one percent of the population (Gabsi, 2003). Even so, it is safe to argue that it would be rare to find monolingual speakers of Berber with no familiarity of Arabic, especially males who are not of preschool age. The majority of Berber speakers are bilingual, with varying levels of competence in both languages, even if they use Berber for communal and family interaction. In northern Morocco, for instance, although a large portion of the population are native speakers of Berber, the majority also speak Arabic even if it is as second language in many cases. The speaker quoted below was interviewed in December 2015. She grew up in the northern Moroccan city of Nador, where she learned Berber as her first language in the 1990s, but because she attended Spanish schools still present in the region, her competence in Standard Arabic is very limited while her competence in vernacular Arabic was developed through contact with Arabic speakers from other areas of Morocco. Her case is not an exception, although Arabized schools and population movement are accelerating the spread of Arabic.
The truth is that when I was little, on the street, almost no one spoke Arabic, hardly anyone, because most people were from Nador. Now that is not the case anymore, there are people coming from the South [of Morocco] speaking Arabic. Now Arabic is spoken a lot. But in my case, it was because of the Arabic teacher [at the Spanish school] and my classmates who were from the South as well. It is simply that. Now I speak Arabic very well, but when I was little I spoke it badly […]. I could not hold a fluent conversation and I had to think what it was that I wanted to say. But as I was growing up between interacting with people and my partner who is from the South, I have become fluent.3
As mentioned above, speaking Arabic refers principally to speaking the L variety, that is, the vernacular of a given region, which is the native language of the great majority of the speakers. The H variety is learned through formal schooling. Most importantly, the role of religion in the maintenance of the prestige and symbolic value of the H variety, and Arabic in general, cannot be underestimated, as is the case in other parts of the Arabic speaking world. In fact, the only place where an L variety of Arabic has risen to gain an official status is Malta, a country where Islam is not the predominant religion. It has to be pointed out, however, that throughout North Africa with increasingly more nationalistic feelings there is a growing pride in and identification with one’s national dialect.
2.2 Lexical and Structural Differences Between the H Variety and the L Variety
While the H variety of Arabic is fully described and its rules closely scrutinized by both language academies and individual language monitors, the L varieties are not standardized and show a wide range of variability from Standard Arabic and amongst themselves. At the lexical level, native Arabic words have been adapted to the morphosyntactic patterns of the dialects and oftentimes show cases of semantic drift and lexicalization. For instance, the verb ‘to bring’ in Standard Arabic—ʒa:ʔa bi (ʒalaba or ʔata: bi)—becomes ʒɛ:b in Tunisian Arabic. In Moroccan Arabic the verb ‘to rely on’ is ʕawwil ʕla, while in Tunisian Arabic it is ʕammil ʕla, and in Standard Arabic it is ʔiʕtamada ʕala:. The North African Arabic dialects also show a substantial amount of lexical borrowing from Berber and the European languages. For example, in Tunisian Arabic there are words such as garʒu:ma ‘throat’, and fakru:n ‘turtle’ which are from Berber, and basku:la ‘scale’ and dibbu: ‘depot’ from French. Some of the older loans have been completely adapted to the structure of Arabic while others continue to preserve their French form, either in full or in part. In a study of lone French lexical insertions in Tunisian Arabic, Poplack et al. (2015) showed that the majority of lone noun insertions behave like Arabic words and should be interpreted as borrowings as opposed to cases of code-switching, even if they do not show full phonetic integration.
With regard to structural differences, they are present both at the phonological level and the morphosyntactic one. Examples include the preference of the Maghrebi dialects for initial complex consonant clusters, limited use of the glottal stop, and variation in the use of the uvular stop (the dialects vary between the voiced and voiceless variants while Standard Arabic has only the voiceless variant). This, in addition to variable stress and intonation, all contribute to set the dialects apart and make it reasonably easy to identify the native dialect of Arabic speakers even when they are using Standard Arabic. At the morphosyntactic level, by way of illustration, the North African varieties are identified as members of the western Arabic dialect zone specifically because of their use of an initial n- in imperfective verb forms when used with first person singular. The dialects also diverge from Standard Arabic in many other features, including personal pronouns, verb paradigms, expression of attributive possession, and discontinuous negative constructions. In sum, the H variety is distinguishable from the L varieties, which in turn vary significantly amongst themselves, to the degree that partial misunderstanding is not uncommon among speakers of Arabic from different parts of the Arabic speaking world and even within North Africa itself (S’hiri, 2002). In (1) below, I illustrate how a simple question such as “What time is it now?” varies significantly in Standard Arabic, Tunisian Arabic, Algerian Arabic, and Moroccan Arabic. All examples are from adult speakers.
2.3 Domains of Use
As a result of these differences, multiple sociolinguistic factors, and speaker awareness of the existence of two varieties belonging to the same historical language, there is a discernible separation in domains of use. Since the L variety is the native language of Arabic-speaking North Africans, it is the system that they acquire and use exclusively before they start school. The situation is of course different for Berber speakers, who may acquire the L variety of Arabic simultaneously with Berber. On the other hand, Standard Arabic is learned formally starting from the first year of schooling. It is true that there is passive exposure to Standard Arabic through mass media before schooling, but that does not translate into active competence in the language, which is developed gradually together with literacy skills. Standard Arabic is not actively used outside the educational context and only later on in life it may be used for professional purposes, if at all.4 Before the spread of electronic communication, Standard Arabic would also have been the chosen code for written personal communication, but with current technologies the L varieties are spreading faster, as will be detailed in section 6.
For all intents and purposes, the domain separation has to do not only with speaker awareness of the artificiality of Standard Arabic for natural communication but also with deep-seated variability in levels of speaker competence and language ideology.5 As in the case of any formally acquired system, learners of Standard Arabic reach different levels of proficiency depending on many factors. In this case, school success and competence in Standard Arabic feed into each other. But in a region where rates of illiteracy remain high and school success fluctuates, competence in Standard Arabic is not always achieved at levels where its use can be normalized easily. As a result, Standard Arabic is not spreading significantly into domains of the vernacular (though it has entered domains dominated by French only a few decades ago, as noted in section 3), but the opposite is true. The functional separation, although it tends to hold relatively well when it comes to the use of Standard Arabic, is more in flux when it comes to the vernacular, which is gradually encroaching upon the domains of use of the Standard variety, as is discussed in sections 4 and 5.
I will now detail the domains of use of the H variety and the L variety and the interaction between the two. The focus will be on the situation in Tunisia with the understanding that its case is similar to what is found across the other western North African nations.
3. Diglossia in School
Following independence from France around the mid-twentieth century, Tunisia, Algeria and Morocco all struggled to find a successful path for long-term language planning policies. Although Arabic is the privileged language when it comes to legal standing, the laws granting it official status are understood for practical matters to refer to the H variety.6 In contrast, the L varieties have not gained any form of recognition and today continue to be officially at the margin of the educational system. Over the last sixty years, Arabization efforts in North Africa focused more on reducing the use of French in public domains, including education, than on coming to terms with any issues generated by the diglossic situation and its implications for school success (Maamouri, 1998). The lack of an official status for the vernacular means that from their first day of school students are required to start developing competence in the H variety of Arabic, and in French later on (Sayahi, 2015). The vernacular may be used for class management tasks and is the de facto variety for communication outside the classroom. It is also used systematically among students, unless they are addressing each other through a read-aloud activity or during classwork when student output is expected to be in the language of instruction, that being Standard Arabic or French. Consequently, the majority of the North African population is able to process different genres of discourse produced in Standard Arabic but not all are able to show advanced oral and written skills in that variety. As a result, there are no native monodialectal speakers of Standard Arabic and no one communicates with their parents in Standard Arabic in a natural setting.
A case in point on the chief role of school in the acquisition of Standard Arabic is the Berber community in Melilla. The city of Melilla belongs to Spain but is located on the Moroccan coast and has a large Muslim population. The second generation of Berbers born in the city speaks Spanish and Berber but generally not Arabic. If they do speak some Arabic, they usually possess a limited knowledge of colloquial Moroccan Arabic and not Standard Arabic. This is so because they attend Spanish schools where Arabic is absent, as illustrated in the excerpt below by a Muslim Berber-Spanish resident of Melilla. The same can be said about other North African diasporas, where the acquisition of Arabic as a heritage language consists in varying levels of competence in the parents’ dialects, which does not translate easily into the ability to communicate with speakers of other Arabic dialects or speakers of Standard Arabic.
I do not speak Arabic, not at all. I’ve always liked it and I regret it. I tell my father that what I would have liked is to learn Arabic, but how? If you don’t have Arab family members and you don’t go to school in a country or a city where Arabic is studied. When there are no resources, then I don’t know how.7
In North Africa, the competition throughout the educational system is not between Standard Arabic and vernacular Arabic but between Standard Arabic and French. As will be discussed below, French kept its presence in the postcolonial Maghreb, and still today continues as a language of instruction or vehicular language for the instruction of math, science, and technology subjects. Despite Arabization of a large part of the school system and public administration, many fields of study at the university level revert back to a monolingual system in French. These include medicine, engineering, and the hard sciences, where Arabic is not used.
To sum up, despite its paramount religious function and unquestionable role in individual and social identity construction in North Africa, Standard Arabic is not transmitted naturally but only formally acquired through the educational system.8 This in turn has resulted in a lack of recognition of the dialects and an absence of any educational materials in/about them. A common view held by many stakeholders, including many language specialists, is that the vernacular is unfit for education and that diglossia will end only if more attention is paid to the Standard variety. These opinions are often based on the belief that the diglossic situation is rather recent and that it was further aggravated by colonial policies and Francophile measures adopted following independence.9 In North Africa, the Arabization and bilingualism debate within the educational system is often set against a larger backdrop of ideological positions regarding modernization and type of relations to be had with the Western world on the one hand and the Arabic speaking world on the other.
4. Diglossia in Politics
In a region where political transparency and freedom of expression have historically been severely restricted, public officials do not usually hold press conferences or give live interviews to the media. As a result, political discourse is often delivered in a carefully scripted style in Standard Arabic. In Tunisia, an early exception was the first president of the Republic, Habib Bourguiba, who governed the country from 1956 to 1987. He was innovative in his use of both varieties of Arabic and French in notoriously lengthy addresses (Boussofara-Omar, 2006). Bourguiba was ousted by a military dictator who during a period of 23 years only read from prepared discourses and consequently used the L variety very rarely in public. Except for these carefully scripted speeches delivered on named occasions, discourse by official figures was all but muted. The Tunisian Revolution of 2010 brought about a total change in the political system which included freedom of speech, recognition of multiple political parties, and the holding of free elections at all levels, including democratic presidential elections for the first time.
With these changes, Tunisian Arabic has seen its use increased significantly in the political arena, including at press conferences and improvised speeches. It is true that speeches which are more formal in nature and those prepared in advance are delivered in Standard Arabic, but with higher frequency political figures are publicly using Tunisian Arabic as they answer questions in press conferences, give interviews, or engage in debate in the Parliament.
The change to a democratic system that requires more extemporaneous speech and an increase in the use of Tunisian Arabic contributes to higher frequency in the occurrence of diglossic code-switching, a concept that describes switching between both varieties of Arabic in the same communicative event (Walters, 1996; Boussofara-Omar, 2006; Sayahi, 2014). A quick look at speeches by the first freely elected president of Tunisia, Beji Caid Essebsi, shows that depending on the event he uses more or less Tunisian Arabic. In cases of recorded addresses to the nation, he reads from a previously prepared speech in Standard Arabic. He uses Standard Arabic and Tunisian Arabic in more formal interviews; while he uses Tunisian Arabic with single word switches to Standard Arabic in more natural interaction.
What the Tunisian Revolution has contributed to the functional distribution of both varieties of Arabic in political discourse is an extension in the use of Tunisian Arabic, which provides more authenticity to the discourse and anchors it in the reality of the Tunisian context. This is different from what appears in several other Arab countries where political discourse is fully scripted and only the H variety is used.
5. Diglossia in Mass Media
Without a doubt, an increased mediatization of society in recent years has resulted in a substantial surge in the use of the L variety in Arab mass media (Eid, 2007; Bassiouney, 2010). In audiovisual media in particular, more and more the use of the L variety is noted in all types of programming. This is especially true following the appearance of multiple private channels at the turn of the 21st century, which did not exist when the only source of audiovisual programming was state-controlled institutions. Today, the number of programs that exclusively use the H variety is shrinking and often they are limited to instances where a presenter is reading from a script. Otherwise, any program may have sections in both varieties and numerous instances of what would be considered diglossic code-switching. Even if a news anchor reads from a teleprompter, sections of the newscast such as interviews on the street are realized in the L variety and guests on the program switch between the two varieties depending on their level of comfort with the H variety. For example, if a French-educated medical doctor is invited to a health program where the host uses the H variety, the doctor might use more Tunisian Arabic with switching to both the H variety and French. This is particularly so in Tunisia, where before the Revolution live interviews were not allowed for fear that criticism of the then autocratic regime could be voiced on air. Moreover, there are private radio stations that appeared after the Revolution which deliver even the news in the L variety.
In example (2), the Director of the Tunisian National Observatory of New and Emerging Diseases is asked about the quality of public water control by the State. She uses primarily Tunisian Arabic, including loanwords from Standard Arabic and French as in the case of the abbreviation SONED (Société nationale d’exploitation et de distribution des eaux), with instances of code-switching to both Standard Arabic (underlined) and French (double underlined). The program host herself uses both Standard Arabic and Tunisian Arabic. The program is broadcast live on the State television channel and is a representative example of the spread of the L variety to unscripted programming, even in the case of programs that are more serious in nature.
(2)Audio 1: Doctor use of TN French
nħib ra:ni nṭamʔin l-muwa:ṭnin wu nʔakkid illi mɛ: s-SONED yṣir taħlilha yawmiyyan ou ʕiddit marra:t fil yawm. maʕnitha, min qbal, min qbal, wu min qbal maʕnitha fi l-wist fi l-muħi:ṭ θamma maʕnitha barnɛ:miʒ mtɛ:ʕ des prélèvements yṣi:ru de façon périodique. wu nħib nʔakkid illi mɛ: s-SONED l-kul fi t-taħa:lil illi ʕmalnɛ:ha l-ħadd al-yawm wu lɛ: taħlil kɛ:n fih l-ʒurθu:ma hɛ:ði. ra:hu lqinɛ:ha kɛ:n fi les stations d’adoucissement xaṭir yɛ:xðu: mɛ: s-SONED wa illa mɛ: ʃtɛ wa illa illi ylimu:ha wu yaʕmlu:lha ʕamaliyɛ:t zaʕma taħliya ɛma hiyya b-ħaʒɛ:t mɛ:hiʃ contrôlées wu iða lqi:na ʃaxṣ maʕnitha muṣa:b bi-l-maraḍ ynaʒʒim yku:n huwwa illi yʕaddi l-ʒurθu:ma hɛ:ði, hɛ:ði:ka aʕlɛ:ch ra:hu mɛ: s-SONED nħib nʔakkid illi θamma ʕamaliyɛ:t mura:qba l-mɛ: s-SONED b-ṣi:fa muntaḍma wu fi l- à froid wu tṣir yawmiyyan.
(Ben Alaya, 2016)
‘I want to reassure the citizens and insist that the water of the SONED [Tunisian National Water Company] is analyzed daily and even several times during the day. I mean, since before, since before, I mean in the middle, in the environment, I mean there is a program for samples to be taken in a regular periodic manner. And I want to stress that all the water of the SONED, in all the analyses we have done until today, no analysis contained that microbe. We’ve found it only at water softening stations because they get the water from the SONED, or they collect rainwater, and realize some processes as if they were softening it but these are things that are not controlled. And if we find a person, I mean, who has this disease it may be him who is passing the microbe. That’s because the water of the SONED, I want to insist, that there are control operations on the water of the SONED on a regular basis, they occur in cold and on a daily basis.’
With regard to the press, both print and online, the tendency is to use the H variety throughout. The high degree of informality associated with the L variety and the fact that literacy in Tunisian Arabic is purposefully not developed, discourages a more prolific use of the L variety in the press as compared to the audiovisual media. Nevertheless, there are instances where the L variety makes its way into the print press, especially when covering Tunisian topics.10 It is revealing, however, that in many cases readers react on media websites to texts written in Standard Arabic with comments written in the L variety.
An illustration of the expanding scope of Tunisian Arabic in the media is in the language of advertising. This is particularly true for wireless phone companies which almost always commercialize their service through ads written in Tunisian Arabic and which at times include code-switching to French. Figure 1 below shows a recent multilingual ad by a major phone company in Tunisia. The text in Arabic script in the orange box is afariyɛ:t ‘business’, an established French loanword in Tunisian Arabic. The other word in Arabic, in the center-right of the image, is part of an adjective phrase di:ma connecté ‘always connected’ where di:ma is a Tunisian Arabic adverb.
In conclusion, it appears that the less state-controlled a mass media source is, the greater the amount of unscripted material delivered, and the higher use of the L variety. This does not mean that the H variety is being displaced in all types of programming, but it indicates that the L variety is more prone to be used in cases of ad-libbing and when there is a strong focus on local issues. Additionally, the ideological leaning of a media organization and its target audience also play a role in how much the L variety is used. It is telling that having programs delivered in their entirety in Tunisian Arabic on some private television channels has led to instances where subtitles for an interview that was conducted in a foreign language are provided in Tunisian Arabic rather than in Standard Arabic. This means that Tunisians are for the first time reading their dialect on TV, which is a significant milestone in the increased recognition of the national variety.
6. Diglossia in Electronic Communication and Social Media
An important step in the process of recognizing the Romance varieties as distinct languages from Latin was the act of writing them, especially following the French Carolingian Reform in the 11th century. In the case of Arabic, and prior to the rapid increase in electronic communication, the writing of the L variety in North Africa and elsewhere was extremely rare and most often limited to folk literature, as rightly noted by Ferguson in 1959. The majority of people wrote in the H variety because, as mentioned in section 3, literacy is tied up with the H variety and anybody who knows how to write must have learned to do so using that variety. Writing was also more restricted to formal domains that overlap with those where Standard Arabic would be the unmarked code anyway. In recent years, however, electronic communication has led to a flood of texts written in the L variety that are publicly available. These include comments on news websites, in social media communication, and in instant messages, making it almost impossible for a literate North African not to have been exposed to the written form of the dialects at levels unknown to previous generations.
Users find ways to circumnavigate the lack of standardization and codification of the dialect by using the Latin alphabet, and the Arabic alphabet, as technological advances allow them to switch back and forth between keyboards. Writing skills from the H variety are easily transferred to the writing of the vernacular with the added advantage that the latter is the native variety, making speaker intuition about its use straightforward and, as a result, much less taxing cognitively. Furthermore, in the case of Tunisia, the popular revolt of 2010 imbibed the dialect with an added legitimacy, validating its use in domains that previously were those of the H variety. These include politics, culture, and many other levels of public life. The role of social media in the Arab Spring movements cannot be underestimated and the use of the L variety as part of these movements is undeniable. It is, indeed, in electronic media where both the H variety and the L variety cohabit much more intimately in the written form.
In example (3) below, the author is putting out a request for ophthalmologists to volunteer to screen potential patients at a short film showing in rural areas. The author shows a deliberate use of Tunisian Arabic and an intended approximation to the spoken language, starting with the use of the [g] sound in the beginning of the message, which is a clear trait of the dialects of the region where the planned activities would take place. At the same time there are instances of switching to Standard Arabic (underlined) and loanwords from French (double underlined).
ayya golna assala:m
[. . .] ayya ʕ:ad golna aʕlɛ:ʃ lɛ:, fi nafs el-waqt maʕa al- ʕur:uḍ al mubarmaʒa fi al-qura:, nʃu:fu kifɛ:ʃ naʕmlu kif al-ʕiya:da al-mutanaqila maʕa al- ʕur:uḍ, illi ʕandu mʃɛ:kil yafiħṣu ṭbi:b mtɛ:ʕ nḍar, wu tawwa aħna min ʃiritna nitkalfu bi- ʔiqtina:ʔ wa tawṣil al-naḍara:tt al-lazima li-ʔaṣħabha. l-ħa:ṣil kɛ:nik ophtalmo (ṭbi:b nḍar) wu illa taʕrif ophtalmo, wu illa taʕrif ʒamʕiyya tunsiyya mtɛ:ʕ ophtalmo wu l-ħkɛ:ya bɛnitlik bɛ:hiya wu tħib tmid fiha yɛddik, sayaku:n min dawa:ʕi: suru:rina: (ma lqitiʃ kilma oxra) anna naḍmu:ha mʕa bʕaḍna. parta:ʒi bɛ:likʃi tu:ṣil l-ʃku:n muhtam wu ynaʒim ysɛ:him.
[…] So we said why not, at the same time as the shows programmed in the villages, we see how we can have something like a mobile clinic together with the shows. Whoever has problems can be examined by an eye doctor, and from our side we will take care of acquiring and delivering eyeglasses to the people who need them.
In sum, if you are an ophthalmologist (eye doctor), or you know an ophthalmologist, or you know a Tunisian association for ophthalmology, and the idea appeals to you and you want to lend a helping hand, it would be our pleasure (I didn’t find another word) that we organize it together. Share in case it reaches someone interested and who can participate.’
This by all means does not exclude the use of the H variety in social media. Depending on the topic, linguistic abilities, and language ideology and attitude, users have a good deal of leeway in deciding in which variety to communicate. It is not uncommon to see educated North Africans make use of colloquial Arabic, Standard Arabic, French, and increasingly English in the messages they post on social media. Writing the dialects and publicly sharing texts in the L variety at such a large scale and with such a frequency represents a major shift in the role of the L varieties in modern North Africa.
7. Diglossia and Language Change
How does the coexistence of two varieties that are from the same historical language and whose increasing overlap in domains of use affect their respective structures? To answer this question, we have to keep in mind that the L variety is the native language and the H variety is the one acquired through conscious effort and formal instruction. This means that while level of competence in the native variety and speaker performance are comparable across users, the same cannot be said about the H variety, since levels of competence and opportunity to use it vary by age, level of education, and occupation.
As the H variety of Arabic is acquired in a formal context and there are no model native speakers, the L variety inevitably influences speaker performance in the H variety. Speakers of Maghrebi Arabic are easily identified as such by other speakers of Arabic when they use the H variety and vice versa. In addition to differences that exist in the target variety based on differences in national educational system objectives and methods, the influence from vernacular Arabic on the H variety is often perceivable at the morphophonological and the syntactic level. On the other hand, influence of the H variety on the L variety oftentimes takes the form of lexical borrowing and minor structural convergence, especially at the phonetic level.
In the case of Tunisian Arabic, stress and vowel quality are among the features that most distinguish Tunisian speakers from other users of Standard Arabic. Word-internally, non-stressed vowels tend to be shortened and even articulated as a schwa (Ghazali et al., 2002). Moroccan speakers of Standard Arabic usually merge the interdental fricatives, /θ/ and /ð/, with /t/ and /d/ respectively. In some cases, deletion of the vowel in the initial syllable leads to a complex consonant cluster that is favored in Maghrebi Arabic in general (Sayahi, 2014, p. 172). At the morphosyntactic level, by way of example, some speakers show increasing use of the vernacular genitive exponent mtɛ:ʕ (instead of the synthetic form for the expression of attributive possession), vernacular question words, and the invariable relative pronoun illi (instead of the variable equivalents in Standard Arabic), but generally any divergence from the target Standard Arabic form is more a case of speaker competence and comfort using the H variety, which does not amount to a permanent influence on the structure of Standard Arabic.
Tunisian Arabic, on the other hand, presents a substantial number of loanwords from Standard Arabic, which in many cases replace earlier borrowed French words, especially in fields related to education and politics. Examples include si:ra ðɛ:tiyya ‘curriculum vitae’ and muwaððif ‘public employee’, among others commonly used by educated speakers in particular. Along with lexical borrowings, and depending on the dialect, some sounds from Standard Arabic may be reinserted (for example use of /ʔ/ and /q/ varies by speaker and context). Although both varieties are in close contact, it is hard to argue for the existence of major changes that could point towards a resolution of the long-standing diglossic situation any time soon.
In sum, the close interaction between both varieties and the increasing frequency of diglossic code-switching lead to a mutual influence without producing a third variety. For a third variety to exist, it has to be stable and possess distinguishable features. Considering code-switching between the H variety and the L variety as a new language is similar to considering code-switching between English and Spanish in the United States as producing a new language, often referred to as Spanglish, which has been shown to be erroneous (Lipski, 2008). This phenomenon is better interpreted as the manifestation of a number of language contact phenomena that include code-switching, lexical borrowing, and structural convergence or even accommodation as opposed to the appearance of a mixed variety. The dialects are still the native languages acquired naturalistically at home, while Standard Arabic is still a sanctioned language that is transmitted via formal education.
8. Critical Analysis of Scholarship
An important characteristic of diglossia in North Africa is its existence alongside bilingualism.11 This situation, where we find a combination of classical diglossia and bilingualism, is what Ferguson (1991) referred to as ‘the bigger picture’ and whose omission he considered to be one of the ‘weaknesses’ of his original outline of what diglossia is. On the other hand, this situation is different from cases where we have two genetically distant languages that are in a functional distribution, which Fishman (1967, 1980) defined as extended diglossia. In North Africa we have a situation of classical diglossia with bilingualism. This situation differs from what Fishman describes, as he was referring to a functional distribution of two separate languages, not two varieties of the same language, that coexist alongside a third (or even a fourth) separate language. Fishman’s (1967) description of a situation of diglossia with bilingualism is in opposition to what he saw as the possibility for bilingualism to exist without diglossia. In his proposal, the latter refers to cases where two separate languages share a similar functional distribution. He argued that: “bilingualism is essentially a characterization of individual linguistic behavior whereas diglossia is a characterization of linguistic organization at the socio-cultural level” (Fishman, 1967, p. 34).
In North Africa the Arabic H variety and L variety coexist with indigenous Berber languages. The situation varies greatly from country to country and from one region or city to another. However, in cases of contact between the two varieties of Arabic with Berber, the one that poses more threat to the maintenance of Berber is the L variety since it competes with Berber in informal domains of communication. Both the Arabic L variety and Berber are oral and at the margin of the educational systems, although efforts to have Berber taught in Morocco and Algeria are starting to bear fruit. On the other hand, Standard Arabic is not in direct competition with Berber and remains separate, and for the reasons listed above is held in the highest regard by all. Language shift among Berber speakers takes place from Berber to vernacular Arabic and not to Standard Arabic. This is another major difference that sets situations of classical diglossia apart from cases of bilingualism (i.e., Fishman’s extended diglossia) where the shift occurs almost always from the L variety to the H variety once the latter becomes nativized.
The Arabic varieties are also in an intense contact situation with French (Figure 2 shows an example of the existence of the three languages in commercial signage). This situation has not led to the nativization of French or the displacement of Arabic, but rather to frequent code-switching by educated North Africans and significant lexical borrowing from French into the Arabic dialects. The interaction between these varieties, in addition to the use of Spanish in Northern Morocco and increasingly English, problematizes the framework of extended diglossia. Likewise, other proposals of alternate classifications of this type of situation, including triglossia (Romaine, 1989; Youssi, 1995), double overlapping diglossia (Mkilifi,1978), and polyglossia (Platt, 1977), are in reality different from what we find in cases of classical diglossia and bilingualism. They fail to call attention to the major role of genetic relatedness and, especially, the nature of the process of acquisition which together condition the type of contact phenomena, the possible convergence between the varieties involved, and, ultimately, the direction of language change and language shift.12 Often in the literature, a situation of contact between two varieties of the same language is equated with that of contact between two separate languages, based only on the broad notion of functional distribution. Extending the concept of diglossia to account for all instances of functional distribution has rendered the term less useful to contact linguistics and limited its potential use as a framework for understanding language change and language genesis.
9. Prospect for the Future
Despite the increasing overlap in domains, Arabic diglossia in North Africa is not going to see a seismic functional rearrangement of the two varieties any time soon. The situation is stable, as originally claimed by Charles Ferguson in his 1959 article, in the sense that the two varieties are clearly distinguishable for speakers and analysts alike. Vernacular Arabic is the native language of Arabic-speaking North Africans and Standard Arabic, on the other hand, is learned through formal instruction and is a variety in which speakers show varying degrees of competence that range from very advanced command to lower levels of competence. This continues to have direct implications for language planning policies, school bilingualism, language ideology, and language change.
At the same time, a qualitative and quantitative change is currently taking place with regard to the functional distribution of both varieties, namely the strong entry of the L variety into public domains, even more remarkably in the written form. In the example of Tunisia, the Revolution of 2010 has allowed for uncensored voices to be heard and less scripting in political discourse and mass media, which in turn has led to the increased encroachment of the vernacular into domains that previously were off limits. Moreover, the spread of the phenomenon of citizen journalism and the appearance of innumerable self-appointed social commentators, whom social media platforms have brought to the forefront of public communication, also contributed to the legitimization of the dialects and their use across domains and media.
Boukous, A. (2008). Globalization and sociolinguistic stratification in North Africa: The case of Morocco. In C. Vigouroux & S. Mufwene (Eds.), Globalization and language vitality: Perspectives from Africa (pp. 126–141). London: Continuum.Find this resource:
Coulmas, F. (2002). Writing is crucial. International Journal of the Sociology of Language, 157, 59–62.Find this resource:
Chakrani, B. (2011). Covert language attitudes: A new outlook on the sociolinguistic space of Morocco. In E. Bokamba, R. Shosted, & B. Ayalew (Eds.), Selected proceedings of the 40th Annual Conference on African Linguistics (Vol. 11, 168–177).Find this resource:
Daoud, M. (2011). The sociolinguistic situation in Tunisia: Language rivalry or accommodation? International Journal of the Sociology of Language, 211, 9–33.Find this resource:
Eckert, P. (1980). Diglossia: Separate and unequal. Linguistics, 18(11–12), 1053–1064.Find this resource:
Eid, M. (1982). The non-randomness of diglossic variation. Glossa, 16(1), 54–84.Find this resource:
Fernández, M. (1993). Diglossia: A comprehensive bibliography, 1960–1990 and supplements. Amsterdam: John Benjamins.Find this resource:
Fishman, J. A. (2002). Diglossia and societal multilingualism: Dimensions of similarity and difference. International Journal of the Sociology of Language, 157, 93–100.Find this resource:
Heath, J. (1989). From code-switching to borrowing: Foreign and diglossic mixing in Moroccan Arabic. London: Kegan Paul International.Find this resource:
Hudson, A. (2002). Outline of a theory of diglossia. International Journal of the Sociology of Language, 157(1), 1–48.Find this resource:
Kaye, A. S. (2001). Diglossia: The state of the art. International Journal of the Sociology of Language, 152, 117–129.Find this resource:
Talmoudi, F. (1984). The diglossic situation in North Africa: A study of classical Arabic/dialectal Arabic diglossia with sample text in “mixed Arabic.” Göteborg, Sweden: Acta Universitatis Gothoburgensis.Find this resource:
Walters, K. (2003). Fergie’s prescience: The changing nature of diglossia in Tunisia. International Journal of the Sociology of Language, 163, 77–109.Find this resource:
Ben Alaya, I. (2016). Interview. Available at https://youtu.be/vQGmiOlk8xM.
Benrabah, M. (2007). The language planning situation in Algeria. In T. Kaplan & R. Baldauf (Eds.), Language planning and policy in Africa (Vol. II, pp. 25–147). Clevedon, U.K.: Multilingual Matters.Find this resource:
Bassiouney, R. (2010). Arabic and the media: Linguistic analyses and applications. Leiden, The Netherlands: Brill.Find this resource:
Boukous, A. (1997). Situation sociolinguistique de l’amazighe. International Journal of the Sociology of Language, 123, 41–60.Find this resource:
Boussofara-Omar, N. (2006). Neither third language nor middle varieties but Arabic diglossic switching. Zeitschrift für Arabische Linguistik, 45, 55–80.Find this resource:
Chaker, S. (2001). Berber challenge in Algeria: The state of the question. Race, Gender and Class, 8(3), 135–156.Find this resource:
Corriente, F. (1975). Marginalia on Arabic diglossia and evidence thereof in the Kitab Al-Agani. Journal of Semitic Studies, 20(1), 38–61.Find this resource:
Eid, M. (2007). Arabic on the media: Hybridity and styles. In E. Ditters & H. Motzki (Eds.), Approaches to Arabic linguistics: Presented to Kees Versteegh on the occasion of his sixtieth birthday (pp. 403–434). Amsterdam: Brill.Find this resource:
Ennaji, M. (2005). Multilingualism, cultural identity, and education in Morocco. New York: Springer.Find this resource:
Gabsi, Z. (2003). An outline of the Shilha (Berber) vernacular of Douiret (Southern Tunisia) (Doctoral dissertation). University of Western Sydney, Sydney, Australia.Find this resource:
Ghazali, S., Hamdi, R., and Barkat, M. (2002). Speech rhythm variation in Arabic dialects. In B. Bel & I. Marlien (Eds.), Proceedings from Speech Prosody 2002 (pp. 331–334). Aix-en-Provence: Laboratoire Parole et Langage. Available at http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.2.6127.Find this resource:
Holes, C. (2004). Modern Arabic: Structures, functions, and varieties. Washington, DC: Georgetown University Press.Find this resource:
Ferguson, C. A. (1959a). Diglossia. Word, 15, 325–340.Find this resource:
Ferguson, C. A. (1959b). The Arabic koine. Language, 35(4), 616–630.Find this resource:
Ferguson, C. A. (1991). Diglossia revisited. Southwest Journal of Linguistics, 10, 214–234.Find this resource:
Fishman, J. A. (1967). Bilingualism with and without diglossia: Diglossia with and without bilingualism. Journal of Social Issues, 23(2), 29–38.Find this resource:
Fishman, J. A. (1980). Bilingualism and biculturism as individual and societal phenomena. Journal of Multilingual and Multicultural Development, 1, 3–15.Find this resource:
Ibn Khaldun, A. (1958). The Muqaddimah: An introduction to history. Trans. Franz Rosenthal. Original Arabic written in 1377. New York: Pantheon.Find this resource:
Krumbacher, K. (1902). Das Problem der neugriechischen Schriftsprache. Munich: K. B. Akademie in Kommission des G. Franz’schen Verlags.Find this resource:
Laroui, A. (1977). The history of the Maghrib: An interpretive essay. Princeton, NJ: Princeton University Press.Find this resource:
Lipski, J. (2008). Varieties of Spanish in the United States. Washington, DC: Georgetown University Press.Find this resource:
Lloyd, P. M. (1987). From Latin to Spanish. Philadelphia: American Philosophical Society.Find this resource:
Maamouri, M. (1998). Language education and human development: Arabic diglossia and its impact on the quality of education in the Arab region. Available at http://www.eric.ed.gov.
Mkilifi, A. M. (1978). Triglossia and Swahili–English bilingualism in Tanzania. In J. Fishman (Ed.), Advances in the study of societal multilingualism (pp. 129–152). The Hague: Mouton.Find this resource:
Owens, J. (2005). Hierarchicalized matrices: Codeswitching among urban Nigerian Arabs. Linguistics, 43, 957–994.Find this resource:
Platt, J. T. (1977). A model for polyglossia and multilingualism (with special reference to Singapore and Malaysia). Language in Society, 6(3), 361–378.Find this resource:
Poplack, S., Sayahi, L., Mourad, N., & Dion, N. (2015). An exception to the rule? Lone French nouns in Tunisian Arabic. University of Pennsylvania Working Papers in Linguistics, 21(2), 177–186.Find this resource:
Posner, Rebecca. (1996). The Romance languages. Cambridge, U.K.: Cambridge University Press.Find this resource:
Psichari, J. (1928). Un pays qui ne veut pas de sa langue. Mercure de France, 207, 63–121.Find this resource:
Romaine, S. (1989). Bilingualism. Oxford: Blackwell.Find this resource:
Sayahi, L. (2014). Diglossia and language contact: Language variation and change in North Africa. Cambridge, U.K.: Cambridge University Press.Find this resource:
Sayahi, L. (2015). A moving target: Literacy development in situations of diglossia and bilingualism. Arab Journal of Applied Linguistics, 1(1), 1–15.Find this resource:
S’hiri, S. (2002). Speak Arabic please! Tunisian Arabic speakers’ linguistic accommodation to Middle Easterners. In A. Rouchdy (Ed.), Language contact and language conflict in Arabic variations on a sociolinguistic theme (pp. 149–174). New York: Curzon.Find this resource:
Youssi, A. (1995). The Moroccan triglossia: Facts and implications. International Journal of the Sociology of Language, 112, 29–44.Find this resource:
Walters, K. (1996). Intrasentential codeswitching in diglossic settings and its implications for linguistic variation and language change. In J. Arnold, R. Blake, B. Davidson, S. Schwenter, & J. Solomon (Eds.), Sociolinguistic variation: Data, theory, and analysis, in selected papers from NWAV23 at Stanford (pp. 401–416). Stanford, CA: CSLI.Find this resource:
Walters, K. (forthcoming). Arab nationalism and/as language ideology. In R. Bassiouney & A. Benmamoun (Eds.), Routledge handbook of Arabic linguistics. Abingdon, U.K.: Routledge.Find this resource:
Wright, R. (1982). Late Latin and Early Romance in Spain and Carolingian France. Liverpool, U.K.: Cairns.Find this resource:
Wright, R. (1994). Early Ibero-Romance: Twenty-one studies on language and texts from the Iberian Peninsula between the Roman Empire and the thirteenth century. Newark, DE: Juan de la Cuesta.Find this resource:
(1.) Marçais (1930, p. 409) wrote: “une sorte d’animal à deux têtes et quelles têtes! Que les programmes scolaires ne savent trop comment traiter, car ils ne sont pas faits pour héberger les monstres.”
(2.) The Muslims founded Kairouan in Tunisia in 670, and by 711 they had already crossed the Gibraltar Strait and entered the Iberian Peninsula.
(3.) My translation from Spanish.
(6.) Article 1 in the 2014 post-revolt Tunisian Constitution states: “Tunisia is a free, independent, sovereign state; its religion is Islam, its language Arabic, and its system is republican. This article [may not] be amended.” (Emphasis added.)
(7.) My translation from Spanish.
(8.) Elsewhere, I refer to the fact that speakers have an unwavering positive attitude towards Standard Arabic but only transmit their native vernacular to their children as the diglossia paradox (Sayahi, 2014, p. 3).
(11.) The spread of English in educational and professional domains in the Middle East is causing Arabic diglossia to coexist with a third language in that part of the world as well.