Version classiqueVersion mobile
OpenEdition Books

Studies on Arabic Dialectology and Sociolinguistics

Catherine Miller
Alexandrine Barontini
Marie-Aimée Germanos
et al.

Issues in Sociolinguistics: Contact and Change, Discourse Analysis, Cultural Practises, Mixed Styles and Written Sources

Mixed Varieties in Political Language in Egypt: the Presidential Debate between ‘Amr Mūsa and ‘Abd al-Min‘im Abu l-Futūḥ

Nina van Kampen


In the diglossic language situation existing in Arabic speaking communities, speakers continuously adapt their language to the context by moving up and down the linguistic continuum between literary Arabic and the various dialects, thus creating mixed varieties of Arabic. This paper deals with the use of mixed language in an Egyptian political debate, broadcasted on national television, between two candidates for the presidential elections of June 2012.

Focusing on the use of three morphosyntactic variables (relatives, demonstratives and negative markers) in the data, it was noted that these features have different degrees of ‘flexibility’: some are more likely to be used in a hybrid context than others. In order to offer an explanation for these differences in flexibility, grammatical and pragmatic methods are combined.

A grammatical analysis of the three variables in the data shows that speakers tend to use the dialectal variants when the syntactic structures for the use of a feature are similar in literary and Cairene Arabic. However, this cannot explain inter- or intraspeaker differences that occur in the data. A pragmatic analysis – looking at the context of enunciation in which the utterances are produced – shows that there seems to be a link between inherent references to the act of enunciation on the one hand, and speakers’ preference for dialectal variants on the other.

Texte intégral

1. Introduction

1Whenever we speak, our goal is to make ourselves understood. We want to transfer a message to our audience, but more importantly, we want this audience to consider what we say as legitimate. When a politician speaks, his goal is to convince his audience of his legitimacy as a leader: he wants support from his people. Thus, he needs to make the right linguistic choices to use a language that is appealing and appropriate.

2The diglossic situation in Arabic speaking communities greatly influences the linguistic choices that speakers continuously make (Ferguson 1959, 1996 [1991]). Whenever they produce an utterance, they situate themselves somewhere on the continuum between literary and dialectal Arabic (Hary 1996), considering the context in which they speak and the symbolic value of both varieties. If a politician uses literary Arabic when speaking to his audience, he could reinforce his image of a cultivated person – a legitimate ruler – but he could also seem too formal or distant. On the other hand, using dialect in political discourse could be perceived as a willingness to use a more ‘democratic’ language, emphasizing closeness to the people. At the same time, the speaker could seem too informal or linguistically unable.

3As a result of these complex symbolic values, politicians often choose to use a mixed language containing features from both literary Arabic and their national dialect (Holes 1993; Mazraani 1997; Bassiouney 2006). The dialectal features in this mixed language should not be considered as errors, but rather as a strategic choice.

4This paper deals with the use of mixed language in an Egyptian political debate broadcasted on national television on the 10th of May 2012. The context of this debate was the Arab Spring: it took place between two candidates for the presidential elections after Mubārak’s resignation. Starting from the idea that mixing between literary and dialectal Arabic does not occur freely but follows grammatical and pragmatic structures, I will analyze the use of mixed Arabic by three speakers of the debate: the two candidates ᶜAmr Mūsa and ᶜAbd al-Minᶜim Abu l-Futūḥ, and the moderator Muna aš-Šazlī.

5The data used here is quite unique in two ways. Firstly, political debates are very scarce in this region where presidents often come into office without a debate or democratic election. This presidential debate was the first one of this kind in Egyptian history, and, as far as I know, the only one in Arabic speaking communities.

6Secondly, most scholars investigating mixed Arabic in political language use speeches as data. However, speeches and debates differ on a number of points. Speeches are often monologues that have been written in advance: they could be considered as oral texts; debates leave more room for improvisation and interaction. The speakers of a debate have to deal with two types of addressees: on the one hand, the viewers of the debate – in this case the Egyptian people –; on the other hand, their fellow speakers on the set with whom they interact directly. This means that political debates emphasize the dialogical dimension of language and can consequently provide us with more information on the unconscious structures of diglossic mixing.

2. Methods

7Mixed varieties in spoken Arabic have been studied using grammatical and pragmatic perspectives. Whereas grammatical methods analyze the morphosyntactic constraints on mixing, the pragmatic methods pay attention to the social motivations and communicative functions of diglossic switching. I will argue that both perspectives pose a number of challenges to analyzing spoken mixed Arabic.

2.1. Pragmatic methods

8As far as the pragmatic analyses are concerned – like Holes (1993) and Mazraani (1997), and partly Mejdell (1999) and Bassiouney (2006) – they focus primarily on intersentential switching. But in actual data of spoken mixed Arabic, phrases are hardly ever entirely in one variety or the other: features from both varieties usually occur in the same phrase or even in the same word. Due to this extensive mixing, it seems almost impossible to define specific places in discourse where switching occurs – the reality of mixed Arabic is often too complex. Another issue encountered when trying to apply a pragmatic method is that the two varieties used in mixed Arabic have many words and morphemes in common. Consequently, a lot of features cannot be classified as either literary or dialectal Arabic.

2.2. Grammatical methods

9The grammatical analyses of spoken mixed Arabic can also pose some challenges. A number of scholars, such as Bassiouney (2006) and Boussofara-Omar (2003, 2006), use the Matrix Language Frame model developed by Carol Myers-Scotton (1993). This model considers roughly that, when speakers switch between two varieties, one variety forms the morphosyntactic frame constituted by system morphemes, whereas the other variety is the ‘embedded language’ constituted by content morphemes.

10Apart from the issue discussed above – namely, that literary and dialectal Arabic share a large amount of features – the main problem with this model is that it considers all morphosyntactic features to be on the same level. However, the reality of the data shows that some features are more likely to be found in a hybrid context than others, as has also been pointed out by Mejdell (2006). I will argue that some morphosyntactic features are more flexible than others. By ‘flexible’, I mean that a feature is often used in a hybrid context with morphosyntactic elements that are not from the same variety as the feature. On a quantitative level, this ‘flexibility’ of a feature means that it is used rather frequently.

2.3. Analyzing flexibility: combining a grammatical and pragmatic method

11In order to take into consideration the differences in the flexibility of morphosyntactic features, it is clear that we need a more nuanced model than Meyers-Scotton’s often-used Matrix Language Frame model. Gunvor Mejdell (2006) develops a very useful and interesting method, offering a quantitative and qualitative analysis of five morphosyntactic variables in several monologues of Egyptian academic discourse. Her analysis shows that some variables are more marked than others.

12The grammatical analysis of my data follows Mejdell’s method. I will focus on three morphosyntactic variables: relatives, demonstratives and negative markers. These variables are analyzed on a quantitative and a qualitative level. Which variants can be defined as flexible?

13However, focusing exclusively on the frequency of the literary and Cairene Arabic variants and their syntactic context cannot explain why there are differences between the three speakers, or, for that matter, why one single speaker alternates between different – literary and Cairene Arabic – variants. In order to provide an answer to these questions, we need to take into consideration the context of enunciation in which the utterances are produced. The results of the in-depth grammatical analysis of the variables will serve as a starting point to examine this context of enunciation.

14The purpose of this paper is not only to examine the flexibility of demonstratives, relatives and negative markers as used in my data, but also to offer an explanation for the differences in flexibility, combining grammatical and pragmatic methods.

3. Results

3.1. The grammatical analysis

15The quantitative analysis of the data shows that out of the three variables, the relatives have on average the largest proportion of Cairene Arabic variants (56%). Speakers tend to prefer the Cairene Arabic variant illi to the numerous literary Arabic variants. On the contrary, the negative markers have the smallest proportion of Cairene Arabic variants (31%). An overview can be seen in table 1.

Table 1: Quantitative results: total use of the three morphosyntactic variables

Table 1: Quantitative results: total use of the three morphosyntactic variables

16These quantitative results clearly show that the different variants do not have the same degree of flexibility and lead us to think that Cairene Arabic relatives are more flexible than the negative markers of the same variety. But how can we explain this difference in flexibility from a grammatical perspective? In order to answer this question, we will need to look at the syntactic structures that mark the use of these variables in literary and Cairene Arabic.


17As far as the relatives are concerned, one should note that the syntactic structures used in literary and Cairene Arabic are identical (Woidich 2006: 199-200; Badawi e.a. 2016: 549):

18The speakers in the debate tend to prefer using the dialectal variant of the relative rather than the literary Arabic variants. illi is very often used in hybrid contexts, as in (1):

19In (1), illi is surrounded by literary Arabic features, such as the verbal morphology of tastagīb (which, however, lacks verbal inflection), and the internal passive and aspecto-temporal values of ḏ̣ulimu and hummišu.

20The dialectal relative is also frequently followed by more hybrid verbal predicates like b-tuṣraf, containing both the dialectal aspectual preverb b- and the literary Arabic internal passive, but lacking verbal inflection. Finally, illi can also be followed by completely Cairene Arabic verbal predicates as ha-yikkallim.

21These examples clearly show the flexibility of the Cairene Arabic relative and how it can be used in literary Arabic, Cairene Arabic or hybrid contexts. The literary Arabic relatives, on the other hand, are less flexible: they are usually followed by a verbal predicate following literary Arabic rules.


22As far as the demonstratives are concerned, one should note that they can have two functions. The first function is that of a determiner of a definite noun. For this function, Cairene Arabic and literary Arabic use different structures (Woidich 2006: 44-45; Badawi e.a. 2016: 149-150):

Cairene Arabic

noun determined by def art


demonstrative determiner



Literary Arabic

demonstrative determiner


noun determined by def art



23The second function of the demonstrative is that of a pronoun. It often has the role of an anaphoric subject, referring to a previous fragment of discourse or to an idea of the speaker. The syntactic structures in Cairene and literary Arabic are identical (Woidich 2006: 44; Badawi e.a. 2016: 151):

Cairene Arabic






Literary Arabic






24When the demonstrative is the determiner of a definite noun, the speakers of the debate prefer using literary Arabic variants: 90% of the demonstrative determiners in the debate are literary Arabic variants.

25The data shows that these literary Arabic variants of demonstratives determining definite nouns are flexible: they can be used in hybrid or dialectal clause level contexts.

26In the above example, we can clearly define several elements as dialectal Arabic, such as the negative marker ma-…-š and the use of the active participle. However, it is more difficult to define the variety of the demonstrative determiner’s local context. As has been noted above, the two varieties share a large number of words which cannot be classified as either literary or dialectal Arabic, such as suʾāl in (2). I have not found any occurrence in the data of a literary Arabic demonstrative determining a noun that could be classified as exclusively Cairene Arabic.

27As for the dialectal variant of the demonstrative determiner, it is not very flexible. It usually needs a dialectal syntactic context, as we can see in the following example:

28As far as the demonstrative pronouns are concerned, one can note that the patterns of use are quite different. The speakers of the debate tend to use dialectal variants for this function: 58% of the demonstrative pronouns are Cairene Arabic variants. These dialectal variants are quite flexible and can be found in hybrid contexts with features from literary Arabic, as in the following example. The same trend has been noted by Mejdell (2006: 229).

Negative markers

29It was noted before that the data contained little dialectal variants of the negative markers, compared to the other morphosyntactic variables. Generally speaking, negative markers of both varieties are not flexible. Morphosyntactic features of the same linguistic variety usually surround them, and there is no switching between the negative marker and the (verbal) predicate. Thus, the data contains, for example: lam ʾakun ‘I wasn’t’, mā waqafna ‘we didn’t stand’, lan yakūna ‘He will not be’, lastu šayxan ‘I am not a religious scholar’, miš cārif ‘I don’t know’ or ma-tiʾdirī-š tiḥlibi baʾara wa-hiyya ma-candahā-š laban ‘you cannot milk a cow that doesn’t have milk’.

30Why are the negative markers not flexible? An explanation can be found in the fact that literary and Cairene Arabic both have different ways of expressing tense, aspect and modality, as has been discussed by Eid (1988). In literary Arabic, these notions are incorporated in the negative markers that can assign modal, temporal or aspectual values in the form of verbal inflections. Adversely, in Cairene Arabic it is the verb that carries the notions of tense, aspect and modality by the presence (or absence) of preverbs. The two systems are not compatible, so switching between them is problematic.

31The only negative marker that proves to be quite flexible in my data is the literary Arabic marker . Used in the lexicalized expression lā budd(a) ‘it is necessary’, or with a following noun expressing the categorical negation, this negative marker is very frequently found in a hybrid syntactic context containing other dialectal features. When functions to negate a verb, this following verb usually follows the literary Arabic way of expressing tense, aspect and mood. However, the verbal predicate can also have a hybrid verbal morphology, as tataṣawwari in (5). This example also shows how can be used alongside many other dialectal morphosyntactic features:

32Why is more flexible than other negative markers? As has been pointed out by Ayoub (1996: 1017-1018), this negative marker does not inherently carry aspecto-temporal or modal value and does not assign an inflection to the verb – unlike, for example, lam and lan. This means that can be used in a more hybrid context. In my data, it seems to be used by speakers as a relatively simple tool to ‘elevate’ their language, as it does not require them to make any fundamental changes in the syntactic structure of the phrase.

33It is clear that the three variables have different degrees of flexibility. The relevant question, now, is why some dialectal variants prove to be more suitable for a hybrid context than others. The ‘compatibility’ between the grammatical systems of literary and Cairene Arabic seems to play an important role in this. When the syntactic structures in Cairene and literary Arabic are similar, the speakers tend to use more dialectal variants and use them even in hybrid contexts. This was the case for the relatives and the demonstratives having a pronominal function: the syntactic structure in both varieties is similar, leading speakers to use Cairene Arabic variants. When the syntactic structures are different, however, speakers tend to use the literary Arabic variants, as was shown by the preference of speakers to use literary Arabic negative markers and demonstratives that determine a definite noun. However, it is not only the (lack of) similarity of the syntactic structures that influences patterns of use, but also the compatibility of variants with their local linguistic environment. Both negative markers and demonstrative determiners have different structures in literary and Cairene Arabic, pushing speakers to use more literary Arabic variants. As has been discussed above, not all negative markers are compatible with any local environment since some of them carry modal, temporal or aspectual value. Demonstrative determiners, however, are compatible with a great number of nouns given the large shared lexicon between both varieties. This can explain why literary Arabic variants represent such a large proportion of the demonstrative determiners in my data.

3.2. The pragmatic analysis

34The abovementioned results largely follow the conclusions of Mejdell and show how a grammatical method can explain the differences in flexibility. However, they fail to explain intra- and interspeaker differences. Table 2 shows these differences:

Table 2: Interspeaker differences

Table 2: Interspeaker differences

35For example, why does Muna aš-Šazlī use a much larger proportion of literary Arabic relatives than the other speakers? We need to consider the context of enunciation in order to provide an answer. The following sections will discuss a number of pragmatic factors influencing code choice in this debate.

The spontaneity of discourse

36Almost all occurrences of literary Arabic relatives used by the moderator are found in the questions that she addresses to the candidates, as in (6). These fragments have been prepared in advance, and often when watching the video we can even see that she reads a written text out loud.

37When the moderator uses the dialectal variant illi, however, it is often when something unexpected happens on the set, as we can see in (7):

The audience

38One can consider that debates have two types of audience: the people watching the debate and the conversation partners on the set. When the speakers of this debate directly address their conversation partners on the set, they tend to globally use more dialectal morphosyntactic features, including negative markers. For example:

39This also means that when the type of discourse is a dialogue rather than a monologue, with speakers alternating turns frequently, more dialectal features will be used. Adversely, when the speakers address the audience rather than their conversation partners on the set, talking for a long time without interruptions, they hardly use any dialectal negative markers.

The topic

40It was noted that when the topic of the utterance is the debate itself, speakers tend to use more dialectal features, including ones that are normally not used frequently, like the dialectal demonstrative determining a definite noun in (9) and the negative marker in (10).

41As far as these pragmatic factors are concerned, one can note one main principle. When the speaker talks spontaneously, when he interacts with his conversation partners directly, when speaker turns alternate quickly, and lastly, when the topic of his utterance is the debate or some technical aspect of it, the speaker inherently refers to the fact that he is on a set, participating in a debate with others, that he is producing utterances, and trying to make himself understood. This pragmatic analysis clearly shows that when the speakers inherently refer to their act of enunciation, they will on a general level use more dialectal features, including ones that were proven to be infrequent in our quantitative analysis.

42Let us get back to the three morphosyntactic variables analyzed before. Could we say that speakers tend to use the dialectal variants of some of these variables, because they inherently refer to the act of speaking? It was shown that speakers tend to use the dialectal variants of the relative and of the demonstrative functioning as an anaphoric subject. These features have anaphoric value: they refer to a previous fragment in discourse. But as this fragment is also product of the enunciation, we could say that it indirectly refers to this act.

43Another dialectal morphosyntactic feature worth mentioning is the Cairene Arabic aspectual preverb bi-. It puts the verb in continuity with the moment of enunciation, which means that it also inherently refers to this act. My data, as well as the analyses of Bassiouney (2006) and Mejdell (2006), show that this feature is particularly flexible and that it is frequently used with a verb that follows literary Arabic verbal morphology. An example is shown in (11).

44There seems to be a link between the inherent reference of a feature to the act of enunciation, and speakers’ preference for dialectal variants of this feature. This possible link should be examined in more detail and for a larger amount of features, especially since previous models seem insufficient in explaining how mixed spoken Arabic works. This research shows some real indications that a further examination of this link could be of great value.

45However, we cannot argue that every single time that a speaker inherently refers to his act of speaking, he will use dialectal features. For example, the demonstrative determining a definite noun also has deictic or anaphoric value, but as was discussed before, the speakers of the debate prefer using its literary Arabic variants.

46To conclude, I would like to argue that in the analysis of data of spoken mixed Arabic, we need to consider both types of ‘flexibility criteria’: the pragmatic factors involving the reference a speaker makes to his act of speaking, and the grammatical factors linked to the compatibility between literary and dialectal Arabic syntax. The question remains to be explored as to whether these flexibility criteria have some kind of universal validity in language or variety contact situations. Should we move towards a general model where the compatibility of language systems and the inherent reference to the enunciation are taken into account? Undoubtedly, further studies and a greater variety of data are needed to determine this.


Ayoub, G. 1996. La question de la phrase nominale en arabe littéraire : prédicats, figures, catégories (Thèse de doctorat d’État). Villeneuve d’Ascq: Presses Universitaires du Septentrion.

Badawi, E., Carter, M. & Gully, A. 2016 [2004]. Modern Written Arabic: A Comprehensive Grammar. London, New York: Routledge.

Bassiouney, R. 2006. Functions of Code Switching in Egypt: Evidence from Monologues. Leiden: Brill.

Boussofara-Omar, N. 2003. “Revisiting Arabic Diglossic Switching in Light of the MLF Model and its Sub-Models: the 4-M Model and the Abstract Level Model”, Bilingualism: Language and Cognition 6. 33-46.

Boussofara-Omar, N. 2006. “Neither Third Language nor Middle Varieties but Diglossic Switching”, Zeitschrift für Arabische Linguistik 45. 55-80.

Eid, M. 1988. “Principles of Code Switching between Standard and Egyptian Arabic”, Al-cArabiyya 21. 51-79.

Ferguson, C. 1959. “Diglossia”, Word 15. 325-340.

Ferguson, C. 1996 [1991]. “Diglossia Revisited”, Elgibali, A. (ed.), Understanding Arabic: Essays in Contemporary Arabic Linguistics in Honor of El-Said Badawi. Cairo: American University in Cairo Press. 49-67.

Hary, B. 1996. “The Importance of the Language Continuum in Arabic Multiglossia”, Elgibali, A. (ed.), Understanding Arabic: Essays in Contemporary Arabic Linguistics in Honor of El-Said Badawi. Cairo: American University in Cairo Press. 69-90.

Holes, C. 1993. “The Uses of Variation: A Study of the Political Speeches of Gamāl cAbd al-Nāṣir”, Eid, M. & Holes, C. (eds.), Perspectives on Arabic Linguistics 5. Amsterdam: John Benjamins. 13-45.

Mazraani, N. 1997. Aspects of Language Variation in Arabic Political Speech-Making. Richmond: Curzon.

Mejdell, G. 1999. “Switching, Mixing: Code Interaction in Spoken Arabic”, Brendemoen, B., Lanza, E., & Ryen, E. (ed.), Language Encounters Across Time and Space. Oslo: Novus Forlag. 225-241.

Mejdell, G. 2006. Mixed Styles in Spoken Arabic in Egypt: Somewhere between Order and Chaos. Leiden: Brill.

Myers-Scotton, C. 1993. Social Motivations for Codeswitching: Evidence from Africa. Oxford: Oxford University Press.

Woidich, M. 2006. Das Kairenisch-Arabische: Eine Grammatik. Wiesbaden: Harrassowitz Verlag.






































first/second/third person






© Institut de recherches et d’études sur les mondes arabes et musulmans, 2019

Conditions d’utilisation :


Open access