Diglossia: A language ideological approach

Helge Daniëls

Abstract

Diglossia is, as far as the Arabic language is concerned, a concept that has been taken for granted, as much as it has been criticized. First, based on Ferguson’s article on diglossia and subsequent interpretations and ramifications of the concept and with a special focus on how language variability is discursively deployed and how it is perceived in the Arab speech community, I will argue that diglossia does not so much describe actual language use, but rather how language variability is ‘read’ in the Arab world. In the second part of the article, an analysis of labeling in a 19th century debate will show how the dichotomy between fuṣḥā and non-fuṣḥā varieties (ʿāmmīya),11.The transliteration of Arabic terms follows Hans Wehr’s system. which is the basis of diglossia, was already taken for granted long before the concept and the term existed, and even before fuṣḥā and ʿāmmīya were used as independent lexical items. The analysis in both parts of the article shows how much diglossia is taken for granted by most native speakers of Arabic, even if it defies linguistic descriptions of actual language use. It is exactly this ‘common-sense-ness’ that suggests that diglossia is a useful tool to describe language ideological attitudes.

Keywords:
Table of contents

1.Introduction

Many linguistic studies of Arabic open with an introduction to the ‘Arabic language situation’ in which reference is made to the diglossic nature of Arabic. It is Ferguson’s article "Diglossia" (1959Ferguson, Charles A. 1959 “Diglossia.” Word 15: 325–40. CrossrefGoogle Scholar) that set the stage for this approach and for students and scholars of Arabic, using it has become almost a rite de passage.22.Before the term "la diglossie arabe" was used by Marçais (1930)Marçais, William 1930 “La diglossie arabe.” L’ensiegnement Public 97: 401–9.Google Scholar and “al-izdiwāğīya” by Frayḥa (1938)Frayḥa, Anīs 1938 “Al-ʻāmmīya wa al-fuṣḥā. ʻawd ilā al-mawḍūʻ [ʻāmmīya and fuṣḥā. Back tot he subject].” Al-muqtaṭaf 93: 292–8.Google Scholar. Even if many aspects of the article received a lot of criticism, diglossia is still often taken for granted when Arabic is concerned. Moreover, the criticisms hardly ever approach the concept as a language ideological one, namely one that is informed, shaped and constructed by linguistic attitudes that were and are prevalent in the Arabic speech community.

If we follow Joseph’s claim (2004Joseph, Earl John 2004Language and Identity. National, Ethnic, Religious. Hampshire and New York: Palgrave Macmillan.Google Scholar, 30–40) that linguistic analysis should not exclusively focus on language production, but that interpretation is at least as important as (if not more than) representation and communication, then diglossia can be a useful concept not to describe the productive aspects of language use (for which it has proven to be invalid), but rather its interpretative aspects. In other words, diglossia can be used to describe the ways in which linguistic variability is ‘read’ in the Arabic linguistic community. This way of reading in itself, however, influences linguistic norms and actual language use, be it in highly complex ways. This is probably the reason why diglossia is so easily misunderstood as a model for actual language use.

An analysis of labeling can offer fruitful insights into how linguistic variability is interpreted, as labeling linguistic varieties is never an exercise in terminology alone and not a matter of merely ‘representing a linguistic reality’. Labeling also implies categorization, drawing boundaries, chopping up the essentially continuous reality of linguistic variation into discontinuous blocks, into 'categories of communication', such as 'language', 'dialect', 'standard' etc. (Gal & Woolard 1995Gal, Susan, and Kathryn Woolard 1995 “Constructing Languages and Publics: Authority and Representation.” Pragmatics 5 (2): 129–138. CrossrefGoogle Scholar, 129) Thus, labeling is not merely giving a name to ‘existing varieties and languages’ but rather involves the construction of linguistic varieties and languages and the ways in which they relate to each other. However, we must admit that if and when the linguistic labels are accepted and become common use, they actually do influence and shape linguistic reality and language use. (see also Joseph 2004Joseph, Earl John 2004Language and Identity. National, Ethnic, Religious. Hampshire and New York: Palgrave Macmillan.Google Scholar)

In the present article, the process of labeling will be analyzed by focusing on a sub-debate of the highly polemical debates concerning ʻāmmīya and fuṣḥā that were conducted since the end of the 19th century. A close reading of the 19th century sub-debate can offer us insight into how variability in Arabic was perceived by 19th-century intellectuals. Even if intellectuals were only a small minority, the debates offer us an entry into the linguistic attitudes of at least one segment of the Arabic-speaking population at the time. The debates are interesting not only because of the explicit and implicit argumentation patterns of those in favor or against the standardization of other varieties than fuṣḥā, but also because a close analysis of the processes of labeling different varieties can offer us deeper insight into how diglossia was constructed as a language ideological concept, even if the term (al-izdiwāğīya in Arabic) as such was not yet used to describe Arabic.33.Basing herself on the work of Mackey (1993)Mackey, W. F. 1993 “Introduction.” In Diglossia: A Comprehensive Bibliography 1960-1990, ed. by M. Fernandez, xiii–xx. Amsterdam: John Benjamins Publishing Company.Google Scholar, Haeri (2000 2000 “Form and Ideology: Arabic Sociolinguistics and Beyond.” Annual Review of Anthropology 29: 61–87. CrossrefGoogle Scholar, 64–5) claims that the term diglossia was first used to describe the Greek language situation in the 1880s. It is interesting to note that this coincides with the early beginnings of the fuṣḥā–ʻāmmīya debate in Al-Muqtaṭaf (1881–1882Ṣarrūf, Yaʿqūb and Fāris Nimr 1881 “Al-luġa al-ʿarabīya wa al-nağāḥ [The Arabic language and success].” Al-Muqtaṭaf 6 (November): 352–4.Google Scholar).

A close reading of the linguistic labels used in debates demonstrates that, despite the variability that is covered by the term ʻāmmīya, the term is used as a collective term for all varieties other than fuṣḥā and as such the wide variety of linguistic characteristics that is covered by the term is overruled by their shared non-fuṣḥā-ness. Thus, the debaters construct a diglossic division,44.Or maybe rather re-construct and reconfirm it, as the early Arab grammarians and philologists are said to have been dealing with this topic from as early the ninth century. (El-Hassan 1977El-Hassan, Shahir A. 1977 “Educated Spoken Arabic in Egypt and the Levant: A critical Review of Diglossia and Related Concepts.” Archivum Linguisticum 8 (2): 112–132.Google Scholar, 113) long before the term was actually coined and started to be used to describe Arabic or the ‘Arabic language situation’. Moreover, the debaters create a contrast that is not symmetrical, as one side (fuṣḥā) is used as the standard of the comparison by which the other side (ʻāmmīya) is measured.

I will first start with a discussion of the most fundamental points of criticism on Ferguson’s article and the ways in which it was interpreted in subsequent research, as well as some of the remedies that have been presented. Subsequently, I will proceed with a close analysis of labeling in a selection of 19th century sub-debates of the fuṣḥā-ʻāmmīya debate.

2.Diglossia: A critical approach

2.1A strict functional division between fuṣḥā and ʻāmmīya? The language continuum, intermediate varieties and code-switching

One of the main points of criticism that were formulated against Ferguson’s argument was that the actual language use of native speakers challenges the strict functional division that Ferguson suggested in his article:

One of the most important features of diglossia is the specialization of function for H [fuṣḥā] and L [non- fuṣḥā]. In one set of situations only H is appropriate and in another only L, with the two sets overlapping only slightly.(Ferguson 1959Ferguson, Charles A. 1959 “Diglossia.” Word 15: 325–40. CrossrefGoogle Scholar, 328, emphasis mine,55.Note the tension between "the specialization of function" which suggests a strict functional distribution of H and L (and which has been taken for granted in much of the literature on diglossia) on the one hand, and the reference to the settings in which H and L are "appropriate" on the other hand. The latter refers to the communicative norms related to the settings rather than to the actual language use in these settings. Moreover, the suggestion of a functional division of language use is confirmed by the examples that Ferguson gives. see also discussion below)

H would then be used for a sermon in church or mosque; a personal letter; a speech in parliament; a political speech; a university lecture; a news broadcast; a newspaper editorial; a news story or the caption on a picture and poetry. L would be used for instructions to servants, waiters, workmen, clerks; a conversation with family members, friends, colleagues; radio soap opera; the caption on a political cartoon and folk literature (Ferguson 1959Ferguson, Charles A. 1959 “Diglossia.” Word 15: 325–40. CrossrefGoogle Scholar, 329). Analyses of fragments of a sermon in a mosque in Upper Egypt (El-Hassan 1977El-Hassan, Shahir A. 1977 “Educated Spoken Arabic in Egypt and the Levant: A critical Review of Diglossia and Related Concepts.” Archivum Linguisticum 8 (2): 112–132.Google Scholar), a university lecture (El-Hassan 1977El-Hassan, Shahir A. 1977 “Educated Spoken Arabic in Egypt and the Levant: A critical Review of Diglossia and Related Concepts.” Archivum Linguisticum 8 (2): 112–132.Google Scholar), speeches in parliament (El-Hassan 1977El-Hassan, Shahir A. 1977 “Educated Spoken Arabic in Egypt and the Levant: A critical Review of Diglossia and Related Concepts.” Archivum Linguisticum 8 (2): 112–132.Google Scholar), political speeches (Holes 1993 1993 “The uses of Variation: A Study of the Political Speeches of Gamal Abd Al-Nasir.” Perspectives on Arabic Linguistics V, ed. by M. Eid, and C. Holes, 13–46. Amsterdam: John Benjamins Publishing Company. CrossrefGoogle Scholar; Mazraani 1997Mazraani, Nathalie 1997Aspects of Language Variation in Arabic Political Speech Making. Richmond Surrey: Curzon Press.Google Scholar) challenge this claim. Counterexamples can also be given for personal letters, news broadcasts,66.For example, the South-Lebanese radio Ṣawt al-ğanūb broadcasted news bulletins in both fuṣḥā and Lebanese Arabic. etc. On the other hand, in conversations with family members, friends, etc., captions on political cartoons, or radio soap operas L is not invariably used either. Many more examples can be given. Therefore, "the language situations Ferguson has cited as the exclusive domains of each variety are not as hermetically separated as he had thought" (Mahmoud 1986Mahmoud, Youssef 1986 “Arabic after Diglossia.” In The Fergusonian Impact: In Honour of Charles A. Ferguson on the Occasion of his 65th Birthday, ed. by Joshua Fishman, Andree Tabouret-Keller, Michael Clyne, Bh. Krishnamurti, and Mohamed Abdulaziz, 239–51. Berlin: Mouton de Gruyter.Google Scholar, 239, op cit. Boussofara-Omar 2011Boussofara-Omar, Naima 2011 “Diglossia.” Encyclopedia of Arabic Language and Linguistics. Managing Editors Online Edition: Lutz Edzard; Rudolf de Jong. Brill 2011 Brill Online. K.U. Leuven – University Library. 05 April 2011 http://​www​.brillonline​.nl​/subscriber​/entry​?entry​=eall​_COM​-0090). Just like speakers of other languages, arabophones, more often than not, do not strictly adhere to one language/variety or another within a certain context, but consciously and unconsciously code-switch and style-shift depending on highly variable factors. As in other speech communities, linguistic interaction is a highly dynamic process defying essentialist views of a one-to-one relationship between language and context.

2.1.1The linguistic continuum and intermediate varieties

The idea that H and L are actually functionally not as separate as is suggested in Ferguson’s article has led to the assertion that there is a continuum of which H and L are the poles.77.Ferguson himself was much aware of this: “A kind of spoken Arabic much used in certain semi-formal or cross-dialectal situations has a highly classical vocabulary with few or no inflectional endings, with certain features of classical syntax, but with a fundamentally colloquial base in morphology and syntax, and a generous admixture of colloquial vocabulary.” (Ferguson 1959Ferguson, Charles A. 1959 “Diglossia.” Word 15: 325–40. CrossrefGoogle Scholar, 332) Subsequently several attempts have been made to divide the continuum in three (triglossia), four (quadriglossia), five or more intermediate varieties. Blanc (1960)Blanc, Haim 1960 “Style Variations in Spoken Arabic – A Sample of Interdialectal Educated Conversation.” In Contributions to Arabic Linguistics. Harvard Middle Eastern Monographs III, ed. by C. A. Ferguson, 81–156. Cambridge Massassuchets: Harvard University Press.Google Scholar makes a distinction between five intermediate varieties: plain colloquial, koineized colloquial, semi-literary or elevated colloquial, modified classical and standard classical (Blanc 1960Blanc, Haim 1960 “Style Variations in Spoken Arabic – A Sample of Interdialectal Educated Conversation.” In Contributions to Arabic Linguistics. Harvard Middle Eastern Monographs III, ed. by C. A. Ferguson, 81–156. Cambridge Massassuchets: Harvard University Press.Google Scholar, 85). Badawi (1973)Badawi, al-Saʻīd 1973Mustawayāt al-ʻarabīya al-muʻāṣira fi miṣr. Baḥt fī ʻalāqat al-luġa bi-al-haḍāra [The Levels of Contemporary Arabic in Egypt. Research on the Relation between Language and Culture]. Caïro: Dār al-Maʻārif.Google Scholar also distinguishes between five levels for Arabic in Egypt: fuṣḥā al-turāt (fuṣḥā of the literary heritage), fuṣḥā al-ʻāṣr (contemporary fuṣḥā), ʻāmmīyat al-mutaqqafīn (the colloquial of intellectuals), ʻāmmīyat al-mutanawwirīn (the colloquial of educated people) and ʻāmmīyat al-ummīyīn (the colloquial of illiterates) (Badawi 1973Badawi, al-Saʻīd 1973Mustawayāt al-ʻarabīya al-muʻāṣira fi miṣr. Baḥt fī ʻalāqat al-luġa bi-al-haḍāra [The Levels of Contemporary Arabic in Egypt. Research on the Relation between Language and Culture]. Caïro: Dār al-Maʻārif.Google Scholar, 89–92). In these labels and their description context, modus (written versus oral) and class are blurred.88.See also Bassiouney (2009Bassiouney, Reem 2009Arabic Sociolinguistics. Edinburgh: Edinburgh University Press. CrossrefGoogle Scholar, 14–16). Finally, Meiseles (1980)Meiseles, Gustav 1980 “Educated Spoken Language and the Arabic Language.” Archivum Linguisticum XI, New Series 118–148.Google Scholar divides the continuum into: literary Arabic or standard Arabic, oral literary Arabic or Sub-standard Arabic, Educated Spoken Arabic and plain vernacular.

However useful the idea of the continuum is (see later), this approach shows various flaws. First of all, all attempts to define intermediate varieties have led to infinite and void discussions about how many varieties could be distinguished and on what basis this could be done. Not only are intermediate varieties vaguely defined, the continuum can be endlessly further divided into yet more varieties. In other words, "the result was the emergence of a constellation of labels to categorize a tentative taxonomy of ‘ill-defined’ middle varieties of Arabic, and hence, a failure to articulate their description in a coherent manner or to relate these sets of practices to a theoretical linguistic model that can account for them" (Boussofara-Omar 2011Boussofara-Omar, Naima 2011 “Diglossia.” Encyclopedia of Arabic Language and Linguistics. Managing Editors Online Edition: Lutz Edzard; Rudolf de Jong. Brill 2011 Brill Online. K.U. Leuven – University Library. 05 April 2011 http://​www​.brillonline​.nl​/subscriber​/entry​?entry​=eall​_COM​-0090). Moreover, Meiseles (1980Meiseles, Gustav 1980 “Educated Spoken Language and the Arabic Language.” Archivum Linguisticum XI, New Series 118–148.Google Scholar, 121) himself concluded that “outlining borders to the different Arabic varieties is not only a very difficult task, but one that seems, prima facie, unnecessary and superfluous.”

Furthermore, these attempts to divide the continuum between shades of so-called "colloquialized fuṣḥā" and "standardized ʻāmmīya" are too one-dimensional, as they, at least implicitly, suggest that the mixing occurs between fuṣḥā and only one non-fuṣḥā variety. But in reality mixing between Arabic and other languages on the one hand, and different non-fuṣḥā varieties of Arabic, on the other, also occur. The definition of Educated Spoken Arabic was an attempt to counter this criticism.

2.1.2Educated Spoken Arabic (ESA)

Mitchell and his colleagues who worked on the Leeds corpus (Mitchell 1978Mitchell, Terence Frederic 1978 “Educated Spoken Arabic in Egypt and the Levant, with Special Reference to Participle and Tense.” Journal of Linguistics 14/2: 227–58. CrossrefGoogle Scholar, 1980 1980 “Dimensions of Style in a Grammar of Educated Spoken Arabic.” Archivum Linguisticum XI: 89–106.Google Scholar, 1982 1982 “More than a Matter of ‘Writing with the Learned, Pronouncing with the Vulgar.” Standard Languages, ed. by W. Haas, 123–55. Manchester: Manchester University Press.Google Scholar, 1986 1986What is Educated Spoken Arabic?” In Aspects of Arabic Sociolinguistics, ed. by M. H. Ibrahim, and B. H. Jernudd, special volume of The International Journal of the Sociology of Language 61: 7–32.Google Scholar, 1994; Mitchell & al-Hassan 1984; El-Hassan 1977El-Hassan, Shahir A. 1977 “Educated Spoken Arabic in Egypt and the Levant: A critical Review of Diglossia and Related Concepts.” Archivum Linguisticum 8 (2): 112–132.Google Scholar, 1978; Sallam 1979Sallam, A. M. 1979 “Concordial Relation within the Noun Phrase in Educated Spoken Arabic (ESA).” Archivum Linguisticum (new series) X (1): 20–56.Google Scholar, 1980 1980 “Phonological Variation in Educated Spoken Arabic: A Study of the Uvular and Related Plosive Types.” Bulletin of the School of Oriental and African Studies, XLIII 43: 1.Google Scholar) argue that the so-called vernacular or colloquial is never pure or unmixed, but always shows some degree of influence by fuṣḥā. They call this variety Educated Spoken Arabic and argue that it is not a static variety but, on the contrary, a highly dynamic one that is defined by the constant interplay between fuṣḥā and the spoken varieties. Moreover, by focusing on cross-dialectal conversations in interactions between speakers of different parts of the Arab world the Leeds corpus also accounts for switches between fuṣḥā and several non- fuṣḥā varieties (not only one) with special attention to these linguistic elements that all non- fuṣḥā varieties have in common. They attempt to describe the general rules of ESA that is the result of this interplay by means of variation grammar.

One may wonder, however, whether it is useful to consider ESA in the way it is defined by Mitchell and his colleagues as yet another variety of Arabic. First of all, more research on cross-dialectal communication is needed. Otherwise, ESA risks remaining as ill-defined as any one of the other intermediate varieties that have been described above. Or as Parkinson states: “Everyone claims to believe that Educated Spoken Arabic is rule-governed, but none seems to be able to come up with the rules. Part of the reason for this may be, of course, that Educated Spoken Arabic may not actually be anything” (Parkinson 2003, 29, op cit. Boussofara-Omar 2011Boussofara-Omar, Naima 2011 “Diglossia.” Encyclopedia of Arabic Language and Linguistics. Managing Editors Online Edition: Lutz Edzard; Rudolf de Jong. Brill 2011 Brill Online. K.U. Leuven – University Library. 05 April 2011 http://​www​.brillonline​.nl​/subscriber​/entry​?entry​=eall​_COM​-0090, emphasis original). Furthermore, describing language use in terms of code-switching allows us better to stress the highly dynamic character of language use, while describing it as a variety implicitly suggests some static stability.

2.1.3Code-switching

Probably, a description in terms of code-switching is then the most suitable and subtle approach to describe language use among arabophones, since it can interpret inter- and intra segmental switches (e.g. switches between and within sentences or words) on all linguistic levels between Arabic and other languages and between different varieties of fuṣḥā and non-fuṣḥā Arabic in one stretch of discourse. Nevertheless, we must add that even the more nuanced approaches that are critical towards the diglossic approach still depart from a basic dichotomy between fuṣḥā and ʻāmmīya that is taken for granted, hence the term "diglossic switching" (Boussophara 2011Boussofara-Omar, Naima 2011 “Diglossia.” Encyclopedia of Arabic Language and Linguistics. Managing Editors Online Edition: Lutz Edzard; Rudolf de Jong. Brill 2011 Brill Online. K.U. Leuven – University Library. 05 April 2011 http://​www​.brillonline​.nl​/subscriber​/entry​?entry​=eall​_COM​-0090) and consistent reference to the "two varieties" of Arabic, e.i. fuṣḥā and ʻāmmīya (or written and spoken Arabic), as is illustrated by the following paragraph:

It is true that the linguistic situation in the Arab world is no longer (if it ever was) characterized by Classical Arabic/Modern Standard Arabic, on the one hand, and the various regional dialects, on the other. Ferguson’s impressionistic and perhaps idealized characterization of the two varieties as being in complementary distribution functionally is removed from the reality of Arabic-speaking communities. It is idealized in the sense that it does not reflect the constant flux and ever-increasing leakage between the two varieties. Nor does it reflect the dramatic social changes that have taken place in the Arab world. “One may ask whether the seal between the two varieties had ever been hermetic”, as Walters (1996a) rightly puts it. The linguistic situation in the Arab world has always been permeated by a state of linguistic flux due to the prolonged contact between the two varieties, on the one hand, and between Arabic and a foreign language (typically a former colonial language, e.g. French or English), on the other. The advocacy of universal education and the accessibility and knowledge of fuṣḥā have increased over the years. Fuṣḥā has increasingly ceased to be used restrictively by a privileged literate elite or to be known passively by a handful of illiterate people. Besides, social changes and growing literacy rates have called for new domains of use of both varieties, resulting in an increasing overlap between the two varieties of Arabic (and hence a significant leakage), and have induced changes in attitude toward the use of either variety in both the written (Daher 1999; Belnap and Bishop 2003) and the spoken mode (Parkinson 1996, 2003; Walters 1996 1996 “Diglossia, Linguistic Variation, and Language Change in Arabic.” In Perspectives on Arabic Linguistics VIII, ed. by Mushira Eid, 157–97. Amsterdam: John Benjamins Publishing Company. CrossrefGoogle Scholar, 2003; Boussofara-Omar 1999, forthcoming).(Boussofara-Omar 2011Boussofara-Omar, Naima 2011 “Diglossia.” Encyclopedia of Arabic Language and Linguistics. Managing Editors Online Edition: Lutz Edzard; Rudolf de Jong. Brill 2011 Brill Online. K.U. Leuven – University Library. 05 April 2011 http://​www​.brillonline​.nl​/subscriber​/entry​?entry​=eall​_COM​-0090, emphasis mine)

This rather lengthy quote suggests that Boussofara-Omar seems to consider switches between only two different codes, rather than more codes, either languages other than Arabic or different varieties of Arabic.

However, a few qualifying remarks must be made. In order to be able to analyze discourse with code-switching as an analytical tool, one has to be able to distinguish at least two codes.99.A similar remark has been made by Bassiouney concerning the attempts at defining intermediate levels: "[…] these intermediate levels cannot be understood unless one presupposes the existence of two 'poles', H and L. It may be that 'pure H' or 'pure L' does not occur very often, and that there are usually elements of both varieties in any stretch of normal speech, but still one has to consider a hypothetical pure H or L in order to presuppose that there are elements that occur from one or the other in a stretch of discourse." (Bassiouney 2009Bassiouney, Reem 2009Arabic Sociolinguistics. Edinburgh: Edinburgh University Press. CrossrefGoogle Scholar, 13) This is not always as straightforward as it might seem at first sight. Even when the two codes are different languages, one might have to deal, for instance, with lexical borrowing and the use of loan words. Needless to say that the relation between borrowing and code-switching is a complex one. This issue becomes even more complex when the codes belong to the same linguistic system, as is the case in code-switching between different varieties of the same language. One has to consider not only borrowings and loan words, but also shared and mixed forms. In his analysis of switching between Black English Vernacular and Standard English Labov (1971Labov, William 1971 “The notion of “system” in creole studies.” In Pidginization and creolization of languages, ed. by D. Hymes, 447–72. Cambridge : Cambridge University Press.Google Scholar, 462) concluded that it is difficult to identify switching sites since both codes have a lot of shared items. The same difficulty was encountered by Gumperz (1982Gumperz, John Joseph 1982Discourse strategies. Cambridge: Cambridge University Press. CrossrefGoogle Scholar, 85) when analyzing switches between Panjabi and Hindi: The two codes shared so many phonological, lexical and syntactical items that they were indistinguishable in many cases. The same observation has been made by researchers dealing with diglossic code-switching in Arabic (e.g. switching between fuṣḥā and a non-fuṣḥā variety). Bassiouney (2009)Bassiouney, Reem 2009Arabic Sociolinguistics. Edinburgh: Edinburgh University Press. CrossrefGoogle Scholar refers to the difficulty to decide whether a linguistic item belongs to the Egyptian variety or fuṣḥā. She refers, among other things to the definite article (p. 49), prepositions (p. 55), morphemes in general (p. 55) and concludes that "ECA [Egyptian Colloquial Arabic] and MSA [Modern Standard Arabic] are different codes but with a lot of shared […] morphemes, and it is almost impossible at times to say whether a certain morpheme belongs to ECA or MSA." (Bassiouney 2009Bassiouney, Reem 2009Arabic Sociolinguistics. Edinburgh: Edinburgh University Press. CrossrefGoogle Scholar, 55) This conclusion can be expanded to code-switching between fuṣḥā and non-fuṣḥā varieties other than Egyptian Arabic. Shared items can be found on all linguistic levels, but one has to keep in mind that concrete examples will differ depending on which non-fuṣḥā variety is taken into account. Shared forms can be the result of the fact that both codes simply belong to the same linguistic system. For instance, the phoneme /q/ is phonologically shared by fuṣḥā and several non-fuṣḥā varieties, such as most urban varieties in the Maghreb, but also several rural varieties in Syria etc. Other examples can be situated on the morphological (e.g. the definite article l-) and lexical levels.

However, shared forms can also be the result of mutual borrowing. Lexical borrowing (with or without phonetic adaptations) is the most obvious example, but borrowing can also take place on all other linguistic levels. In their discussion of general problems of Arabic dialectology, Schippers en Versteegh (1987)Schippers, Arie, and Kees Versteegh 1987Het Arabisch. Norm en realiteit. Muiderberg: Coutinho.Google Scholar refer to the impossibility to describe "pure dialects" because of the continuous influence of fuṣḥā ("Classical Arabic") on the spoken varieties ("dialects"). They mention, for instance, that most descriptive studies of modern dialects refer to the use of both the analytical possessive construction (with an analytical possessive component) and the fuṣḥā genitive construction (without such an analytical component),1010.Note that in fuṣḥā the possessive relation between the possessor and the possessed is expressed by means of flexion (e.g. the genitive for the possessor, which is the second part of the construction, hence the term ‘genitive construction’). In non-fuṣḥā varieties the construction is used without flexion. with the remark that this makes it impossible to know whether both constructions belong to spontaneous language use (by which they mean the non-fuṣḥā variety) or whether the latter is borrowed from fuṣḥā (Schippers & Versteegh 1987Schippers, Arie, and Kees Versteegh 1987Het Arabisch. Norm en realiteit. Muiderberg: Coutinho.Google Scholar, 109)1111.Versteegh (2001) dismisses the idea of diglossic code-switching altogether arguing that the language choices take place on a continuum and therefore "these changes do not take the form of code-switching from one variety to another, but manifest themselves in a larger percentage of features from the opposite variety." (Versteegh 1997Versteegh, Kees 2001 (1997 1st print)The Arabic Language. Edinburgh: Edinburgh University Press.Google Scholar, 194) Note, however, that this approach suffers of what I called one-dimensionality. Hence, "the mixtures of variants" in the quote "[s]ince the colloquial and the standard language are not discrete varieties, but only abstract constructs at the extremes of a continuum, linguistic choice does not a two-way selection, but rather a mixture of variants." (Versteegh 1997Versteegh, Kees 2001 (1997 1st print)The Arabic Language. Edinburgh: Edinburgh University Press.Google Scholar, 195, emphasis mine) should also be interpreted only in terms of fuṣḥā and one non-fuṣḥā variety.. For the purposes of analysis in terms of code-switching, the point is not so much whether the construction without the analytical component is original to non-fuṣḥā varieties or borrowed from fuṣḥā, because we can argue that even in the second case it is a fully integrated borrowing and as such part and parcel of the non-fuṣḥā variety as well. As a result, we cannot but consider it a shared grammatical form and therefore it is – without phonological, morphological or discursive clues – impossible to decide to which code this grammatical construction belongs (and hence whether or not its use is to be considered a code-switch).1212.One could possibly argue that the use of the form with flexion is indication of its fuṣḥā-ness, but in the case of the opposite, the designation is harder because the pausal forms are also used in fuṣḥā. Moreover, linguistic items can be considered to be phonologically, grammatically or lexically shared (e.g. the grammatical or lexical item occurs in both codes having the same semantic load), but can be realized differently on the morphophonological level (with different combinations of morphophonological elements of both the involved codes).1313.See Mazraani (1997Mazraani, Nathalie 1997Aspects of Language Variation in Arabic Political Speech Making. Richmond Surrey: Curzon Press.Google Scholar, 33–38) and Bassiouney (2009Bassiouney, Reem 2009Arabic Sociolinguistics. Edinburgh: Edinburgh University Press. CrossrefGoogle Scholar, 43, 46–7) for a more detailed discussion with examples. However, so-called mixed or hybrid forms do not pose analytically the same problem as completely shared forms, since they can be considered in terms of intra-sentential or intra-word code-switching.

What can be concluded is then that, even when one has to deal only with switches between one non-fuṣḥā variety and fuṣḥā, defining the codes is not as clear cut as it might seem at first sight. The case becomes even more complex if switches between more than one non-fuṣḥā variety involved need to be considered. Only when highly salient markers occur, it is easy to distinguish between codes and hence identify code-switches, but when shared forms are used, discursive and metapragmatic clues including the perception of the interlocutors, rather than linguistic form alone, has to be taken into account as well.1414.One seems not to be able to escape completely from Bassiouney’s (2009Bassiouney, Reem 2009Arabic Sociolinguistics. Edinburgh: Edinburgh University Press. CrossrefGoogle Scholar, 13) observation that at some point "hypothetical pure codes" need to be distinguished in order to proceed with analysis.

2.1.4Diglossia, intermediate varieties and the native speaker

The previous point, namely that not only linguistic form but also the perception of native speakers needs to be taken into account, brings us to another criticism on the attempts to divide the continuum between H and L in distinct varieties. Haeri (2000 2000 “Form and Ideology: Arabic Sociolinguistics and Beyond.” Annual Review of Anthropology 29: 61–87. CrossrefGoogle Scholar, 66) questions the usefulness of the coining of labels for one of more intermediate varieties without considering their meaningfulness in the Arabic speech community: "One of the reasons for the unending disputes about whether there is di-, tri-, or quadriglossia (Meiseles 1980Meiseles, Gustav 1980 “Educated Spoken Language and the Arabic Language.” Archivum Linguisticum XI, New Series 118–148.Google Scholar) is that in practice, stylistic levels are defined purely on the basis of linguistic data. Whether such levels are in fact perceived by and are meaningful to members of the community, or even a part of it, is a question that has not been pursued" (emphasis mine). Or as Jospeh (2004) suggests: the speakers have been left out. This criticism is very fundamental for our purposes, as the main argument of this article is that the concept of diglossia is counterfeited by actual language use (e.g. the linguistic data, linguistic production), but still very persuasive as far as linguistic variability in Arabic is conceived by native speakers (e.g. perception, (meta-)linguistic interpretation).

2.1.5Diglossia as a set of metapragmatic norms

The discussion about the functional division between H and L (that was at least suggested by Ferguson) and the subsequent solutions that have been proposed can maybe be closed by concluding that diglossia is not so much a model for actual language use, but that it should actually be seen as a model of metapragmatic norms (Caton 1991Caton, Steven 1991 “Diglossia in North Yemen: A Case of Competing Linguistic Communities.” Southwest Journal of Linguistics 10 (1): 143–59.Google Scholar, 145, in Haeri 2000 2000 “Form and Ideology: Arabic Sociolinguistics and Beyond.” Annual Review of Anthropology 29: 61–87. CrossrefGoogle Scholar, 66):

Caton (1991Caton, Steven 1991 “Diglossia in North Yemen: A Case of Competing Linguistic Communities.” Southwest Journal of Linguistics 10 (1): 143–59.Google Scholar, 145) responded to the enduring dispute about the perceived inadequacy of Ferguson’s model by pointing out that it was not one of "actual language use" but rather of the metapragmatic norms that prevailed in the speech community (see Ferguson 1991). That is, Ferguson did not offer rules that would account for when and under what conditions one or another of the languages would be chosen by speakers. What he offered was a model of what the community perceives as appropriate usage based on historically and institutionally inculcated norms.(emphasis mine)

However, the confusion between diglossia as a model for norms or as a model for actual language use was caused by Fergusons ambiguous formulation, as is obvious from the quote above.

2.1.6First set of conclusions

The conclusion is then that diglossia is invalid as a model for actual language use, but it still remains an interesting model to understand the overt/explicit norms for language use or the metapragmatic norms in the Arabic linguistic community (see Caton 1991Caton, Steven 1991 “Diglossia in North Yemen: A Case of Competing Linguistic Communities.” Southwest Journal of Linguistics 10 (1): 143–59.Google Scholar; Haeri 1996Haeri, Niloofar 1996The Sociolinguistic Market of Cairo: Gender, Class and Education. London and New York: Kegan Paul International.Google Scholar, 2000 2000 “Form and Ideology: Arabic Sociolinguistics and Beyond.” Annual Review of Anthropology 29: 61–87. CrossrefGoogle Scholar). Most native speakers of Arabic would indeed agree that the norm for Ferguson’s H situations is fuṣḥā while in the L situations it is the use of a non-fuṣḥā variety. We cannot stress enough that these norms are often breached in practice, as is demonstrated in the vast and still growing literature on code-switching in Arabic. However, it is just because of this set of overt metapragmatic norms (which suggest a clear functional distribution of fuṣḥa and ʻāmmīya) that the code-switches in actual language use gain their meaning. For instance, it is exactly because the use of fuṣḥā is the norm in political speeches that Nasser’s switches to Egyptian Arabic are not random or just slips of the tongue, but highly meaningful on the pragmatic level (see Holes 1993 1993 “The uses of Variation: A Study of the Political Speeches of Gamal Abd Al-Nasir.” Perspectives on Arabic Linguistics V, ed. by M. Eid, and C. Holes, 13–46. Amsterdam: John Benjamins Publishing Company. CrossrefGoogle Scholar; Mazraani 1997Mazraani, Nathalie 1997Aspects of Language Variation in Arabic Political Speech Making. Richmond Surrey: Curzon Press.Google Scholar). This is also the case for other examples of so-called diglossic switching. And maybe this is exactly why the term diglossic switching remains interesting, not so much because the switching only occurs between H and L, but rather because it is informed by a sense of dichotomy between H and L that is perceived as very real by most native speakers of Arabic (see also below and Suleiman 2008 2008 “Egypt: From Egyptian to Pan-Arab Nationalism.” In Language and National Identity in Africa, ed. by Andrew Simpson, 26–43. Oxford: Oxford University Press.Google Scholar, 28, Suleiman 2011 2011Arabic, Self and Identity. A Study in Conflict and Displacement. Oxford: Oxford University Press. CrossrefGoogle Scholar, 29–31; Eid 2002Eid, Mushira 2002 “Language is a Choice – Variations in Egyptian Women’s Written Discourse.” In Language Contact and Language Conflict in Arabic – Variations on a Sociolinguistic Theme, ed. by A. Rouchdy, 203–232. London: Routledge-Curzon.Google Scholar, 204).

2.2Diglossia and linguistic prestige: Standard versus prestige

Another aspect of Ferguson’s article that has led to a lot of misinterpretation and that needs to be clarified is the presupposition that only H (the standard language, fuṣḥā) is prestigious and that all other varieties, namely the L varieties, don’t have any prestige whatsoever.

In all the defining languages the speakers regard H as superior to L in a number of respects. Sometimes the feeling is so strong that H alone is regarded as real and L is reported 'not to exist.' Speakers of Arabic, for example, may say (in L) that so-and-so doesn’t know Arabic. This normally means that he doesn’t know H, although he may be a fluent, effective speaker of L.(Ferguson 1959Ferguson, Charles A. 1959 “Diglossia.” Word 15: 325–40. CrossrefGoogle Scholar, 329–30)

In the literature that was inspired by Ferguson’s article, the prestige of the standard language H, in this case fuṣḥā, is most often taken for granted. This is also suggested in Arabic language use by labeling all non-fuṣḥā varieties invariably ʻāmmīya, lahağāt or dāriğa. (see below)

However, even if fuṣḥā is a very prestigious variety in the Arabic speech community, its prestige is blurred in various ways and for different reasons.

Although many if not most of the non-fuṣḥā varieties can be rightly called 'dialects', 'colloquials' or 'vernaculars', other non-fuṣḥā varieties actually do have a lot of (covert) prestige and are used in speech contexts that are much wider than the ones one might expect on a basis of the labels that are used to describe them. On a local or regional level, many (mostly urban) varieties function as local or regional prestige forms. In many Arab states, because of the political and cultural predominance of the capital, the variety spoken by the urban classes of the capital also functions as a national standard or prestige form.1515.The use of the term 'standard' should not be confused with 'standard language', mainly because of their lack of any official recognition whatsoever. This was also mentioned in Ferguson’s article:

In speech communities which have no single most important center of communication a number of regional L’s may arise. In the Arabic speech community, for example, there is no standard L corresponding to educated Athenian dhimotiki, but regional standards exist in various areas. The Arabic of Cairo, for example, serves as a standard L for Egypt, and educated individuals from Upper Egypt must learn not only H but also, for conversational purposes, an approximation to Cairo L.(Ferguson 1959Ferguson, Charles A. 1959 “Diglossia.” Word 15: 325–40. CrossrefGoogle Scholar, 332)

Moreover, often the speech of educated classes in the capital is considered to represent the whole country. Not only do several non-fuṣḥā varieties then actually function as local or regional standard varieties, many examples can be given of cases in which the overt prestige of fuṣḥā is outdone by the covert prestige of these non-fuṣḥā varieties. This is most clearly illustrated when a linguistic feature is shared between fuṣḥā and the stigmatized non-fuṣḥā variety, but deviates from the local prestigious non-fuṣḥā variety. In such cases, the shared fuṣḥā/stigmatized non-fuṣḥā variant will be replaced by the prestigious non-fuṣḥā variant, even if this means from a normative perspective that the 'correct' form will be replaced by an 'incorrect' one. For instance, Holes (1986Holes, Clive 1986 “The Social Motivation for Phonological Convergence in Three Arabic Dialects.” International Journal of Sociolinguistics 61: 33–51.Google Scholar, 1995 1995 “Community, Dialect and Urbanization in the Arabic-speaking Middle East.” Bulletin of the School of Oriental and African Studies 58/2: 270–287. CrossrefGoogle Scholar) has demonstrated that in Bahrain there is a tendency for Bahraini Shiite Muslims (Baḥarna) to adapt to linguistic features of the variety spoken by Sunni Muslims (ʻArab), which is considered to be the locally prestigious variety. One of the characteristics of the ʻArab variety is the realization of the alveolar affricate /ğ/ as a palatal fricative [y]. In the Baḥarna variety /ğ/ is realized as [ğ], just as in fuṣḥā. However, when interacting with ʻArab, Baḥarna will often switch to [y], even when it is in opposition with the fuṣḥā variant. Holes concludes that the educated variety of the ʻArab dialect has become a standard for Bahrain as a whole and that it is used in almost all attempts to represent 'typical' Bahraini speech. (Holes 1995 1995 “Community, Dialect and Urbanization in the Arabic-speaking Middle East.” Bulletin of the School of Oriental and African Studies 58/2: 270–287. CrossrefGoogle Scholar, 276) The same goes for (mostly female) speakers of rural and Bedouin varieties of Jordanian Arabic, (for which the interdental realization of /d/ and /t/ as [d] and [t]is a characteristic shared with fuṣḥā) who tend to switch to the urban pronunciation (respectively [d] and [t] or [z] and [s]) when they have the impression that their own variety is inappropriate. Non-Muslims in Baghdad (e.g. speakers of a qeltu-dialect) tend to switch–among other features–from [q] (shared with fuṣḥā) to [g] in interactions with Muslims (speakers of a gilit-dialect). (Ferguson 1959Ferguson, Charles A. 1959 “Diglossia.” Word 15: 325–40. CrossrefGoogle Scholar, 325; Blanc 1964 1964Communal Dialects in Baghdad. Harvard Middle Eastern Monographs. Cambridge Massachusetts: Harvard University Press.Google Scholar, 9; Abu-Haidar 1991Abu-Haidar, Farida 1991Christian Arabic of Baghdad. Wiesbaden: Otto Harrasowitz.Google Scholar, 6)1616.Note that Abu-Haidar (1991)Abu-Haidar, Farida 1991Christian Arabic of Baghdad. Wiesbaden: Otto Harrasowitz.Google Scholar, a speaker of Christian Baghdadi Arabic herself, refers to and quotes both Ferguson (1959)Ferguson, Charles A. 1959 “Diglossia.” Word 15: 325–40. CrossrefGoogle Scholar and Blanc (1964) 1964Communal Dialects in Baghdad. Harvard Middle Eastern Monographs. Cambridge Massachusetts: Harvard University Press.Google Scholar to substantiate her claim (f.n. 9).

These examples clearly illustrate that in certain contexts the overt prestige that is associated with fuṣḥā is overruled by the covert prestige of the local prestigious non-fuṣḥā varieties. This suggests that when a feature is shared by fuṣḥā and a non-prestigious (or stigmatized) variety, its association with the stigmatized variety seems to overrule its association with fuṣḥā.

2.2.1 Lēš tmaddanti? 1717.Why do you (fem. sing.) talk urban?

However, this does not have to lead us to the conclusion that the so-called stigmatized varieties are never valued by its speakers. They can serve as a strong marker for in-group identity and as such they can be considered to be prestigious within the in-group. This is exemplified by the softly reproachful expression 'lēš tmaddanti' that is directed towards (mostly young female) speakers of rural Jordanian varieties when they insert phonological or lexical items that are associated with urban varieties, such as the glottal stop [ʼ] (urban variant) for /q/ instead of [g] (rural and Bedouin variant). This is also exemplified by Bassiouney reporting the following reaction of an Egyptian student coming originally from Upper Egypt when complimented about his fluency in Cairene Arabic: “I speak Cairene to you. I can never speak it to my mother. If I speak Cairene Arabic to my mother, she will call me a sissy and possibly kill me!" (Bassiouney 2009Bassiouney, Reem 2009Arabic Sociolinguistics. Edinburgh: Edinburgh University Press. CrossrefGoogle Scholar, 3)

2.2.2Foreign languages

Another aspect that must be taken into account is the role of languages other than Arabic, especially French and English, and to a lesser extent Italian and Spanish (the latter two will not be further discussed here).1818.The languages spoken by linguistic minorities will not be taken into account neither, since their prestige is most often limited to being a marker for in group identity of the linguistic minority in question and as such they do not compete for linguistic prestige for speakers who don’t speak them natively. Despite the fact that both languages are related to the colonial past of most of the Arab countries (French in the Maghreb, Syria and Lebanon and English in Egypt, Sudan, Jordan, Palestine, Iraq and some of the countries of the Gulf) and despite the measures for the arabisation (taʻrīb) in the fields of administration and education that were taken in almost all of these countries after independence, English and French continue to be important languages in terms of linguistic prestige. In most Arab countries education offered in primary and secondary public schools is mostly in Arabic. However, with the exception of Syria, in most Arab countries most courses on university level are, depending on the academic field, partly or completely offered in either French or English. Moreover, many private schools in the Arab world tend to offer education in French or English and teach Arabic as a foreign language. As a result, bilingualism in Arabic (mostly spoken varieties) and either English or French, has become a marker for the elite that can afford to send its children to private schools. Moreover, in many cases fluency in either English or French is much more an asset on the labor market than the knowledge of fuṣḥā is.

[…], the use of foreign languages, French and English specifically, is still prevalent in the Arab world, even more so than at the time of colonisation, for different reasons. Some of the reasons are related to economic needs and market forces, as is the case with countries that depend on tourism for their hard currency, such as Morocco, Tunisia and Egypt, or countries that depend on France as their main trading market, such as Morocco. Although most of the reasons why parents in the Arab world are keen on teaching their children a foreign language, and learning one themselves, are to a great extend utilitarian (Shaaban 2006), there are still symbolic connotations of the use of French in North Africa, for example.(Bassiouney 2009Bassiouney, Reem 2009Arabic Sociolinguistics. Edinburgh: Edinburgh University Press. CrossrefGoogle Scholar, 210–2)

Therefore, despite their association with the colonial past, English and French can compete in terms of prestige with fuṣḥā because of their association with modernity and the sciences and upward social mobility:

[…], mastery of a Western language represents an important kind of embodied symbolic capital (Walters 1999b), a fact not lost on future elites across the Arab world. Knowledge of a Western language may be useful in getting a job or helping one’s country modernize, but it also increases one’s social status and may help women, in particular, in the marriage market.(Walters 2011 2013 “Language Attitudes.” In Encyclopedia of Arabic Language and Linguistics. Managing Editors Online Edition: Lutz Edzard, Rudolf de Jong. Brill Online 2013 Reference. K.U. Leuven – University Library. 19 November 2011 http://​referenceworks​.brillonline​.com​/entries​/encyclopedia​-of​-arabic​-language​-and​-linguistics​/language​-attitudes​-COM​_vol2​_0084)

From the above can be concluded that even if fuṣḥā actually does have a lot of overt prestige, this does not necessarily mean that this influences language use in a straightforward way and that speakers will switch invariably to fuṣḥā in contexts in which their own variety is felt to be inappropriate. So, a distinction must be made between different forms of prestige or the symbolic meanings that are attached to different varieties of Arabic.

Furthermore, even the overt prestige of fuṣḥā cannot be fully taken for granted. Haeri (1996Haeri, Niloofar 1996The Sociolinguistic Market of Cairo: Gender, Class and Education. London and New York: Kegan Paul International.Google Scholar, 2000 2000 “Form and Ideology: Arabic Sociolinguistics and Beyond.” Annual Review of Anthropology 29: 61–87. CrossrefGoogle Scholar, 69) argues that in Cairo upper class men refer more often to the political importance of fuṣḥā as a marker of pan-Arab nationalism and cultural and political resistance against colonialism and imperialism than women do. Moreover, to corroborate the remark above, she observed that the overt importance they attached to fuṣḥā did not influence their language use in such a way that it led to an increased use of fuṣḥā forms:

although upper class men in Cairo spoke far more about the political importance of Classical Arabic in forging pan-Arab identity, in resisting colonial domination, and as a cultural and political weapon against more recent forms of foreign domination, their actual use of Classical Arabic forms did not match the overt importance they attached to this language. Such characterizations of Classical Arabic were not as readily present in interviews with women, and they also did not use such forms as often as men did.(Haeri 2000 2000 “Form and Ideology: Arabic Sociolinguistics and Beyond.” Annual Review of Anthropology 29: 61–87. CrossrefGoogle Scholar, 69)

This confirms, first of all, our argument that the overt prestige that is attached to fuṣḥā is not necessarily enacted in actual language use (see also above). Furthermore, this means that not all individuals or groups of individuals are in the same ways attached to fuṣḥā.

2.2.3Second set of conclusions

The above discussion has illustrated that fuṣḥā indeed has a lot of overt prestige in the Arabic speech community. Speakers will very often explicitly refer to fuṣḥā as the language of the Koran and the Islamic heritage, the language of a rich literary heritage, and as a marker of pan-Arab identity. However, this prestige should not be taken for granted, as it is blurred in several ways.

First of all, in contrast to many Western standard languages, fuṣḥā is not the standardized version of the vernacular of a socially prestigious group. Its prestige is not based on the social prestige that is derived from social, cultural, political or religious dominance. In this capacity, the prestige that is related to fuṣḥā is an abstract kind of prestige that is often overruled by the prestige of locally or regionally prestigious varieties in situated oral face-to-face interactions.

Second, fuṣḥā is not necessarily a language variety that is needed for upward social mobility. In many cases, the knowledge of French or English is far more important to find a job than the knowledge of fuṣḥā is.

Third, not all social groups are to the same degree attached to fuṣḥā, neither do they give the same symbolic meanings to it. Research has demonstrated that, for instance, in Egypt women express far less the importance of fuṣḥā as a marker of pan-Arab and religious identity (Haeri 2000 2000 “Form and Ideology: Arabic Sociolinguistics and Beyond.” Annual Review of Anthropology 29: 61–87. CrossrefGoogle Scholar). Neither are Arab Christians religiously attached to fuṣḥā in the same ways as Muslims are, even if they are highly respectful to the religious meanings that Muslims attach to fuṣḥā.

Fourth, the overtly voiced prestige that is attached to fuṣḥā is not necessarily enacted in actual language use. Upper class men in Egypt use fuṣḥā far less than their explicit statements about its religious, political and cultural importance would suggest (Haeri 2000 2000 “Form and Ideology: Arabic Sociolinguistics and Beyond.” Annual Review of Anthropology 29: 61–87. CrossrefGoogle Scholar).

Fifth, the above discussion has also demonstrated that all languages and linguistic varieties can carry prestige, depending on the speaker and the context, be it not always a prestige that is overtly expressed. Even the so-called stigmatized varieties are at least prestigious as markers of an in-group identity. In such contexts, the use of other (overtly more prestigious) varieties is not appreciated.

The discussion above is, as I have done myself, often framed in terms of competing linguistic prestige. Since none of the Arabic varieties seems to have unflawed prestige that is one-dimensionally enacted in linguistic practice, it is probably more fruitful to describe the different ways in which languages and linguistic varieties are overtly and covertly evaluated by speakers in terms of the symbolic meanings that are attached to the different varieties. As has been demonstrated above, these assessments are not necessarily the same for all speakers or categories of speakers, but they can highly differ depending on gender, social, religious, sectarian or national identity. Moreover, the same speaker can attach different meanings to different varieties, depending on the context (in its widest meaning). Therefore, the symbolic meanings of Arabic varieties can also be assessed as markers of identity.

The following table presents a rather generalizing overview of positive and negative symbolic meanings that can be attached to different languages and Arabic varieties in the Arabic speech community.

fuṣḥā:

overt prestige +

  • Islam

  • classical heritage

  • pan-Arab nationalism

  • education

negative assessment −

  • ‘book language’ = ossified

  • pedantic

  • islamism?

A +

(mostly urban varieties)

covert prestige +

  • associated with socio-economic or political dominance

  • effeminate

A −

(mostly rural varieties)

covert prestige +

  • ingroup ("lēš tmaddanti")

  • rural, lower classes, uneducated

foreign language (English, French)

overt prestige +

  • education, science, internationality

  • upper class (access to (private) higher education, travel, etc.)

  • foreign dominance (political = colonial and post-colonial political dominance and economic dominance)

3.Analysis of the Muqtaṭaf debate

3.1Introduction

The main conclusion of the first part of this article is that diglossia cannot be used as a model for actual language use. Neither does it account straightforwardly for the different symbolic values that are attached to the different Arabic varieties. Nevertheless, the idea that there exists a dichotomy between fuṣḥā and non-fuṣḥā varieties, in which fuṣḥā is the prestigious variety and all the other varieties don’t have any prestige whatsoever, in combination with a very clear sense of the domains in which either variety is respectively appropriate, is very real to most speakers of Arabic. In other words, we can say that “[diglossia] accords well with how the Arabs generally conceptualize their language situation” (Suleiman 2011 2011Arabic, Self and Identity. A Study in Conflict and Displacement. Oxford: Oxford University Press. CrossrefGoogle Scholar, 29). In this sense, we can agree with Caton’s (1991Caton, Steven 1991 “Diglossia in North Yemen: A Case of Competing Linguistic Communities.” Southwest Journal of Linguistics 10 (1): 143–59.Google Scholar, 145) conclusion that diglossia can serve as a model for the metapragmatic norms in the Arabic speech community. These norms are based on language ideological attitudes, in which the dichotomy between fuṣḥā and non-fuṣḥā varieties is taken for granted. The above discussion has hopefully clearly demonstrated that the linguistic reality, as far as both actual language use and the distribution of prestige (e.g. the symbolic meanings that are given to different varieties) are concerned, is far more complex than is suggested by this basic dichotomy. Still, both language use and the distribution of prestige are influenced by it, be it in ways that are never straightforward.

“Folk-linguistics1919.Suleiman (2013 2013Arabic in the Fray. Language Ideology and Cultural Politics. Edinburgh: Edinburgh University Press. CrossrefGoogle Scholar, 266), partly quoting himself, defines folk linguistics as “the range of views and attitudes people have about their language, including its origin and the myths surrounding it that “allow us to come closer to the overt or covert orientations, assumptions, and hidden ideologies of the community and how these relate to linguistic repertoire.” (Suleiman 2008 2008 “Egypt: From Egyptian to Pan-Arab Nationalism.” In Language and National Identity in Africa, ed. by Andrew Simpson, 26–43. Oxford: Oxford University Press.Google Scholar, 28)” nativist dualism”, (Suleiman 2011 2011Arabic, Self and Identity. A Study in Conflict and Displacement. Oxford: Oxford University Press. CrossrefGoogle Scholar, 30) captures the language ideological aspects of these diglossic conceptualizations very well, referring at the same time to their shared or “common-sense nature”, meaning that they are taken for granted and as such hardly ever questioned by most speakers of Arabic (1),2020.“The common-sense (basic/normative) nature of ideological meaning is manifested in the fact that it is rarely questioned, in a given society or community, in discourse related to the ‘reality’ in question, possible across various discourse genres.” (Verschueren 2012Verschueren, Jef 2012Ideology in Language Use. Pragmatic Guidelines for Empirical Research. Cambridge: Cambridge University Press.Google Scholar, 12) together with the fact that they can be “highly immune to experience and observation” (2)2121.“Ideology, because of its normativity and common-sense nature, may be highly immune to experience and observation.” (Verschueren 2012Verschueren, Jef 2012Ideology in Language Use. Pragmatic Guidelines for Empirical Research. Cambridge: Cambridge University Press.Google Scholar, 14) and that they are “community-based” (3)2222.“Like paradigms, ideologies (if we can use the plural at all) are community-based. Their relevance, while going beyond the individual, does not extend beyond a given society or community […]. (Verschueren 2012Verschueren, Jef 2012Ideology in Language Use. Pragmatic Guidelines for Empirical Research. Cambridge: Cambridge University Press.Google Scholar, 11) (Verschueren 2012Verschueren, Jef 2012Ideology in Language Use. Pragmatic Guidelines for Empirical Research. Cambridge: Cambridge University Press.Google Scholar, 10–20; 200). Moreover, it is because, “[t]he products of common-sense reflections (mainly descriptive) are turned into norms (both in the sense of what is seen as normal, and in the regulative and prescriptive sense)” (Verschueren 2012Verschueren, Jef 2012Ideology in Language Use. Pragmatic Guidelines for Empirical Research. Cambridge: Cambridge University Press.Google Scholar, 8) that these conceptualizations have evolved into the highly persistent meta-pragmatic norms referred to above.

Furthermore, as Suleiman remarks and as can be concluded from the discussion above, this folk linguistics nativist dualism, or in other words, the language ideological dimension of diglossia, has been largely excluded from the literature in English on the concept. A detailed explanation of the reasons underlying this exclusion would lead us too far here, but we could suggest that the answer partly lies in the prevailing tendency in Arabic sociolinguistics of quantitative variationist and positivist approaches rather than qualitative analysis (Suleiman 2011 2011Arabic, Self and Identity. A Study in Conflict and Displacement. Oxford: Oxford University Press. CrossrefGoogle Scholar, 8–29, 2013 2013Arabic in the Fray. Language Ideology and Cultural Politics. Edinburgh: Edinburgh University Press. CrossrefGoogle Scholar, 264–5). We could also suggest that this exclusion is also language ideologically driven, in the sense that the selection of topics and research methodologies suggests implicit frames and views of what (socio)linguistic research includes and what not, within what we might call a ‘community of researchers’ adhering to the quantitative variationist paradigm.2323.This tendency can be partly explained by the fact that sociolinguistics had to struggle for its place in the linguistics curriculum. (See Joseph 2004Joseph, Earl John 2004Language and Identity. National, Ethnic, Religious. Hampshire and New York: Palgrave Macmillan.Google Scholar, 60–1) Because of the community-based character of ideology (see above), this language ideology obviously differs from the previous one, for one thing because they are prevalent in different communities, respectively a linguistic community and a community of linguists. Nevertheless, it is exactly because of the persistency of diglossia in its language ideological capacity, together with the fact that it does inform actual language use, that it is important to include it in sociolinguistic analysis.

Finally, the ‘diglossic language situation’ is not as stable as is suggested in Ferguson’s article and the subsequent literature:

It might be supposed that diglossia is highly unstable, tending to change into a more stable language situation. Diglossia typically persists at least several centuries, and evidence in some cases seems to show that it can last well over a thousand years.(Ferguson 1959Ferguson, Charles A. 1959 “Diglossia.” Word 15: 325–40. CrossrefGoogle Scholar, 332)

For one, diglossia is implicitly challenged in actual language use. Moreover, Ferguson himself argues that, even if diglossia can be accepted by a community for an extensive period, it tends to become challenged when three trends appear, namely “more widespread literacy”, “broader communication among different regional and social segments of the community” and “desire for a full-fledged standard “national” language” (Ferguson 1959Ferguson, Charles A. 1959 “Diglossia.” Word 15: 325–40. CrossrefGoogle Scholar, 338).

This is exactly what happened in the Arab Middle East, where since the 19th century the diglossic metapragmatic norms have been explicitly challenged in highly polemic debates that were triggered by calls to modernize the Arabic language, be this by completely changing the norms for written and official language use, or just by simplifying grammatical rules or at least the teaching methods of grammar.2424.See Fergsuon: “[L]eaders in the community begin to call for unification of the language […] and tend to support either the adoption of H or of one form of L as the standard, less often the adoption of a modified H or L, a “mixed” variety of some kind.” (Ferguson 1959Ferguson, Charles A. 1959 “Diglossia.” Word 15: 325–40. CrossrefGoogle Scholar, 338) However, in most of the debates the dichotomy between fuṣḥā and non-fuṣḥā is not only kept intact, it is also taken for granted by all the debaters notwithstanding the position they take in the debate. It is exactly for this reason that it forms an interesting case for uncovering and analyzing language ideological attitudes, since ideology lies much more in the unsaid than in what is overtly voiced. (Verschueren 2012Verschueren, Jef 2012Ideology in Language Use. Pragmatic Guidelines for Empirical Research. Cambridge: Cambridge University Press.Google Scholar)

The remainder of this article is an attempt to uncover a part of what is taken for granted and is left unsaid, or only indirectly said, about fuṣḥā and ʻāmmīya and their mutual relation, or in other words to uncover diglossia or folk-linguistics nativist dualism as “common sense with a history, common sense that members of a wider community appeal to in order to be persuasive” (Verschueren 2012Verschueren, Jef 2012Ideology in Language Use. Pragmatic Guidelines for Empirical Research. Cambridge: Cambridge University Press.Google Scholar, 8, emphasis mine).

As will hopefully be demonstrated, labeling, especially when the process goes un(re)marked, when labels are accepted without further discussion and become common use (e.g. are repeated over and over again), proves to be a very fruitful domain to uncover language ideological attitudes.

The data that will be presented here are taken from a short but vivid debate that was conducted between November 1881 and July 1882 in Al-Muqtaṭaf,2525.Al-Muqtaṭaf was published in Beirut from 1876 until 1884 and then in Cairo from 1884 until its last issue appeared in 1951/1952. Its main editors were Yaʻqūb Ṣarrūf (1852–1927) and Fāris Nimr (1856–1952), both prolific intellectuals of Christian descent. a scientific and cultural journal that was widely read. The string of 11 articles of which the debate consists can be considered a debate in the formal sense of the word. The article opening the debate concludes with a request to "all the distinguished writers who vie for the welfare of the fatherland to present their opinions on this issue [of the dichotomy between the written and the spoken language] and to give it due attention." (Al-Muqtaṭaf 1881Ṣarrūf, Yaʿqūb and Fāris Nimr 1881 “Al-luġa al-ʿarabīya wa al-nağāḥ [The Arabic language and success].” Al-Muqtaṭaf 6 (November): 352–4.Google Scholar, 354)2626.All translations from Arabic are mine. Subsequently ten articles appeared reacting upon the first article and previous reactions on it. The debate was also formally closed; the editors of the journal added a footnote to the 11th and last article in which they stated that "this is the end of the debate in this section, without rebuke, nor blame" (Al-Muqtaṭaf 1882, 110). Therefore, we can consider the string of articles as one unit, namely a debate. Besides, all but the first two articles were published in the section "Debate and correspondence" (bāb al-munāẓara wa al-murāsala).

To my best knowledge, this debate was the first time (since the standardization and codification of Arabic) that the issue of linguistic variability in Arabic was explicitly debated on such a large scale. From that moment on, until the present, other, but in many ways very similar debates have continued to erupt in which the main arguments tended to be repeated over and over again.2727.Other manifestations of the debate are the sub-debate in Al-Muqtaṭaf in 1887–1888 (see Diem 1974Diem, Werner 1974Hochsprache und Dialekt im Arabischen: Untersuchungen zur heutigen arabische Zweisprachigkeit. Wiesbaden: Deutsche Morgenländische Gesellschaft.Google Scholar, 129–30, Daniëls 2002Daniëls, Helge 2002Debating Variability in Arabic: Fuṣḥā versus ʻāmmīya. Unpublished Ph.D. Antwerp: University of Antwerp.Google Scholar, 255–6), the publication of The spoken language of Egypt by Selden J. Wilmore in 1901 and the reactions and counter-reactions it triggered in Al-Muqtaṭaf and other prominent cultural journals such as Al-Hilāl and Al-Manār (see Diem 1974Diem, Werner 1974Hochsprache und Dialekt im Arabischen: Untersuchungen zur heutigen arabische Zweisprachigkeit. Wiesbaden: Deutsche Morgenländische Gesellschaft.Google Scholar, 132–3, Daniëls 2002Daniëls, Helge 2002Debating Variability in Arabic: Fuṣḥā versus ʻāmmīya. Unpublished Ph.D. Antwerp: University of Antwerp.Google Scholar, 257–80, Suleiman 2004Suleiman, Yasir 2004A War of Words. Language and Conflict in the Middle East. Cambridge: Cambridge University Press. CrossrefGoogle Scholar, 68–71), the articles that appeared in 1910 in Al-Muqtabas, yet another leading journal at the beginning of the 20th century, the linguistic proposals by Aḥmad Luṭfī al-Sayyid in the 1920s and 30s (see Daniëls 2002Daniëls, Helge 2002Debating Variability in Arabic: Fuṣḥā versus ʻāmmīya. Unpublished Ph.D. Antwerp: University of Antwerp.Google Scholar, 284–8) and by Anīs Frayḥa in 1938Frayḥa, Anīs 1938 “Al-ʻāmmīya wa al-fuṣḥā. ʻawd ilā al-mawḍūʻ [ʻāmmīya and fuṣḥā. Back tot he subject].” Al-muqtaṭaf 93: 292–8.Google Scholar and 1955 (see Daniëls 2002Daniëls, Helge 2002Debating Variability in Arabic: Fuṣḥā versus ʻāmmīya. Unpublished Ph.D. Antwerp: University of Antwerp.Google Scholar, 289–301, Suleiman 2004Suleiman, Yasir 2004A War of Words. Language and Conflict in the Middle East. Cambridge: Cambridge University Press. CrossrefGoogle Scholar). Needless to say that this list is not exhaustive. As such, we can consider the Muqtaṭaf-debate and the other debates as sub-debates of a larger debate that has most often been called the fuṣḥā-ʻāmmīya debate. Moreover, "the continuous repetition of the main arguments in the different sub-debates in each generation and, in the same generation, in different locations, testifies to the perennial nature of the issues that animate them (Daniels 2002). Here, every act of textualization becomes one of retextualization in which society and culture share the right of authorship" (Suleiman 2004Suleiman, Yasir 2004A War of Words. Language and Conflict in the Middle East. Cambridge: Cambridge University Press. CrossrefGoogle Scholar, 94).

However, the analysis in this part will focus only on the Muqtaṭāf sub-debate. But exactly because of the repetitive character of the arguments, the main findings can be extrapolated to other sub-debates and the larger fuṣḥā-ʻāmmīya debate as a whole and to meta-linguistic language use in general. Even if the debate is in a variety of ways interesting as a locus for language ideology, I will limit myself for the purposes of this article to a micro-analysis of the linguistic labels and the ways in which they are used in the debate.

3.1.1Why the 19th century?

In general, the 19th century can be described as a period during which Arab societies were in flux. The important social, political, cultural and religious changes that took place in that era were also felt in the linguistic domain. During the 19th century, the modernization of Arabic was not only debated, it was also actually taking place, as Arabic underwent some drastic changes, mainly under the influence of the development of the printing press, translation from European languages and the modernization of education in several parts of the Arab world, especially Egypt and Lebanon. These linguistic changes were most dramatically felt at the lexical level. For concepts (of all kinds, not only linguistic ones) previously unknown to the Arab world loan words from European languages entered the Arabic language (with or without being adapted to Arabic phonological and morphological rules), new lexical items were created and already existing words were given a new semantic load.2828.For a more elaborate discussion, see Ayalon (1987)Ayalon, Ami 1987Language and Change in the Arab Middle East: The Evolution of Modern Political Discourse. Oxford: Oxford University Press.Google Scholar, Versteegh & Schippers (1987, 146–6), Versteegh (2001, 173–88). One of the implications of re-semanticization is that words tend to have, at least for a certain period of time, a very ambiguous meaning. Semantic ambiguity is of course always to a certain extent a characteristic of language, but when important social, political or other changes take place–as is the case for the 19th century–this characteristic becomes more salient. The inverse is also the case, when language becomes subject to debate, this means that other important changes are taking place in society. The fact that this observation certainly applies to this debate and to the larger fuṣḥā–ʻāmmīya debate in general cannot be dealt with in detail here, but we cannot stress enough that the debate was not only deeply anchored in the social, political and cultural changes that were taking place at the time, it was also one of the factors contributing to the shaping of these changes. Returning to the linguistic changes themselves, it is exactly because of its position between classical and modern language use that 19th century language use proves to be a highly interesting domain for uncovering language ideological attitudes.

Since this is, as far as the modern period is concerned, most probably the first time that the issue of linguistic variability in Arabic was debated on such a large scale, it seems obvious that the debaters, in order to express their ideas, also needed to create new linguistic terms or re-semanticize already existing ones. This process in itself, then, seriously influenced their and later generations’ thinking about language and linguistic variation. An interesting case is, as will be demonstrated below, the term ʻāmmīya itself.

In the analysis special attention will be given to synonymy, as it is yet another interesting aspect of labeling and constitutes an entry into the implicit language attitudes of the debaters. After establishing fuṣḥā and ʻāmmīya as not only different and separate, but also unequal varieties, a wide variety of synonyms is used to refer to either fuṣḥā or ʻāmmīya. Nevertheless, these synonyms carry with them very different connotations and associations by which the inequality is further confirmed. The use of these synonyms is not challenged and goes even un(re)marked. And as such, synonymy is again a strong indicator of the language ideological attitudes of the debaters (Verschueren 2012Verschueren, Jef 2012Ideology in Language Use. Pragmatic Guidelines for Empirical Research. Cambridge: Cambridge University Press.Google Scholar).

3.2The basic dichotomy: Fuṣḥā and ʻāmmīya

One of the first striking elements that comes, not surprisingly, to the forefront is the fact that the debaters divide Arabic variability into two main categories, which are generally labeled respectively al-luġa al-faṣīḥa/al-luġa al-fuṣḥā and luġat al-ʻāmma/al-luġa al-ʻāmmīya. Whereas there are several attempts to define fuṣḥā, either directly or by defining words that are derived from the same tri-radical root f-ṣ-ḥ, such as faṣīḥ, fuṣaḥāʼ, faṣāḥa, no attempts are made to define explicitly what ʻāmmīya means. As a result, ʻāmmīya is defined implicitly and by default as everything that is not fuṣḥā,2929.However, Ṣarrūf and Nimr refer also to the fact that the science books are written in "a language different from the language we speak" (luġa ġayr al-luġa allatī natakallamuhā) (Al-Muqtaṭaf 1881Ṣarrūf, Yaʿqūb and Fāris Nimr 1881 “Al-luġa al-ʿarabīya wa al-nağāḥ [The Arabic language and success].” Al-Muqtaṭaf 6 (November): 352–4.Google Scholar, 353) (hence my own choice to refer to the ʻāmmīya varieties as non-fuṣḥā varieties, see also above). As the discussion of the functional division and the distribution of linguistic prestige has shown, this has serious and far stretching implications for the ways in which non-fuṣḥā varieties are perceived and evaluated. Turning the point upside-down, we can also argue that this implicit default definition is exactly the result of the ways in which non-fuṣḥā varieties are perceived in the Arabic linguistic community.

Apart from the fact that this terminology overrules the internal variability that is covered by both terms, not only in the obvious case of ʻāmmīya, but also that of fuṣḥā (which is maybe less obvious), the terminology veils the fact that both fuṣḥā and the varieties covered by the term ʻāmmīya actually do have a lot of linguistic features in common on all linguistic levels. As mentioned above, this is one of the main difficulties of an analysis of code-switching and as the discussion has illustrated, it depends on a highly complex set of factors (related to both the context in its broadest sense and the identity of the speakers) whether such a shared feature is then considered to be a marker for fuṣḥā or ʻāmmīya.3030.See Parkinson (1991)Parkinson, Dilworth 1991 “Searching for Modern Fuṣḥa: Real-life Formal Arabic.” ʻArabiyya. Journal of the American Association of Teachers of Arabic 24: 31–64.Google Scholar for an interesting discussion of variability within fuṣḥā and how fuṣḥā means different things for different native speakers of Arabic in Egypt.

3.3Clusters of associations Al-fuṣḥā and Al-ʻāmmīya: Synonyms and their connotations

Those who read the biography of George Stephenson which appeared in this issue have seen that this man studied mathematics and other sciences while he was a mere stoker of a steam engine who only had a simple knowledge of reading. Those who succeed best in their efforts can be found among the Westerners who studied the higher sciences, such as algebra, engineering, natural philosophy and mechanics, while they exercise the basest work and have scientific knowledge [based on] simple reading [skills]. This is because the book language (luġat al-kutub) of the Westerners does not differ a lot from the language they speak (al-luġa allatī yatakallamūna bi-hā). So, a commoner among them (al-ʻāmmī minhum) understands a philosophy book just [in the same way] as a commoner among us (al-ʻāmmī minnā) understands the story of Banī Hilāl. Their common people (ʻāmmatuhum) avail themselves of books [in the same way] as their elite ( hāṣṣatuhum) does. Therefore, you can see that the ways to success are open in the same way for their elite as for their commoners and that the qualities of science are widespread among them.(Al-Muqtaṭaf 1881Ṣarrūf, Yaʿqūb and Fāris Nimr 1881 “Al-luġa al-ʿarabīya wa al-nağāḥ [The Arabic language and success].” Al-Muqtaṭaf 6 (November): 352–4.Google Scholar, 352–3, emphasis mine)

With this paragraph, Yaʻqūb Ṣarrūf and Fāris Nimr, the editors of Al-Muqtaṭaf open a prolific debate that will continue for decades. They claim that the main reason that triggered them to write this article, The Arabic language and success (Al-luġa al-ʻarabīya wa-al-nağāḥ), is the observation that Western societies are more successful in pursuing their scientific aims than the Arab world. In their opinion, the main reason is that in the West, all layers of society have access to scientific books because the language of books does not differ from the spoken language (Al-Muqtaṭaf 1881Ṣarrūf, Yaʿqūb and Fāris Nimr 1881 “Al-luġa al-ʿarabīya wa al-nağāḥ [The Arabic language and success].” Al-Muqtaṭaf 6 (November): 352–4.Google Scholar, 353). However important the argumentation concerning success and its relation to science is, with all its social Darwinist implications, we will focus here only on the lexical labels that are used.

As can be seen, two basic interrelated oppositions, a linguistic one and a social one, are referred to in this opening paragraph: “the book language” (luġat al-kutub) versus “the language they speak” (al-luġa allatī yatakallamūna bi-hā); and “the elite” (al-hāṣṣa) versus “the common people” (al-ʻāmma). In the remainder of the article and the ten reactions it triggered, in other words the debate as a whole, these social and linguistic oppositions will be repeated over and over again. The debaters implicitly and explicitly establish connections between the social and the linguistic oppositions, which are further solidified by the continuous repetition. Moreover, the linguistic opposition is not only repeated, but also rephrased by using synonyms for both the "book language" (written language) and the “spoken language”. These synonyms, however, have very different connotations, but by using them interchangeably, the connotations of one label are (implicitly) transferred to the other. Let us first look at the different ways in which synonymy is established.

3.3.1The process of labeling and synonymy: Indirect equation

One of the ways in which the debaters establish synonymy consists of direct or indirect equation. A few examples that are taken from the debate will do to illustrate this process. As the process is repeated over and over again, we will by no means try to be exhaustive.

When Ṣarrūf and Nimr opened the debate on linguistic variability in Arabic, they presented the linguistic problem basically as the existence of a large difference between the written and the spoken language, to which they referred respectively as the "book language" (luġat al-kutub) or the "language we write" (al-luġa allatī naktub bi-hā) and the "language of speech" (luġat al-takallum) and "the language they/we speak" (al-luġa allatī yatakallamūna/natakallam bi-hā). This difference is subsequently compared to the difference between respectively Latin and Italian, and Classical and Modern Greek:

[…] Modern Greek (al-rūmīya) which is related to Classical Greek (al-yūnānīya) in the same way as the Arabic we speak is related to the Arabic we write.(Al-Muqtaṭaf 1881Ṣarrūf, Yaʿqūb and Fāris Nimr 1881 “Al-luġa al-ʿarabīya wa al-nağāḥ [The Arabic language and success].” Al-Muqtaṭaf 6 (November): 352–4.Google Scholar, 353)

Without mentioning it explicitly, written Arabic is then associated with the ancient or the classical form of the language, while the spoken language is associated with the modern form. It follows that implicitly a historical/diachronic dimension is given to the opposition between the written and the spoken language, even if ‘spoken’ and ‘written’ refer to how the language is used or its medium, whereas ‘ancient’ and ‘modern’ refer to the historical phases in the development of a language. Moreover, these adjectives suggest that the written language is historically older than the spoken language, and the idea that the latter is derived from the first.3131.This can be related to the Arab views on the development of the post-Islamic Arabic dialects

Another interesting implicit equation that appears later in the article, is that of the written language with correct Arabic.

Or, we [have] to teach our children to speak correct Arabic (al-ʻarabīya al-ṣaḥīḥa), so that a natural [linguistic] disposition (malaka)3232.For a discussion of the ideological implications of the concept of natural linguistic dispositions or “pristine linguistic intuitions” (fiṭra luġawīya or salīqa luġawīya), see Suleiman (2013 2013Arabic in the Fray. Language Ideology and Cultural Politics. Edinburgh: Edinburgh University Press. CrossrefGoogle Scholar, 53–4). comes in them and that they speak the same as they write.(Al-Muqtaṭaf 1881Ṣarrūf, Yaʿqūb and Fāris Nimr 1881 “Al-luġa al-ʿarabīya wa al-nağāḥ [The Arabic language and success].” Al-Muqtaṭaf 6 (November): 352–4.Google Scholar, 353)

Equating written Arabic in this way with correct Arabic also indicates indirectly that spoken Arabic is incorrect.

On the other hand, the spoken language is associated with general and common language use. When discussing the third solution, namely “to write books in the language we speak” (an naktuba kutubanā bi-al-luġa allatī natakallamu bi-hā), the authors refer to the fact that other peoples/nations replaced “their ancient languages” (luġātuhum al-qadīma) with “the commonly used languages” (al-luġāt al-šāʼiʻa). The authors apply this equation also on the case of Arabic, by arguing that nothing hinders them to standardize "the spoken language that is commonly spoken in the Arabic countries” (luġat al-takallum al-šāʼiʻa fī al-buldān al-ʻarabīya).

3.3.2The process of labeling and synonymy: Direct equation

It is striking that in the article that can be considered to be the opening of the modern debate concerning variability in Arabic, which is most often referred to as the fuṣḥā-ʻāmmīya debate, no reference at all is made to these two linguistic labels. It is only in the first reaction that appeared to the opening article that this happens. This article was written by Halīl al-Yāziğī and introduces us to another way of establishing synonymy. The debaters have the habit of briefly summarizing the arguments that were previously mentioned, before adding their own arguments. The reformulation is done by means of direct quotes, to which very often new terms are added, or by rephrasing the previous argument(s) and using other linguistic key words that are considered to have the same meaning. In the following example we find examples of both methods. Al-Yāziğī rephrases Ṣarrūf’s and Nimr’s solutions as follows:

One of them is the replacement of our language with another language. The second is the replacement of the writing language with the language of speech, meaning the folk language (istibdāl luġat al-kitāba bi-luġat al-takallum ayy luġat al-ʻāmma). The third is the replacement of the folk language in speech with al-faṣīḥa (istibdāl luġat al-ʻāmma fi al-takallum bi-al-luġa al-faṣīḥa).(Al-Yāziğī 1881Al-Yāziğī, Halīl 1881 “Al-luġa al-ʿarabīya wa al-nağāḥ [The Arabic language and success)].” Al-Muqtaṭaf 6 (December): 404–5.Google Scholar, 404)

In this quote, the spoken language (luġat al-takallum) is directly equated with the folk language (luġat al-ʿamma) by means of the particle ayy (meaning), whereas the written language (luġat al-kitāba) is indirectly equated with al-faṣīḥa by replacing "the written language" with this term in the rephrasing of Ṣarrūf’s and Nimr’s third solution. As already mentioned, because of the repetitive and unchallenged character of this process clusters of associations and meanings arise.

In the Muqtaṭaf debate as a whole, the writing language (luġat al-kitāba) is directly or indirectly equated with the book language (luġat al-kutub), the written language (al-luġa al-maktūba), the Arabic we write (al-ʻarabīya allatī naktubuhā), al-luġa al-faṣīḥa or al-luġa al-fuṣḥā, and their elliptic forms al-faṣīḥa and al-fuṣḥā. These labels are associated or equated with the ancient/classical language (al-luġa al-qadīma), correct (Arabic) language (al-luġa (al-ʻarabīya) al-ṣaḥīḥa) or correct Arabic (al-ʻarabīya al-ṣaḥīḥa), the original/authentic language (al-luġa al-aṣlīya) and the language of the Muḍar tribe (luġat muḍar).

The writing language is thus associated with eloquence, but also with an ancient history, correctness, originality and authenticity and the language of the tribal confederation to which also the Prophet’s tribe the Qurayš belonged.

In the same way, the spoken language (luġat al-takallum) or the language they/we speak (al-luġa allatī yatakallamūna/natakallam bi-hā) and the Arabic we speak (al-ʻarabīya allatī natakallam bi-hā) is equated with the language of the common folk (luġat al-ʻāmma or al-luġa al-ʻāmmīya), the common language (al-luġa al-ʻāmma), the current language (al-luġa al-šāʼiʻa) or the current spoken language (luġat al-takallum al-šāʼiʻa) and the Arabic that we suckled with the mother milk (al-ʻarabīya allatī narḍaʻuhā maʻa al-laban). By doing so, the spoken language is associated with the lower (uneducated) social classes, but also with currency (in the sense of being widespread) and naturalness (in the sense of being naturally and effortlessly acquired as opposed to via formal education).

3.3.3The terms and their connotations: al-luġa ʻāmmīya–luġat al-ʻāmma

Above, reference was already made to the fact that in the first article of the Muqtaṭaf sub-debate the key terms fuṣḥā and ʻāmmīya are not used at all. As the debate further burgeons both terms are gradually introduced and, through the processes of direct and indirect equation, they are established as the basic terms to refer to the poles of the perceived linguistic dichotomy. Another striking element is that the terms faṣīḥa/fuṣḥā and ʻāmmīya are not yet used as independent nouns, but rather as adjectives.

This can be explained by the fact that in the 19th century Arabic language use itself was changing. This change can be traced in the Muqtaṭaf sub-debate as well and, as already mentioned, an analysis of the evolution of the use of the linguistic labels constitutes an interesting locus for tracing how the thinking about language and language variability evolved with the evolution of the debate itself. We also mentioned already that 19th century language use can be positioned between Classical and Modern Arabic. An analysis of the term ʻāmmīya and its cognates is an interesting case to illustrate the hinge position between classical and modern language use of 19th century Arabic.

The two terms that are most currently used in the debate in order to refer to al-ʻāmmīya/non-fuṣḥā are luġat al-ʻāmma and al-luġa al-ʻāmmīya. Rabin (1960) mentions that the first one was current in medieval times, whereas the latter, together with al-dāriğa and al-lahağāt,3333.It is important to note that the term 'lahğa (pl. lahağāt)' in the sense of 'dialect' or 'variety' is also used in the Muqtaṭaf debate, however only in order to refer to specific dialects, e.g. "the dialects of the Syrians, the Egyptians, the Iraqi’s and the Maghribians (lahağāt al-sūrīyīn wa al-ʻirāqīyīn wa al-miṣrīyīn wa al-maġribīyīn) (Al-Mumkin 1882a, 494), "the dialect of which province […], city, […] village, […] neighboorhood" (lahğat ayyati muqātaʻa […], madīna […], qariya […], ḥāra." (Dāġir 1882Dāġir, Asʿad 1882 “Istiḥālat al-Mumkin idā amkana [The impossibility of al-Mumkin if it were possible].” Al-Muqtaṭaf 6 (February): 556–60.Google Scholar, 557). Sometimes the term is also used in the sense of 'way of speaking' or 'pronunciation'. is used in modern times.3434.Interestingly enough, Rabin uses the term post-Islamic dialects in reference to what I call the ʻāmmīya or non-fuṣḥā varieties. This can be related to the fact that traditionally pre-Islamic and post-Islamic language variability has been perceived differently. This is mainly related to the fact that, basically, pre-Islamic variants (luġāt) were considered to remain within the realm of correctness, even when deviating from the norm (naḥw), whereas deviations from the norm related to post-Islamic varieties were considered to be mistakes (laḥn). This can be derived to the ways in which the terms luġa, naḥw and laḥn diachronologically evolved. This process of semantic shifts is intrinsically related to the evolution of the concept of faṣāḥa. For a more elaborate discussion see Ayyoub (2011). The observation that ʻāmmīya is only used as an adjective, and not as noun, suggests that at the end of the 19th century ʻāmmīya was not yet considered to be a distinct linguistic concept. This is further supported by the fact that the authoritative dictionary Muḥīṭ al-muḥīṭ, which was published by Buṭrus al-Bustānī in 1870 and which is considered to be one of the first modern Arabic dictionaries, does not refer to the linguistic connotations of ʻāmmī and ʻāmmīya. Nor does ʻāmmīya occur as a separate lemma (Al-Bustānī 1998 (1870)Al-Bustānī, Buṭrus 1998 (1st print 1870). Muḥīṭ al-muḥīṭ. Qāmūs muṭawwal li-al-luġa al-ʿarabīya [Muhit al-muhit. An Arabic-Arabic Dictionary]. Beirut: Librairie du Liban Publishers.Google Scholar, 634). This is also the case for Kazimirski’s translating dictionary (Arabic-French), which is contemporary to Al-Bustānī's (Kazimirski 1860Kazimirski, A. De Biberstein 1860Dictionnaire Arabe – Français (2 tomes). Beirut: Librairie du Liban.Google Scholar, 358–9). Together with the ways in which the terms are used in the debate, this suggests that before the 20th century, the terms could only be used in order to refer to language and language use in combination with luġa (language), e.g. luġat al-ʻāmma (the language of the common people) or al-luġa al-ʻāmmīya (the folk language).

As the larger debate further develops, these lexical items develop, or rather are developed, into independent linguistic concepts, which are mostly used in contrast with (al-luġa) al-faṣīḥa or (al-luġa) al-fuṣḥā and therefore are defined by non-fuṣḥā-ness. Gradually the term ʻāmmīya then obtains the meaning of colloquial, vernacular or dialect. This is corroborated by a quick glance in monolingual and bilingual Arabic dictionaries. The explaining Arabic dictionary Al-muʿğam al-asāsī (2003al-Qāsimī, ʻalī (ed.) 2003Al-muʿğam al-ʿarabī al-asāsī li-al-nāṭiqīn bi-al-ʿarabīya wa mutʿallimīhā [The Elementary Arabic Dictionary for Native Speakers and Learners]. Cairo: ALECSO/Larouse.Google Scholar, 869) explains ʻāmmīya as follows: "the opposite of the official language or the literary language or al-fuṣḥā."3535.In Arabic: “ hilāf al-luġa al-rasmīya aw al-adabīya aw al-fuṣḥā And kalām ʻāmmī and lahğa ʻāmmīya as "the usual speech of the people, the opposite of al-luġa al-fuṣḥā or the literary language."3636.In Arabic: “kalām al-nās al-ʿādī, hilāl [sic] al-luġa al-fuṣḥā aw al-luġa al-adabīya It is interesting to notice that both entries are explained by referring to their contrast with fuṣḥā, which is in itself related to official and literary language use (respectively luġa rasmīya and luġa adabīya). This way of explaining in itself keeps the binary between fuṣḥā and ʻāmmīya neatly intact. In the same way, the explaining Arabic dictionary Al-muʻğam al-wasīṭ (1980Muṣṭafā, Ibrāhīm (ed.) 1980Al-muʿğam al-wasīṭ [The Intermediary Dictionary]. Istanbul: Dār al-daʻwa.Google Scholar, 629) al-ʻāmmīya as "luġat al-ʻāmma, and this is the opposite of al-fuṣḥā"3737.In Arabic: “luġat al-ʿāmma wa hiya hilāf al-fuṣḥā and kalām ʻāmmī as "what the common people (ʻāmma) utter, differing from the habits of Arabic speech."3838.In Arabic: “mā naṭaqa bi-hi al-ʿāmma alā ġayr sunan al-kalām al-ʿarabī The definition of the latter strongly solidifies the relation between the notions of fuṣḥā and the idea that only fuṣḥā is correct and ‘real’ Arabic. These two dictionaries are very popular and widely used in the Arab world. Finally, Hans Wehr (1994 (1979)Wehr, Hans, and J. Milton Cowan (ed.) 1994 (1979 1st print)A Dictionary of Modern Written Arabic: Arabic-English. Wiesbaden: Otto Harrasowitz.Google Scholar, 751), one of the most widely used translating dictionaries, translates al-ʻāmmīya as "popular language, colloquial language."

This brief exercise shows that the use of the term ʻāmmīya underwent some important changes, of which the beginnings already appeared in the Muqtaṭaf sub-debate. We can assume that the term al-luġa al-ʻāmmīya gradually became more frequently used than luġat al-ʻāmma, ultimately replacing it. As the use of al-luġa al-ʻāmmīya became more current, its elliptic form al-ʻāmmīya developed into an independent noun. The second step in this development (e.g. the independent use of al-ʻāmmīya) can be observed already in the Muqtaṭaf debate, be it only in the 11th and last article by Mitrī Qandalaft.3939. Nağāḥ al-umma al-ʻarabīya fī luġatihā al-aṣlīya (Qandalaft 1882Qandalaft, Mitrī 1882 “Nağāḥ al-’umma al-‘arabīya fī luġatihā al-‘aṣlīya [The success of the Arab nation in its authentic language].” Al-Muqtaṭaf 7 (July): 107–10.Google Scholar, 107–110). However, in the same article, Qandalaft also uses frequently the pre-modern label luġat al-ʻāmma.

In conclusion, we must add that the shift in the use of the labels (luġat al-ʻāmmaal-luġa al-ʻāmmīyaal-ʻāmmīya) was accompanied by an important semantic shift from the social connotations of the label, namely luġat al-ʻāmma, and already to a lesser degree al-luġa al-ʻāmmīya (as the language spoken by the common folk) to its linguistic connotations, namely a linguistic variety that is basically defined in contrast with al-fuṣḥā.4040.Only a couple of times faṣīḥā is defined by contrasting it with ʻāmmīya. Ṣarrūf and Nimr refer also to the fact that the science books are written in "a language different from the language we speak" (luġat ġayr al-luġa allatī natakallamuhā) (Al-Muqtaṭaf 1881Ṣarrūf, Yaʿqūb and Fāris Nimr 1881 “Al-luġa al-ʿarabīya wa al-nağāḥ [The Arabic language and success].” Al-Muqtaṭaf 6 (November): 352–4.Google Scholar, 353) and al-Yāziğī refers to faṣīḥa and explains "in the sense that it is not part of the folk language" (bi-maʻnā annahā laysat min luġat al-ʻāmma). (Al-Yāziğī 1881Al-Yāziğī, Halīl 1881 “Al-luġa al-ʿarabīya wa al-nağāḥ [The Arabic language and success)].” Al-Muqtaṭaf 6 (December): 404–5.Google Scholar, 305) However, even if the linguistic connotations of the label became more dominant, its social connotations never disappeared completely. Because fuṣḥā is associated with education, literature and official language use, and because ʻāmmīya is defined as its opposite, ʻāmmīya is, by default, associated to the lack of education.

4.Conclusion

After a careful and detailed analysis of the concept diglossia and its functional and symbolic implications, it was concluded that diglossia suits as a description of metalinguistic norms and language attitudes in the Arabic linguistic community rather than as one of actual language use. In the second part of the article, I described how the dichotomy between fuṣḥā and non-fuṣḥa varieties, that lies at the basis of diglossia, was constructed by 19th-century intellectuals who participated in a short but vivid debate concerning ʿāmmīya and fuṣḥā in Al-Muqtaṭaf. By means of a lexical and semantic analysis of patterns of labeling, direct and indirect equation and the connotations of the terms, I demonstrated how the diglossic dichotomy was constructed and taken for granted long before the term ‘diglossia’ was used. Moreover, even the terms ʿāmmīya and fuṣḥā were not yet used as independent nouns in the ways they are used now. As such, the analysis sustains the argument that diglossia is a useful concept to describe language ideological attitudes concerning linguistic variability in the Arab world and the symbolic values attached to it, rather than actual language use.

Acknowledgements

This article is partly based on research results presented in my Ph.D., Debating variability in Arabic: Fuṣḥā versus ʿāmmīya, which were seriously updated and rewritten during a research stay at the Centre for Islamic Studies at Cambridge University. I would like to thank Yasir Suleiman for welcoming me at the Centre.

Notes

1.The transliteration of Arabic terms follows Hans Wehr’s system.
2.Before the term "la diglossie arabe" was used by Marçais (1930)Marçais, William 1930 “La diglossie arabe.” L’ensiegnement Public 97: 401–9.Google Scholar and “al-izdiwāğīya” by Frayḥa (1938)Frayḥa, Anīs 1938 “Al-ʻāmmīya wa al-fuṣḥā. ʻawd ilā al-mawḍūʻ [ʻāmmīya and fuṣḥā. Back tot he subject].” Al-muqtaṭaf 93: 292–8.Google Scholar.
3.Basing herself on the work of Mackey (1993)Mackey, W. F. 1993 “Introduction.” In Diglossia: A Comprehensive Bibliography 1960-1990, ed. by M. Fernandez, xiii–xx. Amsterdam: John Benjamins Publishing Company.Google Scholar, Haeri (2000 2000 “Form and Ideology: Arabic Sociolinguistics and Beyond.” Annual Review of Anthropology 29: 61–87. CrossrefGoogle Scholar, 64–5) claims that the term diglossia was first used to describe the Greek language situation in the 1880s. It is interesting to note that this coincides with the early beginnings of the fuṣḥā–ʻāmmīya debate in Al-Muqtaṭaf (1881–1882Ṣarrūf, Yaʿqūb and Fāris Nimr 1881 “Al-luġa al-ʿarabīya wa al-nağāḥ [The Arabic language and success].” Al-Muqtaṭaf 6 (November): 352–4.Google Scholar).
4.Or maybe rather re-construct and reconfirm it, as the early Arab grammarians and philologists are said to have been dealing with this topic from as early the ninth century. (El-Hassan 1977El-Hassan, Shahir A. 1977 “Educated Spoken Arabic in Egypt and the Levant: A critical Review of Diglossia and Related Concepts.” Archivum Linguisticum 8 (2): 112–132.Google Scholar, 113)
5.Note the tension between "the specialization of function" which suggests a strict functional distribution of H and L (and which has been taken for granted in much of the literature on diglossia) on the one hand, and the reference to the settings in which H and L are "appropriate" on the other hand. The latter refers to the communicative norms related to the settings rather than to the actual language use in these settings. Moreover, the suggestion of a functional division of language use is confirmed by the examples that Ferguson gives.
6.For example, the South-Lebanese radio Ṣawt al-ğanūb broadcasted news bulletins in both fuṣḥā and Lebanese Arabic.
7.Ferguson himself was much aware of this: “A kind of spoken Arabic much used in certain semi-formal or cross-dialectal situations has a highly classical vocabulary with few or no inflectional endings, with certain features of classical syntax, but with a fundamentally colloquial base in morphology and syntax, and a generous admixture of colloquial vocabulary.” (Ferguson 1959Ferguson, Charles A. 1959 “Diglossia.” Word 15: 325–40. CrossrefGoogle Scholar, 332)
8.See also Bassiouney (2009Bassiouney, Reem 2009Arabic Sociolinguistics. Edinburgh: Edinburgh University Press. CrossrefGoogle Scholar, 14–16).
9.A similar remark has been made by Bassiouney concerning the attempts at defining intermediate levels: "[…] these intermediate levels cannot be understood unless one presupposes the existence of two 'poles', H and L. It may be that 'pure H' or 'pure L' does not occur very often, and that there are usually elements of both varieties in any stretch of normal speech, but still one has to consider a hypothetical pure H or L in order to presuppose that there are elements that occur from one or the other in a stretch of discourse." (Bassiouney 2009Bassiouney, Reem 2009Arabic Sociolinguistics. Edinburgh: Edinburgh University Press. CrossrefGoogle Scholar, 13)
10.Note that in fuṣḥā the possessive relation between the possessor and the possessed is expressed by means of flexion (e.g. the genitive for the possessor, which is the second part of the construction, hence the term ‘genitive construction’). In non-fuṣḥā varieties the construction is used without flexion.
11.Versteegh (2001) dismisses the idea of diglossic code-switching altogether arguing that the language choices take place on a continuum and therefore "these changes do not take the form of code-switching from one variety to another, but manifest themselves in a larger percentage of features from the opposite variety." (Versteegh 1997Versteegh, Kees 2001 (1997 1st print)The Arabic Language. Edinburgh: Edinburgh University Press.Google Scholar, 194) Note, however, that this approach suffers of what I called one-dimensionality. Hence, "the mixtures of variants" in the quote "[s]ince the colloquial and the standard language are not discrete varieties, but only abstract constructs at the extremes of a continuum, linguistic choice does not a two-way selection, but rather a mixture of variants." (Versteegh 1997Versteegh, Kees 2001 (1997 1st print)The Arabic Language. Edinburgh: Edinburgh University Press.Google Scholar, 195, emphasis mine) should also be interpreted only in terms of fuṣḥā and one non-fuṣḥā variety.
12.One could possibly argue that the use of the form with flexion is indication of its fuṣḥā-ness, but in the case of the opposite, the designation is harder because the pausal forms are also used in fuṣḥā.
13.See Mazraani (1997Mazraani, Nathalie 1997Aspects of Language Variation in Arabic Political Speech Making. Richmond Surrey: Curzon Press.Google Scholar, 33–38) and Bassiouney (2009Bassiouney, Reem 2009Arabic Sociolinguistics. Edinburgh: Edinburgh University Press. CrossrefGoogle Scholar, 43, 46–7) for a more detailed discussion with examples.
14.One seems not to be able to escape completely from Bassiouney’s (2009Bassiouney, Reem 2009Arabic Sociolinguistics. Edinburgh: Edinburgh University Press. CrossrefGoogle Scholar, 13) observation that at some point "hypothetical pure codes" need to be distinguished in order to proceed with analysis.
15.The use of the term 'standard' should not be confused with 'standard language', mainly because of their lack of any official recognition whatsoever.
16.Note that Abu-Haidar (1991)Abu-Haidar, Farida 1991Christian Arabic of Baghdad. Wiesbaden: Otto Harrasowitz.Google Scholar, a speaker of Christian Baghdadi Arabic herself, refers to and quotes both Ferguson (1959)Ferguson, Charles A. 1959 “Diglossia.” Word 15: 325–40. CrossrefGoogle Scholar and Blanc (1964) 1964Communal Dialects in Baghdad. Harvard Middle Eastern Monographs. Cambridge Massachusetts: Harvard University Press.Google Scholar to substantiate her claim (f.n. 9).
17.Why do you (fem. sing.) talk urban?
18.The languages spoken by linguistic minorities will not be taken into account neither, since their prestige is most often limited to being a marker for in group identity of the linguistic minority in question and as such they do not compete for linguistic prestige for speakers who don’t speak them natively.
19.Suleiman (2013 2013Arabic in the Fray. Language Ideology and Cultural Politics. Edinburgh: Edinburgh University Press. CrossrefGoogle Scholar, 266), partly quoting himself, defines folk linguistics as “the range of views and attitudes people have about their language, including its origin and the myths surrounding it that “allow us to come closer to the overt or covert orientations, assumptions, and hidden ideologies of the community and how these relate to linguistic repertoire.” (Suleiman 2008 2008 “Egypt: From Egyptian to Pan-Arab Nationalism.” In Language and National Identity in Africa, ed. by Andrew Simpson, 26–43. Oxford: Oxford University Press.Google Scholar, 28)”
20.“The common-sense (basic/normative) nature of ideological meaning is manifested in the fact that it is rarely questioned, in a given society or community, in discourse related to the ‘reality’ in question, possible across various discourse genres.” (Verschueren 2012Verschueren, Jef 2012Ideology in Language Use. Pragmatic Guidelines for Empirical Research. Cambridge: Cambridge University Press.Google Scholar, 12)
21.“Ideology, because of its normativity and common-sense nature, may be highly immune to experience and observation.” (Verschueren 2012Verschueren, Jef 2012Ideology in Language Use. Pragmatic Guidelines for Empirical Research. Cambridge: Cambridge University Press.Google Scholar, 14)
22.“Like paradigms, ideologies (if we can use the plural at all) are community-based. Their relevance, while going beyond the individual, does not extend beyond a given society or community […]. (Verschueren 2012Verschueren, Jef 2012Ideology in Language Use. Pragmatic Guidelines for Empirical Research. Cambridge: Cambridge University Press.Google Scholar, 11)
23.This tendency can be partly explained by the fact that sociolinguistics had to struggle for its place in the linguistics curriculum. (See Joseph 2004Joseph, Earl John 2004Language and Identity. National, Ethnic, Religious. Hampshire and New York: Palgrave Macmillan.Google Scholar, 60–1)
24.See Fergsuon: “[L]eaders in the community begin to call for unification of the language […] and tend to support either the adoption of H or of one form of L as the standard, less often the adoption of a modified H or L, a “mixed” variety of some kind.” (Ferguson 1959Ferguson, Charles A. 1959 “Diglossia.” Word 15: 325–40. CrossrefGoogle Scholar, 338)
25.Al-Muqtaṭaf was published in Beirut from 1876 until 1884 and then in Cairo from 1884 until its last issue appeared in 1951/1952. Its main editors were Yaʻqūb Ṣarrūf (1852–1927) and Fāris Nimr (1856–1952), both prolific intellectuals of Christian descent.
26.All translations from Arabic are mine.
27.Other manifestations of the debate are the sub-debate in Al-Muqtaṭaf in 1887–1888 (see Diem 1974Diem, Werner 1974Hochsprache und Dialekt im Arabischen: Untersuchungen zur heutigen arabische Zweisprachigkeit. Wiesbaden: Deutsche Morgenländische Gesellschaft.Google Scholar, 129–30, Daniëls 2002Daniëls, Helge 2002Debating Variability in Arabic: Fuṣḥā versus ʻāmmīya. Unpublished Ph.D. Antwerp: University of Antwerp.Google Scholar, 255–6), the publication of The spoken language of Egypt by Selden J. Wilmore in 1901 and the reactions and counter-reactions it triggered in Al-Muqtaṭaf and other prominent cultural journals such as Al-Hilāl and Al-Manār (see Diem 1974Diem, Werner 1974Hochsprache und Dialekt im Arabischen: Untersuchungen zur heutigen arabische Zweisprachigkeit. Wiesbaden: Deutsche Morgenländische Gesellschaft.Google Scholar, 132–3, Daniëls 2002Daniëls, Helge 2002Debating Variability in Arabic: Fuṣḥā versus ʻāmmīya. Unpublished Ph.D. Antwerp: University of Antwerp.Google Scholar, 257–80, Suleiman 2004Suleiman, Yasir 2004A War of Words. Language and Conflict in the Middle East. Cambridge: Cambridge University Press. CrossrefGoogle Scholar, 68–71), the articles that appeared in 1910 in Al-Muqtabas, yet another leading journal at the beginning of the 20th century, the linguistic proposals by Aḥmad Luṭfī al-Sayyid in the 1920s and 30s (see Daniëls 2002Daniëls, Helge 2002Debating Variability in Arabic: Fuṣḥā versus ʻāmmīya. Unpublished Ph.D. Antwerp: University of Antwerp.Google Scholar, 284–8) and by Anīs Frayḥa in 1938Frayḥa, Anīs 1938 “Al-ʻāmmīya wa al-fuṣḥā. ʻawd ilā al-mawḍūʻ [ʻāmmīya and fuṣḥā. Back tot he subject].” Al-muqtaṭaf 93: 292–8.Google Scholar and 1955 (see Daniëls 2002Daniëls, Helge 2002Debating Variability in Arabic: Fuṣḥā versus ʻāmmīya. Unpublished Ph.D. Antwerp: University of Antwerp.Google Scholar, 289–301, Suleiman 2004Suleiman, Yasir 2004A War of Words. Language and Conflict in the Middle East. Cambridge: Cambridge University Press. CrossrefGoogle Scholar). Needless to say that this list is not exhaustive.
28.For a more elaborate discussion, see Ayalon (1987)Ayalon, Ami 1987Language and Change in the Arab Middle East: The Evolution of Modern Political Discourse. Oxford: Oxford University Press.Google Scholar, Versteegh & Schippers (1987, 146–6), Versteegh (2001, 173–88).
29.However, Ṣarrūf and Nimr refer also to the fact that the science books are written in "a language different from the language we speak" (luġa ġayr al-luġa allatī natakallamuhā) (Al-Muqtaṭaf 1881Ṣarrūf, Yaʿqūb and Fāris Nimr 1881 “Al-luġa al-ʿarabīya wa al-nağāḥ [The Arabic language and success].” Al-Muqtaṭaf 6 (November): 352–4.Google Scholar, 353)
30.See Parkinson (1991)Parkinson, Dilworth 1991 “Searching for Modern Fuṣḥa: Real-life Formal Arabic.” ʻArabiyya. Journal of the American Association of Teachers of Arabic 24: 31–64.Google Scholar for an interesting discussion of variability within fuṣḥā and how fuṣḥā means different things for different native speakers of Arabic in Egypt.
31.This can be related to the Arab views on the development of the post-Islamic Arabic dialects
32.For a discussion of the ideological implications of the concept of natural linguistic dispositions or “pristine linguistic intuitions” (fiṭra luġawīya or salīqa luġawīya), see Suleiman (2013 2013Arabic in the Fray. Language Ideology and Cultural Politics. Edinburgh: Edinburgh University Press. CrossrefGoogle Scholar, 53–4).
33.It is important to note that the term 'lahğa (pl. lahağāt)' in the sense of 'dialect' or 'variety' is also used in the Muqtaṭaf debate, however only in order to refer to specific dialects, e.g. "the dialects of the Syrians, the Egyptians, the Iraqi’s and the Maghribians (lahağāt al-sūrīyīn wa al-ʻirāqīyīn wa al-miṣrīyīn wa al-maġribīyīn) (Al-Mumkin 1882a, 494), "the dialect of which province […], city, […] village, […] neighboorhood" (lahğat ayyati muqātaʻa […], madīna […], qariya […], ḥāra." (Dāġir 1882Dāġir, Asʿad 1882 “Istiḥālat al-Mumkin idā amkana [The impossibility of al-Mumkin if it were possible].” Al-Muqtaṭaf 6 (February): 556–60.Google Scholar, 557). Sometimes the term is also used in the sense of 'way of speaking' or 'pronunciation'.
34.Interestingly enough, Rabin uses the term post-Islamic dialects in reference to what I call the ʻāmmīya or non-fuṣḥā varieties. This can be related to the fact that traditionally pre-Islamic and post-Islamic language variability has been perceived differently. This is mainly related to the fact that, basically, pre-Islamic variants (luġāt) were considered to remain within the realm of correctness, even when deviating from the norm (naḥw), whereas deviations from the norm related to post-Islamic varieties were considered to be mistakes (laḥn). This can be derived to the ways in which the terms luġa, naḥw and laḥn diachronologically evolved. This process of semantic shifts is intrinsically related to the evolution of the concept of faṣāḥa. For a more elaborate discussion see Ayyoub (2011).
35.In Arabic: “ hilāf al-luġa al-rasmīya aw al-adabīya aw al-fuṣḥā
36.In Arabic: “kalām al-nās al-ʿādī, hilāl [sic] al-luġa al-fuṣḥā aw al-luġa al-adabīya
37.In Arabic: “luġat al-ʿāmma wa hiya hilāf al-fuṣḥā
38.In Arabic: “mā naṭaqa bi-hi al-ʿāmma alā ġayr sunan al-kalām al-ʿarabī
39. Nağāḥ al-umma al-ʻarabīya fī luġatihā al-aṣlīya (Qandalaft 1882Qandalaft, Mitrī 1882 “Nağāḥ al-’umma al-‘arabīya fī luġatihā al-‘aṣlīya [The success of the Arab nation in its authentic language].” Al-Muqtaṭaf 7 (July): 107–10.Google Scholar, 107–110).
40.Only a couple of times faṣīḥā is defined by contrasting it with ʻāmmīya. Ṣarrūf and Nimr refer also to the fact that the science books are written in "a language different from the language we speak" (luġat ġayr al-luġa allatī natakallamuhā) (Al-Muqtaṭaf 1881Ṣarrūf, Yaʿqūb and Fāris Nimr 1881 “Al-luġa al-ʿarabīya wa al-nağāḥ [The Arabic language and success].” Al-Muqtaṭaf 6 (November): 352–4.Google Scholar, 353) and al-Yāziğī refers to faṣīḥa and explains "in the sense that it is not part of the folk language" (bi-maʻnā annahā laysat min luġat al-ʻāmma). (Al-Yāziğī 1881Al-Yāziğī, Halīl 1881 “Al-luġa al-ʿarabīya wa al-nağāḥ [The Arabic language and success)].” Al-Muqtaṭaf 6 (December): 404–5.Google Scholar, 305)

References

Abu-Haidar, Farida
1991Christian Arabic of Baghdad. Wiesbaden: Otto Harrasowitz.Google Scholar
Al-Bustānī, Buṭrus
1998 (1st print 1870). Muḥīṭ al-muḥīṭ. Qāmūs muṭawwal li-al-luġa al-ʿarabīya [Muhit al-muhit. An Arabic-Arabic Dictionary]. Beirut: Librairie du Liban Publishers.Google Scholar
al-Qāsimī, ʻalī
(ed.) 2003Al-muʿğam al-ʿarabī al-asāsī li-al-nāṭiqīn bi-al-ʿarabīya wa mutʿallimīhā [The Elementary Arabic Dictionary for Native Speakers and Learners]. Cairo: ALECSO/Larouse.Google Scholar
Ayalon, Ami
1987Language and Change in the Arab Middle East: The Evolution of Modern Political Discourse. Oxford: Oxford University Press.Google Scholar
Ayoub, Georgine
2011 “Faṣīḥ.” Encyclopedia of Arabic Language and Linguistics. Managing Editors Online Edition: Lutz Edzard; Rudolf de Jong. Brill 2011 Brill Online. University of Cambridge – Cambridge University Library (UK). 01 December 2011 http://​www​.brillonline​.nl​/subscriber​/entry​?entry​=eall​_COM​-vol2​-0014
Badawi, al-Saʻīd
1973Mustawayāt al-ʻarabīya al-muʻāṣira fi miṣr. Baḥt fī ʻalāqat al-luġa bi-al-haḍāra [The Levels of Contemporary Arabic in Egypt. Research on the Relation between Language and Culture]. Caïro: Dār al-Maʻārif.Google Scholar
Bassiouney, Reem
2009Arabic Sociolinguistics. Edinburgh: Edinburgh University Press. CrossrefGoogle Scholar
Blanc, Haim
1960 “Style Variations in Spoken Arabic – A Sample of Interdialectal Educated Conversation.” In Contributions to Arabic Linguistics. Harvard Middle Eastern Monographs III, ed. by C. A. Ferguson, 81–156. Cambridge Massassuchets: Harvard University Press.Google Scholar
1964Communal Dialects in Baghdad. Harvard Middle Eastern Monographs. Cambridge Massachusetts: Harvard University Press.Google Scholar
Boussofara-Omar, Naima
2011 “Diglossia.” Encyclopedia of Arabic Language and Linguistics. Managing Editors Online Edition: Lutz Edzard; Rudolf de Jong. Brill 2011 Brill Online. K.U. Leuven – University Library. 05 April 2011 http://​www​.brillonline​.nl​/subscriber​/entry​?entry​=eall​_COM​-0090
Caton, Steven
1991 “Diglossia in North Yemen: A Case of Competing Linguistic Communities.” Southwest Journal of Linguistics 10 (1): 143–59.Google Scholar
Daniëls, Helge
2002Debating Variability in Arabic: Fuṣḥā versus ʻāmmīya. Unpublished Ph.D. Antwerp: University of Antwerp.Google Scholar
Diem, Werner
1974Hochsprache und Dialekt im Arabischen: Untersuchungen zur heutigen arabische Zweisprachigkeit. Wiesbaden: Deutsche Morgenländische Gesellschaft.Google Scholar
Eid, Mushira
2002 “Language is a Choice – Variations in Egyptian Women’s Written Discourse.” In Language Contact and Language Conflict in Arabic – Variations on a Sociolinguistic Theme, ed. by A. Rouchdy, 203–232. London: Routledge-Curzon.Google Scholar
El-Hassan, Shahir A.
1977 “Educated Spoken Arabic in Egypt and the Levant: A critical Review of Diglossia and Related Concepts.” Archivum Linguisticum 8 (2): 112–132.Google Scholar
Ferguson, Charles A.
1959 “Diglossia.” Word 15: 325–40. CrossrefGoogle Scholar
Frayḥa, Anīs
1938 “Al-ʻāmmīya wa al-fuṣḥā. ʻawd ilā al-mawḍūʻ [ʻāmmīya and fuṣḥā. Back tot he subject].” Al-muqtaṭaf 93: 292–8.Google Scholar
Gal, Susan, and Kathryn Woolard
1995 “Constructing Languages and Publics: Authority and Representation.” Pragmatics 5 (2): 129–138. CrossrefGoogle Scholar
Gumperz, John Joseph
1982Discourse strategies. Cambridge: Cambridge University Press. CrossrefGoogle Scholar
Haeri, Niloofar
1996The Sociolinguistic Market of Cairo: Gender, Class and Education. London and New York: Kegan Paul International.Google Scholar
2000 “Form and Ideology: Arabic Sociolinguistics and Beyond.” Annual Review of Anthropology 29: 61–87. CrossrefGoogle Scholar
Holes, Clive
1986 “The Social Motivation for Phonological Convergence in Three Arabic Dialects.” International Journal of Sociolinguistics 61: 33–51.Google Scholar
1993 “The uses of Variation: A Study of the Political Speeches of Gamal Abd Al-Nasir.” Perspectives on Arabic Linguistics V, ed. by M. Eid, and C. Holes, 13–46. Amsterdam: John Benjamins Publishing Company. CrossrefGoogle Scholar
1995 “Community, Dialect and Urbanization in the Arabic-speaking Middle East.” Bulletin of the School of Oriental and African Studies 58/2: 270–287. CrossrefGoogle Scholar
Joseph, Earl John
2004Language and Identity. National, Ethnic, Religious. Hampshire and New York: Palgrave Macmillan.Google Scholar
Kazimirski, A. De Biberstein
1860Dictionnaire Arabe – Français (2 tomes). Beirut: Librairie du Liban.Google Scholar
Labov, William
1971 “The notion of “system” in creole studies.” In Pidginization and creolization of languages, ed. by D. Hymes, 447–72. Cambridge : Cambridge University Press.Google Scholar
Mackey, W. F.
1993 “Introduction.” In Diglossia: A Comprehensive Bibliography 1960-1990, ed. by M. Fernandez, xiii–xx. Amsterdam: John Benjamins Publishing Company.Google Scholar
Marçais, William
1930 “La diglossie arabe.” L’ensiegnement Public 97: 401–9.Google Scholar
Mahmoud, Youssef
1986 “Arabic after Diglossia.” In The Fergusonian Impact: In Honour of Charles A. Ferguson on the Occasion of his 65th Birthday, ed. by Joshua Fishman, Andree Tabouret-Keller, Michael Clyne, Bh. Krishnamurti, and Mohamed Abdulaziz, 239–51. Berlin: Mouton de Gruyter.Google Scholar
Mazraani, Nathalie
1997Aspects of Language Variation in Arabic Political Speech Making. Richmond Surrey: Curzon Press.Google Scholar
Meiseles, Gustav
1980 “Educated Spoken Language and the Arabic Language.” Archivum Linguisticum XI, New Series 118–148.Google Scholar
Mitchell, Terence Frederic
1978 “Educated Spoken Arabic in Egypt and the Levant, with Special Reference to Participle and Tense.” Journal of Linguistics 14/2: 227–58. CrossrefGoogle Scholar
1980 “Dimensions of Style in a Grammar of Educated Spoken Arabic.” Archivum Linguisticum XI: 89–106.Google Scholar
1982 “More than a Matter of ‘Writing with the Learned, Pronouncing with the Vulgar.” Standard Languages, ed. by W. Haas, 123–55. Manchester: Manchester University Press.Google Scholar
1986What is Educated Spoken Arabic?” In Aspects of Arabic Sociolinguistics, ed. by M. H. Ibrahim, and B. H. Jernudd, special volume of The International Journal of the Sociology of Language 61: 7–32.Google Scholar
Mitchell, Terence Frederic, and S. A. al-Hassan
1994Modality, Mood and Aspect in Spoken Arabic. With special reference to Egypt and the Levant. London and New York: Kegan Paul International.Google Scholar
Muṣṭafā, Ibrāhīm
(ed.) 1980Al-muʿğam al-wasīṭ [The Intermediary Dictionary]. Istanbul: Dār al-daʻwa.Google Scholar
Parkinson, Dilworth
1991 “Searching for Modern Fuṣḥa: Real-life Formal Arabic.” ʻArabiyya. Journal of the American Association of Teachers of Arabic 24: 31–64.Google Scholar
Sallam, A. M.
1979 “Concordial Relation within the Noun Phrase in Educated Spoken Arabic (ESA).” Archivum Linguisticum (new series) X (1): 20–56.Google Scholar
1980 “Phonological Variation in Educated Spoken Arabic: A Study of the Uvular and Related Plosive Types.” Bulletin of the School of Oriental and African Studies, XLIII 43: 1.Google Scholar
Schippers, Arie, and Kees Versteegh
1987Het Arabisch. Norm en realiteit. Muiderberg: Coutinho.Google Scholar
Suleiman, Yasir
2004A War of Words. Language and Conflict in the Middle East. Cambridge: Cambridge University Press. CrossrefGoogle Scholar
2008 “Egypt: From Egyptian to Pan-Arab Nationalism.” In Language and National Identity in Africa, ed. by Andrew Simpson, 26–43. Oxford: Oxford University Press.Google Scholar
2011Arabic, Self and Identity. A Study in Conflict and Displacement. Oxford: Oxford University Press. CrossrefGoogle Scholar
2013Arabic in the Fray. Language Ideology and Cultural Politics. Edinburgh: Edinburgh University Press. CrossrefGoogle Scholar
Verschueren, Jef
2012Ideology in Language Use. Pragmatic Guidelines for Empirical Research. Cambridge: Cambridge University Press.Google Scholar
Versteegh, Kees
2001 (1997 1st print)The Arabic Language. Edinburgh: Edinburgh University Press.Google Scholar
Walters, Keith
1991 “Women, Men, and Linguistic Variation in the Arab World.” In Perspectives on Arabic Linguistics III Papers from the third annual symposium on Arabic Linguistics, ed. By Mushira Eid, and B. Comrie, 199–229. Amsterdam: John Benjamins Publishing Company.Google Scholar
1996 “Diglossia, Linguistic Variation, and Language Change in Arabic.” In Perspectives on Arabic Linguistics VIII, ed. by Mushira Eid, 157–97. Amsterdam: John Benjamins Publishing Company. CrossrefGoogle Scholar
2013 “Language Attitudes.” In Encyclopedia of Arabic Language and Linguistics. Managing Editors Online Edition: Lutz Edzard, Rudolf de Jong. Brill Online 2013 Reference. K.U. Leuven – University Library. 19 November 2011 http://​referenceworks​.brillonline​.com​/entries​/encyclopedia​-of​-arabic​-language​-and​-linguistics​/language​-attitudes​-COM​_vol2​_0084
Wehr, Hans, and J. Milton Cowan
(ed.) 1994 (1979 1st print)A Dictionary of Modern Written Arabic: Arabic-English. Wiesbaden: Otto Harrasowitz.Google Scholar

The corpus (in order of appearance)

Ṣarrūf, Yaʿqūb and Fāris Nimr
1881 “Al-luġa al-ʿarabīya wa al-nağāḥ [The Arabic language and success].” Al-Muqtaṭaf 6 (November): 352–4.Google Scholar
Al-Yāziğī, Halīl
1881 “Al-luġa al-ʿarabīya wa al-nağāḥ [The Arabic language and success)].” Al-Muqtaṭaf 6 (December): 404–5.Google Scholar
al-Mumkin
1882 “Mustaqbal al-luġa al-ʿarabīya [The future of the Arabic language].” Al-Muqtaṭaf 6 (January): 494–6.Google Scholar
al-Ğamʿīya al-adabīya al-dimašqīya
1882 “Nağāḥ al-umma al-ʿarabīya fī luġatihā al-aṣlīya [The success of the Arab nation in its authentic language].” Al-Muqtaṭaf 6 (February): 551–6.Google Scholar
Dāġir, Asʿad
1882 “Istiḥālat al-Mumkin idā amkana [The impossibility of al-Mumkin if it were possible].” Al-Muqtaṭaf 6 (February): 556–60.Google Scholar
al-Mumkin
1882 “Mustaqbal al-luġa al-ʿarabīya. Nağāḥ al-umma al-ʿarabīya fī luġatihā al-aṣlīya [The future of the Arabic language. The success of the Arab nation in its authentic language]. ” Al-Muqtaṭaf 6: 618–21.Google Scholar
Ḥ.Ḥ.
1882 “Kašf al-ğaṭā ʿammā fī kalām al-Mumkin min al-haṭā [Uncovering the mistakes in al-Mumkins words].” Al-Muqtaṭaf 6 (April): 690–4.Google Scholar
Qandalaft, Mitrī
1882 “Nağāḥ al-’umma al-‘arabīya fī luġatihā al-‘aṣlīya [The success of the Arab nation in its authentic language].” Al-Muqtaṭaf 6 (April): 694–6.Google Scholar
al-Ğamʿīya al-adabīya al-dimašqīya
1882 “Nağāḥ al-umma al-ʿarabīya fī luġatihā al-aṣlīya [The success of the Arab nation in its authentic language].” Al-Muqtaṭaf 6 (April): 697.Google Scholar
al-Mumkin
1882 “Mustaqbal al-luġa al-ʿarabīya [The future of the Arabic language].” Al-Muqtaṭaf 7 (June): 42–4.Google Scholar
Qandalaft, Mitrī
1882 “Nağāḥ al-’umma al-‘arabīya fī luġatihā al-‘aṣlīya [The success of the Arab nation in its authentic language].” Al-Muqtaṭaf 7 (July): 107–10.Google Scholar

Address for correspondence

Helge Daniëls

University of Leuven

helge.daniels@arts.kuleuven.be