Grammarians and Diglossia
p. 295-318
Résumés
Collecting linguistic data in a diglossic speech community poses a special challenge for grammarians because of the socially sensitive variation between a High and a Low variety. In Ferguson’s (1959a) original definition only those cases counted as diglossia in which the varieties belonged to the same language. Later definitions expanded the scope to include communities where non-cognate languages serve as High and Low variety, which are much easier to keep apart for both speakers and grammarians. In this paper I explore how Latin and Arabic grammarians dealt with the linguistic variation in their speech community. They regarded all variation as part of one language; yet, in their description of the language they limited themselves to data from the High variety, which was the only one recognized by them as correct speech. Even when they referred to speech errors, they did not target the Low colloquial, but mistakes made in writing by semi-literate users of the High variety.
La collecte de données linguistiques par des grammairiens opérant dans une société diglossique est rendue compliquée par la variation sociolinguistique entre les deux variétés. Selon la définition initiale de Ferguson (1959a), la notion de « diglossie » n’est applicable que dans le cas où les deux variétés, « haute » et « basse », appartiennent à une même langue. D’autres chercheurs élargirent cette définition pour inclure des cas de diglossie de langues non-apparentées, plus aisément distinguables pour les locuteurs et les grammairiens. Cet article explore la façon dont les grammairiens latins et arabes ont abordé la variation linguistique dans leurs sociétés respectives. Cette variation leur est apparue comme relevant d’un seul système linguistique ; ils se sont cependant limités à la collecte de données issues de la variété « haute », la seule qu’ils reconnaissaient comme correcte. Même lorsqu’ils traitent des fautes de langage, ils ne visent pas la variété « basse », mais les erreurs des locuteurs semi-alphabétisés qui tentent d’écrire la variété « haute ».
Entrées d’index
Mots-clés : diglossie, pseudocorrection, grammairiens latins, grammairiens arabes, variation linguistique
Keywords : diglossia, hypercorrection, Latin grammarians, Arabic grammarians, language variation
Texte intégral
1. Grammarians and Diglossia
1Grammarians of Latin and Arabic have in common that they were active in a speech community exhibiting a specific type of sociolinguistic variation with a sharp distinction between two linguistic varieties, a Low variety acquired as a native language and associated with family and informality, and a High variety learnt at school and associated with education and a literary and religious heritage. In a classic paper, Charles Ferguson referred to this type of variation as “diglossia”, which he defines as follows:1
Diglossia is a relatively stable language situation in which, in addition to the primary dialects of the language (which may include a standard or regional standards), there is a very divergent, highly codified (often grammatically more complex) superposed variety, the vehicle of a large and respected body of written literature, either of an earlier period or in another speech community, which is learned largely by formal education and is used for most written and formal spoken purposes but is not used by any section of the community for ordinary conversation (1959a: 435).
2In Ferguson’s treatment of the concept there is a difference between diglossia and what he calls the situation of a standard-with-dialects: in a diglossia no-one uses the High variety informally, since doing so would sound pedantic and would immediately create a distance between speaker and listener. In a standard-with-dialects context, too, using the standard variety in informal contexts can be inappropriate, but at least for some people the standard language is the normal means of communication. Ferguson was certainly right that in a community like the Arabic-speaking world no-one regularly uses the High variety (fuṣḥā) in an informal context, so that it falls squarely within the scope of his definition of diglossia.
3When he revisited the notion of diglossia, Ferguson (1991) explicitly referred to the post-creole continuum in Jamaica, stating that in this situation, too, some speakers regularly use the High variety – which often happens to be the former colonizers’ language – informally, so that the boundary between the two varieties is fluid, while in diglossia the two varieties are strictly separated. Yet, studies such as Mejdell (2006) have provided ample illustration of the occurrence of just as intricate mixed styles in a “classic” diglossia situation, many speakers in Egypt regularly using elements from the High variety in their informal speech (see Boussofara-Omar 2006: 632–635).
4The present paper compares the Latin and Arabic grammarians’ approach to the collection of linguistic data. Both languages served as the High variety in a diglossia context along with spoken Low varieties. The grammarians focused on the High variety exclusively, treating it as a set of grammatical rules within a closed corpus, while ignoring the spoken language entirely. I shall first deal with the historical context in which Latin and Arabic became imperial languages (2), and then separately with the approach of the Latin (3) and the Arabic (4) grammarians. Finally, the comparison between them will bring us back to the notion of “diglossia” and its impact on speakers’ and grammarians’ attitudes (5).
2. Conquests and Empires
5The similarity in approach of the Latin and Arabic grammarians is connected with the linguistic history of the two languages. When their original speech communities expanded dramatically and turned into an empire in which large numbers of new language learners were integrated, grammarians were faced with the task of distinguishing between the language of native speakers and that of new learners.
6It took the Romans quite some time to assemble a large empire, roughly between the third century BCE and the second century CE, but the conquest of each individual province took place rapidly, at most within a few decades. Though they did not impose any formal language policy, it was fairly obvious to anyone wishing to participate in the new order that they had better learn the rulers’ language. Part of Roman culture was a dense network of formal education, but this was not introduced on a large scale in the provinces of the empire, so that only the elite became its beneficiaries. Since Romans formed a minority in the provinces, this local elite played a special role within the grand scheme of Roman expansion.2 For them, education meant a thorough immersion in Roman culture and values, including the language, which came to serve as a powerful means of self-identification, even after the fall of the Western Roman empire in 476 CE.3 Beyond this tiny elite, however, the impact of the educational network was minimal. For the vast majority of the population, acquisition of Latin took place in an unmonitored and imperfect way, leading to the emergence of contact varieties of Latin all over the empire.
7The earliest motivation to study the Latin language from the second century BCE onwards came from an antiquarian interest when scholars started to occupy themselves with the explanation of old texts, both literary and legal, which threatened to become incomprehensible for new generations.4 Early grammar study consisted above all in commentaries on these texts with an emphasis on lexicographical explanation (Zetzel 2018: 15–30). “Real” grammar started with a surge in Greek influence in the first century BCE.5 Zetzel (2018: 24) emphasizes the mythical nature of those stories that attribute every literary or scientific development in Rome to Greek influence, but it is clear that Greek traditions did affect the way grammatical study developed in Rome, starting with Varro (c. 116–c. 27 BCE), as a natural side effect of the prestige Greek culture and language enjoyed in Rome (Valette-Cagnac 2005).
8The Arabo-Islamic empire was founded within a few decades in the seventh century CE, when the Arabs conquered a large territory from Islamic Spain to Central Asia. During the initial period of the conquests, there was no official language policy promoting the use of Arabic. But the simple fact that the new rulers’ language was Arabic and that these rulers did not intend to learn their subjects’ languages explains why, even without being forced to do so, the new subjects soon started to learn Arabic. With respect to education, the situation in the Islamic empire differed from that in the Roman empire, since the Arabs did not possess an educational system of their own, ready to be implemented in the recently conquered territories. Quite the contrary, they themselves were suddenly exposed to cultures in which schools were a normal constituent of civilization. Like the Roman colonists, the Arabs constituted a minority in these territories, but in addition they were dependent on the local elite for the establishment of an educational system, rather than the other way round. Beyond the basic skills of reading and writing, this educational system reached only a select few from the indigenous population. For most people, Arabic was a language they learnt in the street and in sparse contacts with native speakers, which led to the emergence of numerous contact varieties.
9The study of Arabic grammar developed soon after the period of the conquests (Versteegh 1983; Talmon 1987, 2003). Early grammarians in the newly founded cities of Kufa and Basra were motivated by two factors in particular: an antiquarian interest and a religious need. The antiquarian interest was connected with the social changes brought about by the conquests. The new urban culture that flourished in Damascus and later on in Baghdad was tinged with elements of nostalgia for the period when Arabs were still living an (idealized) desert life. The Bedouin way of life, including their language and poetry, became an object of admiration and longing. Grammarians vied with each other in showing off their knowledge of obsolete Bedouin words and lines of poetry. Linguistic data collection was intimately connected with the transmission of poetry.
10The second motive for engaging in the study of Arabic was connected with the recitation and exegesis of the Qurʾān. The new colloquial language of each region diverged sharply from the language of the Qurʾān and the administration. This created a need for grammatical instruction for new converts and for anyone wishing to be involved in administration of the new society. Initially, a considerable proportion of those who taught and transmitted the language consisted of mawālī, new converts, who had had to learn the new language themselves (Bernards 2005). In many cases, they brought with them expertise from a foreign culture, be it Persian or Syriac. The first complete work on Arabic was in fact written by a Persian, Sībawayhi (d. c. 793 CE). After a few generations, however, the difference between ethnic Arabs and converts from other peoples tended to become blurred.
3. Latin Grammarians and Language Use
11Latin grammarians were concerned exclusively with the standard language of culture, religion, and administration. In modern times the purpose of language standardization is to achieve a minimal amount of variation, as Clackson (2015) formulates it, referring to Haugen’s (1966: 931) definition of the term (see Fig. 1).
Fig. 1. Standardization of language (after Haugen 1966)
form | function | |
society | selection | acceptance |
language | codification (minimal variation in form) | elaboration (maximal variation in function) |
12Clackson is certainly right when he states that there is a difference between the modern concept of a standard language and the Ancients’ views on correctness of speech. In the Greco-Latin world, the linguistic norm did not exclude variant forms. In Greek grammatical literature, for instance, the use of different dialects for different literary genres was generally accepted and, at least since the first century BCE, dialect forms were accepted as correct Greek, without invalidating the general rules of an imagined standard language (Versteegh 1987: 156–158; Hintzen 2011). In Latin literature, this kind of dialect diversification did not exist, but Latin grammarians did allow for archaisms from earlier periods of the literary language. This inclusion of variation may not be compatible with modern notions of a standard language, but the existence of a standard language in Antiquity should not be rejected out of hand just because the restrictions in standardization differed from those in modern national languages.
13Within the accepted bandwidth of variation, grammarians sought criteria for the correctness of speech. In Greece, the criteria for correct Greek (hellēnismos) had been set up as early as the time of the grammarian Trypho (first century BCE), who may have been responsible for the introduction of the canonical four criteria: “analogy” (analogia), “dialect” (dialektos), “etymology” (etumologia), “literary authority” (historia).6 The criteria were part of a discourse about the “virtues and vices of speech” (aretai kai kakiai logou), in which the “vices” were categorized as barbarismos or soloikismos (Sandri 2020: 15-27).
14This categorization was taken over at an early date by Latin grammarians (Siebenborn 1976: 53–55). Varro listed four criteria for correct Latin (latinitas): natura, analogia, consuetudo, auctoritas,7 which suggests an opposition between reason (ratio) and language use (usus). The former comprised natura and analogia, while the latter dealt with language use, understood, not as the language of the common people, but that of the acknowledged literary sources (auctoritas) and the literate elite (consuetudo). The virtues of speech are represented by latinitas, which came to denote “correctness of speech” tout court, whose custodians were the grammarians (Kaster 1988).
15Varro condemns the language use of the common people as a consuetudo mala and insists that the way well-informed speakers speak reflects their rational insight into grammar (Siebenborn 1976: 96f.): “Thus, when people call upon us to follow their way of speaking, we’ll follow them if it is the correct way, because there is analogy in it”.8 This means that there is no real conflict between consuetudo and ratio, provided one follows the way of speaking of those who speak correctly, thus ensuring that one’s language will be naturally correct (Zetzel 2018: 47–49). All of this has nothing to do with actually spoken language, which by its nature is disorderly and incorrect. When Cicero apologizes for using a sermo plebeius in some of his letters, he is referring to his own informal use of the language, not to the vulgar Latin of the plebs.9
16Within the Roman empire at large, local realizations of Latin must have differed considerably (Adams 2007). Rustic forms of the language were far less prestigious than the urban language, and provincial varieties did not have much chance of getting accepted as correct language. Chahoud (2007: 44) gives several examples of the condescending way in which Roman intellectuals spoke about provincial realizations of Latin. Cicero (De orat. 3.44) says that one should avoid at all costs the rustica asperitas and the peregrina insolentia, if one wishes to uphold the standard of urbanitas.10 But the rejected deviations from the urban standard are not identical with vernacular forms of Latin within the Roman empire. All over the provinces, and even in Rome itself, contact varieties of Latin ultimately developed into the vernacular languages of the inhabitants of the empire, constituting the Low variety of a diglossia in which Latin served as the High variety.
17Some of the treatises of the post-Classical period address common mistakes in language, but these represent mistakes made by semi-literate people wishing to pose as members of the elite.11 In one of Catullus’ satirical poems, a certain Arrius is mocked for using hypercorrect h: “Arrius said chommoda when he wished to say commoda ‘benefits’/and hinsidias for insidias ‘ambush’”.12 A similar example, arundo non harundo (“reed”, GL IV, 198.33), is found in the so-called Appendix Probi, falsely attributed to Marcus Valerius Probus, a grammarian from the first century CE, but probably dating from the third/fourth century CE.13 Another example of hypercorrection in the Appendix is that of Herculens instead of Hercules (GL IV, 197.25).14 Some errors in the Appendix concern simple spelling mistakes, e.g. virgo non vyrgo (“virgin”, GL IV, 198.20) or ostium non osteum (“front door; river mouth”, GL IV, 198.5), or colloquial forms, e.g. calida non calda (“warm”, GL IV, 198.3) or auris non oricla (“ear”, GL IV, 198.11). But like the hypercorrections, many errors represent misguided attempts by semi-literate speakers to impress their audience, e.g. ermeneumata non erminomata (“interpretations”, GL IV, 199.6) or parentalia non parantalia (“festival for the family dead”, GL IV, 199.4). Quirk (2017) emphasizes that, unlike other “mistakes” in the Appendix, hypercorrections were not taken over in the Romance languages, underscoring the fact that these treatises targeted people trying to write above their station, rather than speakers of vernacular varieties.
4. Arabic Grammarians and Correctness of Speech
18At the beginning of the Arabic tradition, no canon of formal criteria for correct language existed. The grammarians did, however, discuss the relative value of the linguistic sources for establishing the rules of the ʿarabiyya, since they needed to determine which data could be included in their corpus, and which could not. In Sībawayhi’s Kitāb the three main sources are: the speech of the Arabs (kalām al-ʿArab), poetry, and the text of the Qurʾān, in that order.15 The kalām al-ʿArab is represented by anything quoted under headings like “they say” or “the Arabs say”. Sībawayhi opposes the excessive reliance on analogical reasoning (qiyās) by some of the grammarians (naḥwiyyūna) and leaves no doubt that in the end it is the speaker who decides which forms are correct and which are not, by either using them or rejecting them.16 In this discussion the term “speakers” refers to idealized speakers, whose use of the language corresponds to the way grammarians (aspire to) speak (Baalbaki 2007).
19One of the most difficult issues for the grammarians was that of validating those forms in the sources that deviated from the rules of the ʿarabiyya as these had been set up by them. Deviating forms were called šāḏḏ, a term sometimes translated as “irregular”, although this does not quite catch the status of such forms. They belong without question to the ʿarabiyya and can be explained rationally by analogical reasoning (qiyās), but they differ from other forms in that they cannot be used as the point of departure for a grammatical analogy. Thus, certain forms used by one or the other pre-Islamic tribe could be šāḏḏ, while at the same time being recognized by the grammarians as correct Arabic. Even forms occurring in the Qurʾān could be šāḏḏ.
20The sometimes bewildering variation of forms in the pre-Islamic dialects of the Arabian tribes (luġāt) was accepted as part of the corpus of correct Arabic. Larcher (2018) confirms Rabin’s (1951) conclusion that there was a basic dichotomy between “Tamīmī” forms, roughly stemming from tribes in eastern Arabia, and Ḥijāzī forms from western Arabia, including the dialect of Qurayš, Muḥammad’s tribe. The language of the Qurʾān seems to be largely identical with the language used in poetry, which was close to the language of the East-Arabian tribes (Macdonald 2010: 8). Nevertheless, Sībawayhi (Kitāb III, 278; IV, 473, 482) calls the language of the Ḥijāz al-luġa al-ʿarabiyya al-qadīma al-jayyida “the good old Arabic language”, but this appellation is ideologically determined: later scholars believed that the language of the Prophet, and by extension that of the tribe of Qurayš, was the most excellent of all forms of Arabic and must therefore be identical with the language of the Qurʾān.17 Sibileau (2021: 619, n. 103) remarks that the language of the Qurʾān was “classicized” by later scholars, but it remains unclear why they sometimes selected forms for their description of the language that clearly did not belong to the language of Qurayš. In the case of the realization of the glottal stop (hamza), for instance, some reports stress the fact that the people of Mecca, including the Prophet, did not pronounce it. Yet, it was commonly accepted that the Qurʾān had been revealed with the glottal stop (Rabin 1951: 144f.).
21Some Arabic grammarians may have questioned the value of the information they received, but when Ibn Jinnī (d. 1002) devotes a short chapter to this criticism, Bāb fī ṣidq al-naqala wa-ṯiqat al-ruwāt wa-l-ḥamala (Ḫaṣāʾiṣ III, 309-313; Sibileau 2021: 224f.), he sings the praise of the early transmitters of language and poetry: their veracity was such, he claims, that they virtually never transmitted anything untrustworthy. It was not until much later that grammarians started to deal systematically with the criteria for correct language, possibly inspired by similar discussions in legal science. In his Lumaʿ al-ʾadilla, Ibn al-ʾAnbārī (d. 1181) mentions the two main criteria of analogy (qiyās) and transmission (naql). These are identical with legal principles, qiyās representing the element of rationality and naql that of authority.18 The criterion of naql was based on language use, but restricted to that of the Qurʾān and early poetry and that of the idealized Arab speakers, whose kalām al-ʿArab had been transmitted in an uninterrupted line.19 This obviated the need for independent empirical research of the contemporary spoken language.20
22When new generations adopted as their first language the contact varieties that resulted from the early contacts between the relatively small number of Arabic native speakers and the indigenous population, these vernacular dialects served as the Low variety in a diglossic continuum. Yet, they are not the source of the linguistic “mistakes” the grammarians address when they warn against incorrect language. The special genre of literature concerned with the laḥn al-ʿāmma “errors of the general public” (Ayoub 2007) targets those among the general public who seek to emulate the ḫāṣṣa, the elite, when using Arabic as a written medium.21
23At first, the laḥn al-ʿāmma treatises may have dealt specifically with errors committed by new learners of the language aiming to acquire the language of the Arabs. One of the oldest treatises, the Kitāb mā talḥanu fīhi l-ʿawāmm by al-Kisāʾī (d. 805), simply lists the correct forms, citing as evidence a verse from the Qurʾān or a poetic line. It was not intended as a guide for reciters of Qurʾān or poetry helping them to avoid making mistakes, because they would know the relevant quotations by heart (Harvey 2016). When al-Kisāʾī says “it is not a, but b, as in the Qurʾānic verse so-and-so, or in the line by such-and-such poet”, the speakers targeted are those who are uncertain about the correct form and hesitate whether to use, for instance, the faʿala or ʾafʿala form of the verb:
You say qad šaġalanī fulān ʿan ʿamalī “someone distracted me from my work” and šaġaltuhu “I distracted him” without the ʾalif [i.e., not ʾašġaltuhu]; God Almighty says šaġalatnā ʾamwālunā “our wealth distracted us” [Q. 48/11].22
24Typical learners’ errors mentioned in these treatises concern the erroneous usage of declensional endings, as highlighted in ʾAbū l-Ṭayyib al-Luġawī’s (d. 962) comment (Rebhan 1981: 16):
Know that the first thing that gets confused in the language of the Arabs and that is most in need of learning is declension because of the manifest mistakes in the language of converts and the new speakers of Arabic.23
25The confusion hinted at here is caused by learners failing to apply the correct declensional endings, rather than by speakers of a type of colloquial language in which the entire declensional system has disappeared.
26Later treatises shift their attention to errors by semi-literate speakers attempting to write Standard Arabic. In the introduction to his Madḫal ʾilā taqwīm al-lisān, Ibn Hišām al-Laḫmī (d. 1181) states that he aims “to correct the common Arabic expressions that the general public mispronounce and use in a way that the Arabs do not”.24 He does not address the transgressors directly, but calls upon his own students to correct them. In one example someone uses the word ḫirbiz and explains it with baṭṭīḫ, incorrect for biṭṭīḫ “water melon”, prompting the following comment by Ibn Hišām: “This is the worst of the worst when someone uses pretentious words, but fails to use common words correctly”.25
27The ambition of semi-literate speakers of Arabic to speak above their level inevitably led to hypercorrections. Blau (1965, 1970, 1982) shows that such errors were made by people who lacked the knowledge to write Standard Arabic faultlessly. Even though some of the hypercorrections may have been caused by interference from the writer’s vernacular, they do not represent features of the colloquial language. The refusal on the part of the grammarians to deal with the colloquial language even applies to someone like Yūsuf al-Maġribī (d. 1611), an Egyptian author who wrote a treatise about the Arabic of Egypt, entitled Dafʿ al-ʾiṣr ʿan kalām ʾahl Miṣr (Zack 2009). At first sight, it might look as if he defended the Egyptian vernacular dialect, but this turns out not to be the case. He acknowledges that common Egyptians make mistakes but, wherever possible, he tries to explain the forms they use as derived from the lexicon of Standard Arabic.
5. Diglossia and Grammarians
28After the publication of Ferguson’s seminal article on diglossia (1959a), Fishman (1967, 1972) modified his definition and set up a matrix combining the sociolinguistic notion of diglossia with the psycholinguistic one of bilingualism, generating four different linguistic contexts. In this framework, diglossia denotes any kind of sociolinguistic variation (see Fig. 2).
Fig. 2. Matrix bilingualism/diglossia (after Fishman 1967)
+ bilingualism - diglossia (e.g., third generation Italian-speakers, USA) | + bilingualism + diglossia (e.g., Arabic-speaking world) |
- bilingualism - diglossia (e.g., small groups of hunter-gatherers) | - bilingualism + diglossia (e.g., pre-Revolutionary Russia) |
29Hudson (2002: 19–21) applies Gumperz’ (1962) typology of linguistic communities to diglossia, linking its emergence to what Gumperz calls an “intermediate society”, in which a small, powerful elite maintain their power by using the High variety in administration and education as one of the tools of exclusion and dominance. Hudson (2002: 13) states that “it is social consensus rather than structural affinity that confers the status of distinct language on two or more linguistic varieties”. Linguistic distance is not a useful measure for determining sociolinguistic behavior but, at most, the relatedness of the two varieties is a statistical tendency in diglossic societies (see also Schöntag 2022: 57ff.).
30The restrictive definition of diglossia as existing between cognate varieties did make a valid point, however: when the two varieties diachronically and perceptually derive from the same language, this affects the speakers’ attitude as well as the grammarians’ judgment. Even when the linguistic distance between the two varieties is considerable, the speakers believe them to be identical. In the Roman empire, all speakers of Latin were convinced they were speaking the same language. Even those who spoke the vernacular varieties of Latin regarded themselves as speakers of the lingua romana.26 In the modern Arabic-speaking world, speakers cling to the belief that there is but one Arabic language, regardless of diatopic or diastratic variation. Even in Egypt, where spoken ʿāmmiyya is a source of pride, Standard Arabic and the vernacular dialect are treated as part of one linguistic diasystem. In some of the other Arabic-speaking countries, speakers tend to have a more negative attitude towards their vernacular dialect and may even go so far as to deny its very existence. They are highly critical of other speakers’ use of dialect and deny ever speaking anything else but Standard Arabic, all the while freely using dialect themselves.27
31Similar attitudes are unknown in situations where a non-cognate language serves as the Low variety because of a general awareness on the part of the speakers that the varieties are so far apart that they cannot possibly be regarded as one language. In this context, the non-prestigious variety is framed as devoid of grammatical structure and irrelevant for linguistic analysis, grammars of the High variety at most drawing attention to the undesirable intrusion of loanwords from the Low variety (on the difference in attitude see Sayahi 2014: 80-119).
32When the two varieties are related, the fiction that the language is one entails the classification of variants beyond the standard norm as linguistic mistakes. In the case of Latin and Arabic, speakers and grammarians alike treated deviations from the norm as a corruption of speech. Since the High variety is the language of a highly prestigious literary and/or religious heritage, cultivated through education, it is bad form to speak out in favor of the Low variety, which is framed as primitive and uncouth, and as a threat to the cultural and religious values of society. Barrau (2011: 314) points to the difference between monasteries in Romance-speaking countries and those elsewhere with respect to the monks’ relationship with Latin: in the former it takes much more time for the vernacular to become recognized as a means of communication in its own right, compared to, for instance, Old High German, which was used in Sankt Gallen to facilitate the teaching of Latin (Grotans 2006). In Paraguay, the indigenous language Guaraní has remained the Low variety in its relationship with Spanish, but at the same time it is regarded as a viable tool for education and especially for religion, having been provided with a grammar and an orthography by the Jesuit missionaries (Rodríguez 2019).
33Hudson (2002: 28–30, 40) sees the presence of native speakers of the High variety as a defining characteristic of a bilingual society. They tend to use the High variety even informally in everyday interaction, and only when there is a shift in the social balance of power, education may lead to the spreading of education across social levels, so that speakers from lower levels get a chance to develop native competence in the High variety (Schiffman 1997). The High variety then becomes a common language and the sociolinguistic profile shifts to that of a standard language with social and regional dialects. In the case of a “classic” diglossia, development may take a different course. Here, too, the social elite has an interest in maintaining a status quo in which the High variety of the language is only accessible for those who enjoy an education, but when the political situation changes, the Low variety may become more acceptable even in formal contexts. This expansion of its domain may reach the point where the Low variety becomes the common language, the High variety being relegated to the position of a learned language, as happened with Latin.28
34It is beyond the scope of the present article to go into other grammatical traditions arising in a similar sociolinguistic framework. The Tamil and other Dravidian linguistic traditions resemble the Latin and Arabic traditions in that they exclude the colloquial Low variety from their descriptions (see Schiffman 1998; Annamalai 2016). The most striking parallel is that of the Sanskrit grammatical tradition, which dealt only with Sanskrit, ignoring almost completely the colloquial Prakrit languages (Houben 1996: 176). Houben (1996: 182) points out that “Sanskrit is the target, and the degree of realization, indeed, the degree of the capacity to realize the target, is variable”. Significantly, the diglossia between Sanskrit and the Prakrit languages may have emerged in a process of imperial expansion, in which the number of new speakers increased rapidly, just as this was the case in Latin and Arabic (Houben 1996: 175–177). This suggests that, like Latin and Arabic, Sanskrit belongs to the category of what Snow (2013) calls “traditional diglossias”, which are the closest fit to Ferguson’s original proposal.29
35The usual approach to the study of grammatical traditions is one in which the language is viewed as one self-contained, single entity, whose rules are codified in grammar books and whose lexicon is collected in dictionaries. This image corresponds to how the grammarians themselves engaged in describing their language, adopting an idealized speaker/hearer model as the basis for their grammar, while ignoring the linguistic variation in their community. The speakers of Latin and Romance in the first centuries CE and those of Arabic from the seventh century CE onwards believed they all spoke the same language, lingua romana (romance) or ʿarabiyya. The prevalent diglossia forced the grammarians to focus on an artificial norm, while undertaking the arduous task of data collection and codification in order to discover the rules governing their language.
36While the two linguistic traditions briefly discussed here developed under similar circumstances in societies with a sharp dichotomy between (primarily spoken) colloquial and (primarily written) standard language, the ultimate development of the two speech communities turned out quite different. In the case of Latin, the Romance vernaculars were eventually recognized as languages in their own right. For the grammarians this meant a drastic shift in attention: henceforth, they dedicated themselves to the standardization of the vernacular languages (Wright 2016). In the case of Arabic, the old situation has persisted until the present time. Grammarians and politicians alike still do their best to preserve the standard fuṣḥā from being invaded by vernacular features (Boussofara-Omar 2006: 629). Even though in some countries the use of dialect in public life has expanded considerably, modern day language handbooks still focus on errors in written language as opposed to the spoken language. The style book of the Egyptian newspaper Al-Ahram by ʾAḥmad Nāfiʿ, Dalīl al-ʿamal al-ṣuḥufī, contains rules about the correct spelling of the glottal stop (hamza) and the construction of numerals, but has nothing about the use of dialecticisms. Neither the grammarians, nor the Al-Ahram style book are concerned with typical ʿāmmiyya features, because these “errors” rarely make it into the written language, except in captions of cartoons or verbatim quotations.
37The nature of the grammatical tradition in a diglossic situation also affects the collection of linguistic data by modern linguists, who have to be extra careful in using the grammatical literature to evaluate the linguistic situation of the time. When grammarians write about spoken language or about linguistic errors, they are not targeting actually spoken speech: their reproduction of the linguistic situation in their community is by necessity slanted, since the speech of the majority of the population is disregarded. Even nowadays, if one were to describe the language situation in, say, Saudi Arabia exclusively on the basis of the written production in Arabic, whether in grammar books, newspapers, or literature, the resulting picture would not be representative at all of the actual variation in spoken language.30 In the past, the filtering function of written production in diglossia is bound to have been even more restrictive and even more biased in the selection of linguistic data.
Bibliographie
Primary sources
ʾAbū l-Ṭayyib, Marātib = ʾAbū l-Ṭayyib ʿAbd al-Wāḥid ibn ʿAlī al-Luġawī, Marātib al-naḥwiyyīna. Ed. ʾAbū l-Faḍl ʾIbrāhīm. Cairo & Fajjala: Maṭbaʿa Nahḍa Miṣr, 1954.
Appendix Probi = [Marcus Valerius Probus], Appendix Probi. Ed. by Heinrich Keil, Grammatici latini, IV. 193–204. Leipzig: B.G. Teubner, 1864. (Repr., Hildesheim: G. Olms, 2007).
Charisius, Ars = Flavius Sosipater Charisius, Artis grammaticae libri V. Ed. by Carolus [Karl] Barwick. Leipzig: B.G. Teubner, 1954.
Diomedes, Ars = Diomedes, Artis grammaticae libri III. Ed. by Heinrich Keil, Grammatici latini, I. 297–529. Leipzig: B.G. Teubner, 1864. (Repr., Hildesheim: G. Olms, 2007).
Ibn al-ʾAnbārī, Lumaʿ = ʾAbū l-Barakāt ʿAbd al-Raḥmān ibn Muḥammad al-ʾAnbārī, Lumaʿ al-ʾadilla fī ʾuṣūl al-naḥw. Ed. by Attia Amer [ʿAṭiyya ʿĀmir]. Stockholm: Almqvist & Wiksell, 1963.
Ibn Hišām, Taqwīm = ʾAbū ʿAbdallāh Muḥammad ibn ʾAḥmad Ibn Hišām al-Laḫmī, al-Madḫal ʾilā taqwīm al-lisān wa-taʿlīm al-bayān. Ed. by José Pérez Lázaro. 2 vols. Madrid: Consejo Superior de Investigaciones Científicas, 1990.
Ibn Jinnī, Ḫaṣāʾiṣ = ʾAbū l-Fatḥ ʿUṯmān Ibn Jinnī, al-Ḫaṣāʾiṣ. Ed. by Muḥammad ʿAlī al-Najjār. 3 vols. Cairo, 1952–1956. (Repr., Beirut: Dār al-Hudā, n.d.)
Kisāʾī, Mā talḥanu = ʿAlī ibn Ḥamza al-Kisāʾī, Kitāb mā talḥanu fīhi l-ʿawāmm. Ed. by Carl Brockelmann, Beiträge zur Geschichte der arabischen Sprachwissenschaft. I. Al-Kisâ’î’s Schrift über die Sprachfehler des Volkes, Zeitschrift für Assyriologie 13 (1898): 29–46.
Nāfiʿ, Dalīl = ʾAḥmad Nāfiʿ, Dalīl al-ʿamal al-ṣuḥufī/Al-Ahram style book. Cairo: [Al-Ahram], n.d.
Sībawayhi, Kitāb = ʾAbū Bišr ʿAmr ibn ʿUṯmān Sībawayhi, al-Kitāb. Ed. by ʿAbd al-Salām Muḥammad Hārūn. 5 vols. Cairo: al-Hayʾa al-Miṣriyya al-ʿĀmma li-l-Kitāb, 1966–1977.
Varro, De L.L. = Marcus Terentius Varro, De lingua latina. Ed. and transl. by Roland G. Kent. 3rd ed. 2 vols. Cambridge, Mass.: Harvard University Press & London: W. Heinemann, 1958.
Secondary sources
Adams, James N. 2003. Bilingualism and the Latin language. Cambridge: Cambridge University Press.
Adams, James N. 2007. The regional diversification of Latin 200 BC–AD 600. Cambridge: Cambridge University Press.
Adams, James N. 2013. Social variation and the Latin language. Cambridge: Cambridge University Press.
Annamalai, Elay. 2016. Diglossia in Dravidian languages. The languages and linguistics of South Asia, ed. by Hans Henrich Hock & Elena Bashir. Berlin: Mouton de Gruyter. 661–668.
Ayoub, Georgine. 2007. Laḥn. Encyclopedia of Arabic language and linguistics, ed. by Mushira Eid a.o. Leiden: E.J. Brill, II. 628–634.
Baalbaki, Ramzi. 2007. Inside the speaker’s mind: Speaker’s awareness as arbiter of usage in Arabic grammatical theory. Approaches to Arabic linguistics, ed. by Everhard Ditters & Harald Motzki. Leiden: E.J. Brill. 3–23.
Baratin, Marc. 2000. À l’origine de la tradition artigraphique latine entre mythe et réalité. Handbuch für die Geschichte der Sprach- und Kommunikationswissenschaft, ed. by Sylvain Auroux a.o. Berlin & New York: de Gruyter, I. 459–466.
Barnett, Francis J. 2007. The sources of the ’Appendix Probi’: A new approach. Classical Quarterly 57(2). 701–736.
Barrau, Julie. 2011. Did Medieval monks actually speak Latin? Understanding monastic practices of oral communication (Western Europe, tenth-thirteenth centuries), ed. by Steven Vanderputten. Turnhout: Brepols. 293–317.
Barwick, Karl. 1922. Remmius Palaemon und die römische Ars grammatica. Wiesbaden: Dieterich’sche Verlagsbuchhandlung. (Repr., Hildesheim: G. Olms, 1967).
Bernards, Monique. 2005. The contribution of mawālī and the Arabic linguistic tradition. Patronate and patronage in early and classical Islam, ed. by Monique Bernards & John Nawas. Leiden: E.J. Brill. 426–453.
Blau, Joshua. 1965. The emergence and linguistic background of Judeo-Arabic: A study of the origins of Middle Arabic. London: Oxford University Press. (2nd ed., Jerusalem: Ben Zwi, 1981).
Blau, Joshua. 1970. On pseudo-corrections in some Semitic languages. Jerusalem: Israel Academy of Sciences and Humanities.
Blau, Joshua. 1982. Das frühe Neuarabisch in mittelarabischen Texten. Grundriß der arabischen Philologie, I, ed. by Wolfdietrich Fischer. Wiesbaden: F. Steiner. 83–95.
Boussofara-Omar, Naima. 2006. Diglossia. Encyclopedia of Arabic language and linguistics, ed. by Mushira Eid a.o. Leiden: E.J. Brill, I. 629–637.
Carter, Michael G. 2004. Sībawayhi. London: Oxford Centre for Islamic Studies.
Chahoud, Anna. 2007. Alterità linguistica, latinitas e ideologia tra Lucilio e Cicerone. Plurilinguismo letterario, ed. by Renato Oniga & Sergio Vatteroni. Udine: Rubbetino. 41–58.
Clackson, James. 2015. Latinitas, ἑλληνισµός and standard languages. Studi e Saggi Linguistici 53(2). 309–330.
Clackson, James & Geoffrey Horrocks. 2007. The Blackwell history of the Latin language. Oxford: Blackwell.
Desbordes, Françoise. 1995. Les débuts de la grammaire à Rome. Lalies 15. 125–137.
Ech-charfi, Ahmed. 2016. The standardization of a diglossic Low variety: The case of Moroccan Arabic. Saarbrücken: Scholars’ Press.
Ferguson, Charles A. 1959a. Diglossia. Word 15. 325–340.
Ferguson, Charles A. 1959b. Myths about Arabic. Languages and Linguistic Monographs Series, Georgetown University 12. 75–82.
Ferguson, Charles A. 1991. Diglossia revisited. Southwest Journal of Linguistics 10(1). 214–234.
Fishman, Joshua A. 1967. Bilingualism with and without diglossia; diglossia with and without bilingualism. Journal of Social Issues 23. 29–38.
Fishman, Joshua A. 1972. Societal bilingualism: Stable and transitional. Language in sociocultural change, ed. by Joshua A. Fishman. Stanford: Stanford University Press. 135–152.
Garcea, Alexandro. 2007. César et les paramètres de l’analogie. Bilinguisme et terminologie grammaticale gréco-latine, ed. by Louis Basset et al. Leuven: Peeters. 339–357.
Ghersetti, Antonella. 2012. “Dites, mais ne dites pas” : Le Taqwīm al-lisān d’Ibn al-Ǧawzī. Arabica 59. 536–551.
Gouttenoire, Marie-Andrée. 2010. Représentations et écritures du voyage au désert des lexicographes et grammairiens en langue arabe de l’espace iraqien des II/VIIIe et III/IXe siècles. Thèse de doctorat, Aix-Marseille Université.
Grotans, Anna A. 2006. Reading in Medieval St. Gall. Cambridge: Cambridge University Press.
Gumperz, John J. 1962. Types of linguistic communities. Anthropological Linguistics 4(1). 28–40.
Harvey, Ramon. 2016. The preferences of al-Kisāʾī (d. 189/805): Grammar and meaning in a canonical reading of the Qurʾān. International Journal for the Semiotics of Law 29(2). 313–332.
Haugen, Einar. 1966. Language, dialect, nation. American Anthropologist 68. 922–935.
Hintzen, Beate. 2011. Latin, Attic and other Greek dialects: Criteria of ἑλληνισµός in grammatical treatises of the first century BCE. The struggle for identity: Greeks and their past in the first century BCE, ed. by Thomas A. Schmitz & Nicolas Wiater. Stuttgart: F. Steiner. 125–141.
Houben, Jan E.M. 1996. Sociolinguistic atttitudes reflected in the work of Bhartṛhari and some later grammarians. Ideology and status of Sanskrit: Contributions to the history of the Sanskrit language, ed. by Jan E.M. Houben. Leiden: E.J. Brill. 157–193.
Hudson, Alan. 2002. Outline of a theory of diglossia. International Journal of the Sociology of Language 157. 1–48.
Ibrahim, Zeinab. 2000. Myths about Arabic revisited. Al-ʿArabiyya 33. 13–28.
John, Alison. 2022. Cultural memory and Classical education in Late Antique Gaul. Making and unmaking Ancient memory, ed. by Rajiv K. Bhola & Martine De Marre. New York & London: Routledge. 256–272.
Kaster, Robert A. 1988. Guardians of language: The grammarians and society in Late Antiquity. Berkeley, Cal.: University of California Press.
Kramer, Johannes. 1998. Die Sprachbezeichnungen latinus und romanus im Lateinischen und Romanischen. Berlin: Erich Schmidt.
Larcher, Pierre. 2005. D’Ibn Fāris à al-Farrā’ ou un retour aux sources sur la luġa al-fuṣḥā. Asiatische Studien/Études asiatiques 59(3). 797–814.
Larcher, Pierre. 2017. Une formulation ancienne de la diglossie en arabe? Luġat al-qawm vs luġat al-yawm d’Ibn Fāris (IVe/Xe siècle). Philologists in the world: A festschrift in honour of Gunvor Mejdell, ed. by Nora S. Eggen & Rana Issa. Oslo: Novus Forlag. 25–40.
Larcher, Pierre. 2018. La langue du Coran: Quelle influence sur la grammaire arabe? Mélanges offerts à Madiha Doss: La linguistique comme engagement, ed. by Aziza Boucherit, Héba Machhour & Malak Rouchdy. Cairo: IFAO. 91–108.
Macdonald, Michael C. 2010. Ancient Arabia and the written word. The development of Arabic as a written language, ed. by Michael C. Macdonald. Oxford: Archaeopress. 5–28.
Mackridge, Peter. 2009. Language and national identity in Greece, 1766–1976. Oxford: Oxford University Press.
Marçais, William. 1930. La diglossie arabe. L’enseignement public (Paris) 104(12). 401–409.
Mathisen, Ralph W. 2018. “Roman” identity in Late Antiquity, with special attention to Gaul. Transformations of Romanness in the Early Middle Ages, ed. by Walter Pohl. Berlin: de Gruyter. 255–274.
Mejdell, Gunvor. 2006. Mixed styles in spoken Arabic in Egypt. Leiden: E.J. Brill.
Millett, Martin. 1990. Romanization: Historical issues and archaeological interpretation. The early Roman Empire in the West, ed. by Thomas Blagg & Martin Millett. Oxford: Oxbow Books. 35–41.
Nedeljković, Vojin. 2015. Latin vulgaire, latin familier: Quelques aspects d’une distinction parfois difficile à saisir. Lingvisticae Investigationes 38(1). 1–12.
Pagani, Lara. 2015. Language correctness (hellenismos) and its criteria. Brill’s companion to Ancient Greek scholarship, ed. by Franco Montanari, Stephanos Matthaios & Antonios Rengakos. Leiden: E.J. Brill, II. 798–849.
Pezzini, Giuseppe. 2017. Caesar the linguist: The debate about the Latin language. The Cambridge companion to the writings of Julius Caesar, ed. by Luca Grillo & Christopher Krebs. Cambridge: Cambridge University Press. 173–192.
Powell, Jonathan G.F. 2007. A new text of the ’Appendix Probi’. Classical Quarterly 57(2). 687–700.
Putten, Marijn van. 2022. Quranic Arabic: From its Hijazi origins to its Classical reading traditions. Leiden: E.J. Brill.
Quirk, Ronald J. 2017. Hypercorrection in the Appendix Probi. Philologus 161(2). 350–353.
Rabin, Chaim. 1951. Ancient West Arabian. London: Taylor’s Foreign Press.
Rebhan, Helga. 1981. Untersuchungen zu laḥn-Werken des 11. und 12. Jhs. Magisterarbeit, Universität Erlangen-Nürnberg.
Rodríguez, Yliana. 2019. Spanish-Guarani diglossia in colonial Paraguay: A language undertaking. The linguistic heritage of colonial practice, ed. by Brigitte Weber. Berlin & New York: de Gruyter. 153–168.
Sandri, Maria Giovanna. 2020. Trattati greci su barbarismo e solecismo: Introduzione ed edizione critica. Berlin & Boston: de Gruyter.
Sayahi, Lotfi. 2014. Diglossia and Language Contact: Language Variation and Change in North Africa. Cambridge: Cambridge University Press.
Schiffman, Harold F. 1997. Diglossia as a sociolinguistic situation. The handbook of sociolinguistics, ed. by Florian Coulmas. Oxford: Blackwell. 205–216.
Schiffman, Harold F. 1998. Standardization or restandardization: The case for “Standard” Spoken Tamil. Language in Society 27(3). 359–385.
Schöntag, Roger. 2022. Das Verständnis von Vulgärlatein in der Frühen Neuzeit vor dem Hintergrund der questione della lingua. Tübingen: Gunter Narr.
Sibileau, Julien. 2021. Une entreprise de légitimation de la grammaire arabe: Les Ḫaṣāʾiṣ d’Ibn Ǧinnī (m. 392/1002). Thèse de doctorat, INALCO.
Siebenborn, Elmar. 1976. Die Lehre von der Sprachrichtigkeit und ihren Kriterien: Studien zur antiken normativen Grammatik. Amsterdam: B. R. Grüner.
Snow, Don. 2013. Revisiting Ferguson’s defining cases of diglossia. Journal of Multilingual and Multicultural Development 34(1). 61–76.
Talmon, Rafael. 1987. Arabic grammar in its formative age: Kitāb al-ʿAyn and its attribution to Khalīl b. Aḥmad. Leiden: E.J. Brill.
Talmon, Rafael. 2003. Eighth-century Iraqi grammar: A critical exploration of pre-Ḫalīlian Arabic linguistics. Winona Lake, Ind.: Eisenbrauns.
Valette-Cagnac, Emmanuelle. 2005. “Plus attique que la langue des Athéniens”: Le grec imaginaire des Romains. Façons de parler grec à Rome, ed. by Florence Dupont & Emmanuelle Valette-Cagnac. Paris: Belin. 37–80.
Versteegh, Kees. 1983. Arabic grammar and the corruption of speech. Arab language and culture (= al-Abḥāth, 31), ed. by Ramzi Baalbaki. Beirut: American University of Beirut. 117–138.
Versteegh, Kees. 1987. Latinitas, hellènismós, ʿarabiyya. The history of linguistics in the Classical period, ed. by Daniel Taylor. Amsterdam: J. Benjamins. 251–274.
Waquet, Françoise, 1998. Le latin ou l’empire d’un signe: xvie-xxe siècle. Paris: Albin Michel.
Weiss, Bernard G. 1985. Knowledge of the past: The theory of tawātur according to Ghazālī. Studia Islamica 61. 81–105.
Woods, Christopher. 2007. Bilingualism, scribal learning, and the death of Sumerian. Margins of writing, origins of cultures, ed. by Seth L. Sanders. 2nd ed. Chicago, Ill.: The Oriental Institute of the University of Chicago. 95–124.
Wright, Roger. 2016. Latin and Romance in the Medieval period: A sociophilological approach. The Oxford guide to the Romance languages, ed. by Adam Ledgeway & Martin Maiden. Oxford: Oxford University Press. 14–23.
Zack, Liesbeth. 2009. Egyptian Arabic in the seventeenth century: A study and edition of Yūsuf al-Maġribī’s Dafʿ al-iṣr ʿan kalām ahl Miṣr. Ph.D. diss., University of Amsterdam.
Zetzel, James E.G. 2018. Critics, compilers, and commentators: An introduction to Roman philology, 200 BCE – 88 CE. Oxford: Oxford University Press.
Notes de bas de page
1 It is generally assumed that the first one to apply the term “diglossia” to the linguistic situation in the Arabic-speaking world was William Marçais (1930), but the Greek scholar Emmanuel Rhoïdis (1836–1904) used it before him in 1885 to refer to the linguistic situation in Greece (Mackridge 2009: 27; Larcher 2017; Schöntag 2022: 54f.).
2 See Millett (1990), especially p. 38.
3 Roman education survived the longest in those provinces that had old ties with Rome, such as Gaul (see Mathisen 2018; John 2022). On learning Latin in Gaul see Adams (2003: 688–693).
4 On the beginnings of grammar study in Rome see Desbordes (1995); Baratin (2000); Zetzel (2018).
5 Barwick (1922: 95–99, 275ff.) argued that Remmius Palaemon introduced Greek grammatical practices in Latin school grammar. For a critical assessment see Baratin (2000); Zetzel (2018: 181f.).
6 See Pagani (2015); Sandri (2020: 3–15); in this canon, dialektos replaced earlier sunētheia “language use” (Siebenborn 1976: 146–151, 159–162).
7 Varro’s criteria were transmitted by two fourth-century grammarians, Charisius (Ars 63.2) and Diomedes (Ars I, 439), see Zetzel (2018: 53). For Caesar’s ideas about analogia see Garcea (2007); Pezzini (2017).
8 Quare, qui ad consuetudinem nos vocant, si ad rectam, consequemur: in ea quoque est analogia (Varro, De l.l. IX, 18).
9 In a letter to Paetus Cicero writes: “But what do I look like to you in my letters? Don’t you think that I talk to you in the language of the common man?” (Verumtamen quid tibi ego in epistulis videor? Nonne plebeio sermone agere tecum?, Ad Fam. 9, 21, 1); see Clackson & Horrocks (2007: 198–206); Adams (2013: 20); Schöntag (2022: 146–153). About the need to distinguish between colloquial and vulgar Latin see Nedeljković (2015).
10 Chahoud (2007: 51) refers to the use of Greek as a sign of affectation. On the double standard in Roman attitudes towards Greek see Valette-Cagnac (2005).
11 Chahoud (2007: 46f.). The use of such treatises as a source for early forms of Romance is fairly problematic for precisely this reason (Barnett 2007).
12 Chommoda dicebat, si quando commoda vellet/dicere, et insidias Arrius hinsidias (Carmina 84).
13 See Powell (2007); Quirk (2017); Zetzel (2018: 314–316).
14 This form was used by speakers wary of the common “vulgarism” of -ns > -s, as in mesa for mensa (Quirk 2017).
15 Carter (2004: 39–49); Larcher (2018). Less frequently cited sources are: traditions from the Prophet (ḥadīṯ), proverbs and expressions, and made-up examples.
16 On the naḥwiyyūna and their use of qiyās see Carter (2004: 5, 42). On the limits of analogy in the Kitāb see Sibileau (2021: 145–147).
17 Sibileau correctly points out that for Ibn Jinnī mere occurrence in the Qurʾān does not suffice to validate a grammatical form: only those forms are valid that are actually being used by the speakers. On the role of later grammarians in determining the structure of the language of Qurayš and the establishment of al-luġa al-fuṣḥā see also Larcher (2005); van Putten (2022).
18 The third argument, istiṣḥāb al-ḥāl, refers to a special kind of analogical reasoning.
19 Gouttenoire (2010) argues persuasively that grammarians did not actually go into the deep desert to study Bedouin speech, but were content to speak with semi-professional informants among the Bedouin living at the city’s periphery, as the grammarian al-Kisāʾī (d. 805) famously did (Harvey 2016).
20 The point is that the correctness of speech is guaranteed by the unbroken line of transmission of the linguistic data, just as an unbroken line of transmitters of Hadith is needed for the validity of legal regulations (Weiss 1985).
21 A point well made by Ghersetti (2012: 537). On the dichotomy ʿāmma/ḫāṣṣa see Rebhan (1981: 17–19).
22 Wa-taqūlu qad šaġalanī fulān ʿan ʿamalī wa-šaġaltuhu bi-ġayr ʾalif qāla lLāh taʿālā šagalatnā ʾamwālunā (Mā talḥanu 34.17f.).
23 Wa-ʿlam ʾanna ʾawwala mā ḫtalla min kalām al-ʿArab fa-ʾaḥwaja ʾilā l-taʿallum al-ʾiʿrāb li-ʾanna l-laḥna ẓahara fī kalām al-mawālī wa-l-mutaʿarribīna (Marātib 5.9–10).
24 Taṣḥīḥ al-ʾalfāẓ al-ʿarabiyya al-mustaʿmala llatī ḥarrafathā l-ʿāmma ʿan mawḍiʿihā wa-takallamat bihā ʿalā ġayr mā takallamat bihā l-ʿArab (Madḫal preface).
25 Wa-hāḏā min ʾaqbaḥ al-qabīḥ ʾan yastaʿmila l-luġata l-ġarība wa-qad qaṣṣara ʿan taṣḥiḥ al-mustaʿmala al-qarība (Madḫal 12).
26 On the labels romanus or latinus assigned to these varieties, see Kramer (1998).
27 The classic treatment of this topic is by Ferguson (1959b); see also Ibrahim (2000).
28 See Waquet (1998); Barrau (2011). Apparently, a learned language may also develop in cases where there is no dominant elite actually using the High variety, as demonstrated by the case of Sumerian in the Assyro-Babylonian world (Woods 2007).
29 In this category, Snow also includes the case of Chinese. In addition, he distinguishes “revived diglossias”, in which the High variety is a learned language introduced for nationalist purposes, and “modern diglossias”, in which a modern standard languages is imposed as High variety.
30 This is a somewhat simplified account of the facts: in a number of Arabic dialects processes of standardization and expansion are taking place (for the standardization of Moroccan Arabic see Ech-charfi 2016). In Egypt, the initiative for a maṣri Wikipedia started in 2008 (https://arz.wikipedia.org/wiki/, accessed 16/10/2023).
Auteur
University of Nijmegen
Kees Versteegh was full professor of Arabic and Islam at the University of Nijmegen (The Netherlands) from 1989 till 2010, when he retired. He graduated in Classical and Semitic languages and obtained his Ph.D. degree at the University of Nijmegen with a dissertation on Greek elements in Arabic linguistic thinking (1977). He has published on the history of linguistics and contact linguistics. His books include Pidginization and creolization : The case of Arabic (1984), Arabic grammar and Qur’anic exegesis in early Islam (1993), The Arabic linguistic tradition (1997), and The Arabic language (2nd ed. 2014). He co-edited the Handbuch für die Geschichte der Sprach- und Kommunikationswissenschaft (2000-2005) and was the editor-in-chief of the Encyclopedia of Arabic language and linguistics (2006-2009).

Le texte seul est utilisable sous licence Creative Commons - Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International - CC BY-NC-ND 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.
La linguistique et ses formes historiques d’organisation et de production
Actes du colloque de la Société d’histoire et d’épistémologie des sciences du langage et du laboratoire Histoire des théories linguistiques, Paris, 24-26 janvier 2019
Didier Samain et Pierre-Yves Testenoire (dir.)
2022
Eugen Wüster et la terminologie de l’école de Vienne
Actes du colloque de la Société d’histoire et d’épistémologie des sciences du langage et du laboratoire Histoire des théories linguistiques Paris, 3-4 février 2006
Danielle Candel, Didier Samain et Dan Savatovsky (dir.)
2022
Simplicité et complexité des langues dans l’histoire des théories linguistiques
Actes du colloque de la Société d’histoire et d’épistémologie des sciences du langage et du laboratoire Histoire des théories linguistique, Paris, 23-25 janvier 2020
Chloé Laplantine, John. E. Joseph et Émilie Aussant (dir.)
2023
Documenter et décrire les langues d’Asie : histoire et épistémologie
Actes du colloque de la Société d’histoire et d’épistémologie des sciences du langage Paris, 26-28 janvier 2022
Émilie Aussant et Fabien Simon (dir.)
2025