The Arabic Linguistic Tradition

Georges Bohas, Jean-Patrick Guillaume, Djamel Eddine Kouloughli

• Publisher: Routledge
• Year: 2016
• ISBN:
  9781315512754
• Edition: 1
• Topic: Ciencias sociales
• Subtopic: Estudios regionales

The importance and richness of the Arabic linguistic tradition, largely neglected by Western literature, is amply demonstrated by this book, first published in 1990. Written by three experts in the field, it provides us with a comprehensive survey of the historical constitution and theoretical structure of the Arabic linguistic tradition from its beginnings in the eighth century to its mature state around the fourteenth and fifteenth centuries. Besides grammar, the book covers such fields as rhetoric, grammatical semantics, and methodological issues, and pays particular attention to the most representative works of the classical period. It also has the unique benefit of containing the historical background.

1

GENERAL INTRODUCTION

THE GROWTH OF THE ARABIC LINGUISTIC TRADITION: A HISTORICAL SURVEY

Early grammatical thinking to the end of the second/eighth century

As is generally well known, the first grammatical treatise of unquestionable authenticity is Sībawayhi’s Kitāb (this title means ‘The Book’ or ‘Sībawayhi’s Book’). This work, whose author died in or about 177/798, is most probably the first attempt at a comprehensive and systematic description of the Arabic language at every level (phonetics, phonology, morphology, syntax, and semantics). In spite of the great originality of its approach, notably in syntax (see Chapter 2), the breadth of its scope and the depth of its insights clearly point to at least some kind of pre-existent reflection on grammar, even if this reflection had perhaps not yet crystallized into an autonomous discipline.

According to medieval Arabic sources, grammar was first ‘invented’ by Abū l-Aswad al-Duʼalī (d. 69/688?) on the basis of a ‘personnal communication’ (as we would call it nowadays) by ʽAlī ibn Abī Ṭālib (d. 40/660), the Prophet’s cousin and son-in-law. Although it is still accepted by some Arabists (e.g. Mubārak, 1974: 10–37), this account is generally discarded as legend. Another opinion associates, perhaps more plausibly, the emergence of grammar with ʽAbd Allāh ibn Abī Isḥāq (d. 117/734), who is said to have ‘divided grammar and measured it’, farraʽa alnaḥwa wa-qāsa-hu (Abū l-Ṭayyib, Marātib: 12; see Fleisch, 1961: 27; Talmon, 1986), which points to an attempt at a systematical classification of grammatical facts and at building general rules by way of abstract reasoning (qiyās, see next section).

But the important thing is perhaps not so much to discuss the claims of the different candidates to the title of first Arabic grammarian, than to have a reasonably clear picture of the kind of discussions in which the first manifestations of grammatical thought appeared, as the nature of these discussions had an enduring formative influence on the approach and problems of the later tradition. All these discussions, in fact, can be related to a single, major event: the shift of Arabic from a mainly oral language, specific to an ethnically (more or less) homogeneous community of ‘native’ speakers, to a language adapted to a basically written use by an elite of mixed ethnic backgrounds, within a richer and more complex cultural framework.

The first kind of discussion, and perhaps the most ancient, is related to the recension of the Qurʼān and its fixation for ritual recitation. Most of the figures associated with the early developments of grammar and philology are mentioned in connection with the branch of knowledge technically called qirāʼāt (‘readings’ or ‘recitations’, i.e. of the Qurʼān), the purpose of which was to sift the many variant readings which were compatible with the ancient Arabic script in which the oldest copies of the holy text were written. These variants, which seldom carried important differences of meaning, appeared mostly at the morpho-phono-logical and morpho-syntactical levels. Although it was universally admitted that more than one reading could legitimately exist for a given verse, it was also considered necessary to distinguish between ‘acceptable’ and ‘unacceptable’ readings, and, among the former, between ‘current’ and ‘rare’ ones. The basic criterion was, apart from the reliability of the transmitters, the conformity to the ‘speech of the Arabs’, that is the specific linguistic usage of the Beduins of Central Arabia, in the pre-Islamic and early Islamic period.

The second kind of discussion is related to the collection and criticism of ancient poetry, which played perhaps a more decisive and enduring part in the constitution of the philological sciences. The problems which confronted the scholars engaged in this work were basically identical with those relating to the qirāʼāt, but were considerably more complex. On the one hand, ancient poetry provided scholars with an infinitely vaster and more diversified sample of kalām al-ʽArab than the Qurʼān did; poetry made greater use of specialized vocabulary, rare words, difficult constructions, tribal dialectalisms, and so forth. On the other hand, the transmitters of poetry seem to have been, on the whole, much less careful than the transmitters of the Qurʼān. As commonly happens within oral traditions, they tended, more or less consciously, to modify the poems as they transmitted them; some would even interpolate lines of their own in some piece of verse by an older poet, or forge whole pieces outright. The extensive scope and complexity of the problems entailed by this situation goes towards explaining why the body of ‘philological sciences’ (ʽulūm al-ʽArabiyya, literally ‘sciences of Arabity’), which formed the context in which grammar first grew, seems to be mainly focussed on the poetic heritage of ancient Arabia, as it comprised, besides grammar proper (naḥw), lexicography (ʽilm al-luḡa), a specialized field of which was devoted to ‘rare’ words (ḡarīb), metrics (ʽilm al-ʽarūḍ), and even the knowledge of the famous battles and tribal wars of the ancient Arabs (ayyām al-ʽArab), and of their genealogies (ʽilm al-ansāb). These two last branches of knowledge were necessary in order to understand the recondite allusions to tribal feuds and alliances found in the most characteristic kind of ancient poetry.

The third formative factor is the reform initiated by the Umayyad caliph ʽAbd al-Malik ibn Marwān (reigned 65/685–86/705), by which Arabic became the sole administrative language of the Islamic empire. Although we still lack a comprehensive study of the changes that such a reform entailed in the technical and cultural practices associated with language, they cannot but have been quite extensive and, in many ways, decisive. In the long term, their effects were enhanced by the fact that the ‘scribes’ (kuttāb), besides their specific function as administrators, were soon to give the tone to most aspects of classical Islamic high culture. By accepting (probably after some resistance, bureaucratic circles being what they are) that the ‘speech of distinguished people’ (kalām al-xawāṣṣ) had to conform with what was most representative of kalām al-ʽArab, as opposed to the ‘degraded’ vernacular spoken by the populace, they certainly contributed to the general social relevance of grammatical and philological studies.

Such was, then, the context in which appeared the first manifestations of grammatical thought. By the end of the second/eighth century, it was already in a state of considerable advancement and, in some fields, had even evolved its definitive forms. Such was the case of phonetics and metrics, which were codified by al-Xalīl (d. 175/791), Sībawayhi’s teacher. Al-Xalīl is also credited with having devised the basic principles of lexicography in his Kitāb al-ʽAyn, the first Arabic dictionary. As for morpho-phonology (taṣrīf) and syntax (naḥw proper), even if their definitive, canonical form would not be codified until the fourth/tenth century, they had already evolved some of their basic concepts and devices.

Many Arabists have stressed the remarkably swift pace at which the Arabic grammatical tradition had, in so short a period, developed into a complex and sophisticated set of concepts and procedures. According to them, such a precocity can only be accounted for by the effect of some extraneous influence. This influence has been variously identified with Aristotelian logic (Merx, 1889), Islamic law (Carter, 1968 and 1972), or Stoic grammar (Versteegh, 1977). Such a variety of hypotheses sufficiently indicates that no one of them is, in fact, completely satisfying; on the other hand, none can be completely discarded, even if the actual evidence adduced by their respective authors is, to our minds, often unconvincing (that of Merx is, to put it frankly, quite fanciful). In fact, their main failing is that they try to explain the whole of the grammatical tradition in terms of one single factor, which is unnecessary (whatever their respective authors claim to the contrary, these hypotheses are not mutually incompatible), and, indeed, runs counter to the most currently accepted methodology of historical studies. In our opinion, at least, the important thing is that, whatever its model or models can have been, the Arabic tradition developed into something quite different and original; a point on which, moreover, everybody more or less agrees.

From Sībawayhi to al-Mubarrad

Although Sībawayhi’s and al-Xalīl’s contribution to the development of the grammatical tradition was in many ways decisive, it did not result in the constitution of a definitive canonical model for grammatical theory. Actually, such a model was not evolved until the first decade of the fourth/tenth century, its first expression being the Kitāb al-Uṣūl by Ibn al-Sarrāǧ (d. 316/928). The importance of this event has for a long time been underestimated, for many reasons (among others, the fact that the Kitāb al-Uṣūl was not published until quite recently), and the accepted idea has been that Sībawayhi had, in fact, laid down the basic rules and methods of grammar, while the later grammarians’ contribution consisted only in expounding his theory in a more explicit and systematic form, or in finding new applications for it. Such a linear conception of the history of the grammatical tradition led, in fact, to many misrepresentations and false problems.

In the next chapter, we shall try to show that Sībawayhi’s syntactic system is, on the whole, founded on a quite different approach from that of the classical grammarians. But if, in originality and perception, the Kitāb certainly stands alone among the grammatical products of its period, even a perfunctory examination of the few treatises surviving from before the fourth/tenth century shows that they exhibit some common traits which distinguish them collectively from the products of later periods. The most obvious of these is perhaps their extreme heterogeneity in scope, in approach, and even in terminology, together with a strong dependency on what one could call a ‘philological’ outlook. The primary interest of the earlier grammarians is not (as it will be for their successors) in explicitly laying down general rules and principles in order to classify and analyse linguistic facts; it is rather in examining and discussing isolated, specific data, especially when these data exhibit some kind of deviance from the most general behaviour of the class to which they belong. It is, for instance, typical of this approach that al-Mubarrad (d. 285/898), in his Muqtaḍab, devotes a whole chapter to the irregular plural of qaws (‘bow’) qisiyy, this chapter being somewhat longer than the one in which he discusses the much more general and, we should feel, important problem of the assignation of the nominative to the subject of the verbal phrase (Muqtaḍab, I: 8–9 and 39–41, respectively). This kind of approach, in which facts of different nature and rules of different degree of generality are put together in what seems a haphazard order, is also quite perceptible in the two other main grammatical works of the period, the Maʽānī l-Qurʼān by al-Farrāʼ (d. 207/822), who was in his time the leader of the so-called ‘Kūfan’ school (see below, pp. 6–8), and a shorter work bearing the same title by al-Axfaš al-Awsaṭ (d. 221/835), a disciple of Sībawayhi. In these two works, which are grammatical and philological commentaries on the Qurʼān, the order and nature of the problems discussed are more or less governed by the order in which they appear in the text. They show a wide range of interests, but with a stress on lexicology and morpho-phonology, syntax (with the exception of morpho-syntax) receiving a more perfunctory treatment. In fact, it seems that, throughout this period, Sībawayhi was the only grammarian to show a deep and systematic interest in the field of syntax.

This lack of canonical theoretical model does not mean, of course, that grammar was still in a ‘pre-theoretical’ state. As a matter of fact, grammarians systematically used abstract, general rules and principles when analysing and discussing individual facts but these rules and principles were never formally stated, rather, they were taken for granted, as if they were a matter of current knowledge. To put it differently: one could say that they formed a kind of general intuitive background in the light of which the grammarians approached linguistic data: it only became consciously acknowledged when some kind of fact occurred which was felt to need explanation, for instance when qaws, instead of forming a regular plural, quwūs (as, for instance, qalb/qulūb), forms an irregular one, qisiyy; this naturally implies that one has, somewhere in the background, a theory about what the regular plural for this class of nouns should be.

On the other hand, the informal and intuitive nature of the theoretical framework offered a wide scope for individual improvization and interpretation; as long as the basic principles and rules which governed grammatical analysis were not explicitly and systematically defined, even minimally compatible solutions to a given problem could coexist and still be considered as equally legitimate, or indeed substantially equivalent. This property of the grammatical theory in this period is crucial in order to form a clear picture of a much-discussed point of historiography: the rivalry between the Baṣran and Kūfan ‘schools’ of grammar. In fact, many of the difficulties relating to this problem arise, in our opinion, from an inadequate appreciation of the change that intervened in the grammatical tradition when the canonical model was finally evolved.

Philological and grammatical studies first appeared in Baṣra and Kūfa, the two main cities of lower Iraq, which were the principal centres of learning in early Islam, until they were supplanted by Baghdad about the middle of the third/ninth century. In this sense, one can speak of a Baṣran and a Kūfan ‘school’ of grammar, but one should keep in mind that a ‘school’ (maḏhab) in classical Islam refers not so much to a specific body of doctrine as to a channel of transmission of knowledge, by personal contact between master and pupil (there was no other legitimate access to knowledge); on such bases, Zellig Harris and Noam Chomsky would be considered as belonging to the same ‘school’, as the latter was, for some time, the pupil of the former, quite independently of their theoretical divergence. Of course, grammarians of the same ‘school’ could have in common some tenets which distinguished them from others, but these tenets did not necessarily have deep theoretical implications. In fact, the divergences between grammarians of Baṣra and of Kūfa in the pre-canonical period were simply a particular aspect of the general situation of grammar, where the implicit and informal character of the theory made for the coexistence of several potentially conflicting solutions or analyses for the same problem.

In fact, it is even quite possible, as Fleisch (1961) suggests, that the theme of the conflict between the two schools was actually invented after the fact, as a kind of historical justification for the personal rivalry between al-Mubarrad (d. 285/898), the leader of the ‘Baṣrans’, and Ṯaʽlab (d. 291/904), his ‘Kūfan’ counterpart, when they met in Baghdad. But the important thing is that, when the canonical model was evolved, a generation later, by al-Mubarrad’s disciples, they naturally gave it the ‘Baṣran’ label. Now, this model, because of its explicit and systematic character, was naturally more constrained than the former, which had been, in fact, common to Baṣran and Kūfan grammarians alike. In consequence, many views which had hitherto seemed acceptable and legitimate, now appeared incompatible with the new model. An efficient way to make them harmless was to attribute them to the ‘Kūfan school’, which had by then become virtually extinct. From this point on, there was a basic asymmetry between allegedly Baṣran and Kūfan material: whereas the Baṣran views all fell together within a systematic, organized system, the Kūfan, on the contrary, gave the impression of being a haphazard collection of views on points of detail, from which it seems quite difficult to reconstruct any kind of coherent system. The obvious reason is, of course, that they were never meant to have any coherence. A failure to recognize this most crucial point has often misled Arabists into characterizing the Kūfan grammarians as staunch defenders of linguistic ‘usage’ as based on ‘transmitted’ data (samāʽ), while the Baṣrans were described as partisans ...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Original Title Page
• Original Copyright Page
• Table of Contents
• Preface
• Transcription system
• 1 General Introduction
• 2 Sībawayhi’s Kitāb: An Enunciative Approach to Syntax
• 3 The Canonical Theory of Grammar: Syntax (Naḥw)
• 4 The Canonical Theory of Grammar: Morphology, Phonology, and Phonetics (Taṣrīf)
• 5 Major Trends in the Study of Texts
• 6 Rhetoric and Grammatical Semantics
• 7 Metrics
• Bibliography
• References
• Indexes