Free Access
Issue
Histoire Epistémologie Langage
Volume 40, Number 2, 2018
La tradition linguistique arabe et l’apport des grammairiens arabo-andalous
Page(s) 35 - 53
DOI https://doi.org/10.1051/hel/2018019
Published online 22 January 2019

© SHESL/EDP Sciences

In his important dictionary, al-Muḫaṣṣaṣ, Ibn Sīda, the most distinguished Andalusian lexicographer, born in Murcia and died in Denia in 458/1066, presents, as the foundations of his work, a conception of the nature of language and the origin of speech. As observed by Loucel (1964, p. 57), the latter theory is borrowed, almost literally, from al-Ḫaṣāʾiṣ (“The Specificities [of the Arabic Language]” or “The Properties [of the Arabic Language]”) without this being stated explicitly. Al-Ḫaṣāʾiṣ, as we know, is an important work of grammatical epistemology, written by Ibn Ǧinnī, the great Baġdādi grammarian of Greek origin of the 4th/10th century (d. 392/1002). The passage borrowed by Ibn Sīda is famous and already translated and studied by a number of scholars (Loucel 1964, Weiss 1974, Versteegh 1997). It is a rare presentation by a grammarian of the debate on the origin of human speech among Arab scholars: does human speech find its origin in divine Revelation or is it the result of agreement between humans?

In the history of this debate, as presented two centuries later by the theologian and jurist al-Rāzī (d. 606/1209) in al-Maḥṣūl (I, p. 181 sqq.), these two theories had as proponents in the 3rd/9th century, where there were genuine theologian schools with a whole body of doctrines, respectively al-ʾAšʿarī (d. 324/935-6)1 and ʾAbū Hāšim al-Ǧubbāʾī (d. 321/933):2 the Muʿtazilī master al-Ǧubbāʾī for the human convention, and his former brilliant disciple who moved away from the Muʿtazila,3 al-ʾAšʿarī, for the divine institution. Ibn Ǧinnī, in al-Ḫaṣāʾiṣ, does not give names of theologians or jurists. Nevertheless, he mentions a collective, ʾahl al-naẓar, which he warns us immediately that they consider, for the most part, that the origin of language is a human convention:

However, most ʾahl al-naẓar consider that the origin of language is institution and convention [between humans], not revelation and impartation [by God]. (Ibn Ǧinnī, Ḫaṣāʾiṣ, I, p. 40)

Who are these ʾahl al-naẓar? Al-naẓar, “the seeing of a thing”, is used here as related to intellection. It is the “considering and investigating” of a thing and, as a substantive, it is “speculation”, “intellectual examination”.4 The expression ʾahl al-naẓar is very common in al-Ḫaṣāʾiṣ, where it seems to be used in the sense of “speculative people”, “those who know how to investigate and examine questions”, “whose theoretical views are good”. It becomes clearer, though taken with a variant, when Ibn Ǧinnī determines the addressees of his book. This work, says Ibn Ǧinnī, is read and appreciated by ḏawū al-naẓar who are recruited among different categories of intellectuals:

ḏawū al-naẓar among theologians, jurists, philosophers, grammarians, secretaries of chancery and men of letters. (ḏawū al-naẓar min al-mutakallimīn wa-l-fuqahāʾ wa-l-mutafalsifīn wa-l-nuḥāt wa-l-kuttāb wa-l-mutaʾaddibīn, Ibn Ǧinnī, Ḫaṣāʾiṣ, I, p. 67)

If we compare these two quotes, we can say that those who know how to investigate and examine speculative questions among theologians, jurists, philosophers, grammarians, secretaries of chancery or men of letters, consider, for the most part, that language is a human convention. A community with indistinct names, if one can say, identified by a single quality: their love of speculative things. This anonymity seems quite deliberate. On several occasions, while the text might have been appropriate for proper names, it is an indeterminate plural by which the enunciators are designated, especially when Ibn Ǧinnī presents Muʿtazila theses: “They said” (qālū; see I, p. 44-45); or it is an undefined quantifier that is used: “One day I asked one of its partisans” (saʾaltu yawman baʿḍ ʾahli-hi; see I, p. 46).

Nevertheless, in this passage, Ibn Ǧinnī clearly mentions two names, those of two grammarians, his own teacher ʾAbū ʿAlī al-Fārisī (d. 377/987) and a 2nd-3rd/8th-9th century grammarian, ʾAbū al-Ḥasan al-ʾAḫfaš al-ʾAwsaṭ (d. 215/830): both would have hesitated between the two theories, according to Ibn Ǧinnī, considering them both legitimate.

Why this hesitation? Because neither of the two theories, says Ibn Ǧinnī, can be really established. The proponents of the theory of Revelation base their argumentation on a Qurʾanic verse (Kor 2, 31): “He taught Adam all the names” (ʿallama ʾādama al-ʾasmāʾa kulla-hā). This verse seems, reports Ibn Ǧinnī, to have convinced for a moment his master ʾAbū ʿAlī al-Fārisī of the validity of the thesis of the revealed origin of the language. But ultimately, even ʾAbū ʿAlī, in other writings, followed by Ibn Ǧinnī, does not consider it as a decisive argument because it can be interpreted as “He made Adam capable of instituting names” (ʾaqdara ʾādama ʿalā ʾan wāḍaʿa ʿalay-hā).5 A century before, another grammarian, ʾAbū al-Ḥasan, seems to have been unable to decide, believing in the revealed origin, but not condemning the supporters of the human institution.

After discussing why the verse only mentions nouns – and not verbs nor particles, the argument of Ibn Ǧinnī being that it is a kind of what we call metonymy where something as well-known as nouns stands for all language –, Ibn Ǧinnī discusses the arguments of the conventionalist doctrine. Their arguments are not decisive either. The first and most classic one is that God cannot institute anything with humans. Their argumentation is in three movements, punctuated each time by a qālū (“they said”) in the text:

  1. [they said]: it is well established that any institution and convention presuppose, initially, to see with the eye and to make signs to things (al-mušāhada wa-l-ʾīmāʾ)

  2. [they said]: the Lord has no organs (ǧāriḥa) or corporeality by which he can designate or show things.

  3. [they said]: it is therefore wrong to consider that the institution of language is the act of God.6

Although not indicated by Ibn Ǧinnī, this is the classic argument of the Bahšamite7 doctrine, the doctrine of the followers of ʾAbū Hāšim al-Ǧubbāʾī, proponents of the conventionalist theory. This argument is not also a decisive one, says Ibn Ǧinnī: God might as well have put into things sounds that man would have been led to imitate in order to speak. In which case the argumentation would not be conclusive.

Another argument is more personal to Ibn Ǧinnī and would lead him to believe in the revelationist thesis: it is the fascinating “wisdom, subtlety, precision and elegance” of this “noble, excellent and subtle language”, what reminds us “l’amour de la langue” of Jean-Claude Milner (1978) with this strange notion of the wisdom of the language (ḥikmat al-luġa) that needs to be more studied.8

However, the text concludes on a famous final passage where Ibn Ǧinnī describes himself “turned away in defeat” between “the two scales of the balance”, in other words between the two theories. He adopts a non-committal position (waqf): neither theory can be established and they are both equal theoretical possibilities. The leading proponent of the non-committal theory is, according to al-Rāzī, the ʾAšʿarite theologian al-Bāqillānī (d. 403/1013), a contemporary of Ibn Ǧinnī. However, this debate, which agitated different milieux during two centuries, went to an end in the 5th/11th century, according to Weiss (1974), and the non-committal theory became predominant in the milieux of jurists, theologians and grammarians.

Ibn Sīda borrows the whole passage, except for the last paragraph where Ibn Ǧinnī declares himself unable to decide, divided that he is between the two theses. Loucel (1964) wonders if the absence of this paragraph in al-Muḫaṣṣaṣ means that Ibn Sīda adopts the thesis of the divine Revelation. Nevertheless, his hypothesis is rather that the text to which Ibn Sīda would have had access did not contain this final paragraph, which would have been added in a later version by Ibn Ǧinnī. This analysis leaves us rather perplexed because it is neither verifiable nor falsifiable. However, the passage being taken almost literally, it is described by Loucel as a servile borrowing and has not been studied further since the 1960s.9 We propose, in this paper, to show that the relation of Ibn Sīda to Ibn Ǧinnī is much more complex than this first impression of “servile borrowing” suggests. In fact, Ibn Sīda, divided between fascination and reflection, proceeds nonetheless to a reading of the text with significant differences that suggests that another conception of the nature of language could be at stake. But first, let us examine what is the exact relation of Ibn Sīda to Ibn Ǧinnī.

1 Orient/Occident

Ibn Ǧinnī died in 392/1002. This means that almost immediately al-Ḫaṣāʾiṣ was transmitted and well known in the circles of Andalusian grammarians and lexicographers. And indeed, we know that the culture in al-Andalus was totally orientalized at that time. As Blachère (1930, p. 16) writes, “L’Andalousie [était] sous la tutelle intellectuelle de l’Orient”, with many prestigious Oriental scholars there. Ibn Sīda received a part of his education with some of them, in particular with the famous Ṣāʿid al-Baġdādī (d. 417/1026) who was himself the disciple of al-Sīrāfī (d. 368/979) and of ʾAbū ʿAlī al-Fārisī, the master Ibn Ǧinnī had followed for 40 years. In other terms, Ṣāʿid al-Baġdādī and Ibn Ǧinnī had the same master. Moreover, Ṣāʿid was from Mosul, like Ibn Ǧinnī. So, he was probably the transmitter of Ibn Ǧinnī’s work to his pupils.

However, we can say without risk of error that Ibn Sīda knows Ibn Ǧinnī’s works very well and he holds them in great esteem. He cites them as bibliographical sources in his two dictionaries, the alphabetical dictionary, al-Muḥkam, and the thematic dictionary, al-Muḫaṣṣaṣ. Thus, in al-Muḫaṣṣaṣ, he presents an impressive list of sources that he has integrated and quotes, in this list, all the works of Ibn Ǧinnī which, he says, “were accessible to me” (mā saqaṭa ʾilay-ya min-hā),10 some of which are currently lost: al-Tamām, al-Muʿrib,11 Sirr al-ṣināʿa, al-Mutaʿāqib, Šarḥ šiʿr al-Mutanabbī, Tafsīr šiʿr al-Ḥamāsa. Of course, in this list we find al-Ḫaṣāʾiṣ. In the presentation, one hears the high esteem in which Ibn Sīda holds Ibn Ǧinnī’s works: they are mentioned after the grammar works of the greatest, such as Sībawayhi’s (d. 180/796?) Kitāb and the works of al-Fārisī and al-Sīrāfī; their lexical data, he tells us, have adorned his dictionary:

I adorned it with all the well analyzed and illustrative lexical data of the Kitāb of Sībawayhi […] and all the similar data included into every book to which I could have access of the works of al-Fārisī, […] into the book of al-Sīrāfī […] and into all the works of Ibn Ǧinnī that were accessible to me. (Ibn Sīda, Muḫaṣṣaṣ, I, p. 13)

This approach which integrates lexical data from important grammatical works is primarily due to Ibn Sīda’s desire to be as exhaustive as possible. In this way, he includes the data absent in other dictionaries. As observed by Baalbaki (2014, p. 323), al-Muḥkam, his alphabetical dictionary, is described by him as al-muḥīṭ al-ʾaʿẓam (“The greatest exhaustive [dictionary]”). Actually, this expression is a part of its title, the complete title of this alphabetical dictionary being al-Muḥkam wa-l-muḥīṭ al-ʾaʿẓam (“The Greatest Systematic and Exhaustive Dictionary”),12 and indicates explicitly the desire for completeness which animates the lexicographer.

The second objective is that, explicit, to better understand the lexical entry by explaining its morphology. Ibn Sīda criticizes the lexicographers who preceded him for not having mastered grammar (al-naḥw, al-ʾiʿrāb), which was detrimental to their work. And, indeed, the name of Ibn Ǧinnī is cited nearly fifty times in the lexical entries of al-Muḫaṣṣaṣ to shed light on some aspect of the meaning or structure of a word.

Thus, the lack of explicit mention of Ibn Ǧinnī in the passage borrowed is far from being a complete denial of debt. Ibn Sīda acknowledges Ibn Ǧinnī as a great scholar.

2 Affinities

But a plagiarism like that betrays a receptivity, in fact a strong attraction that Ibn Ǧinnī’s work exerts over Ibn Sīda. And the Ibn Sīda’s title, al-Muḫaṣṣaṣ, so close to al-Ḫaṣāʾiṣ in its sounds at least, but also to some of its meaning, appears to us in another light. Both titles have the same root ḫṣṣ, which means to distinguish something or someone, to particularize it, to assign a thing, or make it to belong to it alone, or in particular, exclusively of others.13 However, the exact meaning of al-Muḫaṣṣaṣ is not clear.14 It is the past participle of the verb ḫaṣṣaṣa which has an intensive signification, meaning to render something particular, special; to distinguish it. Does this mean that the title has to be translated as “The Distinguished Dictionary” or is it the dictionary which, by its special organization, is particularly addressed to the eloquent men of letters, the talented poets and orators, thus alluding in this title to both its organization and its addressees as indicated in the introduction? It is this second hypothesis which seems to us more probable. In the latter case, the title is elliptical: this dictionary is [al-Muḫaṣṣaṣ] li-l-faṣīḥ […] wa-l-balīġ […] wa-l-ḫaṭīb […] wa-l-šāʿir […].15

Whatever the meaning of the title, Ibn Ǧinnī’s work exerts undoubtedly a strong attraction over Ibn Sīda. How to explain it? This amounts to questioning Ibn Sīda’s work in order to discover the elective affinities which could have united him to the object of his plagiarism.

Two major points are common to both authors and seem to explain these affinities. The first one is their epistemological concern and their universalism. What I call universalism here is a constant concern to formulate the questions in the most general way that is not limited to the Arabic language. This point is correlated to another one: their great interest in speculative sciences, and in Aristotelian logic in particular. The interest in logic is more visible in Ibn Sīda’s work. The second point that may explain their affinities is their particular way of being fascinated by the Arabic language. Due to the lack of space, I will leave this point aside, despite its great importance.

Ibn Ǧinnī’s interest in speculative sciences and epistemological issues is the very foundation of al-Ḫaṣāʾiṣ. This interest is read on every page and from the first pages. Indeed, from these first pages, Ibn Ǧinnī warns us that he has little intention to treat in this work the questions of nominative or accusative, but his purpose will be to identify the principles of this science and the way whose general principles apply to particular cases.16 The centrality of notions of ʿilla (“cause”),17 of ḥikma, etc., are sufficient to prove it to us.

Ibn Sīda’s interest in speculative sciences is much less known. Contemporary scholars who have studied him insist on its grammatical and lexicographic culture. Talbi (1921-2017), for example, who established an index of al-Muḫaṣṣaṣ and remains the essential reference concerning Ibn Sīda, presents him in EI2 as “a philologist and a lexicographer”, his life being “entirely devoted to philology and lexicography”. Some ancient historians, like Ibn Ḫallikān (d. 681/1282), even consider that he is simply a ḥāfiẓ endowed with a prodigious memory and who, blind, would have kept in memory a multitude of works.

These judgments hardly seem to do justice to his intellectual stature. Already, in the introduction to both of his dictionaries, Ibn Sīda himself insists that his knowledge is far from being limited to lexicography and grammar. He mentions in al-Muḫaṣṣaṣ his mastery of “all the species of ʾādāb: grammar, metrics, rhymes, genealogy, chronicles, and of the sciences of kalām (al-ʿulūm al-kalāmiyya)”.18 These latter sciences, he says, are “exceptional among the men of letters and makes me superior to them”.19 But what are these ʿulūm kalāmiyya? The statement in al-Muḥkam is clearer:

I consider lexicography as the lightest of my intellectual baggage, and the simplest of the arts that I master, if I compare it to my knowledge of the truths of grammar, the difficulties of the metrics, the subtleties of rhyme, the multiple forms of logical figures, the dialectical sciences, sciences that forbids me to transmit the rejection in which I find the children of the century. (Ibn Sīda, Muḥkam, I, p. 49)

We have other testimonies about this mastery: the most important is given by a contemporary of Ibn Sīda, the historian and judge Ṣāʿid al-ʾAndalusī (d. 462/1070), in his Ṭabaqāt al-ʾumam where he classifies the sciences in the nations of antiquity. Ṣāʿid classes Ibn Sīda among the scholars who were particularly concerned with logic, and mentions an important book in logic he wrote following Mattā b. Yūnus’ (d. 328/939) doctrine.20 Mattā b. Yūnus, as we know, is the Nestorian Christian philosopher, translator of and commentator on Aristotle, one of the principal initiators of the reception of the Peripatetic philosophy in the Arab medieval culture. Two centuries later, another Andalusian author, Ibn Saʿīd al-Maġribī (d. 685/1286) considers in al-Muġrib fī ḥulā al-Maġrib, that “he is far from being only a ḥāfiẓ and a scholar”.21 Finally, al-Ḏahabī (d. 748/1348), in his Siyar ʾaʿlām al-nubalāʾ, considers that “he is one of those whose intelligence is proverbial”.22 Two other judgments, reported by al-Ḏahabī, seem to indicate the kind of criticisms addressed to people interested in philosophy: ʾAlīsaʿ/al-Yasaʿ b. Ḥazm (d. 575/1179) considered that “he was a šuʿūbī23 who preferred non-Arabs to Arabs”24 and ʾAbū Zayd al-Suhaylī (d. 581/1185) in his al-Rawḍ argued that al-Muḥkam contains big mistakes, notably in religion.25

And indeed, from the very first lines of al-Muḫaṣṣaṣ, we are faced with a dictionary whose culture and tone are different. Its speculative universalism concern is one of its characteristics.

  • The doxology is a praise of the intellect: Ibn Sīda thanks God for giving men subtle thought and finesse in the speculative examination of things and in the ability to draw lessons.26

  • From the very beginning, there is an immediate reference to Aristotelian logic with its concepts. The nature of language is presented with the help of Aristotelian concepts: species (nawʿ), genus (ǧins) and “differentia” (faṣl):

God who raised in dignity the species called man, who honored it by granting it the virtue of speech/of logos, endowing it with a form that distinguishes it and a differentia which belongs to it and raises it among all species of living beings. (fa-ʾinna Allāh ʿazza wa-ǧalla lammā karrama hāḏā al-nawʿ al-mawsūm bi-l-ʾinsān wa-šarrafa-hu bi-mā ʾatā-hu min faḍīlat al-nuṭq ʿalā sāʾir ʾaṣnāf al-ḥayawān wa-ǧaʿala la-hu rasman27 yumayyizu-hu wa-faṣlan yubayyinu-hu ʿalā ǧamīʿ al-ʾanwāʿ, Ibn Sīda, Muḫaṣṣaṣ, I, p. 2)

We know the formal and enthusiastic adoption of logical terms by the grammarians to present their work from the 4th/10th century onwards, after the translation of Greek philosophy. But this adoption, in its most visible aspect at least, remains formal and, it seems to us, is globally much more a matter of rhetorical organization than of a profound modification of the concepts and methods of the discipline.28 It conforms, in fact, to a model of science, a knowledge whose concepts, postulates and methods have many more numerous sources. Moreover, it is observed in grammar and not in lexicography, which follows a very different path since the 2nd/8th century where many theses of lexicography and grammatical theory corresponded.29 Now, Ibn Sīda is a lexicographer and if we compare his work with that of three great Oriental lexicographers that preceded him immediately, Ibn Durayd (d. 321/933), Ibn Fāris (d. 395/1004) and al-Ǧawharī (d. 400/1010), in their respective works Ǧamharat al-luġa, Maqāyīs al-luġa, and al-Ṣiḥāḥ, we find no trace, among these, of any presentation of the foundations of their work that borrows from logic or philosophy.

Actually, some variations between Ibn Ǧinnī’s and Ibn Sīda’s texts come from the type of organization that we have called rhetoric. Thus, Aristotelian science requires definitions. This is what Ibn Ǧinnī does in chapter 2 of al-Ḫaṣāʾiṣ where he defines his object of study, the language:

As for its definition, [here it is]: language is sounds by which each community expresses its needs. (Ibn Ǧinnī, Ḫaṣāʾiṣ, I, p. 33)

Ibn Sīda, which adopts the same definition, always without mentioning Ibn Ǧinnī, adds to it a formal, purely logical concern, justifying both the place of this definition in its presentation and the good formation of this definition:

As for its definition, and we start there because the definition prevails over the description (wa-nabdaʾu bi-hi li-šaraf al-ḥadd ʿalā al-rasm),30 [here it is]: Language is sounds by which each community expresses its needs. This is a well-formed definition that encircles its object (wa-hāḏā ḥadd dāʾir ʿalā maḥdūdi-hi muḥīṭ bi-hi lā yalḥaqu-hu ḫalal) because any sound which expresses an idea conceived in the soul is language and all [that which] is language is sound expressing an idea conceived in the soul. (Ibn Sīda, Muḫaṣṣaṣ, I, p. 6)

But beyond this formal presentation, more philosophical motifs can be discerned. In particular, that of universalism: Ibn Sīda emphasizes that the definition he puts forward – which is that of Ibn Ǧinnī –, is not peculiar to the Arabic language. It applies to all languages:

Let’s talk about its definition. It is general and applies to all languages. (Ibn Sīda, Muḫaṣṣaṣ, I, p. 5)

His definition of the lexicographic science he calls ʿilm al-lisān is not just for the Arabic language but for all languages:

What we are mentioning here is not limited to the Arabic language but it is a definition that applies to the science of any language. (Ibn Sīda, Muḫaṣṣaṣ, I, p. 6)

Moreover, he distinguishes two branches of lexicography: a branch whose object is the particular terms which constitute the treasure of the language, and another branch which establishes the laws which govern these terms. These laws are general propositions (qawānīn tilka al-ʾalfāẓ wa-hiya ʾaqāwīl ǧāmiʿa). A knowledge of the lexicon could not call itself science if it is not intended to establish these general propositions.31 His dictionary is different from others because he seeks also for these general rules.

But it is above all a theory of knowledge in relation to language, presented in a concise and dense manner from the first pages of the text, which shows the philosophical preoccupation. The text links the ability of men to name things and their ability to acquire knowledge: associating the name to the thing, man can distinguish things, as he can make them present in a situation of enunciation, and possess them by their names. Things are discernible because they are made distinct by the sounds of their names, as they are immediately discernible by their nature (ṭibāʿ) and, first, by their forms (ṣuwar) and their conditions (ʾawḍāʿ):

In spite of this, we do not escape the necessity of naming things in order to keep them present by their names and to distinguish them from each other by their sounds and their timbres, as they are differentiated from the outset by their nature and differ, even earlier, by their forms and conditions. (wa-naḥnu maʿa ḏālika lā naǧidu buddan min tasmiyat ǧamīʿ al-ʾašyāʾ li-tuḥtāza bi-ʾasmāʾi-hā wa-yanmāza baʿḍu-hā min baʿḍ bi-ʾaǧrāsi-hā wa-ʾaṣdāʾi-hā ka-mā tabāyanat ʾawwal wahla bi-ṭibāʿi-hā wa-taḫālafat qabla ḏālika bi-ṣuwari-hā wa-ʾawḍāʿi-hā, Ibn Sīda, Muḫaṣṣaṣ, I, p. 3)

In two lines, Ibn Sīda seems to answer two questions: what is the link between language and knowledge? How does man distinguish things?

The paragraph ends with a strong eulogy of the philosophers (al-ḥukamāʾ) for the subtlety of their reflection and finesse concerning the analysis of this question, the philosophers seeming to be in this case the Greek philosophers, and Aristotle in particular. Indeed, the context that immediately follows this passage, which speaks of discrete and continuous quantity (kammiyyatayn muḫtalifatayan munfaṣilatayn ʾaw muttaṣilatayn) recalling the analytical seconds, suggests it:

How excellent are the finesse and subtlety of the philosophers’ views and their art relative to this question. They have always been concerned with clarity, quick to make the manifest and the explicit prevail. (ibid.)

In fact, if we look at it well, this theory already exists in Ibn Ǧinnī in a less systematized way. It is attributed to the proponents of the human institution, in other words to Muʿtazila. The observation is important because it shows that Ibn Sīda takes Muʿtazila theses on knowledge. Is he Muʿtazilī or is he simply influenced by their theses as it was common before al-ʾAšʿarī, especially if their source is common, namely the Peripatetic philosophy?

Actually, Ibn Sīda does not seem Muʿtazilī and the fascination is not total. The fact that Ibn Sīda does not follow Ibn Ǧinnī in his iʿtizāl, that he does not borrow his iʿtizāl, if one dares to say, is revealed by several clues. We cite two here.

(i) The doxology of Ibn Ǧinnī – as Naǧǧār, the editor of the Ḫaṣāʾiṣ, notes – is a profession of faith:

Praise be to God, the One, the Just, the Eternal. (al-ḥamd li-llāh al-wāḥid al-ʿadl al-qadīm, Ibn Ǧinnī, Ḫaṣāʾiṣ, I, p. 1)

The names he gives to God recall precisely the two real pillars of the Muʿtazilī doctrine, the uniqueness of God and the justice of God, that gave the Muʿtazilīs their name (ʾahl al-ʿadl wa-l-tawḥīd).32

These are hardly the same names that Ibn Sīda gives to God in his doxology:

Praise be to God, Almighty, He who gives death. (al-ḥamd li-llāh al-mumīt ḏī al-ʿizza wa-l-malakūt, Ibn Sīda, Muḫaṣṣaṣ, I, p. 2)

(ii) Similarly, the following is immediately different from Muʿtazilī theses:

God puts man, by His omnipotence, in the incapacity […] to define His essence and to grasp His predicates and attributes. (muʿǧizi-hā […]ʿan taḥdīd ḏāti-hi wa-ʾidrāk maḥmūlāti-hi wa-ṣifāti-hi, ibid.)

In other words, God has attributes, although the intellect cannot apprehend them. Now, in the Muʿtazilī doctrine, God has no attributes. The Muʿtazila “ferociously” deny that God can have attributes, to use the words of al-Maqrīzī (d. 845/1442).

So, it seems that the Muʿtazilī theses about knowledge are adopted by Ibn Sīda, as many other theses which have not remained the exclusive property of the school are adopted by other group within Islam. But if the theory of knowledge is indeed the one developed by the proponents of the human institution, i.e. Muʿtazila, what exactly remains of the reading of Ibn Sīda? Where is his contribution, his own views?

Before answering this question, let us ask another question: does this underlined love of philosophy introduce another lexicographic rationality?

It is customary to state that his training as a logician has left no effect on his dictionaries. This statement should probably be reviewed. He himself tells us the opposite:

[…] although I have included in it logical definitions and brought derived cases back to the canonical ones, and particular cases back to general ones, which no other [lexicographer] has yet done and where no one has surpassed me. (Ibn Sīda, Muḫaṣṣaṣ, I, p. 13)

Naṣṣār (1956, I, p. 374) and Baalbaki (2014, p. 325), in their reference studies on the Arabic dictionary, consider that Ibn Sīda is the first one to have a rigorous system for the internal arrangement of the lemmata. Furthermore, Haywood (1960) and Baalbaki (2014) recognize in the notion of bāb in al-Muḫaṣṣaṣ a fundamental unit that groups together several words, which all refer to a major concept. In other terms, it is the equivalent of the notion of semantic field,33 so as Baalbaki, relying on the terminology of Ibn Sīda, calls mubawwab any thematic dictionary in Arabic.34

Is this organization due to logic or to grammar? In al-Muḫaṣṣaṣ, Ibn Sīda praises the unique arrangement of his dictionary due to logic where the general (ʾaʿamm) precedes the specific (ʾaḫaṣṣ), universals (kulliyyāt) precede particulars (ǧuzʾiyyāt) and substance (ǧawhar) precedes accident (ʿaraḍ). Similarly Grande (2017) considers as a key factor to understanding the methodology of al-Muḥkam the parallel that the Andalusian lexicographer draws between the term bāb and the philosophical taxonomical term ǧins, i.e. the genus: other lexicographers, says Ibn Sīda, assign a given attribute of a bāb to many species, whereas I assign it to a genus [overarching the species]; in this manner I dispense with mentioning specific cases, mentioning the general case.35 All this would suggest that Ibn Sīda is to lexicology what Ibn al-Sarrāǧ has been, one century before, to grammar. This is a hypothesis that deserves consideration.

3 The aim and the structure of the work

But what exactly are the theoretical differences between Ibn Ǧinnī and Ibn Sīda as to the nature of language? Loucel considers that “the variants between the two texts are not lacking”. However, except for the one concerning the missing section on waqf, “on the whole they bring nothing new to the thought of Ibn Ǧinnī” (Loucel 1964, p. 58). We do not agree with this analysis. Additions and entrenchments are both significant. They will help us to determine the reading of Ibn Sīda.

The structure of the first pages of the introduction to al-Muḫaṣṣaṣ shows us two differences between the two authors, one theoretical and the other pragmatic, as to the purpose of their respective works.

As for the theoretical, it is indicated from the first lines. Immediately after the doxology, Ibn Sīda presents us with an apparent paradox: as God has raised in dignity the human species granting it, to the exclusion of other living species, the virtue of logos, he has put man in need of expressing himself. So, expressing oneself is necessary, obligatory (al-luġa iḍṭirāriyya)36 whereas – and this is the paradox – the relation between sound and meaning does not come under any necessity, but falls under the fortuitous conventional choice (mawḍūʿāt ʾalfāẓi-hā iḫtiyāriyya). In sum, it is arbitrary:

God who raised in dignity the species called man, who honored it by granting it the virtue of speech/logos, endowing it with a form that distinguishes it and a differentia which belongs to it and raises it among all species of living beings, has placed man in the need to reveal the ideas (al-maʿānī) conceived in the soul and apprehended by the intellect. So, he loosened tongues with all kinds of sensitive sounds, so that they are the form of what is stirring the soul and which was conceived by it. We know from this that language is necessary, although the meaning/concept that corresponds to sound (mawḍūʿāt ʾalfāẓi-hā) is left to [a conventional] choice/arbitrary. (Ibn Sīda, Muḫaṣṣaṣ, I, p. 2-3)

Ibn Sīda illustrates this thesis by an example: if the first institutor of the language (al-wāḍiʿ al-ʾawwal) had called the white colour “black” and the black colour “white”, if he had called the smallest quantity “totality” and the largest “particle”, he would not have scratched our ears, nor blurred the concept:

For the first institutor (al-wāḍiʿ al-ʾawwal) who named the small quantity ǧuzʾ and the large quantity kull, the color that scatters the light rays perceived by the gaze, diffuses them and extends them bayāḍ and the color that picks them up and restrains them sawād, if he had overthrown these denominations and called the small amount kull and the totality of the amount ǧuzʾ, he would not have detracted from the concept (mawḍūʿ) nor scratched our ears. (ibid., p. 3)

These introductory lines are an unequivocal affirmation of what we call, following Saussure (1857-1913), the arbitrary character of the sign. It is a clear statement that the relation between sound and meaning is a conventional given, the choice of such concept for such sound being arbitrary. This is the first paragraph of al-Muḫaṣṣaṣ and the first thesis of Ibn Sīda in his introduction. One can, of course, imagine that a lexicographer needs to present as a first proposition of his work the relation between sound and meaning. But, at the same time, this first proposition is an implicit refutation of what Ibn Ǧinnī sees as the most probable origin of language, namely the sounds of nature that man would have imitated.

And if we look carefully at what has been removed from Ibn Ǧinnī’s chapter, we immediately find confirmation of our hypothesis, for we see that a fundamental paragraph of this chapter that comes in the immediate aftermath of the exposition of the two theses on the origin of language has been removed and is missing in the introduction of Ibn Sīda. This is precisely the passage in which Ibn Ǧinnī states that all languages can originate from the sounds of nature:

Some people have argued that all languages originate from the sounds of nature heard by [man], such as the howling of the wind, the roaring of the thunder, the murmuring of the water, the braying of the donkey, the croaking of the raven, the neighing of the horse, the belling of the deer, and so on. In my opinion, this is a correct view, and an acceptable thesis. (Ibn Ǧinnī, Ḫaṣāʾiṣ, I, p. 46-47)37

In other words, Ibn Ǧinnī denies – at least partially – the arbitrariness of the sign. We all know that he went further, trying to develop a theory of phonetic symbolism (al-ištiqāq al-ʾakbar). The entrenched paragraph is therefore not an innocuous one. He asserts a fundamental thesis of Ibn Ǧinnī.

How does the arbitrary character of the relation between sounds and meaning influence the conception of an analogical dictionary? This is more understandable if we put Ibn Sīda’s theory in perspective, in the context of the debate on the origin of speech. Al-Rāzī, for instance, in his book of ʾuṣūl al-fiqh, presents four theories on the origin of speech in Muslim thinking, the first one, the naturalist theory, being that language has its origin in the natural affinity (munāsaba ṭabīʿiyya) between expressions and the things they signify.38 “Language, on this theory, is born of a natural human inclination to imitate the sounds of nature”, as Weiss (1974, p. 34), who summarizes the debate on the origin of speech and tries to reconstruct its history, says. In other words, it is by its own structure that the language in this case signifies. A name tells us something about the thing it refers to: al-lafẓ mufīd li-l-maʿnā li-ḏāti-hi (Rāzī, Maḥṣūl, I, p. 181). The leading proponent of this position is ʿAbbād b. Sulaymān al-Ṣaymarī (d. 250/864). Al-Suyūṭī (d. 911/1505) links the naturalist theory with the early philologists: they all recognize a natural correspondence between sounds and meaning:

As for lexicographers and grammarians, almost everyone agrees on the existence of a correspondence/harmony between the sounds (al-ʾalfāẓ) and the concepts (al-maʿānī) to which these sounds refer. However, the difference between their doctrine and ʿAbbād’s one is that the latter considers this correspondence intrinsic to the language and obligatory, unlike the first. (Suyūṭī, Muzhir, I, p. 47)

In al-Ḫalīl’s (d. 175/791?) introduction to Kitāb al-ʿAyn, this natural correspondence between sound and meaning is precisely called ḥikāya (“imitation”). But, as al-Suyūṭī observes, there is an important difference: this relation between sound and meaning is inherent to the language and obligatory for ʿAbbād, while its domain is limited for grammarians and lexicographers. For al-Ḫalīl, for instance, this phenomenon is restricted in its scope and does not constitute the structure of the language. The best proof of it is that these onomatopoeic words have more flexible phonological rules than the other words (Ḫalīl, ʿAyn, p. 54-55). Al-Rāzī and al-Suyūṭī claim that the naturalist theory was abandoned by Muslim scholars in general (ʾankara al-ǧumhūr hāḏihi al-maqāla, Suyūṭī, Muzhir, I, p. 47). Two arguments were given against it: all languages would be known by everyone if that was true; and there would be no enantiosemy (ʾaḍdād) in language.

If we now return to al-Muḫaṣṣaṣ and look at the plan of the introduction, we immediately understand why the paragraphs 2 and 3 are devoted to synonymy, homonymy and enantiosemy. As Ibn Sīda explains later in the introduction, this dictionary is thematically arranged in order to provide the orator or poet with several nouns or qualificatives relating to a certain semantic field, thus allowing him to choose the one that best suits his needs or fits the required rhyme. In other words, the lexicographer bases his whole dictionary on the existence of synonymy, homonymy/polysemy and enantiosemy in the language. And he begins by asserting it. But if there was a natural relation between the form and the thing expressed, there would be no enantiosemy: and as enantiosemy is a special case of homonymy/polysemy, according to Ibn Sīda, there would be no homonymy either. In sum, for Ibn Sīda, the existence of its own dictionary depends on the refutation of the naturalist theory, adopted – partially – by Ibn Ǧinnī: the form of word says nothing about things and concepts.

The link between this debate and the debate of theologians about the origin of speech is explicitly made by Ibn Sīda himself. We find in this same paragraph a new praise of the ḥukamāʾ, either those of the revelationist or the conventionalist theories, for their perspicacity and the clarity of their debates. Actually, he defends them against those who accuse them of exaggeration in their disputes, thus sowing confusion and doubt.39 Moreover, Ibn Sīda contributes himself to the debate, as he announces in his introduction, and in a special section inside the dictionary, that he develops an argumentation seeking to prove the existence of enantiosemy. We know in this book that at his time there were still scholars who denied their existence – so was one of his teachers. The point is that, according to Ibn Sīda, the existence of enantiosemy means that the relations between sound and meaning, between sound and thing, are arbitrary.

***

Loucel is right when he says that the most important thing in this introduction is not the passage borrowed from Ibn Ǧinnī. This borrowing is a way for Ibn Sīda to say that he agrees with Ibn Ǧinnī on the presentation of the debate. And even if Ibn Sīda does not mention the final passage of the non-committal position, we are not sure that he does not share this position. Many clues suggest the opposite. We simply quote two: (i) as we saw above, he speaks in his introduction about “the first institutor” (al-wāḍiʿ al-ʾawwal) and not God, as if the question was open; (ii) the most striking clue is the assertion that he added to Ibn Ǧinnī’s text and inserted in it. Actually, he says in this passage, in his own philosophical language, the necessity to stick to what the intellect can grasp:

It would be appropriate for the one who examines the question with finesse and impartiality, wishing to give justice to the debate […], to stand where the understanding stopped. He then has the obligation to abstain […]. (wa-qad yanbaġī li-l-mutaʾammil al-munṣif wa-l-daqīq al-naẓar ġayr al-mutaʿassif […] ʾan yaqifa ḥayṯu waqafa bi-hi al-ʾidrāk fa-waǧaba ʿalay-hi ʿinda ḏālika al-ʾimsāk […], Ibn Sīda, Muḫaṣṣaṣ, I, p. 5)

Why does Ibn Sīda cut the passage of Ibn Ǧinnī and adds his own? Question difficult to answer. We can nevertheless observe that his own passage is more difficult to understand. It is as if he wanted to say his opinion without being heard. Is it a courtesy for the conservative views of his ‘milieu’ around him?40 Is it the same reason that leads him to end Ibn Ǧinnī’s text by the affirmation of the divine Revelation of the language? Undecidable questions but still likely hypotheses.

Whatever the answer, the text itself shows us that the main aim of Ibn Sīda is elsewhere: it is to assert the propositions on which his dictionary is based. After the passage borrowed from Ibn Ǧinnī and the definition of the language, Ibn Sīda comes back to the first idea of the necessity of expression. So, dictionaries are necessary and he presents his own approach.

In order to conclude, I would like to ask why Ibn Ǧinnī mentions, besides ʾAbū ʿAlī al-Fārisī, ʾAbū al-Ḥasan [al-ʾAḫfaš al-ʾAwsaṭ]41. In fact, al-ʾAḫfaš was a Muʿtazilī theologian too. He was a pupil of ʾAbū Šamir (d. between 204/819 and 218/834), a Muʿtazilī theologian, as mentioned by al-Suyūṭī (Muzhir, II, p. 454). The point is that al-ʾAḫfaš was a pupil of Sībawayhi too. It means that Sībawayhi knows all these debates, and the clear relation between this question and theologians, as shown by Ibn Sīda’s introduction, sheds a new light on the epistle of Sībawayhi’s Kitāb, where chapter 4 hāḏā bāb al-lafẓ li-l-maʿānī treats of synonymy and homonymy. The reader does not understand why this chapter is in the epistle, as the Kitāb does not come back to it. Would this be an assertion of the arbitrary character of the relation between sounds and meaning in the general case and a criticism of those who profess the naturalist theory before ʿAbbād b. Sulaymān? This is a hypothesis that deserves consideration.

Primary Sources

  • Ḏahabī (al-), Siyar = Šams al-dīn Muḥammad b. ʾAḥmad b. ʿUṯmān al-Ḏahabī, 1401-1409/1981-1988. Siyar ʾaʿlām al-nubalāʾ, ed. Šuʿayb al-ʾArnāʾūṭ et al., Beirut, Muʾassasat al-risāla, 25 vol. [Google Scholar]
  • Ǧawharī (al-), Ṣiḥāḥ = ʾAbū Naṣr ʾIsmāʿīl b. Ḥammād al-Ǧawharī, 1990. Tāǧ al-luġa wa-ṣiḥāḥ al-ʿarabiyya, ed. ʾAḥmad ʿAbd al-Ġafūr ʿAṭṭār, Beirut, Dār al-ʿilm li-l-malāyīn, 6 vol. [Google Scholar]
  • Ḫalīl (al-), ʿAyn = ʾAbū ʿAbd al-Raḥmān al-Ḫalīl b. ʾAḥmad al-Farāhīdī, 1980-1985. Kitāb al-ʿAyn, ed. Mahdī al-Maḫzūmī and ʾIbrāhīm al-Samarrāʾī, Bagdad, Dār al-rašīd, 8 vol. [Google Scholar]
  • Ibn Durayd = ʾAbū Bakr Muḥammad b. al-Ḥasan Ibn Durayd, 1987-1988. Ǧamharat al-luġa, ed. Ramzi Munīr Baalbakī, Beirut, Dār al-ʿilm li-l-malāyīn, 3 vol. [Google Scholar]
  • Ibn Fāris, Ṣāḥibī = ʾAbū al-Ḥusayn ʾAḥmad Ibn Fāris, 1963. Al-Ṣāḥibī fī fiqh al-luġa wa-sinan al-ʿArab fī kalāmi-hā, ed. Mustafā al-Šuʿaymī, Beirut, Muʾassasat Badrān li-l- ṭibāʿa wa-l-našr. [Google Scholar]
  • Ibn Fāris, Maqāyīs al-luġa = ʾAbū al-Ḥusayn ʾAḥmad Ibn Fāris, 2001. Muʿǧam maqāyīs al-luġa, ed. Muḥammad Murʿib and Fāṭima ʾAṣlān, Beirut, Dār ʾiḥyāʾ al-turāṯ al-ʿarabī. [Google Scholar]
  • Ibn Ǧinnī, Ḫaṣāʾiṣ = ʾAbū al-Fatḥ ʿUṯmān Ibn Ǧinnī, 1952-1956. Al-Ḫaṣāʾiṣ, ed. Muḥammad ʿAlī al-Naǧǧār, Cairo, 3 vol. [repr.: Beirut, Dār al-hudā, s.d.]. [Google Scholar]
  • Ibn al-Nadīm = ʾAbū al-Faraǧ Muḥammad b. ʾIsḥāq al-Warrāq Ibn al-Nadīm, 1997. Al-Fihrist, ed. ʾIbrāhīm Ramaḍān, Beirut, Dār al-maʿrifa. [Google Scholar]
  • Ibn Saʿīdal-Maġribī, Muġrib = Ibn Saʿīd al-Maġribī, 1993-1995 [1953-1955]. Al-Muġrib fī ḥulā al-Maġrib, ed. Šawqī Ḍayf, Cairo, Dār al-maʿārif, 2 vol. [Google Scholar]
  • Ibn Sīda, Muḫaṣṣaṣ = ʾAbū al-Ḥasan ʿAlī b. ʾIsmaʿīl Ibn Sīda al-ʾAndalusī, s.d. Al-Muḫaṣṣaṣ, Damas, Dār al-fikr, 5 vol. [Google Scholar]
  • Ibn Sīda, Muḥkam = ʾAbū al-Ḥasan ʿAlī b. ʾIsmaʿīl Ibn Sīda al-ʾAndalusī, 2000. Al-Muḥkam wa-l-muḥīṭ al-ʾaʿẓam, ed. ʿAbd al-Ḥamīd Hindāwī, Beirut, Dār al-kutub al-ʿilmiyya, 11 vol. [Google Scholar]
  • Rāzī, Maḥṣūl = Faḫr al-dīn ʾAbū ʿAbd Allāh Muḥammad b. ʿUmar b. al-Ḥasan b. al-Ḥusayn al-Rāzī, 1997. Al-Maḥṣūl fī ʿilm ʾuṣūl al-fiqh, ed. Ṭaha Ǧābir Fayyād al-ʿAlwānī, Beirut, Muʾassasat al-risāla, 6 vol. [Google Scholar]
  • Ṣāʿid al-ʾAndalusī, Ṭabaqāt = ʾAbū al-Qāsim Ṣāʿid b. ʾAḥmad b. Ṣāʿid al-ʾAndalusī, 1912. Ṭabaqāt al-ʾumam, ed. Louis Cheikho, Beirut, al-Maṭbaʿa al-kāṯulikiyya. [Google Scholar]
  • Suyūṭī (al-), Muzhir = Ǧalāl al-dīn ʿAbd al-Raḥmān b. ʾAbī Bakr al-Suyūṭī, s.d. Al-Muzhir fī ʿulūm al-luġa wa-ʾanwāʿi-hā, ed. Muḥammad ʾAbū al-Faḍl ʾIbrāhīm et al., Cairo, Dār ʾiḥyāʾ al-kutub al-ʿarabiyya, 2 vol. [Google Scholar]
  • Ayoub, Georgine, 2010. “al-fiʿl wa-l-ḥadaṯ : la description sémantique du verbe dans le Kitāb de Sībawayhi” , Jerusalem Studies in Arabic and Islam 37, 1–52. [Google Scholar]
  • ― 2016. “L’emprunt dans le dictionnaire arabe des premiers siècles”, Sartori, Manuel, Giolfo, Manuela E.B. and Cassuto, Philippe (ed.), Approaches to the History and Dialectology of Arabic in Honor of Pierre Larcher, Leiden/Boston, Brill, 289–326 [Google Scholar]
  • ― 2017a. “Langue, mythe et histoire : La ʿArabiyya, langue du Paradis, des prophètes et des tribus de l’Arabie antique”, Ayoub, Georgine and Guetta, Alessandro (ed.), La langue et le sacré, Paris, Geuthner, 73–126. [Google Scholar]
  • Baalbaki, Ramzi, 2014. The Arabic Lexicographical Tradition: From the 2nd/8th to the 12th/18th Century, Leiden/Boston, Brill. [Google Scholar]
  • Blachère, Régis, 1930. “ Un pionnier de la culture arabe orientale en Espagne au xe siècle : Ṣāʿid de Baġdād ”, Hespéris 10, 15–36. [Google Scholar]
  • ― 1973. “ Réflexions sur le développement de la lexicographie arabe ”, Revue de l’Occident musulman et de la Méditerrannée 13-14, 125–129. [Google Scholar]
  • Bohas, Georges, 1981. “ Quelques aspects de l’argumentation et de l’explication chez les grammairiens arabes ”, Arabica 28/2-3, 204–221. [CrossRef] [Google Scholar]
  • EI2 = Bearman, Peri et al. (ed.), 1960-2005. Encyclopaedia of Islam, 2nd ed., Leiden, Brill, 12 vol. [Google Scholar]
  • Enderwitz, Suzanne. “al-S̲h̲uʿūbiyya”, EI2. [Google Scholar]
  • Gardet, Louis.“al-Djubbāʾī”, EI2. [Google Scholar]
  • ― “Kasb”, EI2. [Google Scholar]
  • Gimaret, Daniel. “Muʿtazila”, EI2. [Google Scholar]
  • Grande, Francesco, 2017. “ The Arabic Lexicographer Ibn Sīdah and the Notion of Semantic Field ”, Journal of Arabic and Islamic Studies 17,415–433. [CrossRef] [Google Scholar]
  • Guillaume, Jean-Patrick, 1986. Recherches sur la tradition grammaticale arabe (unpublished), thèse de 3e cycle, Paris, Université de la Sorbonne Nouvelle, 2 vol. [Google Scholar]
  • Haywood, John A., 1960. Arabic lexicography: Its history, and its place in the general history of lexicography, Leiden, E.J. Brill. [Google Scholar]
  • Hindāwī, ʿAbd al-Ḥamīd, 2000. “Muqaddimat al-muḥaqqiq”, Ibn Sīda, al-Muḥkam wa-l-muḥīṭ al-ʾaʿẓam, Beirut, Dār al-kutub al-ʿilmiyya, 3–10. [Google Scholar]
  • Lane, Edward William, 1863. Arabic-English Lexicon, London, Willams and Norgate. [Google Scholar]
  • Loucel, Henri, 1963-1964. L’origine du langage d’après les grammairiens arabes”, Arabica 10, 188-208, 253-281; Arabica 11, 57–72, 151-187. [CrossRef] [Google Scholar]
  • Milner, Jean-Claude, 1978. L’amour de la langue, Paris, Seuil. [Google Scholar]
  • Naǧǧār (al-), Muḥammad ʿAlī, 1952. “Muqaddima”, Ibn Ǧinnī, al-Ḫaṣāʾiṣ, Cairo, vol. I, 5–73 [repr.: Beirut, Dār al-hudā, s.d.]. [Google Scholar]
  • Naṣṣār, Ḥusayn, 1956. Al-Muʿǧam al-ʿarabī, Cairo, Maktabat Miṣr, 2 vol. [repr.: Cairo, 1968]. [Google Scholar]
  • Talbi = Ṭālibī (al-), Muḥammad, 1956. Al-Muḫaṣṣaṣ li-Ibn Sīda. Dirāsa − dalīl, Tunis, al-Maṭbaʿa al-ʿaṣriyya. [Google Scholar]
  • ― “Ibn Sīda”, EI2. [Google Scholar]
  • Thiele, Jan, 2015. “Notes sur l’ašʿarisme d’Abū l-Walīd al-Bāǧī”, Hentati, Nejmeddine (ed.), Contribution d’al-Béji et d’al-Lakhmi à l’évolution du malikisme, Tunis, Université Ez-zitouna, 411–433. [Google Scholar]
  • Versteegh, Kees, 1981. “ La conception des ‘temps’ du verbe chez les grammairiens arabes ”, Analyses Théorie 3, 47–68. [Google Scholar]
  • ― 1997. Landmarks in Linguistic Thought III: The Arabic Linguistic Tradition, London/New York, Routledge. [Google Scholar]
  • Weiss, Bernard G., 1974. “ Medieval Muslim Discussions of the Origin of Language ”, Zeitschrift der Deutschen Morgenländischen Gesellschaft 124/1, 33–41. [Google Scholar]
  • Watt, William Montgomery. “al-Ashʿarī, Abū l-Ḥasan”, EI2. [Google Scholar]
  • ― “ʿAbbād b. Sulaymān”, EI2. [Google Scholar]

Secondary Sources

  • Ḏahabī (al-), Siyar = Šams al-dīn Muḥammad b. ʾAḥmad b. ʿUṯmān al-Ḏahabī, 1401-1409/1981-1988. Siyar ʾaʿlām al-nubalāʾ, ed. Šuʿayb al-ʾArnāʾūṭ et al., Beirut, Muʾassasat al-risāla, 25 vol. [Google Scholar]
  • Ǧawharī (al-), Ṣiḥāḥ = ʾAbū Naṣr ʾIsmāʿīl b. Ḥammād al-Ǧawharī, 1990. Tāǧ al-luġa wa-ṣiḥāḥ al-ʿarabiyya, ed. ʾAḥmad ʿAbd al-Ġafūr ʿAṭṭār, Beirut, Dār al-ʿilm li-l-malāyīn, 6 vol. [Google Scholar]
  • Ḫalīl (al-), ʿAyn = ʾAbū ʿAbd al-Raḥmān al-Ḫalīl b. ʾAḥmad al-Farāhīdī, 1980-1985. Kitāb al-ʿAyn, ed. Mahdī al-Maḫzūmī and ʾIbrāhīm al-Samarrāʾī, Bagdad, Dār al-rašīd, 8 vol. [Google Scholar]
  • Ibn Durayd = ʾAbū Bakr Muḥammad b. al-Ḥasan Ibn Durayd, 1987-1988. Ǧamharat al-luġa, ed. Ramzi Munīr Baalbakī, Beirut, Dār al-ʿilm li-l-malāyīn, 3 vol. [Google Scholar]
  • Ibn Fāris, Ṣāḥibī = ʾAbū al-Ḥusayn ʾAḥmad Ibn Fāris, 1963. Al-Ṣāḥibī fī fiqh al-luġa wa-sinan al-ʿArab fī kalāmi-hā, ed. Mustafā al-Šuʿaymī, Beirut, Muʾassasat Badrān li-l- ṭibāʿa wa-l-našr. [Google Scholar]
  • Ibn Fāris, Maqāyīs al-luġa = ʾAbū al-Ḥusayn ʾAḥmad Ibn Fāris, 2001. Muʿǧam maqāyīs al-luġa, ed. Muḥammad Murʿib and Fāṭima ʾAṣlān, Beirut, Dār ʾiḥyāʾ al-turāṯ al-ʿarabī. [Google Scholar]
  • Ibn Ǧinnī, Ḫaṣāʾiṣ = ʾAbū al-Fatḥ ʿUṯmān Ibn Ǧinnī, 1952-1956. Al-Ḫaṣāʾiṣ, ed. Muḥammad ʿAlī al-Naǧǧār, Cairo, 3 vol. [repr.: Beirut, Dār al-hudā, s.d.]. [Google Scholar]
  • Ibn al-Nadīm = ʾAbū al-Faraǧ Muḥammad b. ʾIsḥāq al-Warrāq Ibn al-Nadīm, 1997. Al-Fihrist, ed. ʾIbrāhīm Ramaḍān, Beirut, Dār al-maʿrifa. [Google Scholar]
  • Ibn Saʿīdal-Maġribī, Muġrib = Ibn Saʿīd al-Maġribī, 1993-1995 [1953-1955]. Al-Muġrib fī ḥulā al-Maġrib, ed. Šawqī Ḍayf, Cairo, Dār al-maʿārif, 2 vol. [Google Scholar]
  • Ibn Sīda, Muḫaṣṣaṣ = ʾAbū al-Ḥasan ʿAlī b. ʾIsmaʿīl Ibn Sīda al-ʾAndalusī, s.d. Al-Muḫaṣṣaṣ, Damas, Dār al-fikr, 5 vol. [Google Scholar]
  • Ibn Sīda, Muḥkam = ʾAbū al-Ḥasan ʿAlī b. ʾIsmaʿīl Ibn Sīda al-ʾAndalusī, 2000. Al-Muḥkam wa-l-muḥīṭ al-ʾaʿẓam, ed. ʿAbd al-Ḥamīd Hindāwī, Beirut, Dār al-kutub al-ʿilmiyya, 11 vol. [Google Scholar]
  • Rāzī, Maḥṣūl = Faḫr al-dīn ʾAbū ʿAbd Allāh Muḥammad b. ʿUmar b. al-Ḥasan b. al-Ḥusayn al-Rāzī, 1997. Al-Maḥṣūl fī ʿilm ʾuṣūl al-fiqh, ed. Ṭaha Ǧābir Fayyād al-ʿAlwānī, Beirut, Muʾassasat al-risāla, 6 vol. [Google Scholar]
  • Ṣāʿid al-ʾAndalusī, Ṭabaqāt = ʾAbū al-Qāsim Ṣāʿid b. ʾAḥmad b. Ṣāʿid al-ʾAndalusī, 1912. Ṭabaqāt al-ʾumam, ed. Louis Cheikho, Beirut, al-Maṭbaʿa al-kāṯulikiyya. [Google Scholar]
  • Suyūṭī (al-), Muzhir = Ǧalāl al-dīn ʿAbd al-Raḥmān b. ʾAbī Bakr al-Suyūṭī, s.d. Al-Muzhir fī ʿulūm al-luġa wa-ʾanwāʿi-hā, ed. Muḥammad ʾAbū al-Faḍl ʾIbrāhīm et al., Cairo, Dār ʾiḥyāʾ al-kutub al-ʿarabiyya, 2 vol. [Google Scholar]
  • Ayoub, Georgine, 2010. “al-fiʿl wa-l-ḥadaṯ : la description sémantique du verbe dans le Kitāb de Sībawayhi” , Jerusalem Studies in Arabic and Islam 37, 1–52. [Google Scholar]
  • ― 2016. “L’emprunt dans le dictionnaire arabe des premiers siècles”, Sartori, Manuel, Giolfo, Manuela E.B. and Cassuto, Philippe (ed.), Approaches to the History and Dialectology of Arabic in Honor of Pierre Larcher, Leiden/Boston, Brill, 289–326 [Google Scholar]
  • ― 2017a. “Langue, mythe et histoire : La ʿArabiyya, langue du Paradis, des prophètes et des tribus de l’Arabie antique”, Ayoub, Georgine and Guetta, Alessandro (ed.), La langue et le sacré, Paris, Geuthner, 73–126. [Google Scholar]
  • Baalbaki, Ramzi, 2014. The Arabic Lexicographical Tradition: From the 2nd/8th to the 12th/18th Century, Leiden/Boston, Brill. [Google Scholar]
  • Blachère, Régis, 1930. “ Un pionnier de la culture arabe orientale en Espagne au xe siècle : Ṣāʿid de Baġdād ”, Hespéris 10, 15–36. [Google Scholar]
  • ― 1973. “ Réflexions sur le développement de la lexicographie arabe ”, Revue de l’Occident musulman et de la Méditerrannée 13-14, 125–129. [Google Scholar]
  • Bohas, Georges, 1981. “ Quelques aspects de l’argumentation et de l’explication chez les grammairiens arabes ”, Arabica 28/2-3, 204–221. [CrossRef] [Google Scholar]
  • EI2 = Bearman, Peri et al. (ed.), 1960-2005. Encyclopaedia of Islam, 2nd ed., Leiden, Brill, 12 vol. [Google Scholar]
  • Enderwitz, Suzanne. “al-S̲h̲uʿūbiyya”, EI2. [Google Scholar]
  • Gardet, Louis.“al-Djubbāʾī”, EI2. [Google Scholar]
  • ― “Kasb”, EI2. [Google Scholar]
  • Gimaret, Daniel. “Muʿtazila”, EI2. [Google Scholar]
  • Grande, Francesco, 2017. “ The Arabic Lexicographer Ibn Sīdah and the Notion of Semantic Field ”, Journal of Arabic and Islamic Studies 17,415–433. [CrossRef] [Google Scholar]
  • Guillaume, Jean-Patrick, 1986. Recherches sur la tradition grammaticale arabe (unpublished), thèse de 3e cycle, Paris, Université de la Sorbonne Nouvelle, 2 vol. [Google Scholar]
  • Haywood, John A., 1960. Arabic lexicography: Its history, and its place in the general history of lexicography, Leiden, E.J. Brill. [Google Scholar]
  • Hindāwī, ʿAbd al-Ḥamīd, 2000. “Muqaddimat al-muḥaqqiq”, Ibn Sīda, al-Muḥkam wa-l-muḥīṭ al-ʾaʿẓam, Beirut, Dār al-kutub al-ʿilmiyya, 3–10. [Google Scholar]
  • Lane, Edward William, 1863. Arabic-English Lexicon, London, Willams and Norgate. [Google Scholar]
  • Loucel, Henri, 1963-1964. L’origine du langage d’après les grammairiens arabes”, Arabica 10, 188-208, 253-281; Arabica 11, 57–72, 151-187. [CrossRef] [Google Scholar]
  • Milner, Jean-Claude, 1978. L’amour de la langue, Paris, Seuil. [Google Scholar]
  • Naǧǧār (al-), Muḥammad ʿAlī, 1952. “Muqaddima”, Ibn Ǧinnī, al-Ḫaṣāʾiṣ, Cairo, vol. I, 5–73 [repr.: Beirut, Dār al-hudā, s.d.]. [Google Scholar]
  • Naṣṣār, Ḥusayn, 1956. Al-Muʿǧam al-ʿarabī, Cairo, Maktabat Miṣr, 2 vol. [repr.: Cairo, 1968]. [Google Scholar]
  • Talbi = Ṭālibī (al-), Muḥammad, 1956. Al-Muḫaṣṣaṣ li-Ibn Sīda. Dirāsa − dalīl, Tunis, al-Maṭbaʿa al-ʿaṣriyya. [Google Scholar]
  • ― “Ibn Sīda”, EI2. [Google Scholar]
  • Thiele, Jan, 2015. “Notes sur l’ašʿarisme d’Abū l-Walīd al-Bāǧī”, Hentati, Nejmeddine (ed.), Contribution d’al-Béji et d’al-Lakhmi à l’évolution du malikisme, Tunis, Université Ez-zitouna, 411–433. [Google Scholar]
  • Versteegh, Kees, 1981. “ La conception des ‘temps’ du verbe chez les grammairiens arabes ”, Analyses Théorie 3, 47–68. [Google Scholar]
  • ― 1997. Landmarks in Linguistic Thought III: The Arabic Linguistic Tradition, London/New York, Routledge. [Google Scholar]
  • Weiss, Bernard G., 1974. “ Medieval Muslim Discussions of the Origin of Language ”, Zeitschrift der Deutschen Morgenländischen Gesellschaft 124/1, 33–41. [Google Scholar]
  • Watt, William Montgomery. “al-Ashʿarī, Abū l-Ḥasan”, EI2. [Google Scholar]
  • ― “ʿAbbād b. Sulaymān”, EI2. [Google Scholar]

1

See EI2, art. “al-Ashʿarī, Abū l-Ḥasan” (W.M. Watt).

2

See EI2, art. “al-Djubbāʾī” (L. Gardet).

3

The name of one of the most important theological schools in Islam, founded in Baṣra in the first half of the 2nd/8th century and known for its rationalism. See EI2, art. “Muʿtazila” (D. Gimaret).

4

See Lane 1863, p. 2811.

5

See Ibn Ǧinnī, Ḫaṣāʾiṣ, I, p. 40-41.

6

See for this argumentation Ibn Ǧinnī, Ḫaṣāʾiṣ, I, p. 44-45.

7

The label “Bahšamite” is derived from the name ʾAbū Hāšim, the founder of the school.

8

See in particular Bohas 1981, p. 215, where reference is made to the notion of ḥikma in relation to the ancient Arabs who intuitively understood the perfection of the harmony of the language.

9

Actually, the studies devoted to Ibn Sīda are very few.

11

We owe the correct reading of this title to an anonymous reviewer, whom we thank. He also drew our attention to the fact that the book is cited in some sources under another title, al-Muġrib, as well as al-Mutaʿāqib can be found in other sources under the title al-Taʿāqub. In fact, as indicated by Naǧǧār (1952, p. 66), Ibn Ǧinnī himself mentions these two works in al-Ḫaṣāʾiṣ: the first under the title al-Muʿrib (Ḫaṣāʾiṣ, I, p. 84), and the second under the title al-Taʿāqub (Ḫaṣāʾiṣ, I, p. 264 and 266). The last title is also mentioned by al-Suyūṭī. Finally, Naǧǧār considers that the title al-Muġrib is a reading error (taṣḥīf).

12

We follow Haywood (1960, p. 65) in his translation.

13

See Lane 1863, p. 746.

14

We did not find a translation of the title in the works that mention or study al-Muḫaṣṣaṣ.

15

See Ibn Sīda, Muḫaṣṣaṣ, I, p. 10, where the exact expression is: kitāban ʾaḍaʿu-hu mubawwaban ḥīna raʾaytu-hu ʾaǧdā ʿalā al-faṣīḥ […] wa-l-balīġ […] wa-l-ḫaṭīb […] wa-l-šāʿir […].

16

Ibn Ǧinnī, Ḫaṣāʾiṣ, I, p. 32.

17

See for this notion Guillaume 1986.

18

Ibn Sīda, Muḫaṣṣaṣ, I, p. 14.

19

Al-ʿulūm al-kalāmiyya al-latī bi-hā ʾabuḏḏu al-muʾallifīn wa-ʾašuḏḏu ʿan al-muṣannifīn; Ibn Sīda, Muḫaṣṣaṣ, I, p. 14.

20

ʿUniya bi-ʿulūm al-manṭiq ʿināyatan ṭawīlatan wa-ʾallafa fī-hi taʾlīfan kabīran mabsūṭan ḏahaba fī-hi ʾilā maḏhab Mattā b. Yūnus; Ṣāʿid al-ʾAndalusī, Ṭabaqāt, p. 77.

21

Huwa ʿindī fawqa ʾan yūṣafa bi-ḥāfiẓ ʾaw ʿālim; Ibn Saʿīd al-Maġribī, Muġrib, II, p. 259.

22

ʾAḥad man yuḍrabu bi-ḏakāʾi-hi al-maṯal; Ḏahabī, Siyar, XVIII, p. 144.

23

A movement within the early Muslim society which denied any privileged position of the Arabs. See EI2, art. “al-S̲h̲uʿūbiyya” (S. Enderwitz).

24

Kāna šuʿūbiyyan yufaḍḍilu al-ʿaǧam ʿalā al-ʿarab; Ḏahabī, Siyar, XVIII, p. 145.

25

Ḏahabī, Siyar, XVIII, p. 145.

26

Laṭīf al-fikra wa-daqīq al-naẓar wa-l-ʿibra; Ibn Sīda, Muḫaṣṣaṣ, I, p. 3.

27

We interpret rasm in the technical sense of “form”, and we hence distinguish it from faṣl.

28

Though fundamental concepts of grammatical theories were changed in the 4th/10th century after the adoption of the Aristotelian model. A good example of this is the apprehension of the category of time; see Versteegh 1981 and Ayoub 2010.

29

See for some of these correspondences Baalbaki 2014 and Ayoub 2016.

30

We understand rasm here in its technical sense in logic, in opposition to ḥadd.

31

Ibn Sīda, Muḫaṣṣaṣ, I, p. 14.

32

See EI2, art. “Muʿtazila” (D. Gimaret).

33

This notion exists in many other Arabic thesauri. See the references mentioned above.

34

See Baalbaki 2014, p. 47-48, 276 and 323: “It has been pointed out that Ibn Sīda makes a clear distinction between mubawwab and muǧannas lexica – two terms we owe to him in referring to onomasiological and semasiological lexica respectively.”

35

Ibn Sīda, Muḥkam, I, p. 37-38 and Grande 2017, p. 426.

36

We thank an anonymous reviewer for having drawn our attention to the proximity of this notion with that of ʿilm ḍarūrī. Indeed, since its formative phase, Muslim theology (ʿilm al-kalām) distinguishes between two kinds of “certain science”: a so-called “constraint” or “forced” one (ʿilm ḍarūrī), that man possesses without intervening actively, and the other “acquired” (muktasab, kasbī/iktisābī); see EI2, art. “Kasb” (L. Gardet), and Thiele 2015. This division is already adopted by al-ʾAšʿarī. Al-Bāqillānī defines the ʿilm ḍarūrī in these terms (we borrowed the definition from Thiele 2015, p. 8): ʿilm yalzamu nafs al-maḫlūq luzūman lā yumkinu-hu maʿa-hu al-ḫurūǧ ʿan-hu wa-lā al-infikāk min-hu. If we compare this division to what Ibn Sīda does, we can say that like the ʿilm ḍarūrī, the necessity of expression, i.e. the language ability, is “forced” (iḍṭirāriyya). It is a given whereas this or that determined language is muktasab since the ʾalfāẓ are subject to choice (iḫtiyāriyya).

37

See, for a translation and a discussion of this passage, Versteegh 1997, p. 110-111 sqq.

38

Rāzī, Maḥṣūl, I, p. 181 sqq.

39

Mubarriʾan al-ḥukamāʾ al-mutawāṭiʾīn ʿalā al-luġa ʾaw al-mulhamīn ʾilay-hā min al-tafrīṭ wa-munazzihan la-hum ʿan raʾy man wasama-hum fī ḏālika bi-l-ʾilbās wa-l-taḫlīṭ; Ibn Sīda, Muḫaṣṣaṣ, I, p. 3.

40

See for instance Ayoub 2017 for the theological opinion on the Arabic language and its origin of an Andalusian contemporary of Ibn Sīda, Ibn ʿAbd al-Barr (d. 463/1070).

41

Ibn Ǧinnī, Ḫaṣāʾiṣ, I, p. 41.

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.