ARABIC LITERATURE: GRAMMAR AND LEXICOGRAPHY
The intellectual activity of the early Muslims stemmed directly from their devotion to religion. The Arabs had throughout been sensitively proud of their language; contacts with foreigners were regarded by them as derogatory to pure Arabism. However, before Islam any corruption of the dialect was but a social drawback; after Islam any lapse from the norm inevitably led to distortion of the sacred text with dire consequences both in this as well as in the next world. Curiously enough, it was Islam itself which brought about the commingling of the Arabs with the non-Arabs on a vast and unprecedented scale. In the very second decade of the Hijrah the Arabs were carried on the crest of a wave of military conquests across the bounds of their homeland to settle down in the neighbouring countries of
A History of Muslim Philosophy
Arabic Literature: Grammar and Lexicography
of these neophytes in the use of the Arabic tongue excited the laughter of the younger folk in Arab households; it also shocked the elders as it amounted to inadvertent profanity and distortion of the Qur'anic verses.' The corruptive effects on the new generation of the Arabs-the townsmen among themwere no less disconcerting; the daily usages marked a sharp decline from the Qur'anic idiom. Thus, there is little doubt that about the middle of the first century of the Hijrah the Muslims were squarely face to face with their foremost literary problem, viz., the need for the preservation of the Qur'an. The Arabs needed reinforcing their own natural way of speech with a discipline of conscious effort; they were also eager, in keeping with the true spirit of Islam, to pass on to the myriads of non-Arabs, who daily swelled the ranks of the faithful, not only the religion and the practices of Islam but also the language as a key to a first-hand knowledge of its primary source or sources.2 Actually, however, only a few of the Arabs concerned themselves with those branches of studies which involved the use of the method of qiyds, i. e., analogy and deduction.3 Such creative intellectual activity was notably a flair of the non-Arab inhabitants of Iraq, which province occupied a unique position in the incipient literary life of Islam. It is worthwhile recalling that the province had been the cradle of ancient civilizations and the nursery of cultural currents from the Hellenes, including those relayed from the important academy at Jundi-Shapur; hence, the mental attitudes of its inhabitants bore the stamp of philosophical and scientific discipline. Still more remarkable was the spirit motivating the political relationship of these "intellectuals" with their proud and unlettered masters, the Arabs, and their peculiar religious and cultural propensities towards Islam and the Arabic language. In contrast with Syria and Egypt, it will be seen that the 'Ajamis of Iraq were from the very beginning determined to assert their own individuality, albeit only within the pale of Islam and on the ground of Arabs' own devotion to the Arabic language. Even the Shu'ubiyyah movement, the outburst of an outraged sense of superiority of the Persians over the Arabs, involved no resilience from loyalty to the language of the Qur'an. It was a clear parallel to early Shi'ism, which was calculated to work out the political ascendancy of the Persians but only under the supreme and authoritarian overlordship of the House of the Arabian Prophet. Basrah and Kiifah, the two cantonments of the Arabs, provided ideal conditions for fruitful contact between the Arabs and the non-Arabs. Of particular importance was the proximity of the two towns to the northern
4 This is amply borne out by the different versions of what prompted abu
al-Aswad al-Du'ali to turn to grammar.
2 It is noteworthy that abu al-Aswad al-Du'ali, who showed himself genuinely
anxious to help the non-Arabs learn Arabic and Islam, did so in spite of his jealousy of their prosperity and influence. There was not the slightest trace of
any tendency among the Arabs to sit Brahman-like over the treasures of religious knowledge.
4 This applies equally to grammar and to al-rd i in the realm of Fiqh.
norm, and the market-place of al-Mirbad-on the outskirts of Basrah-was no less a close-by rendezvous of the A'rdb (Bedouin Arabs of the desert) and the literati until the former, becoming aware of the demand, themselves came to offer their linguistic materials to the elite of Iraq and western Persia.
According to the classical tradition, it was abu al-Aswad (Zalim b. 'Amr) al-Du'ali (or al-Dili), a poet, warrior, and teacher (died in 69/688-89 at the age of 85), who took the first step to stem the tide of growing laxity and error in the use of the Arabic tongue. He was an active partisan of 'Ali in politics and actually fought against Mu'awiyah at Siffin. It is, therefore, no surprise that he should take pride in claiming that the rudiments of Arabic grammar were confided to him by 'Ali. This assertion can safely be dismissed as only an instance of the too frequent attempt to trace all learning to 'Ali, the "Gateway of the City of
(i) the vowel "a," the pronunciation of which needs a full upward opening (fathah) of the mouth, to be marked with a dot above a letter.
(ii) the vowel "i," the pronunciation of which needs a little downward movement (kasrah) of the, mouth, to be marked with a dot below the letter.
4 Encyclopaedia o/ Islam, "Abu al-Aswad."
I In the Islamic literary tradition, the written book long continued to serve merely as an aide memoire-a copy of what was preserved in memory and not vice versa.
A History of Muslim Philosophy
Arabic Literature: Grammar and Lexicography
(iii) the vowel "u," the pronunciation of which needs a rounded closing (dammah) of the lips, to be marked with a dot in front of the letter.6
This system of dots is to be seen in one of the oldest copies of the Qur'an dated 77/696, now preserved in the National Library at
the usual practice. It has been noted that a similar system of dots was in use in the writing of Syriac, and, though abu al-Aswad's contacts with the Syrians
are not expressly alluded to, it stands more than probable that having realized the urgency he turned round and took the cue from his compatriots of the Syrian
It is also possible, as some reports make out, that abu al-Aswad went a
step further to propound some broad distinctions in the main parts of a sentence such as the subject and the predicate. On the whole, however, his contribution was merely to focus attention on the usage of vowel-endings as the distinctive
characteristic of Arabic. Hence, observation of vowel-endings was desig
nated al-'Arabiyyah, i.e., the art of speech in the correct and characteristic Arab way. The use of vowel-endings itself was known as al-i'rab, i, e., rendering
into the proper Arabic way.8 The al-'Arabiyyah was undoubtedly an embryonic
form of Arabic grammar.
The emphasis on al-'Arabiyyah grew in proportion to the need for saving
the Qur'an from being consigned to antiquity. So far the method used was mere taigin, i.e., putting the particulars in the mouth of the pupil. Only the necessary terms and signs for indicating the different vowels in speech and writing had been devised. As yet there was no ta'lil or reasoning on the basis of general principles governing the incidence of the i'rdb. But certainly the
i'rab was under intense and searching observation, from which it was not a far step to collecting a number of analogous examples and inducting from them some rules for general guidance. This was the beginning of the discovery
6 It will be remarked that the other synonymns such as nasb, jarr, and rat' also refer to the same varied movement of the mouth. Closely parallel to the Arabic terms are the Persian equivalents: zir, zabar, and piis_k.
v The Syrian Christians of the West had another system, first introduced in second/eighth century, in which letters of the Greek alphabet (five altogether: Y, E, H, 0, A), instcad of the dots, were used as vowel-marks. At some later date, not exactly ascertained, the Arabs also replaced the dots with letters of their own alphabet albeit in an abbreviated form: _ from 1, ; from S (somewhat doubtful), and L from J. Obviously, the change must have been necessitated by the use of dots for diacritical marks along with their use for vowel-marks. The diacritical marks are said to have been brought into somewhat systematic use at the behest of al-Hajjaj b. Yiisuf, the Governor of Iraq, by Nasr b. 'Asim (d. 89/708), who, remarkably enough, is also reckoned as one of the founders of Arabic grammar. For some time the two kinds of dots were distinguished by the different colours of the ink. The replacement of the vowel-dots with abbreviations of 1, S, and , is sometimes ascribed to al-Khalil b. Ahmad, which is supported by the title Kitdb al-Nagt w-al-Sakl among his works.
8 Al-Suyati, al-ALhbah w-al-Naza'ir,
of the logical structure of the language which, in the words of Sarton, was as much a scientific discovery as, for example, the discovery of the anatomical structure of the human body. This scientific discovery, the Nahw proper, reached the proportions of a separate branch of study at Basrah with 'Abd Allah b. abi Iskiaq al-13adrami (d. 117/736) and his pupil, abu 'Am, 'Isa b. 'Umar al-Thaqafi (d. 149/767). Both the teacher and the pupil were non Arab clients (the latter being the client of none other than Khalid b. alWalid) who relished putting the Arabs to shame on the score of incorrect speech. They had a reputation for boldness in 'ilal w-al-giyas, i.e., induction of causes from an array of analogous examples. Even in the first flush of discovery, they were so confident of the principles arrived at that they did not mind criticizing on their basis the ancient model poets such as alNabig_hkah, not to speak of the contemporary al-Farazdaq. When the latter composed a 'vitriolic satire against his dogmatic critic, ibn abi Isbaq would only retaliate by pointing out a grammatical mistake even in the satirical verse.9 The pupil elaborated the method explicitly, as in discovering principles which held good generally and in listing the deviations as lug4at, i. e., exceptional usages. And it was he who embodied the results in two books said to have been the first on the subject.
It must be noted that al-la4n, i.e., incorrect speech, which gave stimulus to the thought of abu al-Aswad, had by the turn of the first/seventh century assumed alarming proportions. It had percolated to the ranks of the Mite of the Court and the administration as well as the circles of the learned such as the traditionists and the jurists. But the deterioration, far from inducing an attitude of toleration, gave rise to a strong reaction against what was regarded almost as a sin, and there was a determined effort not so much to preserve the purity of the Qur'anic text as to make the ordinary speech conform to the standards of its idiom.10 It was at this very time that al-Nahw, the science of "the proper way of the speech of the Arabs" (ibn Jinni), was fully recognized as an independent branch of study and the term al-nahwi became widespread in popular parlance."
The Basrah school reached its perfection in the following age, which produced such giants as al-Khali1 and Sibawaihi. Al-Khalil b. Abmad, a truly versatile genius of Arab descent (al-Furhiidi/al-Farahidi, al-Azdi), whose contribution alone would outweigh the achievement of the host of non-Arabs, was born in 100/718-19 and died some time between 170/786 and 175/791. There can be no greater testimony to his high powers of originality than the discovery of Arabic prosody without any previous pattern, taking his cue merely from the rhythmic beats of the smith's hammer. No surprise that after benefiting from the teachings of 'Isa b. 'Umar, he should have been able to elaborate the framework of Arabic grammar, a framework within which
9 Al-Jumahi, Tabagat, Dar al-Ma'arif,
10 J. Fuck, al-'Arabiyyah (Arabic translation),
A History of Muslim Philosophy
Arabic Literature: Grammar and Lexicography
al-i'rab could be explained and reasoned out. But al-Khalil cared neither for fame nor for material gain; it is said of him that he lived in a state of abject penury while his pupils made a fortune with the learning imbibed from him. It fell to the lot of his Persian pupil Sibawaihi,15 who also had direct contact with Isa b. 'Umar, to complete the work of al-Khalil and to arrange and produce his findings in concrete book form. Sibawaihi (abu Bis_hr 'Amr b. 'Uthman b. Qanbar), a native of Shiraz who died at the young age of about forty years in the last quarter of the second century of the Hijrah, really proved to be another genius for comprehensiveness, if not so much for originality. His Kitdb has throughout the ages been regarded as the final word on Arabic grammar and has become proverbial for its unique position in the field. Those who followed Sibawaihi right down to the present time could only comment upon, remove obscurities from, and arrange and rearrange the materials furnished in the "Book" without adding much to'it.
It has been a vexed question as to whether the main concepts of Arabic grammar are an indigenous growth or they are traceable to some external pattern. Modern scholars have stumbled upon casual resemblances such as those with the Indian Praticakhyas, but they offer no secure ground. for any assumption of borrowing. It must be remembered that the Arabic grammar is concerned mainly with the i'rab, which is a peculiarity of the Arabic language and was actually realized and proudly asserted to be so by the early grammarians. Hence, it is no less misleading to make much of the similarity between the division of a word into "ism," " i'l," and "tar/" in Arabic and the analogous categories in Syriac or Greek. Obviously, the Arabic grammarians had to chalk out and proceed on their own lines and, in fact, they have given us a "fair idea of how they applied their efforts to the problem, which was peculiarly their own. As hinted earlier, they began by observing the various positions of the words in a sentence and the particular i`rab taken by them in those positions.'3 These positions came to be designated by distinctive terms and certain rules were laid as to the i'rab appropriate for those positions. These rules went on developing in the direction of reducing further and further the number of exceptions which would not admit of their general application. What helped the people of Iraq in this undertaking was a flair for `ilal and giyas, which was exhibited in an equal measure in grammatical and literary studies as well as in Fiqh and jurisprudence 14 This flair certainly
11 The reading "Sibuyah" is not supported by comparison with "Niftawaihi," which latter is in no doubt because of its occurrence in the rhyme of a verse. Vide ibn Khallikan, Wetaydt, "Ibrahim b. Muhammad b. 'Arafah."
13 Cf. Fuck, op. cit., pp. 11-12.
14 What distinguished the Filth of abu Hanifah was exactly the same: the probing into the "efficient cause" ('iUah) governing a number of given instances and then applying the same to unforeseen circumstances. The people of the Hijaz were extremely chary of such reasoning and it is no mere chance that they came
to be notorious for their ignorance of grammar. It is remarkable that the opponents of abu Hanifah, who wanted to run down his school of Fiqh, thought it necessary
bears the impress of Hellenism. Nevertheless, it remains a mere conjecture that the early Muslims took over anything specific from Greek sources in grammar, in the same way as it is a mere wishful thought that Fiqh is indebted to anything specific in the Roman Law.'5
The cornerstone of Arabic grammar is the correlation of the i'rab of the different parts of a sentence based on the theory of an 'amid, i.e., an efficient cause supposedly resident in one of the parts and governing the whole. The earliest trace of it is perhaps in the Kitdb al-'Awamil of al-Khalil-a work known to us only by its title. But there is no reason to suppose that al-Khalil diverged in any way from the general line pursued thitherto by 'Isa b. 'Umar and others. Unless, therefore, this 'dmil theory is proved to have been formulated on a familiar pattern, the indebtedness of Arabic grammarians to any external source will remain highly problematic.
There is, however, yet another development of Arabic grammar which is clearly and directly traceable to Greek influence. The most notable and lasting effect of the assimilation of Greek logic and philosophy in the 'Abbasid period was a general tendency to remould into logically defined systems almost all the nascent branches of learning, which until then lacked a rigid order. So far as Arabic grammar is concerned, this development took place when a Mu'tazilite Mutakallim and a natvi were combined in the person of abu alHasan 'Ali b. 'Isa al-Rummani (d. 384/994). Actually, the process must have started with the Kfifan grammarian, al-Farra' (d. 207/822), who was also a Mu'tazilite. Under the patronage of al-Mamfin he produced the Kitab alHudud, which must have been the first attempt, so to say, to "philosophize" Arabic grammar. However, the process reached its culmination in alRummani so as to justify his being credited with that highly conventional logical reasoning which has since formed such a notable feature of Arabic grammar. This new development is amply borne out by a saying that out of the three contemporaries the words of al-Sirafi (abu Sa'id al-Hasan b. 'Abd Allah) were thoroughly understood without a teacher, those of abu 'Ali (al-I3asan b. Ahmad) al-Farisi were only partly so, whereas those of al-Rummani were not intelligible at all.16 Even abu 'Ali al-Farisi, who, according to the above testimony, was himself partly affected by the innovation, is reputed to have commented that if Natw be what was expounded by al-Rummani, then he had nothing to do with it, and vice versa. Undoubtedly, al-Rummani did not bring out a new system of grammar; he only applied the methods and the jargon of Aristotelian logic to the adumbration of those nebulous conceptions which, in the simple language of the old tradition as represented by al-Sirafi, were easily comprehended by the average student. There was a similar
to make fun of the application of his methods to grammar. Cf., Fuck, op.
cit., p. 65.
15 M. Hamidullah, "Influence of Roman Law on Muslim Law"-a paper read before the All-India Oriental Conference, December 1941.
16 Yaqut, Mu'jam al-Udabd', " 'Ali b. 'Isa."
A History of Muslim Philosophy
Arabic Literature: Grammar and Lexicography
transformation in Arabic rhetorics too. Further, it will be noted that by this time the Arabs had acquired some familiarity with Greek grammar, which warranted their indulging in a comparison of its merits with those of Arabic grammar. But the latter was considered to have already possessed a separate entity with a different development.
While the general trend at Basrah was to go ahead with the formulation of general rules, there also developed a reaction against the scant attention paid to the angularities of actual usage, which, however, came to the fore only when abu Ja'far (Mul}ammad b. abi Sarra 'Ali) al-Ru'asiyy took it over as the basis of the rival school of Knfah founded by him in the later half of the second/eighth century. The Kiifans would assiduously collect such instances as violated the general rules of the Basrans and would treat them not as exceptions but as the basis of another general rule opposed to that of the Basrans. This school achieved a meteoric rise in importance under the favour of the 'Abbasid Caliphs. Two of its very influential representatives at the Court were: (a) al-Kisa'iyy (abu al-Iiasan'Ali b. IIamzah), the Persian pupil of both al-Khalil and abu Ja'far al-Ru'asiyy, who came to be regarded as the compeer of Imam abu Yasuf under Harun al-Ras_hid, and (b) al-Farra' (Yabya b. Ziyad), the Dailamite, who was appointed tutor to al-Mamiin's sons and was designated as Amir al-Mil'minin in the realm of al-Nahw. Ultimately, however, Baghdad proved a veritable crucible for the gradual fusion of the two schools through interchange. From the end of the third/ ninth century onwards there flourished at the metropolis scholars who were free from prejudice for or against any particular town or tribe and were actuated by sheer academic interest and reasonableness.
Just one more development may. be noted. Abu 'Ali al-Farisi, who has been mentioned above, had an illustrious pupil called 'Uthman b. Jinni (d. 392/1002), the son of a Greek slave, regarded as the last of the philosophergrammarians. But ibn Jinni did not help in clothing the 'dmil theory with the armoury of logic; rather he submitted the 'dmil theory itself to the scrutiny of reason. The result was a scathing attack on the false notion that one particular word in a sentence governed the whole. The hint dropped by ibn Jinni was picked up in far distant Spain by "ibn Madda'," the Zahirite Qadi of Cordova under the Muwabbids, who in his al-Radd 'ala al-Nuhat attempted something in grammar akin to al-Ghazali's Taha/ut in philosophy. However, his attack, though not lacking in flashes of brilliance, remained a cry in wilderness as no alternative formulation of Arabic grammar on a basis other than the 'amid theory was ever achieved, far less accepted.
To sum up, the inspiration for Arabic grammar came from religion; the need for it was created by the commingling within Islam of the Arabs and the non-Arabs. The methods of observation and induction yielded the discovery of the main body of "laws" in the working of language; the only snag was that the laws of language are not so uniform and immutable as the laws of nature. The older school of grammarians at Basrah suffered from an immature
pedantry which was aggravated by the desire of the non-Arabs among them to outdo the Arabs. At a very early time 'Isa b. 'Umar had the temerity to
boast in the presence of the Arab philologist, abu 'Amr b. al-'
progenitor of the Arabs! And both 'Isa b. 'Umar and abu 'Amr b. al-'Ala' exhibited a tendency to prefer such readings of the Qur'an as, in their opinion, were more in consonance with the general rules of grammar.17 This authori
tarianism on the part of the "wisdom of the school seeking to improve upon
the facts" (Noldeke) was checked by the rise of the rival school of Kiifah. Rather the latter erred on the other extreme; it is said of al-Kisa'iyy that in
his avid search for the unusual and the exceptional he would not pause to
test the reliability of his sources. None the less, a relieving feature of the situation was that dogmatism always felt compelled to bow before actual usage,
as typically exemplified in the contest between Sibawaihi and al-Kisa'iyy at
the Court of Hariin al-Ras_hid.'s Ultimately, Baghdad provided the necessary atmosphere for the gradual shedding of prejudices and the engagement of all
in a joint effort to erect a common edifice large enough to accommodate the conflicting viewpoints on most, if not all, of the established usage. The final
success was vitiated by sporadic attempts at putting possible constructions on actual usage. This tendency was decried at the very start by 'Isa b.'Umar,19
but it reappeared prominently later on and is justly parodied by abu al'Ala' al-Ma'arri in his Risalat al-Ghu/ran.S0 The instruments of Aristotelian logic helped to hammer out the crudities of enunciation and adumbration.
Finally, there was an attempt to rebuild the entire system on a simpler basis other than the 'amid theory, which, however, did not fructify. On the whole, the Arabic grammar remains a magnificent achievement-religious in spirit,
17 It must be pointed out that it was merely a choice from among the various current readings; there was no attempt to "correct" the Qur'an in line with usage elsewhere. As pointed out by Wolfensen, it is an entirely wrong and unscientific approach on the part of some Western scholars to judge and criticize the Qur'an on the basis of pre-Islamic poetry. Apart from any religious sentiment, the Qur'an is the oldest and the most reliable book; other sources, though relating to anterior times, are posterior to it in point of actual compilation. Tarikh al-Lughhat al-Samiyyah, Cairo, 1926, pp. 169 et seq.
11 The reference is to what is known as "al-Mas'alah al-Zunburiyyah." When Sibawaihi challenged al-Kisa'iyy on a point of grammar, the matter had to be referred for decision to the Arabs. It is alleged that the Arabs were bribed to save the face of the royal tutor. The incident affected Sibawaihi so deeply that perhaps it caused his death prematurely.
11 Once when al-Kisa'iyy began giving the various grammatically correct readings of a particular phrase, 'Isa b. 'Umar rebuked him saying: "I want the actual way in which it is spoken by the Arabs." Yaqut, op. cit., "'Isa b. 'Umar," last paragraph. This tendency is to be compared with the hiyal-permissible tricks for evading the Law-in which some of the legists exhibited their acumen.
S0 Al-Ma'arri contrives to bring the grammarians and the poets in the heaven together when the latter protest at the former's purely speculative interpretation of verses, e.g., p. 152 of the Riadlah, ed. Bint al- Shati, Cairo.
A History of Muslim Philosophy
linguistic in material, scientific in methods, and logical in form-which has been eminently successful in preserving the Qur'an and keeping its idiom unchanged yet alive throughout the centuries.
The preservation of the Qur'an involved the institution of such disciplines as would effectively safeguard not only the authentic rendering of the text but also the warranted understanding of its import against error, corruption, or ignorance overtaking those for whom it was "plain Arabic" at the time. The former purpose was achieved through al-'Arabiyyah, which later on developed into a full-fledged science under the name of al-Nahw. The next concern was naturally the meaning conveyed by the text. In the beginning, there could have been little difficulty about it in the same way as about the vocalization of the text which was just a matter of natural aptitude.%' However, with the lapse of time and the changes in the social pattern, uncertainties began to creep in around words and expressions which had gradually assumed an air of rarity. Obviously, the way to clearing such doubts and uncertainties was to search for the occurrence of those words and expressions in the speech of the Arabs elsewhere 22 In doing so, care had to be taken that the citations should faithfully reflect the idiom of the time of the Prophet during which the Qur'an was revealed. That is to say, either the citations should belong to the period contemporary with, or immediately antecedent to, the Qui an or be culled from the current usage of those whose social pattern had continued unchanged and who, therefore, could be relied upon to have preserved the idiom from that time uncorrupted and untainted by extraneous influences. Consequently, a zealous hunt was afoot to collect and preserve as much of pre-Islamic poetry, proverbs, and orations as could be salvaged from the memories of the people together with the current idiom of the A'rab, i.e., the people of the desert impervious to influences from outside. The method of collection was identical with that of the collection of the Hadit_h.
The end of the first/seventh century witnessed the rise of a band of scholars specially noted for their profundity in the field of al-lug-hah (Arabic usage)
21 In the words of abu 'Ubaidah introducing his Majaz al-Qur'dn: "The Qur'an was revealed in clear Arabic language and those who heard it recited by the Prophet had no need of asking for its meaning. ..."
22 Cf. the saying attributed to ibn 'Abbas: When you be in doubt about any
rare expression of the Qur'an, seek it in poetry. Al-Suyuti, al-Muzhir, ed. Muhammad Abmad Jad al-Maula and others, Cairo, II, p. 302. It was in
consideration of this that linguistic studies were regarded an obligation on a par with the obligation of prayer;cf. the verses (ibid.). Ibnal-Qatta' (al-Af'al,Hyderabad,p.3) went so far as to declare that anyone who decries the poetry of the Arabs is a sinner, and the one who runs down their language is an infidel (kafir).
Arabic Literature: Grammar and Lexicography
with its ancillary branches of al-shi`r (poetry), al-akhbar (historical annals), al-ayydm (accounts of tribal wars), and al-ansab (genealogies). The most prominent name among these scholars is that of abu 'Amr b. al-'Ala' (70/689154/770), an Arab nobleman of Basrah and an associate of 'Tsa b. 'Umar. His collection of Arabic philology, when piled up, touched the ceiling of his room. He set fire to this vast collection when he was overwhelmed by a fit of asceticism towards the end of his life. Yet he continued to be the primary source of knowledge for the next generation.
While the process of collecting the vocabulary and the illustrations of its diverse uses was still going on, the genius of al-Khalil, whom we have mentioned before, burst with the idea of arranging and fitting the vocabulary into the orderly scheme of a lexicon. Actually, al-Kbis known as the author of the first Arabic lexicon called the Kitdb at-'Airs, but the authorship is a bit disputed. This much, however, is certain that even if the actual compilation was not exclusively or partially the work of al-Khalil, the idea of a lexicon and the scheme thereof were first conceived by him. Let us now examine what the scheme is like.
Al-Khalil starts with (a) reducing all words to their roots, i.e., the radical letters (al-usril) which form an immutable kernel in contradistinction to those that are added (al-zawd'id) in the course of derivation and inflexion. Next (b) he classifies the roots according to the number of letters comprised in them: 2, 3, '4, and 5. Each class of words is then arranged in a separate part and even within each class special treatment under distinctive heading is resorted to in the case of words containing one or more of the vowels, double letters, or,a hamzah.
The above framework is in line with al-Khalil's attempt at a computation of Arabic vocabulary, which is a further proof of his originality. This quest he pursued on the same structural basis in a mathematical way. By multiplying the 28 letters of the alphabet by 27 (28 minus 1, to drop out double letters) he got 756 forms of the biliteral (there being no uniliterals in Arabic). Dividing this number by 2, he had 378 combinations irrespective of the order of the two letters. Taking these biliteral forms as one unit and adding a third letter to them, he worked out the number of triliteral forms and so on. It will be observed that the above method yielded the theoretically possible combinations of letters, all of which are not in actual use (musta'mal). Consequently, al-Khalil had to mention each and every possible combination and indicate if any specific forms were unused (muhmal). A further peculiarity, which made reference so difficult and cumbersome, was that in the arrangement of the lexicon he concerned himself merely with combinations of letters and mentioned all the forms yielded by a change of order of the letters under one and the same heading. For example, under MY one will find both MY and its reverse (magliib), YM.
Within the above framework, intrinsically scientific but practically unhandy, the order was according to the opening letter of the alphabet in the
A History of Muslim Philosophy
words. But the order of the alphabet observed by al-Kcl alil was not free from novelty; the grouping was according to the part of the mouth, from down the throat right out to the lips, which produced the sound. This novelty has been aptly noted and the similarity between it and the practice of the Sanskrit lexicographers has aroused a good deal of speculation. There is no doubt that the present-day arrangement, based on grouping of words according to the shape of the letters in writing, was the one in common use even at that distant date, though the Arabs were also familiar with the order according to the abjad system, which was originally taken over from the Syriac (and Hebrew) along with the art of writing.23 The phonetico-physiological system of al-Khalil was neither common at the time nor did it achieve popularity afterwards. But any significance which its similarity to that of Sanskrit might suggest is whittled down by due consideration of the fact that in all probability it developed indigenously out of the practice of the recitation of the Qur'an. With the emphasis on recitation it was but natural that phonetics should receive special attention and that there be a grouping of letters on that basis. Actually, evidence is not wanting that the linguists did engage themselves in such a study; there were some differences too between the Basrans and the KSfans as to the order of the alphabet on the basis of phonetics.24 Moreover, al-Khalil also paid some regard to the frequency of the letters in use; otherwise 'ain would not have come first in order 25
No doubt, the general lexicon of al-Khalil represented an idea much in advance of his time; for the following one century or so no one dared imitate, far less improve upon, his scheme. In the meantime, however, much valuable work was done in the form of small tracts comprising words, synonyms, and cognates with their fine shades of meaning grouped around particular subjects. Typical of such subjects are: al-ibil (the camel), al-matar (the rain), al-silah (the weapons), and the like. Similarly, special features of the Arabic usage were also singled out for monographic treatment: (a) al-muthallat_hdt, (b) at magsilr w-al-mamdiul, (c) al-itba' w-al-muzdwajah, (d) al-ajnds, and (e) atnawddir. Some philologists wrote running commentaries (concerned merely with the meaning of selected words and phrases supported with illustrations from other sources) on the Qur'an and the sayings of the Prophet under such captions as Qharib al-Qur'an, Oharib at Hadith, Majaz al-Qur'dn, Ma'ani al. Qur'dn, etc. The veterans in this field were two Arabs and two non-Arabs, one of the latter being from the distant province of Sind:
23 It was in that original source that numerical values were assigned to the letters in that order, which is still adhered to in Arabic and other Islamic languages for
purposes of chronograms. The assertion by later Arab philologists that abjad, hauwaz, etc., were the names of the inventors of the art of writing (al-Suyuti,
op. cit., II, p. 342) should be taken merely as a recollection of the old borrowing.
94 Ibn Duraid, Jamharat al-Lughhah, Hyderabad, cf. the Preface; cf. also al. Suyuti, op. cit., I, p. 85.
2s Al-Suyuti, op. cit., I, p. 90.
Arabic Literature: Grammar and Lexicography
(1) Al-Asma'i (abu Sa'id'Abd al-Malik b. Quraib), an Arab of Basrah, was born in 122/739 or 1231740 and died in about 217/832. He amused Hgrim alRaWid with his stock of interesting anecdotes about the life of the Arab.
(2) Abu Zaid (Sa'id b. Aus) al-Ansari was another Arab of Basrah who reached Baghdad during the time of al-Mahdi and died about 215/830, then over ninety years of age. He was not inhibited by partisanship and eagerly learnt from al-Mufaddal and other Kufans. By common agreement, he is regarded as thoroughly trustworthy, though his pedantry is often a source of amusement.
(3) Abu 'Ubaidah Ma'mar b. al-Muthanna, a maukt, said to have been of Persian Jewish descent, was born in 110/728 at Basrah where he spent most of his life. He was patronized by the Baramikah and was summoned to Baghdad by Harun al-Rashid to read his works to him. While rendering yeoman service to the Arabic philological studies, he collected the maLalib or the vices of the Arab tribes and caused such offence to tribal pride that at his death in 210/285 nobody attended his funeral.
(4) Ibn al-A'rabi (abu 'Abd Allah Muhammad b. Ziyad) was the son of a Sindian slave and the foster-child of the famous Kufan philologist, al-Mufaddal al-Dabbi. His prodigious memory was a storehouse of Arabic philology and folklore. Remarkably enough, he relied on his own independent sources and questioned not without success the authority of al-Asma'i and abu 'Ubaidah. He died about 231/845.
The special treatises referred to above naturally swelled to a considerable extent the volume of material which lay ready at hand for incorporation in a general lexicon. Another such lexicon was produced, rather dictated mostly from memory, by ibn Duraid (abu Bakr Mul}ammad b. al-Hasan, born at Basrah in 223/837 and died 321/933) who enjoyed the patronage of the Mikalids of Fars. Though ibn Duraid claims that his work is much easier for reference than that of al-Khalil, the fact is that there is little improvement so far as the scheme, particularly the break-up of the vocabulary into structural categories, is concerned. Even the irksome device which jumbles up all the orders, forward and reverse, of a combination of letters under one and the same heading, continues to be there. Only the phonetic order of the alphabet is discarded. Much of the confusion was caused by the nebulous state in which al-tasril (etymology) happened to be at that time. There was so far no clarity as to the roots of words, particularly those containing a vowel, a double consonant, or a hamzah. Similarly, lack of clarity as to the distinction between al-usisl and al-zawa'id caused the different categories to be mixed up. As a matter of fact, it was this uncertainty which made it expedient for ibn Duraid to insert a miscellany here and there, apart from the nawadir or peculiar usages and expressions listed under appropriate captions at the end.
There is indeed one important point of difference which is indicated by the very name, Jamharat al-Lu &ah. Ibn Duraid included in it only the familiar and the useful and eschewed the obsolete and the discordant. This was the
A History of Muslim Philosophy
Arabic Literature: Grammar and Lexicography
beginning of a process of subjecting to criticism and sifting out the useful and the dependable from the large mass of material left behind by the early
scholars, who were concerned with collecting and recording whatever they came across. At the time when the mistakes were being corrected, an attempt
was also made to supply the omissions in the works of the earlier authors. These, in short, are the new features noticeable in the lexicographical pro
ductions of the fourth/tenth century. Particularly notable in this respect is the Tahdib, whose author, abu Mansur Muhammad b. Atimad al-Azhari
(d. 370/980), a pupil of ibn Duraid, was urged to wanderlust in the desert for the collection of al-lughdt. Incidentally, he fell a captive into the hands of
a Bedouin tribe; this provided him with the desired opportunity. Equally
important is al-Muhit of al-Sahib b. 'Abbad, who died in 385/995.
The culmination of the critical activity of the fourth/tenth century aiming
at authenticity and comprehensiveness, was reached in the Sihah of al-Jauhari, abu Nasr Isma'il b. Hammed (died about 398(1007), a native of Farab who
settled down at Nishaapur. The very name Sidah reminds one of the Sahih of al-Bukhari. It has already been hinted at that the method of collecting
al-lug_hat was essentially the same as the one applied to the collection of the traditions, only a higher degree of stringency was observed in the case of the latter than in that of the former. This is aptly illustrated by the example of al-Asma'i, who is held to be trustworthy in regard to Hadit_h, but he risks conjectures
in matters pertaining to the lughdt and even embellishes anecdotes for the sake of amusement 2e Anyway, it is worthwhile to note that even the nomenclature
of the Hadit_h such as the mutauatir and the shad was applied to the lug_hdt and the degree of reliability of any particular usage determined accordingly. In the beginning it was not uncommon even to mention the isnad or the chain
of narrators and to discuss the personal character and reputation of the
transmitters.27 Thus, a compendium of the Sahih was sought to be arrived at in the field of lughhdt parallel to a similar, though much more scrupulously worked out, effort in the field of religious tradition.28 It has, however, to
be noticed that the Sidah suffered grievously from an unfortunate circumstance: the author was overtaken by a fit of melancholy which rendered
him incapable of revising the manuscript. Further, due to the absence of any authentic copy of the text, a good deal of corruption also set in. All
this necessitated a re-examination of the work in glosses and commentaries by later writers.
26 Cf. Ahmad Amin, Duha al-Islam, Cairo, 1952, II, p. 301. Abu 'Ubaidah once ridiculed al-Asma'i's extreme cautiousness in the interpretation of the Qur'an
by asking him whether he was sure of the meaning of at-Lhhubz (bread). Cf. Yagiut, op. cit., "Ma'mar b. al-Muthanna"
27 Al-Suynti, op. cit., I, pp. 118 et seq.
26 Just because the sciences of al-Hadith and al-lugjah were recognized as twins, the highest academic title for the learned in either was the same, al-p&. Ibid.,
II, p. 312.
The work of al-Jauhari was still more remarkable in another way. In it the entire vocabulary was integrated (instead of being split up into structural categories) and arranged in alphabetical order with first reference to the last letter and a second reference to its combination with the first. This new scheme at once became popular and was highly appreciated as particularly suited to a language in which the endings of the words had a unique importance for purposes of rhyme (gdiyah and saj'). Apart from the merits of this integrated scheme, the development and standardization of al-tasri/ (etymology) at the hands of al-Mazini (abu 'Uthman Bakr b. Muhammad, d. 249/863), ibn Jinni, and al-Rummani during the course of the fourth/tenth century removed a good deal of the confusion which marred the works of al-Khalil and ibn Duraid.
We have now reached a time when the Arabic vocabulary was supposed to have been exhaustively collected and the meanings of words established with reasonable certainty. Henceforth, efforts were directed at collecting the material scattered in the previous works either (a) in the form of large comprehensive dictionaries or (b) in concise handy volumes designed for the ordinary student. Naturally, the latter often dispensed with illustrations and citations. The most important works of the former category are:
(1) Al-Muhkam by the blind Spanish scholar, ibn Sidah (abu al-Hasan 'Ali b. Abmad?, d. 460/1068), was held in great esteem for comprehensiveness and absolute reliability. But perhaps the author did not like innovations; hence he went back to the earliest model of al-Khalil for its arrangement.
(2) Al-'Ubdb (incomplete) was composed by Radi al-Din Hasan al-Sagh_ani, born in
(3) The Lisdn al-'Arab was compiled by ibn Mukarram/ibn Man~ur (Jamal al-Din Muhammad), who was born in 690/1291, and died at Cairo in 771/1369. It is expressly based on the works of ibn Duraid, al-Azhari, al-Jauhari, and ibn Sidah.
Of the latter category, the work which achieved a high degree of popularity is the Qdmis of Maid al-Din al-Firiizabadi (Muhammad b. Ya'giib) who died
in 816/1413. It draws upon al-Muhkam and al-'Ubdb.
Yet another work which deserves special mention is the Asds al-Balaghah of the well-known Mu'tazilite al-Zamakhs_hari (abu al-Qasim Jar Allah Mabmud b. 'Umar, born 467/1074 and died 538/1143). The author was a native of Khwarizm who spent a long time in Makkah and Baghdad. He realized that the mere recording of meanings was an insufficient guide to the practical use of words. He, therefore, would give the occasions and the contexts in which the words were employed. What is still more remarkable is the arrangement of the Asas, which is in the alphabetical order with reference to the first (and then the second and so forth) letter of a word. That is to say, its arrangement
A History of Muslim Philosophy
Arabic Literature: Theories of Literary Criticism
is exactly the same as has come into vogue in modern times since the impact of Western literary influences.
It is interesting to note that the early trend towards compiling treatises dealing with words grouped around particular subjects did not die with the appearance of the general lexicons; it had an uninterrupted development on parallel lines. The greatest work of this kind is al-Muk_hassas, a twin of the general lexicon, al-Muhkam, by the Andalusian ibn Sidah. In al-Mukhassas, the vocabulary is grouped under subject headings, e.g., the hair, the eye, etc., which are classified into "books such as that oc "human body." Even if the position of al-Muhkam is not wholly unsurpassed, that of al-Muk_hassas is definitely so.
Once the framework of a general lexicon was fixed, the running commentaries on the rare and difficult words in the Qur'an and/or the Hadith were also brought under that form.20 Similarly, no time was lost in extending the facility and the benefits of a general dictionary to the other specialized branches such as zoology, botany, biography, geography, bibliography, and finally the encyclopedias (al-mausii`at). It may be observed in this connection that interest in language and literature, which the scheme of a lexicon was originally designed to subserve, seldom disappeared in any of the works, however specialized and limited the scope of their treatment. It would, for example, be really odd to conceive of a zoologist or a geographer who was not familiar with the references in the Qur'an and the Hadith or who would be unable to recall poetry, proverbs, and pithy sayings concerning animals or towns. This all-pervading interest in humanities is perhaps the most valuable asset of Islamic culture.
In conclusion, it will be recalled that the early philologists were fully conscious of the sanctity of their task; they showed themselves to be scrupulous in method and honest in purpose. But the scope of the linguistic studies was bound in course of time to extend beyond what was strictly relevant to the Qur'an and the Hadith. As the bounds of the sacred faded into those of the profane, the common failings of vanity, mere guess or conjecture, or even unguarded reliance on genuine misunderstandings, contributed to the interpolation of the spurious. Also, as these studies came to be held in high esteem and patronized with abundant monetary gifts, the veterans in the field were sometimes tempted to window-dress faked rarities in their shop. But the probe into their personal weaknesses, so characteristic of Islamic religious and literary tradition, and the severe tests subsequently applied to their statements served to a large extent to clear the chaff from the grain. On the whole, there is no doubt that a fair degree of reliability was achieved. In the same way it is impossible to claim that the entire vocabulary and usage were exhausted, yet there is no gainsaying the fact that an enormous part of them was actually encompassed. The charge that the Arabic philologists
29 Cf. Kashf al-Zunun, II, pp. 1204-06.
concerned themselves too exclusively with the idiom of the Qur'an and showed no interest in contemporary deviations from the same, tantamounts to questioning their objective or purpose, which has been steadily confirmed throughout the ages. In regard to the scheme and the arrangement of a lexicon, the early pioneers proceeded on the basis of a scientific etymological analysis of the structure of the vocabulary. Practical convenience was achieved later in the superbly original plan of al-Jauhari, which remains the one specially suited to the genius of the language. Even the model which has become so popular in modern times is traceable to al-Zamakhs_hari
Al.Suyuti, al-Muzhir; Yaqut, Mu'jam al-Udaba'; ibn KhallikAn, Wafaydt alA'ydn; ibn Khaldnn, Muqaddimah; Hajj Khalifah, Kaa_hf al-Zunun; Jurji Zaidan, Tarik_h al-Adab al-`Arabiyyah; Ahmad Amin, Duh a al-Islam, Part II, Chap. VI;
Zuhr al-Ialdm, Part II, Chaps. III and IV.