A LITERARY HISTORY OF THE ARABS

BY

REYNOLD A. NICHOLSON

CAMBRIDGE AT THE UNIVERSITY PRESS 1966

PUBLISHED BY THE SYNDICS OF THE CAMBRIDGE UNIVERSITY PRESS

Bentley House, 200 Euston Road, London, N.W. 1 American Branch: 32 East 57th Street, New York N.Y. 10022, West African Office: P.O. Box 33, Ibadan, Nigeria

First edition (T. Fisher Unwin) 1907, reprinted 1914, 1923 Reprinted (Cambridge University Press) 1930, 1941, 1953, 1962, 1966

First printed in Great Britain at the University Press, Cambridge Reprinted by offset-litho by Latimer Trend & Co. Ltd, Whitstable

To Professor A. A. BEVAN In grateful recollection of many kindnesses

ix

PREFACE

A Literary History of the Arabs, published by T. Fisher Unwin in 1907 and twice re-issued without alteration, now appears under new auspices, and I wish to thank the Syndics of the Cambridge University Press for the opportunity they have given me of making it in some respects more accurate and useful than it has hitherto been. Since the present edition is printed from the original plates, there could be no question of revising the book throughout and recasting it where necessary; but while only a few pages have been rewritten, the Bibliography has been brought up to date and I have removed several mistakes from the text and corrected others in an appendix which includes a certain amount of supplementary matter. As stated in the preface to the first edition, I hoped "to compile a work which should serve as a general introduction to the subject, and which should be neither too popular for students nor too scientific for ordinary readers. It has been my chief aim to sketch in broad outlines what the Arabs thought, and to indicate as far as possible

the influences which moulded their thought.... Experience has convinced me that young students of Arabic, to whom this volume is principally addressed, often find difficulty in understanding what they read, since they are not in touch with the political, intellectual, and religious notions which are presented to them. The pages of almost every Arabic book abound in allusions to names, events, movements, and ideas of which Moslems require no explanation, but which puzzle the Western reader unless he have some general knowledge of Arabian history in the widest meaning of the word. Such a survey is not to be found, I believe, in any single European book; and if mine supply the want, however partially and inadequately, I x shall feel that my labour has been amply rewarded.... As regards the choice of topics, I agree with the author of a famous anthology who declares that it is harder to select than compose (ikhtiyáru ’l-kalám aṣ‘abu min ta’lífihi). Perhaps an epitomist may be excused for not doing equal justice all round. To me

the literary side of the subject appeals more than the historical, and I have followed my bent without hesitation; for in order to interest others a writer must first be interested himself.... Considering the importance of Arabic poetry as, in the main, a true mirror of Arabian life, I do not think the space devoted to it is excessive. Other branches of literature could not receive the same attention. Many an eminent writer has been dismissed in a few lines, many well-known names have been passed over. But, as before said, this work is a sketch of ideas in their historical environment rather than a record of authors, books, and dates. The exact transliteration of Arabic words, though superfluous for scholars and for persons entirely ignorant of the language, is an almost indispensable aid to the class of readers whom I have especially in view. My system is that recommended by the Royal Asiatic Society and adopted by Professor Browne in his Literary History of Persia; but I use ẓ for the letter which he denotes by dh. The definite article al, which is frequently

omitted at the beginning of proper names, has been restored in the Index. It may save trouble if I mention here the abbreviations 'b.' for 'ibn' (son of); J.R.A.S. for Journal of the Royal Asiatic Society; Z.D.M.G. for Zeitschrift der Deutschen Morgenländischen Gesellschaft; and S.B.W.A. for Sitzungsberichte der Wiener Akademie. Finally, it behoves me to make full acknowledgment of my debt to the learned Orientalists whose works I have studied and freely 'conveyed' into these pages. References could not be given in every case, but the reader will see for himself how much is derived from Von Kremer, Goldziher, Nöldeke, and Wellhausen, to mention only a few of the leading authorities. At the same time I have constantly gone back to the native sources of information."

xi

There remains an acknowledgment of a more personal kind. Twenty-two years ago I wrote—"my warmest thanks are due to my friend and colleague, Professor A. A. Bevan, who read the proofs throughout and made a number of valuable remarks which will be found in the footnotes." Happily the present occasion permits me to renew those ties between us; and the book which he helped into the world now celebrates its majority by associating itself with his name.

REYNOLD A. NICHOLSON November 1, 1929

xii

Frontispiece Litigants before a Judge (British Museum Or. 1200)

xiii

Contents

xv

Introduction

The Arabs belong to the great family of nations which on account of their supposed descent from Shem, the son of Noah, are commonly known as the 'Semites.' The Semites. This term includes the Babylonians and Assyrians, the Hebrews, the Phœnicians, the Aramæans, the Abyssinians, the Sabæans, and the Arabs, and although based on a classification that is not ethnologically precise—the Phœnicians and Sabæans, for example, being reckoned in Genesis, chap. x, among the descendants of Ham—it was well chosen by Eichhorn († 1827) to comprehend the closely allied peoples which have been named. Whether the original home of the undivided Semitic race was some part of Asia (Arabia, Armenia, or the district of the Lower Euphrates), or whether, according to a view which has lately found favour, the Semites crossed into Asia from Africa,1 is still uncertain. Long before the epoch when they first appear in history they had branched off from the parent stock and formed separate nationalities. The relation of the Semitic languages to each other cannot

be discussed here, but we may arrange them in the chronological order of the extant literature as follows:—2

xvi

1. Babylonian or Assyrian (3000-500 b.c.).

2. Hebrew (from 1500 b.c.).

3. South Arabic, otherwise called Sabæan or Ḥimyarite (inscriptions from 800 b.c.).

4. Aramaic (inscriptions from 800 b.c.).

5. Phœnician (inscriptions from 700 b.c.).

6. Æthiopic (inscriptions from 350 a.d.).

7. Arabic (from 500 a.d.).

Notwithstanding that Arabic is thus, in a sense, the youngest of the Semitic languages, it is generally allowed to be nearer akin than any of them to the original archetype, the 'Ursemitisch,' from which they all are derived, just as the Arabs, by reason of their geographical situation and the monotonous uniformity of desert life, have in some respects preserved the Semitic character more purely and exhibited it more distinctly than any people of the same family. From the period of the great Moslem conquests (700 a.d.) to the present day they have extended their language, The Arabs as representatives of the Semitic Race. religion, and culture over an enormous expanse of territory, far surpassing that of all the ancient Semitic empires added together. It is true that the Arabs are no longer what they were in the Middle Ages, the ruling nation of the world, but loss of temporal power has only strengthened their spiritual dominion. Islam still reigns supreme in Western Asia; in Africa it has steadily advanced; even on European soil it has found in Turkey

compensation for its banishment from Spain and Sicily. While most of the Semitic peoples have vanished, leaving but a meagre and ambiguous record, so that we cannot hope to become intimately acquainted with them, we possess in the xviicase of the Arabs ample materials for studying almost every phase of their development since the sixth century of the Christian era, and for writing the whole history of their national life and thought. This book, I need hardly say, makes no such pretensions. Even were the space at my disposal unlimited, a long time must elapse before the vast and various field of Arabic literature can be thoroughly explored and the results rendered accessible to the historian.

From time immemorial Arabia was divided into North and South, not only by the trackless desert (al-Rub‘ al-Khálí, the 'Solitary Quarter') which stretches across the Arabs of the North and South. peninsula and forms a natural barrier to intercourse, but also by the opposition of two kindred races widely differing in their character and way of life. Whilst the inhabitants of the northern province (the Ḥijáz and the great central highland of Najd) were rude nomads sheltering in 'houses of hair,' and ever shifting to and fro in search of pasture for their camels, the people of Yemen or Arabia Felix are first mentioned in history as the inheritors of an ancient civilisation and as the owners of fabulous wealth—spices, gold and precious stones—which ministered to the luxury of King Solomon. The Bedouins of the North spoke Arabic—that is to say, the language of the Pre-islamic poems and of the Koran—whereas the southerners used a dialect called by Muḥammadans 'Ḥimyarite' and a peculiar script of which the examples known to us have been discovered and

deciphered in comparatively recent times. Of these Sabæans—to adopt the designation given to them by Greek and Roman geographers—more will be said presently. The period of their bloom was drawing to a close in the early centuries of our era, and they have faded out of history before 600 a.d., when their northern neighbours first rise into prominence.

It was, no doubt, the consciousness of this racial distinction Ishmaelites and Yoqṭánids. xviiithat caused the view to prevail among Moslem genealogists that the Arabs followed two separate lines of descent from their common ancestor, Sám b. Núḥ (Shem, the son of Noah). As regards those of the North, their derivation from ‘Adnán, a descendant of Ismá‘íl (Ishmael) was universally recognised; those of the South were traced back to Qaḥṭán, whom most genealogists identified with Yoqṭán (Joktan), the son of ‘Ábir (Eber). Under the Yoqṭánids, who are the elder line, we find, together with the Sabæans and Ḥimyarites, several large and powerful tribes—e.g., Ṭayyi’, Kinda, and Tanúkh—which had settled in North and Central Arabia long before Islam, and were in no respect distinguishable from the Bedouins of Ishmaelite origin. As to ‘Adnán, his exact genealogy is disputed, but all agree that he was of the posterity of Ismá‘íl (Ishmael), the son of Ibráhím (Abraham) by Hájar (Hagar). The story runs that on the birth of

Ismá‘íl God commanded Abraham to journey to Mecca with Hagar and her son and to leave them there. They were seen by some Jurhumites, descendants of Yoqṭán, who took pity on them and resolved to settle beside them. Ismá‘íl grew up with the sons of the strangers, learned to shoot the bow, and spoke their tongue. Then he asked of them in marriage, and they married him to one of their women.3 The tables on the opposite page show the principal branches of the younger but by far the more important family of the Arabs which traced its pedigree through ‘Adnán to Ismá‘íl. A dotted line indicates the omission of one or more links in the genealogical chain.4

xix

xxIt is undeniable that these lineages are to some extent fictitious. There was no Pre-islamic science of genealogy, so that the first Muḥammadan investigators had only confused and scanty traditions to work on. They were biassed, moreover, by political, religious, and other considerations.6 Thus their study of the Koran Character of Muḥammadan genealogy. and of Biblical history led to the introduction of the patriarchs who stand at the head of their lists. Nor can we accept the national genealogy beginning with ‘Adnán as entirely historical, though a great deal of it was actually stored in the memories of the Arabs at the time when Islam arose, and is corroborated by the testimony of the Pre-islamic poets.7 On the other hand, the alleged descent of every tribe from an eponymous ancestor is inconsistent with facts established by modern research.8 It is probable that many names represent merely a local or accidental union; and many more, e.g., Ma‘add, seem originally to have denoted large groups or confederations of tribes. The theory of

a radical difference between the Northern Arabs and those of the South, corresponding to the fierce hostility which has always divided them since the earliest days of Islam,9 may hold good if we restrict the term 'Yemenite' (Southern) to the civilised Sabæans, Ḥimyarites, &c., who dwelt in Yemen and spoke their own dialect, but can hardly apply to the Arabic-speaking 'Yemenite' Bedouins scattered all over the peninsula. Such criticism, however, does not affect the value of the genealogical documents regarded as an index of the popular mind. From this point of view legend is often superior to fact, and it must be our aim in the following chapters to set forth what xxithe Arabs believed rather than to examine whether or no they were justified in believing it.

'Arabic,' in its widest signification, has two principal dialects:—

1. South Arabic, spoken in Yemen and including Sabæan, Ḥimyarite, Minæan, with the kindred dialects of Mahra and Shiḥr.

2. Arabic proper, spoken in Arabia generally, exclusive of Yemen.

Of the former language, leaving Mahrí, Socotrí, and other living dialects out of account, we possess nothing beyond the numerous inscriptions which have been collected South Arabic. by European travellers and which it will be convenient to discuss in the next chapter, where I shall give a brief sketch of the legendary history of the Sabæans and Ḥimyarites. South Arabic resembles Arabic in its grammatical forms, e.g., the broken plural, the sign of the dual, and the manner of denoting indefiniteness by an affixed m (for which Arabic substitutes n) as well as in its vocabulary; its alphabet, which consists of twenty-nine letters, Sin and Samech being distinguished as in Hebrew, is more nearly akin to the Æthiopic. The Ḥimyarite Empire was overthrown by the Abyssinians in the sixth century after Christ, and by 600 a.d. South Arabic had become a dead language. From this time forward the dialect of the North established an almost universal supremacy and won for itself the title of 'Arabic' par excellence.10

The oldest monuments of written Arabic are modern in date compared with the Sabæan inscriptions, some of which take us back 2,500 years or thereabout. Apart The oldest specimens of Arabic writing. from the inscriptions of Ḥijr in the northern Ḥijáz, and those of Ṣafá in the neighbourhood of Damascus (which, although written by northern Arabs before the Christian era, exhibit a peculiar character not unlike the xxiiSabæan and cannot be called Arabic in the usual acceptation of the term), the most ancient examples of Arabic writing which have hitherto been discovered appear in the trilingual (Syriac, Greek, and Arabic) inscription of Zabad,11 south-east of Aleppo, dated 512 or 513 a.d., and the bilingual (Greek and Arabic) of Ḥarrán,12 dated 568 a.d. With these documents we need not concern ourselves further, especially as their interpretation presents great difficulties. Very few among the Pre-islamic Arabs were able to read or write.13 Those who could generally owed their skill to Jewish and Christian teachers, or to the influence of foreign culture radiating

from Ḥíra and Ghassán. But although the Koran, which was first collected soon after the battle of Yamáma (633 a.d.), is the oldest Arabic book, the beginnings of literary composition in the Arabic language can be traced back to an earlier period. Probably all the Pre-islamic poems which have come down to us belong to the century preceding Islam (500-622 a.d.), but their elaborate form and technical perfection forbid the hypothesis that in them we have "the first sprightly runnings" of Arabian song. It may be said of these magnificent odes, as of the Iliad and The Pre-islamic poems. Odyssey, that "they are works of highly finished art, which could not possibly have been produced until the poetical art had been practised for a long time." They were preserved during hundreds of years by oral tradition, as we shall explain elsewhere, and were committed to writing, for the most part, by the Moslem scholars of the early ‘Abbásid age, i.e., between 750 and 900 a.d. It is a noteworthy fact that the language of these poems, the authors of

which represent many different tribes and districts of the xxiiipeninsula, is one and the same. The dialectical variations are too trivial to be taken into account. We might conclude that the poets used an artificial dialect, not such as was commonly spoken but resembling the epic dialect of Ionia which was borrowed by Dorian and Æolian bards. When we find, however, that the language in question is employed not only by the wandering troubadours, who were often men of some culture, and the Christian Arabs of Ḥíra on the Euphrates, but also by goat-herds, brigands, and illiterate Bedouins of every description, there can be no room for doubt that in the poetry of the sixth century we hear the Arabic language as it was then spoken throughout the length and breadth of Arabia. The success of Muḥammad and the conquests made by Islam under the Orthodox Caliphs gave an entirely new importance to this classical idiom. Arabic became the sacred language of the whole Moslem world. This was certainly due to the Koran; but, on The Koran. the other hand, to regard the dialect of Mecca,

in which the Koran is written, as the source and prototype of the Arabic language, and to call Arabic 'the dialect of Quraysh,' is utterly to reverse the true facts of the case. Muḥammad, as Nöldeke has observed, took the ancient poetry for a model; and in the early age of Islam it was the authority of the heathen poets (of whom Quraysh had singularly few) that determined the classical usage and set the standard of correct speech. Moslems, who held the Koran to be the Word of God and inimitable in point of style, naturally exalted the dialect of the Prophet's tribe above all others, even laying down the rule that every tribe spoke less purely in proportion to its distance from Mecca, but this view will not commend itself to the unprejudiced student. The Koran, however, exercised a unique influence on the history of the Arabic language and literature. We shall see in a subsequent chapter that the necessity of preserving the text of the Holy Book uncorrupted, and of elucidating its obscurities, caused xxivthe Moslems to invent a science of grammar and lexicography,