How far such a work as the Assemblies is to be ascribed to any influence of Byzantine scholarship is a question of some difficulty. There is no reason to believe that Ḥarîri took directly any ideas from the Greeks. He probably knew no language but his own, and all that he produced is to be traced to exclusively Arabic, or, at least, Moslem sources. But indirectly he, and the rest of his countrymen, could not fail to be influenced by the pedantry of Constantinople, which came to confirm and increase their own. In almost every age of the world we find a certain similarity between the pursuits of existing nations, as if the same informing spirit possessed them all, in spite of diversities of race, position, and his­tory. In the time of Ḥarîri, Greek, Jew, Syrian and Arab had become grammarians, lexicographers, archæologists; the race of translators and commentators abounded in the east, and to the furthest limits of the Moslem west. Each race, except perhaps the haughty and exclusive Greeks, did indeed learn something from the others. The Syrians took their notions of grammar from the Greeks. All that the Arabs knew of Greek philosophy and science was from translations made into their lan­guage by Syrians, either from the Greek originals, or from former Syriac translations. On the other hand the Jews, in the tenth century, founded or re-cast their grammar under Arab influence, and even adopted the word as their paradigm, though, from its containing a medical guttural, it insufficiently expresses the changes of the Hebrew verb. We find the philologers of the various languages vying with each other in industry; we find Maydâni, the contemporary of Ḥarîri, collecting and explaining the national proverbs precisely in the manner of Suidas. No one can pretend to say that such an influence as this may not have extended from Con­stantinople to Basra, but of direct imitation of Greek authors there is nothing. The work with which the Assemblies will be most readily compared is the Cassan­dra , or Alexandra, of Lycophron. That was a poem com­posed in an age when the demon of pedantry had entered into poets and orators; when even the exquisite genius of Theocritus did not disdain to produce a poem in the form of a Syrinx. The Cassandra is an iambic poem of fourteen hundred and seventy lines, very famous in anti­quity, though utterly neglected now. It is an almost unintelligible rhapsody, supposed to be uttered by the prophetical daughter of Priam, who from the top of a tower, where she is confined by her father’s order, sees the fleet of Paris depart, and knows that it is to bring destruction on Troy. In her sacred fury, she pours forth a torrent of obscure and oracular verse. The epithets are gigantic and monstrous compounds, the language is ran­sacked for rare words, metaphor is heaped on metaphor, every person and every place are described, not by their ordinary names, but by some accidental relation, or by allusion to some obscure event; the prophetess passes from calamity to calamity with a Pythian enthusiasm which scorns the distinctions of paragraphs, and ends by lamenting that mankind will not believe the predictions she has uttered. This poem, if it can be called so, is but a linguistic and geographical puzzle. The war of Troy, the return of the Greek heroes, their dispersion over the shores of the Mediterranean, the planting of Greek and Trojan colonies, the settlement of Æneas in Italy, the foundation of Rome, the wars between Europe and Asia, the expedition of Xerxes, and the conquests of Alexander, the descendant of Æacus and Dardanus, give opportunity for the display of boundless but useless learning. This production, which gained for the author the epithet of the “dark” Lycophron from the despairing scholars of antiquity, was not fully elucidated till the time of the brothers Tzetzes,* whose labours on it are a remarkable proof of the learning and industry of the Byzantines. Except in the display of erudition, there is no resem­blance between Lycophron and Ḥarîri, nor do we find in the latter that impenetrable and perverse obscurity which distinguished the Greek. But it is possible that the fame of this work, at which successive generations of schoolboys had been made to labour, may have reached Arabic writers, and encouraged their tendency to learned display.

If we allow some share in the formation of such a taste to the predispositions of race and language, and some­thing to the influence of foreign culture, we must also trace some elements of it to the peculiar studies of Islam. A strange feature in Ḥarîri’s work is the number of com­positions in which the merit is supposed to consist in the alternation of pointed and unpointed letters, or the ex­clusive use of one of these two kinds. It is not difficult to see how any tendency to these caprices which may have been innate in the learned Arabs must have been increased by the long controversies which had been held on the orthography of the Koran. On the proper use of the diacritical points, on the question whether a letter should be with or without a point, whether it should be pointed above or below, depended the meaning of many passages of God’s word. The early copies of the Koran were in a character singularly incomplete. There was hardly a chapter in which doubts were not raised as to the meaning of important passages. In the lifetime of the Prophet, when the sacred fragments were read publicly at Medina by those who knew their purport accurately, and were committed to memory by numbers of the faith­ful, these deficiencies were unnoticed; but as early as the time of Abû Bekr many of those who knew the Koran by heart were slain in the campaign against Musaylimeh the Liar, and ‘Omar counselled the Khalif to cause a stan­dard copy to be written.

This task was executed by Zayd ibn Thâbit, who first collected the Koran into a book. But great varieties of readings made the sense doubtful, and in the year 30 the Khalif ‘Othmân undertook the recension which has ever since been the sacred text of the Moslems. He com­mitted the work to Zayd ibn Thâbit, ‘Abd Allah ibn Az Zobayr, and other distinguished persons, with orders to choose the best reading of each passage, and not to give the various conflicting readings, as appears to have been done in the first edition. All who could throw any light on the Prophet’s meaning were consulted, and ‘Âyisheh in particular had great influence in determining the text; her testimony as to the Prophet’s oral recitations being respectfully received as conclusive. The Koran was brought more completely into accord with the dialect of Ḳoraysh, since the Khalif commanded that in doubtful cases the language of the Prophet’s family and country­men should be preferred. Hence there remain very few dialectical peculiarities in the present Koran. But, when all was done, the text still remained very uncertain; and the difficulties of the Moslems were increased by the sudden corruption of the pure Arabic speech after the first conquests. Ḳoraysh or Temîm might have been able to detect the precise form and pronunciation of a word under its obscure orthography, but the settlers of the conquered lands, who in fifty years after the Hijra had almost lost the use of the classic desinences, required a clearer guidance. The efforts of Ḥajjâj ibn Yûsuf to settle the text of the Koran have been already referred to. Ibn Khallikân, in his life of the enlightened tyrant, says that the people used ‘Othmân’s copy for forty years, but in the days of ‘Abd al Melik ibn Merwân erroneous readings had become numerous in Irak, so that Ḥajjâj bade his scribes set distinctive marks on the words of uncertain pronunciation, and Naṣr ibn ‘Aṣim placed single or double dots to certain words. But as this did not give sufficient accuracy, the , or system of diacritical pointing now in use, was invented. The honour of giving fixity of meaning to the Koran belongs chiefly to Abû ’l Aswad and Khalîl ibn Aḥmed, though the latter is said to have superseded the system of point­ing adopted by his predecessor, as well as that of the scribes of Ḥajjâj.* Abû ’l Aswad is said to have marked the vowels and the tanwîn, Khalîl ibn Aḥmed the hamzeh and teshdîd, the rowm and ishmâm. Thus the science of pointing, or determining which of two or more letters, similarly formed in ancient writing, was to be preferred, became of the highest interest. In our languages such accuracy is not needed, but in the Arabic, where every letter has and must have a value, where a change in a root letter is sufficient to transform one word into another, and a change in an augmentative letter may affect gender, number, and the other accidents of the root, the scholar was obliged to employ himself on the primary elements of speech. The formation of passives and plurals by internal vowel change, unexpressed in the usual ortho­graphy, must have tended still further to promote this minute scholarship. The compositions of pointed or unpointed letters, such as Ḥarîri delights in, may have been suggested, or at least encouraged, by a desire to embody the correct orthography of doubtful words. The author would fix them with their right pointing in a poem or address, of which the artifice would insure that they should not be misspelt.

There was, furthermore, a religious sanction for this careful study of the alphabet. One of the most mysterious attributes of the Koran was its essential and eternal con­nection with the Arabic language. In that language it had not only been revealed, but had existed before the world and time. “This (the Koran),” said Mohammed, “is a clear Arab tongue.” Koran xvi. 105 (compare xxvi. 195). If a confirmation of this unity of the sacred book and the sacred language were needed, it would be found in the prefixion to some of the Suras of isolated letters, whose meaning no man knew, but which had un­doubtedly a divine quality. If Arabic words, which have equivalents in foreign tongues, had alone been revealed, it might be held that there was no essential connection between the revelation and the language; but when sepa­rate letters, having no idea affixed to them, were brought down from on high, it could not be doubted that the ineffable sanctity of the Arabic language was thereby figured. In course of time a mystic interpretation was given to these letters, which is carried to the extreme in the commentary of Bayḍâwi. At the beginning of the second Sura he comments the letters elif, lâm, mîm, which are prefixed to it, and takes occasion to show that the presence of these monograms in the Koran is plainly miraculous. He points out that the letters are used in such a manner as to indicate a profound knowledge of scientific grammar. As Mohammed was uninstructed, such a knowledge could only have come from God him­self. For instance, there are fourteen letters thus pre­fixed to Suras of the Koran, and fourteen is just the half of twenty-eight, the total of the letters when elif is not distinguished from hamzeh. These fourteen let­ters, moreover, are to be found in twenty-nine Suras, which is just the whole number of letters in the al­phabet, if elif be reckoned. Again, the half of each of the different classes of letters are thus employed. Of the ten weakly articulated letters, namely those which make up the words , five are to be found in these monograms; while of the other eighteen, which are strongly articulated, nine are to be so found. The gram­matical forms which take an augment never exceed in length the number of seven letters, hence the Almighty has introduced into these monograms seven out of the ten augmentative letters. These are but a few of the dis­coveries which in the time of Bayḍâwi had been made concerning the monograms of the Koran, and which the orthodox commentator explains with undoubting faith. In the age of Ḥarîri these interpretations were in high favour; indeed, Bayḍâwi does little but reproduce the comments of Zamakhshari in the Keshshâf, who himself followed other doctors of an earlier time. Zamakhshari tells us in his preface that being asked by his admirers to produce a commentary on the Koran, his first care was to write something on these monograms, as a specimen of the loftiness and the difficulty of the enterprise. It will thus easily be perceived that exercises in the arrange­ment of letters had not among the Arabs that character of futility which they bear with us.