Some Proposed Emendations to the Text of the Koran

In this article, eleven difficult passages in the Koran, which have defied the efforts of both Muslim commentators and orientalists to explain them, are interpreted as corruptions resulting from faulty copying by scribes. Emendations of the text are proposed to bring it as close as possible to the form it had when first spoken by the prophet Muhammad. At the end, a few changes are made in the author's old hypothesis that the Mysterious Letters at the head of some of the suras are old abbreviations of the basmalah.

A curious feature of studies on the Koran in the West over the last 150 years is the scant attention paid by scholars to the Koranic text as such. Orientalism has many excellent works on the Koran to its credit, but one seeks in vain for a systematic application of the techniques of textual criticism to the textual problems of the Koran, although classicists and biblical scholars have for centuries made continuous efforts to improve the quality of the texts that are the bases of their disciplines. It is difficult to see why this should be so. Early Koran scholars such as Fleischer, Noldeke, and Goldziher were good textual critics; they were all well educated in classical and biblical studies, and they made good editions of later Arabic texts that are still in use today.

Whatever the reasons, Western scholarship, with very few exceptions, 1 has chosen to follow the Muslim commentators in not emending the text. When faced with a problem, the Westerners have resorted to ety mologizing and hunting for foreign words and foreign influences. They have produced a great deal of valuable scholarship important for our study of the Koran and the origins of Islam, but where they exercised their skill on corrupt texts, they, of course, produced only fantasies.

The Arabs, on the other hand, tend to paraphrase, stating in different terms what they think the passage must mean. However, their Arabic was very good, so we find sometimes that they sensed the correct meaning of a problematical passage, and then defined, or, better said, "redefined," the crucial word accordingly even when lexically it was impossible. This is of great help to the modem textual critic, who has only to carry the process one step further and make the necessary emendation. We shall see below several instances of this sort of redefinition.

The earliest generation of Muslim commentators, although they did not emend the text, had no doubt that it did contain mistakes. Our sources list several acknowledged errors, and-if we are to believe the Arab tradition-the first textual critics of the Koran were 'Uthman, 'Ali, and 'A'ishah. The caliph 'Uthman, when the recension of the Koran that he had sponsored was presented to him, looked it over and noticed some mistakes (lahn), and said, "Don't correct them for the Bedouin Arabs will correct them with their tongues."

Ibn Abbas, cousin of the Prophet and a famous early commentator, is credited with detecting and correcting several errors in the text. In XIII.31 we find a -fa-lam yay'asi lladhina amanu, "Have not those who believed despaired?" Ibn Abbas, following Ibn Masud, read yatabayyan, "Have they not seen clearly?" and said that the copyist must have been sleepy when he wrote yay'as. In XVII.23, wa-gada rabbuka alla ta'buda illa iyyahu, "Your lord has decreed that you should not worship any except him," he read wa-wassa, "Your lord advised," explaining that the copyist had taken up too much ink in his calamus, and that the waw had flowed into the sad, turning it into a gaf.4

Probably none of the anecdotes cited above is really true, but they are important in that they show that about a generation after the promulgation of the Uthmanic recension, some readers noted that there were mistakes in the Koran, and suggested corrections, though they prudently did not try to alter the official text. They also show us that the Arab commentators were well acquainted with drippy pens and copyists' errors brought on by fatigue.

In addition to the errors noted above, there are in the Koranic text many variant readings (gira'at), which do not involve errors, but each of which is evidence that a mistake was made in the tradition at some time or other. Otherwise we must admit that the Prophet may have recited a passage in a certain way when it was first revealed, but then changed it in a subsequent recitation-not impossible, but this could not account for all the variants. Most of the gira'at derive ultimately from the fact that the Uthmanic recension was published without diacritics-though they did exist at that time-and without vowel signs, which were not invented until some years later. They are important to us here because they prove that there was no oral tradition stemming directly from the Prophet strong enough to overcome all the uncertainties inherent in the writing system.

Given the fact that the Koran contains acknowledged errors and, in the gira'at, evidence of many more, it is impossible to deny that still more mistakes, as yet undetected, may lie hidden in the text. In this article I shall attempt to isolate several errors and then to emend the text in order to restore it as nearly as possible to its original form. In the Koran "original form" means, of course, the form the word or phrase had when it was first uttered by the prophet Muhammad.

The first step in this process is the isolation of possible errors. The most important clue that an error may have been made is the lack of good sense in the word or passage and the resulting variety of opinion among scholars as to what it means. Another clue is when the word is transmitted in more than one form. In general, different views about the meaning and/or form of a particular word make it likely that the word is wrong. Still another clue is when the word in question is said by the lexicographers to be dialectal or foreign. Some such claims may be the result of academic pretentiousness, but others indicate that the word was not known to the Meccans and the Medinese and hence is probably a mistake.

In proposing emendations, I shall follow rules laid down by classi cists. In order to be acceptable, an emendation must make good sense, better than the received text; it must be in harmony with the style of the Koran; it must also be palaeographically justifiable; and finally, it must show how the corruption occurred in the first place.

The cases examined below share a common feature; each occurs in a context of simple, everyday words, which makes it most unlikely that the difficult word represents something mysterious, arcane, or foreign. Indeed. in some cases, as noted above, the meaning required is obvious, or nearly so, so all we have to do is search for a simple, everyday word that will fill the slot and, at the same time, meet the requirements for emendation listed above. The results are likely to be dull and commonplace, since they will lack the ambiguity of the mistakes that allows the imagination of scholars to soar.


We shall begin with a case in which, by a lucky accident, both the original and the error have been preserved. In XXI.98 we read: innakum wama ta`buduna min duni lldhi hasabujahannama, "You and what you worship other than God shall be the fuel of hell." However, Ubayy read hatab instead of hasab, as did 'Ali and'A'ishah.5 Bell translates, "coals," but in a note says it literally means "pebbles";6 Paret has "Brennstoff" with a query. 7

Hasab, in the meaning of fuel, is found only here. The basic meaning of the verb hasaba is "to pelt with pebbles" or "to scatter pebbles." From this sense the lexicographers redefine it to mean "to throw pebbles (i.e., fuel) on a fire"; others limit it to fuel that is thrown into an oven, or used as kindling, but they offer no shawahid in support of any of these meanings. In order to explain its strangeness, they hold that hasab is Ethiopic, or in the dialect of Nejd or the Yemen;8 the word is also said to mean "the fuel of hell" in Zanjiyah.9 All this only goes to show that it was not known to the Meccans and Medinese. Rabinl0 apparently takes the Yemeni ascription seriously, but does not mention Nejd or Ethiopia. He relates it to the Hebrew hasabh, the agent noun of which, hosebh, occurs in Isa. 10:15, as the hewer or chopper with an ax. However this is the only occasion on which the word "apparently" refers to cutting wood, the other instances refer to hewing stone.I I We note, too, that the regular Old Testament verb for cutting or gathering firewood is hatabh = Arabic hataba.

Obviously correct is hatab; it is the regular word in Arabic for firewood and occurs elsewhere in the Koran (CXI.4 and LXXII.15) in that meaning. Closely parallel to XXI.98 is LXXIL15: wa-amma l-gasituna fa-kanu li-jahannama hataban, "As for the unrighteous, they shall be fuel for hell." It is easy to see how the mistake occurred; in copying hatab, the scribe forgot to write the vertical stroke of the t, turning it into a s. This is much like our forgetting to cross a t or dot an i, something that everyone does from time to time.


The word ummah appears twice in the Koran in the apparent meaning of "while, time": XI.8 reads wa-la-in akhkharna `anhumu l- `adhaba ild ummatin ma `dudatin la-yaqulunna ma yahbisuhu, "And if we postpone for them the punishment for a reckoned (amount of) time, they will surely say, `What is holding it back?"' And in XII.45: gala lladha naja minhumd wa-ddakara ba `da ummatin, "And the one of them who was saved remembered after a time and said."

These two occurrences have not attracted much attention from Western scholars. Paret, in a note on XI.8 says only that here and in XII.45, ummah means "Frist, Weile," thus accepting the meaning given by the majority of the commentators.12 Blachere translates XI.8 by "jusqu'a un moment compte,"13 and XII.45 by "s'amendant apres reflexion," and notes that he translates by intuition, and that the commentators take it to mean "apres un temps," which has little relation to the sense of the root.14

Ummah, of course, cannot mean "time, while," but this is one of the cases in which the commentators instinctively grasped the meaning necessary and went on to redefine the word accordingly. In XI. 18 they all assert that the word means "time, while" (hina, zaman), but there is some variety of opinion on XII.45. In addition to "time," some suggest immah "favor"15 (poorly attested), and amah or amh "forgetting."16 Ummah makes its way into the list of dialect words in the meaning sinin "years" (Azd Shanu'ah) and as "forgetting" (Tamim).17

The meaning plainly must be "time, while," as the majority of the commentators held, and this we can restore simply by emending h to d, and reading amad, which means "time, term, period of time." The addition of the feminine ending to macdud would occur naturally to anyone reading ummah for amad; the copyist may have thought he was correcting the text, but he may have done it instinctively without being aware of it. Amad occurs four times elsewhere in the Koran, 111.30, XVIII.12, LVIL 16, and LXXII.25.


In a brief passage in Surah LXXX:26-32 God enumerates some of the blessings-specifically foodstuffs-that he has bestowed on mankind. Thumma shagagna l-ardor shaqqan 26, fa-anbatna fiha habban 27, wa`inaban wa-qadban 28, wa-zaytunan wa-nakhlan 29, wa-hada'iga ghulban 30, wa-fakihatan wa-abban 31, mata'an lakum wa-li-an'amikum 32; "Then we split the earth and caused to grow in it grain, grapes, and clover, olives and date-palms, and luxurious orchards, and fruit and abb, as a benefit for you and your livestock."

The crux in this passage is the word abb in verse 31, though there is some uncertainty about qadb in verse 28 as well. Blachere translates the latter as "cannes" (canes; possibly he means sugar-cane), apparently tacitly emending qadb to gasab.18 Paret translates: "Futterpflanzen" and marks the Arabic word with a query.19 Neither annotates the word. I believe that qadb is correct here; the word is well attested in the dictionaries where it is defined as "clover" (ratbah, fisfisah), "lucern" (qatt), or anything that is cut and eaten while it is green. There are several shawahid and several other words with related meanings derived from the same root.20 It may be that Blachere preferred gasab to gadb because he felt that the needs of the livestock were taken care of by abb, but this is not so. They were taken care of by gadb.

The word abb is glossed by the commentators as "fodder, pasturage" (mar'an, kala'), as "grass" ('ushb), "straw" (tibn), and "dried fruit." 21 They were doubtless influenced by verse 32, and since they could not know what abb really meant, "fodder, pasturage" was the best choice under the circumstances. There are hints, however, in our sources that some were not sure of the meaning and admitted their ignorance. Abu Bakr, when questioned about abb, exclaimed, "What heaven will cover me, or what earth will carry me, if I say about the book of God something I do not know?"22'Umar after reciting the verses remarked, "I know what fakihah is, but what is abb?" Then at once he checked himself and exclaimed that this was presumptuousness (takalluf). In another version of the same story he states, "Sufficient for us is what we already know."23 Abb was assigned by some to the dialect of the People of the West (ahl al-gharb), presumably the Berbers!24

Among commonplace words such as grain, olives, and date-palms, abb was very cryptic, so scholars felt obliged to work hard to give it similar currency. In addition to redefining the word, they invented shawahid, both prose and verse, trying to show that abb meant pasturage. An anonymous poet is quoted as saying: "Our tribe is Qays and our home is Najd; we have there pasture (abb) and a watering place."25 In the list of poetic shawahid falsely ascribed to Ibn Abbas we find another anonymous verse: "You see in it pasturage (abb) and gourds mingled together, on a way to water beneath which willows run."26 Zamakhshari cites the following expression: Fulanun rd'a lahu 1-habbu wa-ta`a lahu l-abbu,27 which Lane translates: "Such a one's seed-produce [or grain] increased and his pasture became ample."28 Another statement is ascribed to the legendary Quss b. Sacidah: Fa ja`ala yarta`u abban wa-asidu dabban, "And he proceeded to graze on abb while I hunted for lizards."29 The prose expressions may not have been invented to deceive, but may have been coined after abb as "pasture" had been absorbed into the vocabulary of educated people. One should not underestimate the power of the Koran to generate new expressions such as these.

A. Jeffery, following earlier scholars, relates abb ultimately to Hebrew 'bb "to be green," but assumes that it came into Arabic directly from Syriac 'b',30 which means "fruit" = fakihah.

Despite these attempts at redefinition and etymologizing, the fact remains that abb was not understood by the first commentators on the Koran. The word is not found in Arabic literature before or after its occurrence here (except the spurious verses and the proverbial expressions cited above) and it stands in the midst of common words that everyone could understand. Stylistically it is disturbing. What could be the purpose of reminding people of God's blessings using a word that not even the experts could understand? Everything points toward its being a word as commonplace as grain, olives, fruit, and so forth. In short, abb has to be a mistake.

We can restore the text with a very simple emendation, by reading lubban instead of abban. The copyist's pen, as it turned to the left after the lam, for a split second ceased to flow, thus breaking the connection with the following ba' and converting the lam into alif. Lubb is a common word meaning "kernel" or, according to the dictionaries, anything of which the outside is thrown away and the inside eaten; specifically mentioned are pistachio nuts and almonds. Today, if one buys libb from a street vendor in the Near East, he gets sunflower seeds or pumpkin seeds. Stylistically, fruit and nuts go together much better than fruit and pasturage.


In XXI.104 God describes how he is going to proceed on the last day: yawma natwi l-sama'a ka-tayyi l-sijilli lil-kutubi, "The day on which we shall fold up the heavens as the sijill folds up the writings."

The meaning of sijill, a well-known word in Arabic, is "document," consequently the "document" could not do any folding or rolling up of other documents. This problem has been approached from two directions. Some of the commentators realized that sijill had to be the subject of the masdar tayy, so they interpreted it as the name of an angel, a man's name, or the name of the prophet's scribe. Others, however, held that sijill was a sheet of vellum or papyrus (sahifah) and redefine the phrase to mean: ka-tayyi 1-sijilli `ala ma fiha mina l-kitdb, "as the sijill is folded over the writing that is on it." Tabari prefers the latter explanation since he says sijill is well known, and that there is no angel or scribe known by this name.31 The redefinition of the function of the preposition li-, however, is too drastic to be credible.

The Westerners generally follow Theodor Noldeke's opinion that Muhammad mistakenly took the name of the document for the writer of it.32 This idea, however, is untenable. Although he may have been illiterate, the Prophet was nevertheless surrounded by writing. He was a merchant and so was his wife. He dictated portions of the revelations to scribes, and he doubtless dictated his correspondence as well, and must have received letters that were opened and read before him. He had a share in the drafting of two important legal documents, the Constitution of Medina, and the Treaty of Hudaybiyah. In short, writing was so widely employed at that time that Muhammad could not have confused the document with its writer.

Those commentators who saw in al-sijill the writer and the subject of fayy were correct, although they could not take the last step necessary for reaching the correct reading. This problem can be solved with a simple emendation, by changing al-sijill to al-musjil or al-musajjil. The loss of the min is easy to explain. In older hands the mim after the definite article does not turn back under the lam as it does in later hands, but is often no more than a thickening of the connecting line between the lam and the letter following. Here, too, a leaky pen may have run the mim into the first tooth of the sin, causing the mim to lose its identity; and possibly one of the teeth was indistinct, thus facilitating the misreading.


This and the following emendation are of mistakes that arose from the inability of the writing system to indicate all the hamzahs that had been lost in the Hijazi dialect33 but were added at a later time when it was decided that the Bedouin pronunciation should prevail. Usually the absence of hamzah is indicated by one of the consonants alif, waw, or ya', but not always. In these two cases the absence of a possible carrier for the hamzah had already resulted in the erroneous readings that we find, so the revisers did not suspect that hamzahs were etymologically justified.

In Sura 11.58 God recalls that he told the children of Israel to enter the village and eat from it wherever they wished in ease, and then says: udkhulu l-baba sujjadan wa-qulu hittatun naghfir lakum khatayakum, "Enter the gate prostrating yourselves and say hittatun' and we shall forgive your sins." In VII. 161 we find essentially the same phrase repeated.

Bell and Paret leave the word untranslated,34 but Bell says it may come from Hebrew her "sin," and Blachere translates "dites, Pardon!"35 On page 645 he refers it to the Hebrew hatta, "sinners." Of the translators Blachere comes closest to the Muslim exegetes, who take the word to mean "forgiveness," that is, a "pulling down" of the burden of sin. Some commentators say that hittah means "speak the truth" (imptv., masc. pl.) in Zanjiyah.36

The word, however, must surely be the Arabic khitah, the Hijazi form of the Classical khit'ah, which is a masdar of khati'a, "to commit a sin." The spelling is like that of shth = shat'ahu, "its sprout, shoot" (XLVII.29); cf. GdQ, 3:43. The people, of course, are appealing for forgiveness but to obtain this they must first confess their sins. Khitatan < khit'atan with the implied omission of the verb khatind < khati'na is the equivalent of "We have sinned!" The word may have been pronounced kittatan, since some readers read al-marri for al-mar'i (VIII.24) and juzzun for juz'un (XV.44).37 Usually, however, this doubling is limited to wdw and ya'.38

We note finally that hittah is the only word in the Koran derived from the root htt, which means basically "put down," that is, from a higher to a lower level. There are twenty-two words, however, derived from kht', all of which have some meaning related to sin.


In 11.260 Ibrahim asks God to show him how he raises the dead. At first God doubts that Ibrahim really believes, but he insists that he wants to see the process only to ease his heart, so God gives him the following instructions: fa-khudh arba`atan mina 1-tayrifasurhunna ilayka thummaj'al `ala kulli jabalin minhunna juz'an thumma d'uhunna ya'tinaka sa`yan, "Take four birds and incline them toward yourself, then put a part of them on each mountain, then call them, and they will come to you flying."

The crux lies in the words fa-surhunna ilayka, which is the reading of the seven canonical readers without exception, but one finds also sir; rare and late seem to be surrahunna, "tie them," and sirrahunna, 14 shout at them."39

Blachere translates: "et serre-les contre toi (pour les broyer),"40 and says that he translates by intuition. Bell has "incline them to thyself," noting that the sense is uncertain.41 Paret: "richte sie (mit dem Kopf?) auf dich zu (und schlachte sie?)."42 In Kommentar, he notes that the commentators either read "incline them," which is not understandable, or "cut them up," with which the following "to yourself' does not fit.43 In short, neither of the accepted readings makes good sense. The meaning "cut up" is said to be Nabataean; others take it to be Greek.44

Tabari devotes several pages to these words 45 He cites the two major views on the meaning of sur, "incline" and "cut up," and decides emphatically for the latter, because the overwhelming majority of the exegetes hold this opinion, and he takes issue with a few Kufan lexicographers who insist that sara, yasu ru never means "cut up" in the language of the Arabs.

Both these groups are right, each in its own way. The lexicographers are right in denying that Sara means "cut up"; the shawahid are late or suspicious, so it looks as if the exegetes had redefined the word in the way we have noted before. However, the context clearly demands that the phrase read "cut to pieces," so the exegetes are "right" as well. One of them even goes so far as to insist that the pieces of the birds are all mixed up: "The wing of this one is with the head of that one, and the head of that one is with the wing of this one."46 Others say that the flesh and feathers are mingled.47

Since the meaning must be "cut them to pieces and mix them up," we can restore the text as follows: fa-jazzihinna (wa-)lbuk, which, not surprisingly, means "make them into pieces and mix (them) up." The emendation of sad to jam is no problem since the two letters resemble each other closely enough for such a misreading to occur. Jazzi, of course, is the classical jazzi'; the change of final-hamzated verbs to final-ya' verbs is well known, and was doubtless universal in the Hijazi dialect, where, as noted above, all the hamzahs had been lost. The meaningless ilayka is removed by reading ulbuk without any change in the rasm at all; the wawas dropped when the word was misread as ilayka. Another possibility is that this phrase originally read wa-labbik, which has the same meaning, on the assumption that the waw was mistaken for an alif. This is not impossible if the handwriting was small.


This and the following two emendations are of special interest since, in addition to correcting the text, they depend on assuming the same mistake. One could argue from this that all three were copied by a single scribe with a certain peculiarity in his handwriting.

The mysterious word mathani occurs twice in the Koran, first in XV:87: wa-la-qad ataynaka sab`an mina 1-mathani wal-qur'ana l-`azim, "We have given you seven mathani and the mighty Koran." It is found in a group of verses (86-97) in which God comforts the prophet in his disappointment at the doings of those who pay no attention to his message. The verse seems to be a reminder that God has favored him above all others with these special gifts.

Mathani is also found in XXXIX.23: Allahu nazzala ahsana 1-hadithi kitaban mutashabihan mathaniya tagsha`irru minhu juludu lladhina yakhshawna rabbahum thumma talinu juluduhum wa-qulubuhum ila dhikri llah, "God has sent down the best account, a book alike (in its parts), mathani, at which the skins of those who fear their Lord creep, then their skins and hearts become soft to the remembrance of God."

The problem of the mathani has generated much scholarly writing, most of which shall be ignored here.48

Muslim authorities derive the word from the root thny, and most of them assign it the meaning of something repeated, as can be seen in two of the translations above. However, the verb thana means "to double, to fold, to make something twofold by adding a second element to the first," but the idea of repetition easily follows. One of the meanings ascribed to Form II of the verb is "repeat, iterate."49 Others suggest that it comes from the Form IV verb athna "to praise."

An early suggestion by A. Geiger that the word is derived from the Hebrew Mishnah, or, as preferred by Noldeke, the Aramaic mathnitha, has been accepted by many Western scholars, but the word as used in the Koran does not reflect the character of the Mishnah.

The number seven has caused as much trouble as the word mathani itself. The exegetes say that the seven mathani are the seven longest suras of the Koran, or the seven verses of the Fatihah, which is the most popular view, or they are the suras that have less than a hundred verses but more than the shortest, which are called al-mufassalat; all of these definitions reflect the idea of repetition in some way. We note further that sab`an is a masculine numeral, so it demands a feminine singular. The only word that approaches the idea of repetition would be the past participle of Form II, muthannah, but this would take a plural muthannayat, not mathani.

Western scholarship has mostly accepted the theory that they are seven punishment stories that are scattered throughout the Koran because of the effect they have on the hearers. This is only speculation, but it is not refuted by our emendations.

I believe that the word in XV.87 should be emended to read al- mataliyi, and in XXXIX.23 to mataliya, meaning "recitations," literally, something that has been, or is to be, recited. This is the broken plural of matluwun, as in maktubun, makatibu "writings" and mazmurun, mazamiru "psalms," and others. The copyist mistook the lam for a nun because it was too short; having accepted nun the only other word that could be formed from the rasm was mabani, which could not be right, so he had to decide for mathani, and so initiated the idea of redoubling or repeating.

One reason that the scribe failed to read matali is that the word does not appear in the Koran nor does the singular matlaw. The verb talc, "recite," however, is very common, occurring more than sixty times in a variety of forms both active and passive, so the past participle is surely possible. It is likely that the verb talc was not much used in the common speech of Mecca, but came suddenly into extensive use only in the Koran, so the copyist was not alert to the possibility here. In the dictionaries, the space allotted to talc in this meaning is quite small and there are no shawahid. Additional evidence that the word was little used is provided by a mukhadram poet, Khufaf b. `Umayr al-Sulami, who in describing his beloved's campsite misuses the agent noun under the impression that it means "scribe": ka'annaha suhufun yakhuttuha tali, "as if they were pages written by a tall."so

Tala is usually used with ayat "signs" as its object or passive subject, but we find it with other words as well, including kitab (seven times, see concordance), so there is no discrepancy in XXXIX.23 in equating the kitab and the matali. The distinction, however, between kitab and qur'an (XV.87) is found elsewhere in the same sura (v. 1): tilka ayatu l-kitabi wa- qur'anin mubin, "Those are the signs of the book and (of) a clear Koran."

It is also necessary to emend sab`an, which I believe should be read shay'an. The mistake occurred when the scribe carelessly wrote a small loop resembling an 'ayn instead of the minim of the ya'. This is comparable to our writing a small e when we intend to make the shaft of an i. The next copyist, seeing s", could hardly do anything other than add the ba'. Seven was also doubtless congenial to him; it is virtually a sacred number in the Near East, and many things come in sevens. Since he did not know what mathani meant, he must have felt that the number seven was appropriate for such a mystery.

So XV.97 should read: wa-lagad ataynaka shay'an mina l-matalayi wal-qur'ana l-`azim, "We have given you some recitations and the mighty Koran."


In XXII.52 God points out that Satan distorts the message brought by messengers and prophets: wa-ma arsalna min qablika min rasulin wa-la nabiyin illa idha tamanna alga l-shaytanu fi umniyatihi fa-yansakhu llahu ma yulgi 1-shaytanu thumma yuhkimu llahu ayatihi, "We have not sent down before you any messenger or prophet but that when he desired, Satan injected (something) into his desire, but God cancels what Satan injects, then God makes his signs strong."

Tamanna and umniyatihi in the meaning "desire" (verb and noun) have caused problems for the translators. Bell has "but when he formed his desire Satan threw (something) into his formulation," with a note saying that the meaning is doubtful.51 Paret has "ohne dass ihm, wenn er etwas wunschte, der Satan (von rich aus etwas) in seinen Wunsch unter- schoben hatte."52 Blachere has "sans que le Demon jetat [l'impurite (?)] dans leur souhait, quand ils (le) formulaient."53 All three rely on the dictionary definition of tmanna, but none of them annotates the passage.

Tabari devotes most of his commentary on this verse to the reason for its revelation;54 it was sent down as a comfort to the prophet for having inadvertently, because of Satanic interference, spoken favorably of the pagan goddesses Allat, cUzza, and Manat. But on pages 113 f. he quotes from exegetes who hold that tamanna here means qara'a, talc, and haddatha. Ibn Hisham reports on the authority of Abu cUbaydah that the Arabs used tamanna in the meaning of gara'a, and cites two shawahid, obviously spurious since both refer to the recitation of the book of God.55

This is another example of the redefinition by the exegetes and/or lexicographers of the crucial word in a problematic passage in which the redefinition is correct. One should emend tamanna to read yumli, "dictates" and ft umniyatihi to ft imld'ihi, "in his dictation"; the latter was originally written 'mlyh, with no alif for the long a, a common feature of Koranic spelling. The nun was written for lam because the latter was too short, as in mathani, and one of the minims was lost. The word was probably pronounced imlayihu or imldyhu.56 After reading tamanna, umniyatihi was, of course, inevitable. The copyist may have felt more comfortable with the perfect tamanna, since idha yumli does not appear in the Koran; idha tutla, however, is found a number of times, and the two words mean much the same thing.


Sura 11.74-79 is a polemic against the Jews but directed to Muslim listeners. The Jews are denounced for perverting the true scriptures and for pretending to believe when they really do not. In verse 78 we read: wa- minhum ummiyuna la ya`lamuna l-kitaba illy amaniya wa-in hum illd yazunnuna, "And among them are ummiyuna who do not know the book except desires and they can only guess." The passage then ends with an imprecation against those who write a book with their own hands and say that it is from God just to make a small profit.

The meaning of ummiyuna has been much discussed by scholars and need not delay us here, since in this context it must mean ignorant people who do not know the scriptures. The problem for us is the meaning of illd amdniya. Bell, p. 11, translates, "except as things taken on trust, and who only cherish opinions," and notes that the meaning of the word is uncer- tain.57 Blachere has "qui ne connaissent point l'Ecriture [mais] seulement des chimeres, et [qui] ne font que conjecturer."58 Paret translates "Unter ihren [i.e., the Jews] gibt es Heiden (ummiyun), die die Schrift nicht kennen, (ihren Ansichten and Behauptungen) vielmehr (eigene) Wi nsche (zugrunde legen) and nur Mutmassungen anstellen."59 It is very unlikely that this one word can carry all the nuances that are heaped on it in the last translation.

Some exegetes define amaniya as lies (kadhib), talk (ahadith). Others cite the phrase yatamanna `ala llahi al-batila wa-ma laysa lahum, which seems to mean "and they want to get vain things from God and what is not due them." Tabari himself prefers the meaning "lies, falsehood" and in arguing for it has to maintain that tamannd here cannot mean tall ( _ amla), which as we have seen, was derived from XXII.52, nor have its usual meaning "desire," but must mean takhallaqa, to kharrasa, and ifta`ala, all of which mean "falsify, fabricate." He accuses the ummiyuna of committing such sins because of their ignorance of the book, that is, the Torah.60 It seems, however, that the meanings other than "to desire" and "to ask, i.e., someone to satisfy one's desire" all derive from this passage in the Koran. Here we get more help from the lexicographers than the exegetes, since the former redefine the word as "recitation." Abu Ishaq al-Zajjaj (d. 311/923), in discussing this verse, says plainly: ma`ndhu 1-kitaba illd tilawatan, "its meaning is: (they do not know) the book except by recitation."61

I believe that amanrya, like mathani and tamanna, is a result of the copyist's mistaking lam for nun, and should be emended to read amaliya "dictations." So the passage should run: "And among them are ummiyuna who do not know the book except dictations (from it) and so they can only guess." These poor ignorant people know of the scriptures only what the evil perverters of the word and the forgers mentioned in the following verse will let them know. They are victims to be pitied and not reproached. Since they had no scriptures at all, they could not be the perverters of it, nor could such ignorant people be so effective as forgers as to write out the book with their own hands, as mentioned in verse 79. The perverters and the forgers must be the same group and the people they deceive are the ummiyuna, certainly not the Muslims, who now have the true scriptures.


Sura 11. 134-41 is a segment in which God answers the Christians and Jews who urge people to be Christian or Jewish in order to be rightly guided: qul bal millata Ibrahima hanifan wa-ma kana mina l-mushrikina 135, qulu amanna billahi wa-ma unzila ilayna ... 136, fa-in dmanu bi-mithli ma dmantum bihi fa-qadi htadaw wa-in tawallaw fa-innama hum fi shigagin fa-sa-yakfikahumu llahu wa-huwa l-sama`u 1'alim 137, sibghata llahi wa man ahasanu mina lldhi sibghatan wa-nahnu lahu 'dbidun 138. "Say (sg., addressed to Muhammad), no, rather the community of Ibrahim, a hanif, for he was not one of the polytheists. Say (pl. addressed to the Muslims), we believe in God and what has been sent down to us ... (here follow the names of all the prophets whose messages the Muslims believed in).... And if they believe in the same things you believe in, they are rightly guided, but if they turn away, they are in schism, but God will take care of them for you (sg.), for He hears and knows; the sibghah of God! and who is better at sibghah than God? so we worship Him."

Bell translates: "The savour of Allah, and in savour who is better than Allah? Him are we going to serve," and notes that the exact meaning is uncertain. ("Savour" is singularly ill chosen.)62 Blachere has "Onction (?) d'Allah! qul donc est meilleur qu'Allah en [Son] onction? [Dites] Nous sommes Ses adorateurs." In a note he rejects the explanations offered by the exegetes (see below), but admits that "onction" is not satisfactory, and suggests it might mean: "L'allure procuree par Dieu a l'homme converti au Monotheisme d'Abraham."63 Paret translates: "Das baptisma (? sibga)"; in Kommentar he cites the commonly held views, and adds the opinion of E. Beck (from Le Museon 65 [1952]: 92) that the word, which means baptism (Taufe) is used here in a more general sense for religion, which agrees with the exegetes' views.64 Jeffery derives it from the Syriac but does not discuss its meaning in the Koran.65

The word gave considerable trouble to the exegetes. They knew it meant the Christian baptism, but because in the passage the Jews are referred to as well, some of them expanded its meaning to include cir- cumcision.66 However, it is the Muslims who receive the sibghah of God and so neither baptism or circumcision can apply-the Jews and the pagan Arabs already practiced circumcision. The exegetes therefore redefine the word asfitrah, din "religion," imdn "faith," or they equate it with the millata Ibrahim in verse 135, which they take to mean Islam. Thus Tabarl paraphrases: bali ttabi`u millata Ibrdhima sibghata llah; and Qatadah says: wa-inna sibghata lldhi l-Isldm.67 With this interpretation, however, the comparison at the end makes little sense; can one really ask, "Who is better at Islam or imdn than God?" Other redefinitions of sibghah are shariFah "law," and khilqah "constitution, make-up."68 Grammatically, most of the commentators take sibghah to be in apposi tion with millah, even though the two are rather far apart. Those who take sibghah to mean iman take it as the acc. internal object of amanna in verse 136.

In this case I believe that the exegetes were far off track. It is to me inconceivable that one should find in the Koran the name of a Christian sacrament used-even metaphorically-for Islam or iman. The whole idea runs counter to the general attitude toward Christianity and Judaism in the Koran, and is so disturbing that the word practically announces itself as a mistake.

Neither the exegetes nor the orientalists have considered that sibghata llah might refer simply to the words immediately preceding: fasa-yakfikahumu llah. Taken thus, sibghah is an exclamatory acc., used in praise of God's action in sparing the prophet the trouble of dealing with his own enemies. There are two emendations that would give this sense. The first is to read sanilah, "favor." This emendation can be effected without altering the rasm at all if we assume that the original sad did not have the little nub on the left-this is often omitted in manuscripts-but that the next copyist took the nun to be the nub. Otherwise we can add a minim to the rasm, a minor change which is easily acceptable.

The second possibility is to read kifayah, the masdar of kafa, which would have been spelled kfyh, the long a without alif. In older mss and inscriptions the initial kaf is often written without the diagonal stroke that we add separately. The line of the letter runs parallel to the line of writing so that it sometimes closely resembles sad and ddl. The copyist first misread kaf as sad, and then carelessly took the loop of the fa' as a minim. Kifayah is what we should most likely expect grammatically, given fa-sa-yakfikuhum above, but on the whole I prefer sani'ah since fewer changes are required to bring it into line. Both "favor" and "sufficiency" are stylistically better in this position than any of the other meanings proposed, and the comparison at the end of verse 138 makes good sense with either of them.


Sura VII.46 and 48 speak of a group of men who are situated in some coign of vantage from which they can observe both the blessed in heaven and the damned in hell: wa-baynahuma hijabun wa-`ala l-a`rafl rijalun ya'rifana kullan bi-simahum wa-nadaw ashaba 1-jannati an salamun 'alaykum lam yadkhuluha wa-hum yatma'una 46, "Between them is a curtain and on the a`raf are men who know each by their mark and they call to the people of heaven, Peace be with you; they have not entered it but they hope to." However, these same men, when they look at the people of hell, pray to God not to put them with the sinners, and we then read: wanada ashabu l-a`rafi rijalan ya`rifunahum bi-simahum gala ma aghna 'ankumjam'ukum wa-ma kuntum tastakhbiruna 48, "and the people of the a'raf call to men whom they know by their mark; they say, Your collecting (of money) has not helped you nor has your arrogance."

The word alraf is the plural of `urf, which means "mane" or "comb of a cock," so if correct here it must be used metaphorically. Bell, however, translates "men of recognition," reading i`raf instead of a`raf.69 However i'raf does not mean "recognition" but only "inform someone of his misdeeds and forgive him," and "to have a sweet odor" (from 'arf), and "to have a long mane" (from 'urf).70 Bell's i'raf is rejected by Blachere, who leaves the word untranslated, but has a long note in which he reviews the opinions of some of the exegetes, he makes no suggestion as to the lexical meaning of the word.71 Paret translates simply, "auf den Hohen" and "die Leute der H6hen."72

The problem in this passage is both textual and eschatological. The eschatological problem concerns who the ashab al-a`raf really are. Some orientalists, notably Bell and Tor Andrae, think that they are the inhabitants of the highest realm of heaven, but in order to get this out of the text they have to take the people of heaven as the subject of lam yadkhuluha wa-hum yatma'una 46. This results in very clumsy Arabic and the exegetes are doubtless correct in keeping the ashab al-a`raf as subject here and in the following verse (47). The ashab al-a`raf are men who are not yet sure whether they are going to heaven or to hell.

I would first point out that a`raf may not be incorrect. The word might be used here metaphorically of some high place on which these observers are located. What makes it a bit suspicious is that the metaphor does not appear to have been used in Arabic either before or after the revelation of this passage. Furthermore, if the word refers to the top of the hijab, as some think, one would expect 'ala a`rafthi. We can propose two emendations here, neither of which has to be metaphorical, though the second one may be.

The first is ajraf, plural of juruf or jurf, which means "bank," specifically of a wadi that has been undercut by the current, or simply, "a bank or ridge that rises abruptly from the bed of a torrent or stream."73 Such a position would allow the observers an unimpeded view of what was going on below. Palaeographically there is no difficulty. Sometimes in early manuscripts and papyri initial ha' begins with a lead-in line like a small arc with the concavity facing right, which then continues downward to the right completing the main body of the letter. If this arc is exaggerated, the whole letter can easily be taken for an `ayn.

The other suggestion is ahruf, plural of hart, which means, among many other things, "point, ridge, brow, ledge, of a mountain."74 The same emendation, `ayn to ha', is needed here as in ajraf, and the alif presents no problem. It might have been introduced at the time of the Uthmanic recension, or it could have been added by `Ubaydalldh b. Ziyad, who, during his governorship of Kufah (53-59/673-679), instituted a reform in Koranic orthography that consisted of the introduction of about two thousand alifs into the text.75 Taken this way, ahruf is not metaphorical, but we find the singular, harf, used metaphorically in Sura XXII. 11: wa-mina l-nasi man ya`budu llaha `ala harfin fa-in asabahu khayruni tma'anna bihi wa-in asabahu fitnatuni nqalaba 'ala wajhihi khasira l-dunya wal- akhirata, "And among the people there are those who serve God on a harf, and if good comes to them they are at ease with it, but if trouble comes to them, they turn back to their (old) ways. They lose both this world and the next." These people who serve God "on a ridge" are fencesitters and summer soldiers who are not sure which way they will jump, since circumstances can vary. The same is true of the ashab al-a'raf, who are not sure whether they will end up in heaven or hell, since it depends on God's will, which they do not yet know. The two usages are not exactly parallel since al-a`raf is plural and def. and harf is singular and indefinite; nevertheless, the similarity is striking. In general, I prefer the reading ahruf, but would suspend judgment on whether it should be taken metaphorically or not.


Some years ago I wrote an article76 in which I argued that the Mysterious Letters (the fawatih al-suwar or al-huruf al-mugatta'ah) of the Koran were old abbreviations of the basmalah. The argument was based on the assumption that these abbreviations, like the words studied above, had been corrupted through copyists' errors, so it is not inappropriate here to add a few additional observations on the fawatih, and, in particular, to record a change of opinion with regard to some of them.

At that time I was anxious to avoid any suggestion that the emendations proposed might be arbitrary, so I left out of account those groups of letters that might, as they stood, be considered abbreviations of the basmalah. In so doing I relegated HM to a footnote (p. 280, no. 72), although I was convinced that it derived from an original BM or BSM. I think now that I was somewhat overcautious, since HM-to be read BSM and not BM-is the best evidence in favor of the hypothesis.

The derivation is well supported palaeographically. The ba' of the basmalah often begins with a flourish, which in some cases, especially in carelessly written manuscripts and papyri, starts above the line to the left, proceeds to the right and then turns under to form the rest of the letter, giving it a form that can easily be mistaken for ha'. Today in printed texts the ba' is written taller than usual and bends slightly to the left. This practice probably descends from the ancient practice, which in handwriting could be exaggerated.

The sin of the basmalah is often flattened out to such an extent that it appears to be omitted altogether. Tradition tells us that Zayd b. Thabit disapproved of writing the bsm of the basmalah without the sin, and Ibn Sirin did not like people to stretch the ba' to the mim until the sin had been written. The caliph 'Umar is said to have beaten a scribe for omitting the sin from the basmalah.77

These anecdotes date from a time when interest was growing in how the Koran should be written, and in which the Kufic hand was in the course of development. In fact, Ibn Sirin (ha. 110/728) might well have taken an interest in such matters.

Tables 3 and 4 (p. 282) can now be largely ignored since they make the process of corruption much more complicated than it really was. In HM ZSQ, I would now keep the two "words" separate as they regularly appear in the Koran. Both segments I believe represent an original BSM. The first to be written was the second segment, which was eventually corrupted to read CSQ; this was not understood by a subsequent copyist or editor who added at the beginning another BSM, which was later misread as HM. The copyist may have been the same one who wrote BSM (> HM) in all the suras where the latter appears.

The original BS was misread as `S by the Uthmanic editors and as simple S by Ibn Masud because of uncertainty as to the number of minims. The first two were probably badly written as well since they resembled an initial `ayn. Ibn Mascud's SQ is closer to the original than the Uthmanic SQ.

KHYCS turns out to be less of a problem than I had originally thought. The real crux is in the ha', but this can be solved by dividing the letters into two segments, KH and Y'S, following the example of HM SQ.

In discussing the word kifayah, we pointed out how kaf closely resembles dal and sad in some early hands; it may also resemble the carelessly written ba' that we have seen in HM. I believe that this kaf was originally a ba', and with this reading all the other difficulties vanish. The resulting BH-which could have been an original BSM-is a good abbreviation of the basmalah, and in Y'S we can see how the yd' and the open-topped `ayn were miscopied from an original sin in which the teeth were not clearly written. There may even have been a fourth minim representing ba', which could have been swallowed up when the two segments were combined later on. The original form was like Ibn Maslud's variant SQ (= SM < BSM).

Similarly, I conclude that Y'S was written first, then not understood by a later copyist, who added BH or BSM (> KH) to represent the basmalah.

The final point concerns those abbreviations in which the letter ta' is found. In the article referred to (p. 280), I assumed that these ta's all went back to an original BA. This, however, is not satisfactory, for since the basmalah in XXVII.30 at the head of Solomon's letter to the Queen of Sheba is spelled without alif, it is not likely that any abbreviation of the phrase would contain that letter. I now believe that the vertical strokes of the td's were originally cancellation marks, added by some copyist when he went through his old surahs to write out the basmalah in full. The vertical cancellation mark is well known from later manuscripts and there is no reason why it should not have been employed here. One should keep in mind that the Arabs at the time of Muhammad were not an ignorant people struggling toward literacy; writing was widely used, though not for literature apparently, especially in urban centers such as al-Hirah, where a chancery style must have been employed. The heads of the ta's, now unencumbered by alifs, become simple ba's, written in the same careless way as the others that are concealed under HM, and the resulting BS, BSM, and BH are all good abbreviations of the basmalah. Although it is not necessary for the argument, I believe further that BS and BH also go back to an original BSM. The final flourish of the sin and the final ha' could both easily have been miscopied from a mim.

I am now more than ever convinced that the fawatih are indeed old abbreviations of the basmalah that suffered corruption at the hands of later copyists. And, after all, what can more properly stand before a surah than the basmalah?

It should not be assumed that in making these emendations I am in any way trying to diminish the remarkable achievement of Zayd b. Thabit and his colleagues in producing the Uthmanic recension of the Koran. When one considers that the Arabs at that time had no literary culture based on written texts, their accomplishment becomes truly monumental, and one can readily believe that Zayd really said when ordered to do the editing, "By God, if they had charged me with carrying a mountain, it could not have been heavier for me than this." Without any experience of editing or, indeed, of reading a book of similar size and content, they were able to publish a work that has taken its place as one of the three or four greatest books that mankind has produced. It remains for modern scholarship to correct the few mistakes that they overlooked, and to restore the text to the form it had when first spoken by the prophet Muhammad.


1. In the cruces discussed below I have found only one proposed emendation, that of R. Bell, who wanted to read i`raf for a'raf; see section 11; this does not effect the rasm.

2. See T. Noldeke et al., Geschichte des Qorans [hereafter GdQ] (Hildesheim: Georg Olms Verlagsbuchhandlung, 1961), 3:2f.

3. I. Goldziher, Die Richrungen der islamischen Koranauslegung (Leiden: E. J. Brill, 1952), p. 36.

4. Abd al-Rahman al-Suyuti, Al-Itgan fi 'ulum al-Qur'an, ed. Muhammad Abu 1-Fadl Ibrahim (Cairo: Maktabat al-Mashhad al-Husayni, 1387/1967), 2:275, where other mistakes are noted. The scribe who wrote yay'as was probably not sleepy but confused by similar consonantal outlines. The words yay'as and yatabayyan are so different that such a mistake could not have occurred in the oral tradition, so we have to look to the written tradition for an explanation. However, the Uthmanic rasm of yay'as is y'ys, so it is equally difficult to see how it could be a mistake for yatabayyan, or vice versa. My guess is that yay'as was originally written y'ys, and so the two words are virtually identical. Each has four minims: yys (probably pronounced yayyas) with the two yd's and the first two teeth of the sin, and yatabayyan with its ytby. The final flourish of the sin was mistaken for a nun, or vice versa. For the loss of hamzah in the Hijazi dialect and compensatory lengthening of a preceding waw or yd' with sukun, see section 5.

A minim-the term is borrowed from medieval Latin palaeography-is the shortest vertical stroke in any given hand. The word is not wholly suited to Arabic, since in good Arabic manuscripts adjacent minims are often written with slightly differing heights to show that they belong to different letters. It is convenient, however, since it can be used of the teeth of the sin, the nub of the ba', ta', etc., and also of those nubs that are mistakes, even those that are omitted: Next to the omission or misplacement of dots, minim errors, that is, copying more or fewer minims than are in the original, are the most common mistakes in Arabic mss.

5. R. Bell, The Qur'an: Translated with a Critical Re-arrangement of the Surahs (Edinburgh: T. & T. Clark, 1937), p. 313.

6. R. Paret, Der Koran: Ubersetzung (Stuttgart: Kohlhammer, 1962), p. 269.

7. A. Jeffery, Materials for the History of the Text of the Qur'an (Leiden: E. J. Brill, 1937), p. 147.

8. Muhammad Murtada al-Husayni al-Zabidi, Taj al-'arus min jawahir al- Qamus, ed. `Abd al-Sattar Ahmad Farraj. Kuwayt: Matba`at Hukumat Kuwayt, 1385/1965), 2:283; E. W. Lane, Arabic-English Lexicon (London: Williams and Norgate, 1863-1893), p. 581

9. Suyuti, Itgan, 2:111.

10. C. Rabin, Ancient West Arabian (London: Taylor's Foreign Press, 1951), p. 26.

11. F. Brown et al., Hebrew and English Lexicon of the Old Testament (Oxford: Clarendon Press, n.d.), p. 345.

12. Paret, Ubersetzung, p. 23.

13. R. Blachere, Le Coran: Traduction nouvelle (Paris: G. P. Maisonneuve, 1949), p. 433.

14. Ibid., p. 471.

15. `Abdallah b. 'Umar al-Baydawi, Anwar al-tanzil wa-asrar al-ta'wil, ed. H. Fleischer (Osnabruck: Biblio Verlag, 1968), 1:462.

16. Muhammad b. Jarir al-Tabari, Jami `al-bayan fi tafsir al-Qur'an (Bulaq, 1323; Beirut: Dar al-Ma`rifah, 1409/1989), 11:135.

17. Suyuti, Itgan, 2:97, 101.

18. Blachere, Le Coran, p. 36.

19. Paret, Ubersetzung, p. 500.

20. Zabidi, Taj, 4:49-52; Lane, Arabic-English Lexicon, p. 2538.

21. Zabidi, Taj, 2:5 f.; Lane, Arabic-English Lexicon, pp. 3 f.

22. Suyuti, Itgan, 2:4.

23. Tabari, Jami `al-bayan fi tafsir al-Qur'an, 30:38.

24. Suyuti, Itgan, 2:108.

25. Zabidi, Taj, 2:5.

26. Suyuti, Itgan, 2:84.

27. Mahmud b. 'Umar al-Zamakhshari, Asas al-balaghah (Beirut: Dar Sadir, 1399/1979), p. 9.

28. Lane, Arabic-English Lexicon, pp. 3 f.

29. Zabidi, Taj, 2:6.

30. A. Jeffery, The Foreign Vocabulary of the Qur'an (Baroda: Oriental Institute, 1938), p. 43.

31. Tabari, Jami 'al-bayan fi tafsir al-Qur'an, 17:78 f.

32. Theodor Noldeke, Neue Beitrage zur semitischen Sprachwissenschaft (Amsterdam: APA Philo Press, 1982), p. 27; Jeffery, Foreign Vocabulary, p. 164.

33. This feature of Hijazi Arabic is discussed at length by Rabin, Ancient West Arabian, pp. 130 if.

34. Bell, The Qur'an, pp. 9, 143; Paret, Ubersetzung, pp. 12, 137.

35. Blachere, Le Coran, pp. 645, 742.

36. Suyuti, Itgan, 2:111.

37. Rabin, Ancient West Arabian, p. 134.

38. See no. 2, above.

39. Zabidi, Taj, 12:361 f.

40. Blachere, Le Coran, p. 309.

41. Bell, The Qur'an, p. 39.

42. Paret, Ubersetzung, p. 39.

43. R. Paret, Der Koran: Kommentar and Konkordanz (Stuttgart: Kohlhammer, 1971), p. 56.

44. Suyuti, Itqdn, 2:114.

45. Tabari, Jami 'al-bayan fi tafsir al-Qur'an, 3:36 f.

46. Tabari, Jami `al-bayan fi tafsir al-Qur'an, 3:37 ult.

47. Ibid., 3:38.

48. For a summary and extensive bibliography, See Paret, Kommentar, pp. 279 f.; also Jeffery, Foreign Vocabulary, p. 26. Paret (Ubersetzung) translates XV.867: "Sieben Erzahlungen"; Blachere (Le Coran, p. 223), "Sept des repetitions"; Bell (The Qur'an, p. 243), "Seven of the repetitions."

49. Lane, Arabic-English Lexicon, p. 360.

50. Abu Ghalib b. MaymUn, Muntaha 1-talab, ms Laleli 1941, facsim. ed. by F. Sezgin. (Frankfurt am Main, 1986), p. 23.

51. Bell, The Qur'an, p. 322.

52. Paret, Ubersetzung, p. 276.

53. Blachere, Le Coran, p. 1043.

54. Tabari, Jami `al-bayan fi tafsir al-Qur'an, 17:131-34.

55. 'Abd al-Malik Ibn Hisham,. K. Sirat rasul Allah, ed. F. Wastenfeld (Frankfurt am Main: Minerva, 1961), pp. 370 f.

56. For the Hijazi suffix -hu, where Classical Arabic has -hi, see Rabin, Ancient West Arabian, pp. 99, 151; for the loss of hamzah in Hijazi, see above, pp. 783-84.

57. Bell, The Qur'an, p. 11.

58. Blachere, Le Coran, p. 748.

59. Paret, Ubersetzung, p. 14.

60. Tabari, Jami `al-bayan fi tafsir al-Qur'an, 1:297 f.

61. Ibn Manzur al-Ifrigi, Lisan al-`Arab (Beirut: Dar Sadir. 1374-1376/ 1955-1956), 15:294.

62. Bell, The Qur'an, p. 18.

63. Blachere, Le Coran, p. 767.

64. Paret, Ubersetzung, p. 21; Kommentar, p. 34.

65. Jeffery, Foreign Vocabulary, p. 192.

66. Ifrigi, Lisan, 8:438.

67. Tabari, Jami 'al-bayan fi tafsir al-Qur'an, 1:444.

68. Ifrigi, Lisan, 8:438.

69. Bell, The Qur'an, pp. 141 f.

70. Lane, Arabic-English Lexicon, p. 2014.

71. Blachere, Le Coran, pp. 618 f.

72. Paret, Ubersetzung, p. 126 f. For further bibliography on this much-discussed point, see Paret, Kommentar, p. 160, and Jeffery, Foreign Vocabulary, p. 65.

73. Lane, Arabic-English Lexicon, p. 411.

74. Ibid., p. 550.

75. Noldeke, GdQ, 3:255 f.

76. A. Fischer, "The Mysterious Letters of the Koran: Old Abbreviations of the Basmalah," JAOS 93 (1973): 267-85.

77. Suyuti, Itgan, 4:159.

If you find an error please notify us in the comments. Thank you!