Sura I of the Koran bears on its face evidence that it was not originally part of the text, but was a prayer composed to be placed at the head of the assembled volume, to be recited before reading the book, a custom not unfamiliar to us from other sacred books of the Near East. The Koranic style, as is well known, is that in it from beginning to end, Allah is addressing man. In the Fatiha, however, it is man addressing Allah, and the common explanation that the word "Say!" is to be understood at its beginning, is obviously due to the desire to bring this first sura into harmony with the style of the rest of book. The sura, moreover, when we examine it, proves to be more or less a cento of ideas and expressions taken from other parts of the Koran. It is possible, of course, that as a prayer it was constructed by the Prophet himself, but its use and its position in our present Koran are due to the compilers, who placed it there, perhaps on the fly leaf of the standard codex. Its division into seven members in orthodox Muslim tradition has suggested the idea that it was put together as an Islamic counterpart to the Lord's Prayer.

The peculiar nature of the Fatiha has been recognized by Western scholars¹ from Nöldeke downward, but it is not merely a hostile Western opinion, for Fakhr ad-Din ar-Razi² quotes Abu Bakr al-Asamm (313)³ as saying that he considered it not to be part of the Koran and apparently the oldest commentaries began with Surat-al-Baqara. It is also well-known that the Fatiha was not included in the codex of Ibn Mas'ud.⁴ It is said that some early Kufic manuscripts of the Koran are to be found which commence with the second sura, and if they have the Fatiha, have it only at the end; but the present writer has never seen such an examplar.

It should not surprise us then if the Fatiha should have been handed down in somewhat different forms. One such variant form has for long circulated in Shi'a circles. In the Tadhkirat al-A'imma of Muhammad Baqir Majlisi (edition of Tehran, 1331, p. 18) it is given:

which we may translate:

Last summer in Cairo, I came across a similar variant version. It is given in a little manual of Fiqh, whose beginning, unfortunately, is missing, so that we do not know the name of the author. It is a quite unimportant summary of Shafi'i Fiqh, written, if one may venture a judgement from the writing, about one hundred and fifty years ago, perhaps a little earlier, in a clerkly hand, and the variant version is written on the inside cover under the rubric - qira'a shadhdha li 'l - Fatiha. The manuscript is in private possession, and though the owner was willing to let me copy the passage, and use it if I saw fit, he was not willing that his name be revealed, lest he come into disrepute among his orthodox neighbors for allowing an unbeliever to see such an uncanonical version of the opening sura of their Holy Book.

The text of this variant has certain similarities to that already given, and runs:

which, being interpreted, means:

Under the text follows the statement: Riwayat Abi 'l- Fathi 'l-Jubba'i 'an shaikhihi's -Susi 'an an-Nahrazwani 'an Abi 's Sa' adati 'l - Maidani 'an al - Marzubani 'an al - Khalil b. Ahmad.

On the readings in the two texts we may note: Sayyid for Rabb is merely a case of replacement by synonym. Sayyid is used in Sura xii: 25 for Joseph's master down in Egypt, and in iii: 34 of John the Baptist, who is announced as a sayyid, a chaste one, and a prophet, and the plural form is used in xxxiii: 67 for the chiefs whom the infidels followed and were led astray. It is not, however used of Allah.

Ar-razzaq occurs as a title of Allah in li: 58 - inna 'llaha huwa 'r - razzaq.

Mallak is a reading attributed to the third Kufan Reader among the Seven, al-Kisa'i (180), cf al-Alusi, Ruhu'l - Ma'ani, I, 78 and Abu Hayyan, Bahr, I, 20. It is curious that both the variant texts agree in this reading. Mallak is perhaps more precise and emphatic than the alternative forms malik, m¯alik and mali 'k, the first of which is perhaps the best attested reading, and the second is the TR [textus receptus "accepted text".]

Inna laka. This, and hiyyaka, wiyyaka, ayyaka, iyaka and the iyyaka of the TR, seem all to be independent attempts to interpret the unvoweled, unpointed skeleton term that stood in the original codex. Hiyyaka or hayyaka was the reading of Abu's-Sawwar al-Ghanawi (c. 180) and Abu'l Mutawakkil (102); wiyyaka or wayyaka was read by Abu Raja' (105).

Arshidna means much the same as the ihdina of the TR and was the reading in Ibn Mas'ud's codex (az-Zamakhshari in loc., and Ibn Khalawaih, p. 1) This imperative does not occur elsewhere in the Koran, but other forms from the root are commonly used, and the Shi'a variant is uses the imperfect of Form IV.

Sahil is a commoner word than the sirat of the TR, and is much more commonly used in the Koran, though both are foreign words, borrowed through the Aramaic. Sirata'l-mustaqim, taking it as in idafa, where al Mustaqim is a title of Allah, i.e.-, "the Upright One", was the reading of Ubai, Ja'far as-Sadiq and Abdallah b. 'Umar, so that it has very early and good attestation. It is a possible and appropriate reading, even though Mustaqim is not one of the Ninety-nine Names. That sabi la'l - mustaqim should occur in both these texts is curious.

Mananta and na' 'amta are simple replacements by synonym for they do not affect the meaning. Form IV of n'm is more common in the Koran than Form II, which is used only once in lxxxix: 14, but manna, with much the same meaning, is used still more often.

Siwa or ghair is a similar replacement by synonym, though siwa is not used elsewhere in the Koran.

Ghair for la was the reading of 'Umar, Ali, Ubai, Ibn, az-Zubair, 'Ikrima and al-Aswad among the early codices, and was supported by Ja'far as-Sadiq and Zaid b. 'Ali, so that it has respectable authority for a claim to be the original reading. It makes no change in the sense.

It will have been noticed that the sense of the Fatiha is precisely the same whether we read the TR or either of these variants. There is no ascertainable reason for the variant readings. They are not alterations in the interests of smoother grammatical construction or of clarity, nor do they seem to have any doctrinal significance. They are just such variants as one might expect in the transmission of a prayer at first preserved in an oral form, and then fixed later when the Koran was assembled.

The second variant form comes from Khalil b. Ahmad, who as a Reader belonged to the Basran School though he is said to have taken huruf from both 'Asim of Kufa and Ibn Kathir of Mecca, among the seven, and is even noted as the one who transmitted the variant ghaira from Ibn Kathir (Abu Hayyan, Bahr, 29; Ibn al-Jazari Tahaqat I, 177, 275; Ibn Khalawaih, p. 1). But he was also known to have transmitted from 'Isa b. 'Umar (149) (Ibn Khallikan, II,420) and was a pupil of Ayyub as-Sakhtiyani (131), both of whom were Basrans and famous for the transmission of uncanonical readings. It is thus quite possible that Khalil had access to good old tradition as to the primitive reading of the Fatiha. I can make nothing of the rest of the isnad from Kalil to al-Jubba'i, and possibly it is much later than the matn from Khalil.

¹ Nöldeke-Schwally, Geschichte des Qorans, I, 110.

² Matafatib al-Ghaib, V. 281.

³ Ibn al-Jazari Tabaqat, No. 3943 (vol. ii, p. 404). He was Imam of the mosque at Wasit, and a great authority on the isnads of the Kufan reader 'Asim, and one of the teachers of Abu Bakr an-Naqqash.

⁴ 'Abu 'Ubaid, Fada 'il, fol. 434. That Ibn Mas'ud knew the Fatiha as used liturgically, however, is clear not only from the fact that we have several variants in it from him (see the present writer's Materials fo the History of the Text of the Qur'an, p. 25), but also from the story coming from al-A'mash (148) that Ibn Mas'ud was asked why he did not include the Fatiha in his codex, and he answered that if he had included it he would have put it in front of every sura (Qurtubi, Al-Jami'li Ahkam al-Qur'an, I, 115). This statement shows quite clearly that he considered it to be a liturgical piece to be recited before reading the Koran. Late copies of Ibn Mas'ud's codex, made in the next generation or two, added the Fatiha at the beginning (Itqan, 152, 187; Fihrist, 26).