In Search of the Original Koran: The True History of the Revealed Text by Mondher Sfar

as "In the name of Allah, the Merciful Benefactor" (Blachere), or as "In the name of God lenient and merciful" (Paret). It is evident, after what I have just said on the subject of the gradual formation of surahs, that this formula could not have been considered as being part of the surahs until after their composition.
    This might be proved by other considerations. First of all, as Welch demonstrated, the first Koranic revelations referred to God as rabb (Lord). It was only in the second period that the names "Allah" and "al-Rahman" appeared (with even a preference for the latter, for example, in surah 19, where "al-Rahman" is cited sixteen times). Verse 17:110 authorizes Muslims to use the two names of God: "Say: call on Allah or call on al-Rahman! By whatever name you call on Him, His are the most gracious names." This verse might be explained by the debate (related in Tradition) on the subject of the rejection by certain Qurayshite companions of the use of the name al-Rahman, who preferred the name Allah. No doubt it was after these incidents that al-Rahman became less and less used as a name for God in the revealed texts.28

    Let us note here that the name "Rahman" is the name of a south Arabian divinity. It was assimilated in the west Semitic domain to Hadad, the god of thunder. Musaylima, who pretended to prophesize in the time of Muhammad, was directly inspired by this same god alRahman. The debate over the name of God might have had a direct or indirect relation with the dissidence of Musaylima, who was executed by the prestigious military leader Khalid ibn al-Walid during a military expedition ordered by the first caliph of Islam, Abu Bakr, after the death of the Prophet. In addition, we find the story of Musaylima curiously mentioned in the context of the collecting of the Koran, since certain accounts explain that the decision to undertake it was taken as a result of the death of a great number of readers of the Koran in the course of the battle waged against this "false" prophet. This battle might perhaps have been the culmination of a disagreement that went back to the debate over the god al-Rahman and that would have needed sanction from the new powers to refine an official text that might ratify this victory and prevent the revelations of Musaylima being considered as canonical. This danger was all the more real because Muhammad had adopted an attitude, as we have seen, of reconciliation with al-Rahman.
    And so there might have been a certain relation between the elimination of Musaylima and the commencement of the formation of a Koranic canon. Al-Kindi, a Christian Arabic author of the eighth and ninth centuries of the Common Era, replies to a Muslim opponent on the subject of the inimitability of the Koran that he had held in his hands a compilation of the revelations of Musaylima: "You cannot dis regard that men such as Musaylima al-HanafI, al-Aswad al-`Anasi, Tulayha al-Asadi, and so many others, produced works similar to that of your master. I attest, for my part, that I have read a collection by Musaylima that, had it appeared, might have brought several of your friends to renounce Islam. But these men had no support, as was the case for your master."29

    Quite evidently, we do not know if the basmala was present in Musaylima's compilation, but Muslim Tradition was reluctant to count the basmala as a verse, even if the current vulgate considers it such. According to Tradition, some compilations of the Koran assimilated this formula into a verse, thus increasing the total number of Koranic verses by 114. In his Kitab al-kashf, al-Qaysi, who relayed this fact, rejected this practice as not conforming to the consensus among the companions of the Prophet or to the view of their immediate successors .3' In fact, two important legal schools defended opposing doctrines: the legal advisers of Medina, Basra, and Syria refused to grant the basmala a verse status, reducing it to a simple editorial technique serving