rel="nofollow" href="#fb3_img_img_077c6ee5-5deb-5491-bdf2-517a67eee076.jpg"/> Bakr archive from Hafsa. However, the ambition was not merely to reproduce Hafsa’s document exactly as it appeared, as ’Uthmn assembled a team of experts to examine and confirm the text. He appointed Zayd ibn Thbit as overseer and also recruited three Meccans to assist the process—one because he was an expert in the Arabic language, the other two because they were from Muammad’s tribe, the Quraysh—and reportedly asked that if the three ever contested Zayd’s opinion regarding a verse, that they should write the verse in the “original” Quraysh dialect.20 In a further innovation, ’Uthmn ordered that copies of the state-supported Qur’n be sent to major cities, and that competing local versions be destroyed. Some of these versions reportedly differed from ’Uthmn’s codex in the inclusion or exclusion of particular verses or entire sras. Similar to the way in which reports of Muammad’s statements would be authenticated, these rival collections of the Qur’n were associated with the prestige of specific Companions. In Kfa, where the Companion Ibn Mas’d’s collection had been established as the official version, there was brief resistance to ’Uthmn’s state codex. What we can gather from the accounts of these variants, explains Estelle Whelan, is that early Muslims were willing to base their arguments against each other “on the premise that the Qur’n had not been given definitive form by the Prophet to whom it had been revealed.”21 There is also a report that after Ab Bakr’s collection was returned to Hafsa, the governor of Medina demanded that she hand it over to him for destruction; even if her archive had provided the foundation for ’Uthmn’s project, it was not equal to the finished, official codex, and the governor feared that it would undermine the new caliphal archive. Hafsa refused; but after her death, the governor seized her collection and ordered that the pages be torn up.22 This could illustrate a point about the ironies of preserving tradition: To safeguard the Qur’n’s integrity and unity, the oldest complete and “official” copy of the Qur’n had to be destroyed.
What has been called the Qur’n’s “still-fluid pre-canonical text”23 did not instantly become solid with ’Uthmn’s codex, which remained capable of variation. It appears that the copies that ’Uthmn sent to cities such as Mecca, Damascus, Bara, and Kfa did not match each other perfectly, perhaps including copyists’ mistakes. ’Uthmn is reported to have allowed the imprecise copies, assuming that any mistakes would be corrected by knowledgeable Arabs.24 On top of these challenges, there was the problem of instability in Arabic writing. At the time, the Arabic script was not fully developed, lacking vowel marks or dots. The government’s Qur’n, therefore, provided only a bare consonantal skeleton, which allowed for multiple vocalizations and changes in meaning. Without voweling, the word mlk in the first sra could undergo a subtle shift in interpretation, recited as either malik (“king”) or mlik (“owner”). Without dots to properly distinguish the letters, the word fl (“elephant”) could be read in numerous ways, such as ql (“it is said”), qatala (“he killed”), or qabala (“he kissed”). While the variations themselves did not produce major controversies over meaning—at no point, for example, were there debates over 105:1’s mention of an elephant—the mere fact of difference nonetheless enabled competing schools to discredit each other through the charge of faulty readings.25 Roughly fifty years after the establishment of the ’Uthmnic codex, the Qur’
