SHĪ'A LAW 265

set free—a kind of primitive systematic reasoning—and on the other, the conservative resistance to this innovation. We find both doctrines expressed in Iraqian traditions, the first, which was the ancient Kufian doctrine, ascribed to ʿUmar (Āthār A.Y. 872), the second to ʿAlī.²

There were, furthermore, compromises suggested between these two extreme doctrines. One, attributed to Ibn Masʿud but not with the standard isnād of the Kufian school,³ set the umm al-walad free at the death of her master as a charge on the share of her child; another set her free at the death of her master as a first charge on the whole estate. This last compromise, with the provision that the umm al-walad could not be sold or otherwise alienated, became the common doctrine of the ancient schools of law; the Kufians interpreted the statements of their earlier doctrine accordingly,⁴ and the Medinese, who entered the discussion only at this stage, expressed it in a tradition, through Nāfiʿ, from Ibn ʿUmar.⁵ Later, the isnād of this tradition grew backwards to the Prophet.⁶

Because the ancient resistance to an improvement in the status of the umm al-walad happened to be expressed in a tradition from ʿAlī, in pointed opposition to traditions from ʿUmar and Ibn ʿUmar, the Shiites, when they came to elaborate their own legal doctrines, insisted on considering the umm al-walad as a slave who could be sold or otherwise alienated by her master. This was not directly derived from the corresponding ancient doctrine, but introduced as a modification into existing Sunni law which the Shiites borrowed. We therefore find traces of the opposite Sunni doctrine in Shiite law: the 'Twelver' Shiites teach that the umm al-walad can be sold but becomes free at the death of her master provided she is still in his possession and her child is still alive; and the Zaidīs allow the sale of the umm al-walad but forbid the sale of the mudabbar slave.⁸