ARGUMENTS FOR AND AGAINST TRADITIONS 51

even Shafi'i, who argues strongly for the acceptance of traditions even if they are transmitted by single individuals only, has to acknowledge it to a certain extent.¹ He points out, among other things, that the number of witnesses demanded for legal evidence is not always two. This fact is used in favour of the khabar al-wahid in two traditions which make 'Uthman and Zaid b. Thabit respectively accept the information of one woman on certain decisions of the Prophet (Ris. 60). For these decisions concern feminine matters, and a widely held doctrine admitted the evidence of one woman on such subjects.

The disparagement of the khabar al-wahid was, in fact, so typical of the ancient schools of law that Shafi'i, using a synonym, could refer to them as 'those who reject the khabar al-khassa'.² According to them, it is ignorance to accept the khabar al-infirād (Tr. IV, 256, at the end). Abu Yusuf warns against isolated traditions³ and says: 'We consider an isolated tradition irregular, and do not follow it' (Tr. IX, 9). Shaibanī points out that a certain tradition is isolated, and states that the majority of scholars do not follow it (Muw. Shaib. 148). According to Țahawi, ii. 280, an isolated tradition cannot serve to establish matter additional to the Koran and to generally recognized traditions, or prove their repeal. The Medinese reject isolated traditions from the Prophet (Tr. III, 148, p. 242), and hold that their own consensus takes precedence over them (Ris. 73). They are not consistent, however, and Shafi'i can say to them: 'If Malik objects that this is an isolated tradition,⁴ then what does he think of all those cases where he relates isolated traditions and relies on them? Either the isolated tradition is a reliable argument . . . or it is not; and if not, you must discard all those cases in which you rely on isolated traditions' (Tr. III, 148, p. 249). The same applies to the Iraqians.

The ahl al-kalam go farther and demand that a tradition, to be accepted, must be transmitted by many from many (mā rawāh al-kāfa 'an al-kāffa) or widely spread (khabar al-tawātur).⁵ In defining this condition they disagree: 'They disagree as to how a tradition becomes certain. Some say: through one