MUSLIM USE OF INDUCTION IN LEGAL THOUGHT
In light of these fundamental problems of induction, I propose to investigate a number of central issues in Sunni legal thought. My concern centers on the use which Muslim theoreticians of law made of the theory of induction, and the extent to which they diverged in their juridical formulations from the precepts of this theory.
It must be immediately stated, however, that Muslim jurists were acutely conscious of the precarious logical status of induction, and had no illusions concerning its probabilistic nature. The oft-quoted ruling about watr, a type of prayer whose performance is recommended rather than obligatory, is particularly illustrative. On the basis of an inductive survey, it was found that, unlike watr, no obligatory prayers examined could be performed while on a journey. The ruling that the performance of watr is only recommended and not obligatory was deemed probable (zanni) rather than certain (qat'i) precisely because the inductive survey of prayers was admittedly incomplete, since watr was not included in the survey.8 The judgment that the performance of obligatory prayers while on a journey was impermissible was reached on the basis of an examination of a number of individual types of prayer. In other words, there was a need for a reasonable number of prayers in order to establish this ruling as being highly probable. Reaching a ruling on the basis of examining a single prayer would have resulted in a ruling of a considerably7 Ayer, Probability, pp. 54-88; Max Black, “Induction,” in Encyclopedia of Philosophy, 8 vols. (New York & London: Macmillan, 1972), 4:169-181.
8Fakhr al-DTn al-RazT, Al-Mahsul fi ‘Um usul al-fiqh, ed. Taha J. ‘UlwanT, 2 vols., 6 parts (Riyadh: Lajnat al-Buhuth wa-al-Ta’lTf wa-al-Tarjamah wa-al-Nashr, 1979-1981), 2.3:217-218; Abu Hamid al-GhazaiT, AlrMustasfa min ‘Um al-usul, 2 vols.
(Cairo: al-Matba‘ah al-AmTrlyah, 1324 A.H.), 1:51; Shihab al-DTn al-QarafT, Shark tanqih at-fusiil, ed. T. Sa‘d (Cairo: Maktabat al-Kulliyah al-Azharlyah wa-Dar al-Nashr, 1973), p. 445.lower degree of probability. Therefore, increasing the number of cases relevant to a certain issue constitutes added corroboration that raises the level of probability.
The notion of inductive corroboration has indeed played a major role in the elaboration of a number of central concepts in Islamic legal theory. On the most basic level, this notion appears clearly in the discussions concerning the rationale behind the number of witnesses. Although the number of witnesses in a court of law was considered a divine prescription subject to no human stipulation, jurists argued· for a rational explanation of the divine command.9 As stated in the case of the watr prayer, reasoning on the basis of a single case would yield a conclusion with a rather low degree of probability. This principle was used to justify the unacceptability of one witness whose testimony without any corroboration is nearly as good as none (testis unus, testis nullus).10 At least one other corroborating witness was thought to constitute a reasonably sufficient basis for deciding a case. While one witness’s testimony would yield doubtful knowledge, the testimony of two trustworthy witnesses would enhance the credibility of testimonial evidence enough to render it highly probable. In the words of SarakhsT, this highly probable knowledge assures our hearts and minds (tuma ’ninat al-qalb) that the probability of an error is quite slim.11 To phrase it according to modern legal jargon, the testimonies of two genuine witnesses bring up the level of probability near enough to certainty for the verdict to be “beyond reasonable doubt.”
In what may be termed an extraordinary legal case, where the Qur’an stipulates punishments for committing adultery (hu- dud al-zina), four witnesses are required to satisfy the criteria
9 Abu Ya‘la al-Farra’ al-BaghdadT, APUddah p usul al-fiqh, ed. A.
MubarakT, 3 vols. (Beirut: Mu’assasat al-Risalah, 1980), 3:856. For a contrary view see Shams al-DTn Ibn Sahl al-SarakhsT, Usui, ed. Abu al-Wafa al-AfghAnT, 2 vols. (Cairo: Dar al-Kitab al-ArabT, 1372 A.H.), 1:332.10 Mu waffaq al-DTn Ibn Qudamah, Rawdat ahnazir wa-junnat ahmunazir, ed. Sayf al-DTn al-Katib (Beirut: Dar al-Kitab al-ArabT, 1981), p. 91, 11. 7-8.
11 SarakhsT, Usui, 1:290, 331, 1.10. For the definition of qalb, see Farra’, ‘Uddah, 1:89, 11.3-6.
of evidence.[221] The importance attached to this particular case makes further corroboration necessary. The more serious the litigated issue is, the more witnesses the court requires. Although two trustworthy witnesses are normally sufficient in ordinary cases, and four in cases of adultery, the judge has the prerogative of requiring additional witnesses if he sees fit to do so. According to one prominent jurist, seeking testimonial corroboration beyond the required standard evidence is permitted by the SharT ah.[222]
Corroboration, however, does not depend merely on the number of witnesses required in a case. Rather, it must have as a prerequisite the condition of independence. What this means is that all testimonial statements relevant to a particular case must be given by independent witnesses. There must exist an intact means of transmission between the primal event and the subsequent utterance of the testimony. If one witness is told by another witness what to say, or is influenced by the testimony of another witness, or is involved in any kind of collusion with him, then the validity of his testimony becomes highly questionable, if not altogether worthless. For example, in a case where a man had a quarrel with the inhabitants of a village regarding a certain matter, the testimonial statements of the villagers in favor of each other were INDUCTIVE CORROBORATION, PROBABILITY, AND CERTAINTY 9 rejected on the grounds that they were made in collusion.14 On the same grounds, the testimony of a father in favor of his son is deemed unacceptable in a court of law.15 Thus, to yield a genuine corroboration, the testimonial statements must correspond with each other on points of fact and must be given independently of one another. They must not be adulterated by extrinsic factors. In fact, the quality of rectitude (‘adalah) required of every Muslim witness may be said to have been intended as a safeguard against hampering the independent character of the testimonial evidence.16 In due course, we shall observe the importance of the role which the principle of independence, as a validating condition of inductive corroboration, played in Islamic legal theory.