Ill LOGIC, FORMAL ARGUMENTS AND FORMALIZATION OF ARGUMENTS IN SUNNI JURISPRUDENCE

Some decades before the translation movement from Greek into

Arabic began, a battle over the ultimate source of Law had already been fought and won by Traditionalist Islam, represented by the well-known Ibn Idris al-Shafi^ci (d.

204/820). The triumph of the theory of the divine origins of law compelled Islamic jurisprudence to dissociate itself from all rationalist tendencies involved in philosophical ventures. Chief among these tendencies were the discussions on metaphysics in which logic served as the guiding spirit of philosophical argumentation. Associated with such heretical enterprises, logic as an organon of philosophy could hardly be distinguished from substantive metaphysical doctrines, and in such an environment Traditionalist jurisprudence kept suspected logic at bay. It was not until the appearance of the Shafi^ci-Ash^cari intellectual Abu Hamid al-Ghazali (d. 505/1111) that logic as a method was sifted out of Aristotelian philosophy and thus cleared of its «non-theistic» tendencies. Ghazalfs endeavor in bringing logic into law, and the persistence of what may be called the Ghazalian thesis among some of his influential ufulist successors, will form the subject of this essay. While the following section attempts, in general terms, to analyze Ghazali’s attitude towards logic in the law, it is not until Part II of this essay that we will encounter, in the form of a translation, what may be described as one of the most revealing statements which Ghazali made on the logic of legal argument.

Until the fifth/eleventh century, Sunni jurisprudence succeeded to a significant extent in resisting the influences of Aristotelian formal logic. In Shafi^cT’s theory, for instance, no formal arguments were recognized; his typology of arguments was confined to analogy, argumentum e contrario and a fortiori arguments.

Any rudimentary deductive arguments that may be discerned in his positive legal system were, for him as well as for his successors, but linguistic arguments.^[115] Even towards the end of the fourth/tenth century, long after the translation of Greek philosophical works was undertaken, Sunni jurisprudence was still little affected by the introduction of Greek formal logic. While several nonformal methods of legal inquiry seem to have been developed and refined in light of the new material that these translation offered, only one important formal argument was assimilated; namely, the conditional disjunctive syllogism, known to usulists as al-sabr vua l-taqsim or, briefly, taqsim.^[116] In this argument the possible effects of a legal fact are enumerated with the view of establishing the true effect(s), or the absence thereof. If P, for instance, is the legal fact, and Q and R are phenomena which we think to be the possible effects of

P, then the course of reasoning may yield either of two conclusions:^[117] ^{^[118]}

P is either Q or R

Not Q

Then, R (Or not R; then Q)

Or,

P is either Q or R Not Q and R

Then, Not P

In the fourth/tenth and fifth/eleventh centuries, this argument was subsumed under the broad category of istidldl, generally conceived by jurists and theologians as an umbrella term for inferential methods of inquiry. Although taqsim played an important role in the verification of the cause (^illa) in the methodology of qiyas, Abu Ishaq al-Shlrazi (d. 476/1083), for instance, seems to have considered taqsim as one of the istidldl arguments which constitute an offshoot rather than an integral part of qiyas* On the other hand, the theologian-jurist Abu Bakr al-Baqillani (d. 403/1012) subsumed the two separate sets of arguments qiyas and taqsim under the heading istidldl.^[119] ^{^[120] Taqsim and istidldl then were not unqualifiedly integrated into, or totally identified with qiyas.^5a}

The subsumption of non-qiyds arguments under istidldl seems, as we can glean from later sources, to have contributed to the segregation of istidldl from the traditional qiyas arguments.

After the fifth/eleventh century, istidldl comprised, in addition to taqsim and other formal arguments, such principles as istihsan, istisldh, and istishdb al-half* Grouping these particular arguments and juridical principles together under one designation, namely, istidlal, was not without significance. All of these shared an important characteristic; that is, they existed in the realm of the controversial within the systems of the four schools of law. But while istihsdn, istislah. and istishdb were always subject to discussion and potential approval of the opponent, formal arguments in istidlal were either unqualifiedly embraced by the usdlist or totally ignored. Thus, unlike istihsdn, for instance, formal arguments were expounded in certain usul works without the usual mention of the opponents’ claims or the proponents’ vindication. They were briefly set forth in the manner logical principles were exposed in manuals of logic.

The inclusion of formal arguments in the chapter on istidlal in a number of usul al-fiqh works was, to be sure, a later development. At first, particularly during the sixth/twelfth century, it was in the introductory pages of those usul works which admitted the Greek logical element that such arguments appeared. In other words, prior to incorporating formal arguments in the chapter of istidlal (which was ordinarily placed towards the end of the treatise), there existed, and continued to exist, a preliminary attempt at introducing Greek, and especially Aristotelian logic, into legal theory in the form of an introduction to this discipline. This attempt can be clearly seen in such influential works as those of Ibn Qudama (d. 620/1223), Ibn al-Hajib (d. 646/1248), Ibn al-Humam (d. 861/1456) and others.^[121]

To trace this trend from its start, we must begin with Ghazali, who was, as intimated above, the first jurist in Sunni Islam to have integrated logic into legal theory to a large extent.^[122] In the beginning of his Mustasfa* he delivers a manual on logic, being the shortest in a trilogy whose other two parts are Mihakk^iQ and Mi^cyar^u.

Although he makes the reading of this introductory treatise entirely voluntary, he asserts in the most unequivocal manner that he who has no knowledge of logic has, in effect, no knowledge of any science.¹² When he proceeds into the legal part of al-Mustasfd, however, no sign of any formal logical analysis can be detected and his treatment stands perfectly within the convention of classical usul al- fiqh. What Ghazal! obviously intended in this treatise was not to revolutionize legal analysis but rather to insist on the necessity of logic as the only meaningful tool by which all inferences can be tightly molded according to a rational design. This is a remarkable retreat in comparison with his approach to the subject in his earlier works al-Mi^yar and, particularly, ShifiP al-Ghalil in which he had analyzed legal arguments in terms of syllogistics.¹³ It is instructive to note that the first of these two treatises is essentially an introduction to Aristotelian logic, and the second is a legal work treating of causation in law.

Apart from a number of brief notes made throughout al-Mifydr on legal theory, Ghazali illustrates the three figures of categorical syllogism together with their moods with examples drawn not only from philosophy and theology but also from law. This is also done in the case of conjunctive and disjunctive syllogisms, reductio ad absurdum and induction.¹⁴ It is quite obvious that with these logic with multiple legal examples. Such examples were not advanced merely for the purpose of illustration. Logic, Ibn Hazm asserted, is most useful in interpreting the Quran and the Sunna, and in drawing legal inferences. Those who do not comprehend such a simple fact have indeed distanced themselves from religion, and should not be allowed to decide in legal matters. See the significantly entitled al-Taqrib li-Hadd al-Manfiq wa l-Madkhal Ilayhi bi-l-Alfdz al-^cAmmiyya wa l- Amthila l-Fiqhiyya, ed. Ihsan ^cAbbas, Beirut, 1959, pp.

9-10; A. Chejne, «Ibn Hazm of Cordova on Logic,» The Journal of the American Oriental Society, 104, 1 (1984), pp. 52-72, especially at 62-63.

⁹ 2 vols., Cairo, 1322 H., I, 10-55.

¹⁰ Eds. M. Na^csam and Mustafa al-Qabbani, Cairo, n.d.

¹¹ Ed. Sulayman Dunya, Cairo, 1961.

¹² Ghazali, Mustajd, I, 10, 1. 17. This is reminiscent of Ibn Hazm’s aforementioned assertion that those who do not come to grips with logic must not be trusted with religious and legal matters. See note 8 above.

¹³ ShifiP al-Ghalil ft Bayan al-Shabah wa l-Mukhil wa-Masdlik al-Tahiti, ed. Hamd al-Kabisi, Baghdad, 1971, pp. 435-455, translated below.

¹⁴ Ghazali, Mi^yar, pp. 134-165. See also Hassan Abdel-Rahman, «La place du syllogisme juridique dans la methode exegetique chez Gazali,» in Le Raisonnement Juridique, ed. H. Hubien, Bruxelles, 1971, pp. 187 ff.

examples GhazalT was merely trying to bring closer to the mind of jurists an understanding of the structure of these inferences. There is no attempt at analyzing legal cases through the medium of these arguments. Nor is there any effort at identifying, in terms of the established theory of logic, the distinctive structure of legal logic. The sole exception, however, concerns analogy, which Ghazali, following the tradition of Aristotle, insisted that in order for it to be logically valid, it must be converted to first figure syllogism.^[123] In a chapter in ShifiP al-Ghalil he goes further in this direction towards the formalization of legal logic and attempts an analysis of a wider variety of legal arguments in terms of syllogistics (see Part II).^[124]

As it will become evident, GhazalT’s conception of the relationship between logic and law as expressed in his Mustasfa seems to have put the final stamp on the attitudes of a number of his successors towards the role of logic in law. These successors, however, exercised a great deal of caution in introducing to their usul works the principles of logical theory as expounded by GhazalT.

While following GhazalT faithfully, they have, with remarkable discretion, chosen the most relevant parts of the theory to introduce it to their jurisprudence. Despite GhazalT’s declared effort^[125] ^{^[126] at keeping the size of the logical introduction in the Mustasfa to a minimum, the end result remained a clearly independent treatise on logic. And like any standard Arabic manual on logic, this introduction begins with the exposition of the theory of definition (haddf* and the rules governing its use and validity. The theory of syllogistics (jburhari) is then stated with the familiar three figures and their moods, conditional hypothetical syllogism and finally conditional disjunctive syllogism (al-sabr iva l-taqsim or ta^anud). In a later section of the treatise, he discusses the types of premises in demonstrative arguments with special emphasis on the criteria according to which a premise is determined to be certain or probable. And like any other standard manual on logic, the discussion of induction and analogy comes subsequent to the exposition of demonstrative arguments. In addition to the point which he already made in al- Miydr about the validation of analogy by means of its conversion to the first figure, Ghazali follows the Aristotelians in relegating induction to the ranks of inconclusive sciences. He asserts that when it is perfect, induction as such loses its identity since it necessarily reverts to categorical syllogism.^[127] The fundamental issue which Ghazali does not fail to address in more than four of his aforecited works, particularly al-Mustasfd and Shifa? al-Ghalil, is the differing terminology concerning two arguments common to both disciplines usul and logic.^[128] When the Hila is efficient in the major term, the argument is known to philosophers as burhdn al-lima and to jurists as qiyds al-Hlla. When the Hila is non-causal, the argument is known to philosophers as burhan IPanna and to jurists as qiyds al-daldla. The form of juridical argument is therefore identical to the logicians’ formal argument. The difference between them, however, lies in quality of material premises (muqaddimat) used in legal and rational arguments. «The premises which are suitable for rational qiyds (demonstration) are suitable for juridical qiyds, but not all the premises which are suitable for juridical qiyds are suitable for rational qiyds.»^[129] The conversion of a probable legal premise to a conclusive one requires its ‘universalization’; that is, the extention of the rule governing a particular case to the species in which this case is only an instance. This granted, the form of legal argument will be no different from the formally rational argument. In fact, aside from the thorny probelm of the logical validity of the ‘universalized’ major premise, this is the very process by which nonformal legal arguments are formalized (see part II and the introduction thereto).
II
Ghazali’s singular contribution to legal logic was certainly instrumental in bringing about certain structural changes in a good number of authoritative usul al-fiqh works.^[130] Just as he conceived of logic as the valid organon of any inferential procedure, and thus prefaced his Mustasfa with a manual on formal logic and insisted upon the conversion of analogy into first figure syllogism, we find many of his successors employing logic to ground legal theory in what is fundamentally an Aristotelian conception of knowledge. The eighth/fourteenth century Hanbali intellectual Taqi 1-Din Ibn Taymiyya disapprovingly testified that those jurists and theologians who treated of law in this manner did so under the influence of Ghazali.^[131] ^{^[132]}
The Hanbali jurist Muwaffaq al-Din Ibn Qudama (d. 620/1223), a staunch advocate of Traditionalist Islam,^[133] is the earliest legal theorist known to us to follow in the footsteps of Ghazali in integrating logic into law. While Ghazali makes the reading of his introduction optional, Ibn Qudama, like many of his younger contemporaries and successors, incorporated the fundamentals of logic as part and parcel of his legal treatise. In his Rawtfa, logic is introduced into law in a manner bolder than that adopted by Ghazali. Following the Aristotelian (-Stoic) division of knowledge into tasawwur (conceptio) and tasdiq (verificatio or judgment),^[134] Ibn Qudama briefly expounds in the first pages of his book the theory of definition (Jiadd) and categorical and conditional syllogism.^[135] Perception of things, according to him, obtains either through definition or demonstration. The intellect first perceives individual matters, such as the knowledge of the meaning of ‘world’, ‘eternity’ and ‘createdness’. Further knowledge then obtains by predicating these individual matters on each other, thus formulating, for instance, the knowledge expressed in the statement ‘the world is created,’ or ‘the world is eternal.’ Unlike the knowledge of the meaning of ‘world’ and ‘createdness’ which is not subject to affirmation or denial, the knowledge which results in a judgement, e.g., ‘the world is created’, may be true or false. The first type, known as tasawwur, is acquired through definition, and the second, known as tasdiq, is acquired through demonstrative syllogism (burhan) ^[136]
In order to relate the inferences through which tasdiq obtains to legal matters, Ibn Qudama explains that the term burhdn is used for inferences in which the two premises are certain, while probable premises make up what he terms qiydsfiqht (juridical inference). He hastens to add, however, that assigning the term qiyds to legal argument is merely a metaphorical usage, for the form and structure of this argument does not differ from that of demonstrative syllogism in that they both require the subsumption of a particular under a general (idraju khufiisin tabita ^zumum). Qiyds, on the other hand, literally means measuring a thing by another, an inference which for Ibn Qudama seems incapable of attaining the knowledge defined under the category of tasdiq^[137]* Thus, despite the given difference with regard to the certitude of the premises, the form of juridical inference is identical to that of demonstrative syllogism.
Syllogistics, Ibn Qudama argues, stand as the methodological foundation of the knowledge of any science, be it rational or legal.^[138] They are the only sound criteria for any inference. Accordingly, he proceeds to delineate these syllogisms, the conditions for their validity, and the manner in which they may serve in legal matters. For him then there are five types of syllogisms, three of which are categorical, one hypothetical and one disjunctive.^[139] Using no technical terminology for categorical syllogism, Ibn Qudama begins by casting an example of the first type of syllogism:
All intoxicants are forbidden
All wines are intoxicants
Therefore, all wines are forbidden
Explaining that the category of‘wine’ is the thing being judged and prohibition is the judgement, he then excludes from the conclusion the term which is repeated in the minor and major premises. This term, he remarks, is known as the 4lla which represents the answer to the question ‘Why is wine prohibited?’ An affirmative major and a universal minor are prerequisites for the validity of such an inference.
The second and third categorical syllogisms are in the second and third figures, respectively. As expounded by Ibn Qudama, however, the first type of syllogism does not conform to the structure of the first figure. For, although the major and minor premises are universal, the position of the middle term in the premises follows the arrangement of the fourth figure. This poses the question of whether or not our author confused the first figure with the fourth. This is likely to have happened since the reducibility of figures was a common practice.^[140] At any rate, the same argument can well be cast in the first figure as illustrated above.
In the second figure, the 4lla plays the role of the predicate in both premises. The conclusion that ‘A Muslim must not be given the death penalty if he kills a non-Muslim’ emerges from the argument that non-Muslims are not equal to Muslims and that punishing by death a killer indicates that the man killed is a Muslim. For the validity of this mode, Ibn Qudama observes, one of the premises must be negative and the other affirmative, and the major must always be universal. In such a syllogism the conclusion is necessarily negative.
The 4lla in the third figure, known to jurists as naqd, is the subject of both premises. Ibn Qudama gives the following example:
All wheat is usurious
All wheat is an edible foodstuff
Therefore, some edible foodstuffs are usurious
The fourth argument in Ibn Qudama’s arrangement is conditional hypothetical syllogism, which he terms talazum. Again without using the technical terminology of the logicians or the usuli- logicians, he distinguishes two types, the modus ponens and the modus tollens. An example of the first type is:
If this prayer is valid then ablution was performed This prayer is valid
Therefore, ablution was performed
In the modus tollens the consequent of the major premise is denied, i.e., ‘Ablution was not performed,’ thereby yielding the conclusion ‘The prayer is not valid’. In this argument, the consequent, ablution, is predicated upon the validity of prayer; when ablution is performed prayer is valid, and when it is not performed prayer is invalid. Therefore, the consequent is a prerequisite for the antecedent; to have the second exist requires the existence of the first, and the absence of the first indicates the absence of the second, not vice versa. To argue that
If this prayer is valid then ablution was performed
Ablution was performed
Therefore, this prayer is valid
would be fallacious, for prayer may be rendered invalid by another factor unrelated to ablution. The same fallacy may be committed if the minor premise were ‘This prayer is not valid’. Such an arrangement does not yield a valid conclusion. The governing principle of this syllogism is that the consequent must be inclusive of, or at least equal to, the antecedent. When the particular (antecedent) is affirmed the general (consequent) is also necessarily affirmed, and when the general is denied the particular is also denied. By the same token, the affirmation of the general does not lead to the affirmation of the particular. Nor does the denying of the particular lead to the denying of the general. In the statement ‘All animals are corporeal,’ affirming the existence of animality would necessarily affirm the existence of corporeality, and denying corporeality would certainly deny animality. An argument affirming the contrary cannot serve.^[141]
The aforementioned al-sabr wa l~taqsim ranks as the fifth and last syllogism in Ibn Qudama’s exposition. Having the characteristics of conditional disjunctive syllogism, al-sabr wal-taqsim presupposes the decomposition of a given matter into all of its possible constituent elements, and by affirming one or a set of elements the rest of the elements are denied. For example, one may argue that the universe is either eternal or created. But the universe is created; therefore it is not eternal. Or that it is eternal; therefore, the universe is not created. To ascertain the validity of the conclusion about the given matter, an exhaustive account of all the possibilities which may constitute or affect that particular matter is deemed a prerequisite. In other words, leaving out of consideration a third possibility in the example of the universe may nullify the conclusion since the truth of the matter of creation may rest with the third excluded possibility.^[142] ^{^[143]}
Arguments in all fields of knowledge, Ibn Qudama asserts, must conform to the rules herein set. When they do not, it is only because the proper arrangement of the premises has been neglected, or because a premise has been suppressed. Suppressing a premise, Ibn Qudama avers, is quite common in legal argument, because being so evident, the premise is taken for granted. A lawyer may thus argue that a man must be stoned to death because he committed adultery while enjoying the status of muhsan (being a free man and married, or having been married, to a free woman). Suppressed here is the major premise, viz., 'All muhsan adulterers must be stoned to death.’ Similarly, lawyers often analogically argue that since grape-wine is intoxicating, it is forbidden, just like date-wine. Following in the footsteps of Ghazali, Ibn Qudama insists that analogy cannot be valid unless reduced to first figure syllogism.³* This last observation must by no means be underrated, for it not only illustrates the persistence of the Ghazalian thesis of the paramount position of syllogistics, but also demonstrates that for Ibn Qudama there are no valid arguments in law unless they can be reduced to one or the other of the formal arguments which he has just expounded.
Ill
Though Ibn Qudama went a step beyond Ghazal! in integrating the formal precepts of logic into the body of legal theory,^[144] he remained within the general boundaries which Ghazal! had already set. We thus find no attempt on Ibn Qudama’s part to alter what is otherwise a conventional treatment of legal theory.
The early beginings of a somewhat different approach towards integrating logic into law can already be detected in Ibn Qudama’s younger Shafi⁰! contemporary Sayf al-Din al-Amid! (d. 631/1233), who was heavily involved in the study of logic alongside the traditional sciences.^[145] ^{^[146] Amid! begins his work al-lhkdm by defining uful al· fiqh and its relationship to the sciences upon which it draws. The absence of an exposition of formal logic in Amidi’s opening pages must not obscure the fact that espistemologically he was committed to the formal conception of legal science. This is evident even in the introductory pages where he asserts, in the tradition of Aristotle (and the Stoics), that knowledge of any science obtains either through tafawwur or through tafdiq.³¹ Obviously it is by means of the latter that unspecified legal rulings are inferred from the revealed premises. An indispensable instrument in this process of inference is the legal indicant, the da til, which presents itself in three forms: rational, revelational and a combination of both.^[147] The rational indicant, mostly used in theological discussions, leads to knowledge on grounds of rational premises, e.g., ‘The universe is Composite,’ and ‘All composite things are created’; therefore, ‘The universe is created.’ The revelational indicant is that which is derived either directly from the Quran, the Sunna or consensus, or indirectly through qiyds. But that which combines both the purely rational and the purely revelational is what can be characterized as a formal argument taking its premises from revelation.^[148] ^{^[149]}}
The argument cast in the first figure about the prohibition of wine is a classic example of an inference grounded in a rational- revelational dalil. This type of dalil constitutes for AmidT the basis of a category of inferences independent of qiyds, a category which he terms istidlal.*^Q AmidT discusses several inferences under the heading of is ti dial, the third of which concerns itself with the dalil that consists of propositions which, if admitted to be true, result in a true conclusion. Unlike Ibn Qudama who does not use the technical terminology of the logicians, AmidT freely employs such terminology in his discussion of formal arguments.
Following the Aristotelian division of syllogistics, AmidT recognizes categorical conjunctive syllogism (iqtirani), conditional disjunctive (sharti munfasil) and conditional hypothetical syllogism (shartimuttasil). Categorical conjunctive syllogism, consisting of two premises and a conclusion, must have one term repeated in both premises. This term, known as the middle term (al-hadd al-awsat), determines, depending upon its position in the minor and major premises, the figure of syllogism. In the first figure, the middle term stands as the subject of the major and the predicate of the minor. In the second, it takes the position of the predicate in both premises, and in the third the subject. The fourth figure has the middle term as the predicate of its major and the subject of its minor. AmidT finds the fourth figure superfluous, capable of providing nothing that is not already found in the first three figures.^[150] In fact, just as Aristotle made his assertion that the first figure is the source of all other figures, and that the fourth is a mere variation on the theme of the first, AmidT insists that the first figure is the most demonstrative (abyanuha) and constitutes the real basis for the construction of all other figures.
The first figure, AmidT states, lends itself to four moods (durub), the first of which has two universal affirmative premises and a universal affirmative conclusion. He gives the following example:
All acts of worship require intent (niyya) [All] ritual ablution is an act of worship
Therefore, [all] ritual ablution requires intent
In the second mood, the major and the conclusion are universal negative. Thus:
No act of worship is valid without intent
[All] ritual ablution is an act of worship
Therefore, no ritual ablution is valid
without intent
When the major, however, is universal affirmative, and the minor is particular affirmative, the syllogism acquires a different mood with a particular affirmative conclusion:
All acts of worship require intent
Some ritual ablution is an act of worship
Therefore, some ritual ablution requires intent
Amidi’s fourth mood consists of a universal negative major, a particular affirmative minor and a particular negative conclusion. Thus, using the same basic legal example:
No act of worship is valid without intent
Some ritual ablution is an act of worship
Therefore, some ritual ablution is invalid without intent
The four moods of the second figure have one quality in common, namely, a universal major. When the major is negative and the minor is universal affirmative, the conclusion is universal negative, as in the following example:
No sale is valid if the identity of the sold commodity is unknown All sales of non-existent commodities are sales of commodities which are unknown
The notable characteristics of the six moods of the third figure are their having in common an affirmative minor and at least one universal premise. Their conclusions are always particular, whether negative or affirmative. It suffices here to illustrate these moods by presenting examples of the first and the fourth moods, as these two share, respectively, the same conclusions as those of the second and third, and the fifth and sixth. The following is an example of the first mood which consists of two universal affirmative premises and a particular affirmative conclusion:
AH wheat is an edible foodstuff
All wheat is usurious
Therefore, some edible foodstuffs are usurious
The fourth mood consists of a negative universal major and an affirmative universal minor with a particular negative conclusion:
No wheat can be bartered for an unequal quantity (of the same kind of wheat)
All wheat is an edible foodstuff
Therefore, no edible foodstuff can be bartered for an unequal quantity (of its own kind)
Premises structurally different from those used in categorical syllogism must, in order to be valid, be cast in a different type of syllogism. In Amidi’s arrangement, the next two important syllogisms after the categorical are the conditional conjunctive (‘If... then...’) and conditional disjunctive (‘Either... or...’). Ibn Qudama has already expounded the conditions for their validity.
IV
The Maliki jurist Jamal al-Dln Ibn al-Hajib (d. 646/1248), a young contemporary of Amidi, conceives the logical and epistemological foundation of the discipline of u$ul al-fiqh in a like manner. In his legal works he firmly anchors the methodology of legal analysis and interpretation in the dichotomous conception of tafawwur and tasdiq. This, of course, is no innovation over his two predecessors Ibn Qudama and Amidi, but it appears from the extensive commentaries and super-commentaries on his Mukhtasar by Iji(d. 756/1355), Taftazani(d. 791/1388), Jurjani (d. 816/1413) and Harawi (probably died after the eighth/fourteenth century) that Ibn al-Hajib managed to secure a greater role for this dichotomy of the typology of the acquisition of knowledge in legal theory. The ultimate goal of this theory is the discovery of the law of God from the primary sources, and the means of such discovery, Ibn al-Hajib asserts, is the dalli. The process of reaching the rule of a case which had not been stipulated in the sources belongs to the ta$diq category of acquiring knowledge.^[151] In this context of dealing with inferences from things known about the things unknown, Ibn al-Hàjib expounds, in the same relative detail as Amidi and Ibn Qudàma, the types of syllogisms, including the fourth figure of categorical syllogism which Amid! deemed superfluous.^[152]
Unlike Àmidi who touches only briefly on the concept of knowledge and the means of its acquisition in the beginning of his work, and who leaves syllogistics to be discussed later under the heading istidlàl, Ibn al-Hàjib sets out by expounding syllogistics and the concept of knowledge while elaborating a different set of formal arguments in the chapter of istidlàl. Both Àmidi and Ibn al-Hàjib generally define istidlàl as a means of inference other than qiyàs, consensus or the two primary sources. But Ibn al-Hàjib qualifies this definition by limiting qiyàs here to the qiyàs al-Hlla, that which is based on an efficient common cause. With this he asserts that istidlàl is the fifth source of Islamic law,^[153] thus implying, as the commentator Taftazàni observes, that other types of qiyàs, such as the qiydses known as nafy alfàriq and talàzum, come under istidlàl. (Like Àmidi, he also subsumes istishàb and the pre-Islamic laws—shar^c man qablanà—under istidlàl).^[154] The qiyàs of nafy alfàriq, known to Ibn Idris al-Shafi^ci as qiyàs fi ma^cnq al-as I, may be described as an analogical inference in which the differences between the two cases are shown to be irrelevant to the cause (filla) and the rule (hukm) of the case. The sexual difference between a concubine and a male slave, for example, must be dismissed as irrelevant in the case of contractual manumission (fatq). Likewise, the impurity brought about by a mouse falling in oil or clarified butter is effective in both foodstuffs in spite of the differences in their texture.^[155] ^{^[156]}
The chief formal argument of istidlal, however, is the talazum (concomitance) between the rules of two cases without the specification of an efficient ^cilla. The inference of concomitance lends itself to a fourfold division, the first of which is what may be termed as coextensiveness (fard) and coexclusiveness (Wj), known together as the method of dawaran.™ In the example of corporeality and compositeness it may be argued that all corporeal things are composite and all composite things are corporeal; likewise, it may be argued that whenever there is no corporeality there is no compositeness and whenever there are no composite things there are no corporeal things either. In the case of divorce and temporary separation (ziAar), the course of reasoning is as follows: He whose divorce is valid, his temporary separation is valid and he whose temporary separation is valid, his divorce is valid. The argument gains additional strength by negatively reasoning that he whose divorce is invalid, his temporary separation is invalid and vice versa.^[157]
The second consists of coextensiveness only, as in the example of created corporeality. Here, it is argued that all corporeal things are created, but not all created things are corporeal. Likewise, if wutfiP (ablution) is accepted as valid without niyya (intent), then tayammum (ablution with soil or sand) is also accepted as valid without intent. The third is essentially the coexclusiveness of contradictions; what is true cannot be false and what is prohibited cannot be permissible. The fourth corresponds to the third but is negative in its postulate; e.g., what is not permitted is illicit.^[158]
Although these inferences differ from the strictly syllogistic arguments expounded by Amidi in the chapter of istidldl, they can be reduced, as the commentator Iji remarks, to disjunctive and conjunctive conditional syllogisms.^[159] In fact, these arguments and the principles of coexistence and coexclusion according to which they operate are identical to the procedures employed in verifying the Hila of the original case (aj/) in qiyas.^[160]
V
All this goes to confirm Ibn al-Hajib’s (and his commentators’) commitment to formal arguments. In the long-standing tradition of usulists, he affirms that usul al-fiqh derives from the five legal norms (al-ahkam al-khamsa), the established general principles of kalam and the Arabic language. TaftazanT observes that Ibn al-Hajib, when dealing with kalam as one of the fundaments of usul, had nothing to say about central theological issues such as belief, the truthfulness of the Prophet, miracles, etc. Instead, he discussed under kalam problems of knowledge and reasoning which are as pertinent to kalam as they are to any other science.^[161] JurjanT further explains that Ibn al-Hajib dwelled on logic under the guise of kalam because law, being a religious science, cannot be grounded in a science as alien as logic, and since kalam, the crown of religious sciences, has come to be inclusive of logic, it is used as a cover under which logic is subsumed. This, he says, makes obvious the fact that what is really at stake here is the need for logic in legal theory, a need which «al-Imam al-Ghazall alluded to in the Muslasfd.»^[162]* In his supercommentary on JurjanT, Harawi resents this play on words and seems to demand that the issue be addressed even more openly and directly. He remarks that
These scholars went out of their way and placed restrictions on the principles of logic, restrictions which (aim at) removing these principles from the domain of logic. Furthermore, they claimed them to be part of kalam, in order that legal science, the chief of religious sciences, would not look as if it were in need of a non-religious science. But the fact is that all these (religious) sciences are in need of logic.
Those who argue that the theological principles which are needed in usul are (an intrinsic) part of kalam have erred, because these principles are of no service except in theological issues. But what is of service in legal theory is nothing (of these issues) but logic.^[163]
This bold argument affirming the centrality of logic in legal theory is stated even more forcefully by the Hanafi jurist Muhammad b. ^cAbd al-Wahid Ibn al-Humam (d. 861/1456) who insists that aside from the Arabic language and legal norms, usul al-fiqh draws heavily upon logic. «Calling this a kaldm principle is farfetched.»^[164] Kaldm and usul al-fiqh equally depend on logic. Logic is necessary for the knowledge of all sciences; it is the only tool for discerning the true from the false. The only question concerned with kaldm in legal theory, Ibn al-Humam asserts, is perhaps that which pertains to whether good and bad acts are conceived as such by reason or by revelation.^[165] Thus, what is termed by other theoreticians as kaldm principles is not but principles of logic. In line with this conception, Ibn al-Humam, like Ibn Qudama, Amidl and Ibn al-Hajib, makes sure to delineate the theory of definition and the various types of syllogism (including the fourth figure).^[166] In his commentary on Ibn al-Humam’s Tahrir, Ibn Amir al-Hajj follows suit.^[167]
Similarly, the two later jurists Muhammad b. Ni?am al-Din al- Ansari (d. 1119/1707) and Muhibb Allah Ibn ^cAbd al-Shakur (d. 1225/1810) view logic as one of the three fundamental fields of knowledge upon which legal theory is founded.^[168] Their predecessors, they note, have assigned the term kaldm to this field,
because logic is the tool through which theological conclusions are reached. Without elaborating much further on this point they proceed with a brief discussion of theological, epistemological and logical principles, including categorical and conditional syllogisms.^[169]
VI
The infiltration of logic into an important strand of Sunni jurisprudence heralds a significant change in the attitudes of jurists towards the epistemological status of usul al-fiqh. Such an infiltration represents a natural (though not logically necessary) state of development the origins of which may be traced down to the history of the concept of Him (knowledge). The interest of usulists in the notion of Him as the point of departure in the study of their field made its first appearance in the second/eighth century. But it was not until the third/ninth and the beginning of the fourth/tenth century that the concept was developed in the arenas of kalam to encompass a variety of categories and methods which soon became part and parcel of the kalam oriented usul al-fiqh. To the usulists, Him meant the knowledge of the law and the methods and procedures which lead to this end. It could not have, therefore, been more fitting than to introduce logic and its sophisticated tools for the enhancement of the existing knowledge about knowledge. It would seem that what logic—as understood, for instance by Ghazal!—had to offer stood in total consonance with the established doctrine of Him. Logic was seen as instrumental in grounding legal theory within a larger system of knowledge, a system which was already recognized in the equally authoritative discipline of kalam.
Thus in the theories of a number of influential jurists of all schools, logic and its formal arguments served not only as the general epistemological framework of legal theory but also as the guiding methodology of the highest order in reasoning about individual positive legal cases. The theories of the acquisition of knowledge, definition and syllogistics were viewed as universal organons of control through which the precarious position of legal certitude can be strengthened. By means of logic legal argument could be constructed in the most formalized form possible. The details of legal analogy, one of the most prevalent forms of legal argument, had to fit indirectly in this scheme of formal logic. Insofar as logical construction is concerned, analogy was deemed as a consequent or secondary to formal arguments. In the final analysis, analogy, in order to be formally valid, had to conform to the laws of logic. This is perhaps the pivotal point in the process which Ghazal! had started when he affirmed time and again that analogy must ultimately revert to first figure syllogism.
Part II. Formalization of Arguments: Ghazali on the Modes of Demonstration in Law
Ghazal! regarded legal logic as that part of logic in which legal arguments are subjected to formalization rather than as a logical explication of a peculiar system of arguments. While discarding all Greek philosophical formulations, he tenaciously dinged to formal Aristotelian logic and made it the methodological foundation of all enquiries. His conception of formal logic as an indispensable instrument for all areas of knowledge is evidenced in the fact that the examples which he provides in his logical works extend over a wide spectrum of religious sciences. But in these same works, law and legal examples are often no more than an «illustration» of how a demonstrative argument must be constructed and validated; for, after all, as Ghazal! tells us, «reasoning about legal matters does not differ from reasoning about rational matters... except in what concerns the premises.»^[170] Although for Ghazal! the forms of legal and rational arguments are identical, one looks in vain in his logical works—as well as in the logical-legal treatise al-Mustasfd—for an analysis of legal arguments from the standpoint of logic. A noteworthy exception, however, is found in his otherwise traditional legal work ShifcP al-Ghalil, where in one chapter^[171] he analyzes, mostly in terms of syllogistics, three major legal arguments commonly subsumed under juridical qiyds\ namely, causal demonstration, non-causal (indicative) demonstration and reductio ad absurdum.
The central question with which Ghazal! had to grapple was the validity of the essentially inductive character of legal argument. The underlying issue in his discussion of the above mentioned causal and non-causal demonstrations was the cause (filla), the term common to the minor and the major premises. In analogical inference, the fundamental problem lies in the obscurity or uncertainty of the distinction between essential and nonesssential elements with regard to two similar matters (or cases). Viewed from the standpoint of formal logic, analogical inference does not rest on logical necessity; that is, the similarity between the two cases is not a fundamental logical relation. A typical analogical argument precludes the syllogistic procedure:
All A is (has the characteristic) C All B is (has the characteristic) G
Therefore, all B is (must have the characteristic) A
Rather, the similarity in analogical inferences does not necessarily entail the conclusion. As Ghazall points out repeatedly, the argument for a particular similarity may well be controversial because the typical form of analogical inference is as follows:
All A is (has the characteristic) C All B is (has the characteristic) D
Therefore, all B is (must have the
characteristic) A
In this argument it is assumed that A is similar to B in all essential respects. With this assumption the argument is rendered valid, for if we assume that C and D are sufficiently similar to justify our treating them as one, say C-D, the argument reverts to its strict syllogistic form:
All A is (has the characteristic) C-D All B is (has the characteristic) C-D
Therefore, All B is (must have the characteristic) A
It is in this light that one must view Ghazali’s insistent claim that in order to be valid analogy must adopt the form of categorical syllogism. But to do so, it is of utmost importance to establish, as Ghazall repeatedly attempts to do, that the ‘characteristic’ (kha^siyyd) common to both premises (cases) is an essential characteristic in all respects. The crux of the problem of formalizing inductive legal arguments thus lies in the circle of similarity and the extent to which it can be proven to logically entail a conclusion.
Ghazali’s concern about the formal validity of legal argument was not wholly shared by his successors. As far as I know, no later usulist attempted a formal reconstruction of inductive legal arguments. Ghazali’s real legacy, however, appears most clearly in the attempt of later usutists at grounding legal theory in a formal conception of knowledge.^[172] This attempt manifested itself in emulating Ghazali’s Mustasfd, in which the introduction to the fundamentally formal Aristotelian logic set the grounds for legal theory, but had little effect on the practical construction of legal argument.
TRANSLATION^[173]
//435// Concerning The Explication Of The Modes of Analytical Demonstration Used in Legal Matters^[174]
The aim [of this chapter] is the exposition of the methods [of demonstration] used in cases which jurists consider to be inferential rather than revelational.^[175] These demonstrations (barahiri) are
three: Causal (iHildl) demonstration, indicative (is ti dial)
demonstration⁶⁸ and reductio ad absurdum (khulf) demonstration.⁶⁹ and self-evident textual premises on the other. Muslim jurists conceive of the texts of the primary sources as linguistically consisting of two basic categories, one is self-evident and the other ambiguous, capable of varying interpretations. Texts in the first category yield necessary knowledge and thus reasoning about them becomes superfluous. It is only in cases where God did not clearly stipulate the ruling of a case that qiyas is allowed. In the beginning of Shifd^ al-Ghalil, Ghazali asserts that in any case where there is no Hila (cause) qiyas cannot be undertaken. This, of course, precludes from qiyas all textual statements which are so clear as to yield in the mind necessary knowledge of the case, including its species. Thus, in Quran V:3 «Forbidden to you are carrion, blood, pork (lahm al-khinzir)...» the jurist takes the term khinzir to cover all types of swine meat, including that of wild boars (khinzir barn). Though reasoning in this case can be reduced to a syllogistic inference, jurists, including Ghazali, insisted that reaching the conclusion ‘The meat of wild boars is forbidden’ needs no inference since it is understood from the language of revelation (naqlan, or as otherwise stated min jihati al-lqfz). Ibn Rushd, Biddyat al-Mujtahid wa Nihayat al-Muqtasid, I, 3; Tahanawi, Kashshdf, s.v. »mqfhum», II, 1153-1154.
⁶⁸ In his Mi'yar al-^Ilm, pp. 243-245, Ghazali dwells on the distinction between qiyas al-Hlla and qiyas al-dalala in a chapter entitled «On the Supplements to Qiyas». The subtitle, «On Matters the Knowledge of which Complements the Knowledge of Demonstration,» is quite telling. Here, he establishes the terminological connection between the juridical qiyases of Hila and dalala and the equivalent philosophical types of demonstration burhdn al-lima and burhdn al-li’anna. As we have previously mentioned, when the middle term of a syllogism is an efficient cause, the inference is called by jurists qiyas al-Hlla and by logicians burhan al-lima. But when the middle term constitutes non-causal evidence in the relationship between the major, and minor premises, jurists call this inference qiyas al-dalala and logicians burhdn al- li’anna. An example of causal demonstration in the realm of sense perception is ‘This piece of wood was burnt because it was subjected to fire.’ In the inference ‘This man is satiated because he has just eaten,’ the major and minor premises are ‘All men who eat become satiated,’ and ‘This man has eaten’, with the conclusion ‘This man is satiated.’ However, in the example ‘This man is satiated; therefore, he has just eaten,’ the effect (natija) points out the so-called cause (muntij). The effect in the qiyas al-dalala (burhdn al-lPanna) demonstrates the existence rather than the efficiency of the cause. Non-causality in this inference comes through clearly in the syllogism ‘All those who write poetry are learned,’ ‘This man writes poetry’; therefore, ‘This man is learned.’ Learnedness here is not a cause, but the customary co-existence of learnedness and the ability to write poetry enables us to infer that this man is learned. See also Ibn Taymiyya, Radd, p. 90, II. 15-20. This distinction between causal and non-causal demonstration seems to have originally been made by Aristotle who termed these types of inferences demonstrative and dialectical syllogisms, respectively. The latter is an argument from signs (quite synonymous to the Arabic dalalat) rather than causes, for it yields no understanding of the logical connection between the two premises. Further on the Aristotelian conception see H. W. B. Joseph, An Introduction to Logic, Oxford, 1906, pp. 398-399.
⁶⁹ In the Mi^yar, pp. 159-160, Ghazali intimates that qiyas al-frulf is used in scholarly disputations (mundzarat). «You take the claim of the opponent and make it a premise [in a syllogism]. Then you add to it another premise whose
Causal demonstration (burhan is bringing together the assimilated case (far^ and the original case (a?l) through a cause (S’/Za), as it has been mentioned in the [discussion about] qiyds. The form of this demonstration reverts to two premises and a conclusion. You illustrate this by saying: ‘A usurped object must be restituted’—this is a premise—and ‘The property is usurped’—this is a second premise—and their conclusion is ‘The property must be restituted’. You also argue: ‘The usurped must be restituted,’ and ‘The son of a usurped woman is usurped’; therefore, ‘He must be restituted’. You also argue: ‘Property is restitutable if destroyed,’ and ‘Fructuous assets are property,’; therefore, ‘Fructuous assets are restitutable if destroyed’. You also argue: ‘The thiefs hand must be amputated/ and ‘The body-snatcher is a thief ; therefore, ‘The body-snatcher must be amputated’. You also argue: ‘Foodstuffs are usurious,’ and ‘Quince is a foodstuff,’ therefore, ‘Quince is usurious.’^[176] ^{^[177]}
To begin with, all this reverts to the principle of subsuming a particular under a given general.^[178] If you will you may argue: ‘Edibility is the cause of usury,’ and ‘Edibility is found in quince’; therefore, ‘[Quince] is usurious’. Likewise, ‘Usurpation is the cause of restitution,’ and ‘Usurpation is found in real estate’;^[179] therefore, ‘Restitution [in usurped real estate] is obligatory’. The jurists’ reasoning in this category is that quince is edible just like wheat, or that analogous to wheat it is usurious. Or that immovable property is capable of being usurped and therefore it is restitutable like movable property.
All this reverts to the principle of subsuming a particular under a universal. In rational demonstration it takes the form of two premises and a conclusion, as it was previously mentioned.
//436// Disagreement may be assumed to arise concerning the first premise, while the second premise is agreed upon. The adversary may argue: «1 agree that quince is edible but I do not agree that edibility is the cause of [prohibiting] usury, or that edible objects are usurious. Rather, some edible objects are usurious but not due to their being edible.»^[180] The adversary may agree upon the first premise but may disagree on the second. He may argue: «1 agree that usurpation is the cause of restitution, but do not agree that [restitution in] usurpation applies to real estate, to the son of a usurped woman, or to fructuous assets. Likewise, I agree that theft is the cause of amputation, but do not agree that the bodysnatcher is a thief.»
If there is disagreement concerning the first premise, it cannot be settled except through legal evidence, for what is being argued, namely, edibility being the cause, is legal.^[181] This cause may be established by the revealed, unambiguous texts, by textual allusions (ima³),^[182] by the setting of the case,^[183] by the occurance [of the effect] with the occurance of the property, or by effectiveness (taWiir). The latter, as has been previously mentioned, is establishing the effect of the cause in the essence of the judgment in another case through a text, consensus, relevance (munasaba), or coextensiveness and coexclusiveness.^{76 ^[184] Or through consensus reached on the indispensability of a sign (Wama), or, subsequently, through investigation (sabr). This is in order to negate all signs aside from that which has been stipulated—as we have mentioned concerning the premises of argumentum a simile. [The stipulated sign] is also termed a cause (Hila) according to the majority of legal theoreticians. Shafi^cI also pointed to this cause when speaking of edibility and cash.}
If there is disagreement concerning the second premise—that is, concerning the existence of the cause in the assimilated case subsequent to admiting that the [existing] property is the cause—the cause may be identified either through sensory perception if the property is sensory, or through custom or language. It may also be identified through seeking definition (hadd) and conceptualizing (tasawwur) the inner reality of the thing, or through revealed scriptural evidence.
An example of the sensory is the large body of water which is spoiled by an impurity. //437// If dust falls into this water, the water is not spoiled because dust does not render water impure, just like wind when it blows [carrying with it, e.g., tree leaves] or when water is stand-still for a lengthy period of time. The form of the argument is: ‘That which does not spoil water [by changing its characteristics] does not render it impure,’ and ‘Dust does not [change water]’; therefore, ‘Dust does not render water impure.’ The adversary may argue: «I agree that what does not spoil [water] does not render it impure, but I do not agree that dust does not change [the characteristics of water]. Rather, dust conceals the characteristics of water, jpst like saffron and musk.»^[185] This can be known through natural sensory perception.
An example of the customary is our argument: «Since the sale entailing risk is prohibited, and since the sale of a non-existent commodity is a sale entailing risk, the sale of a non-existent commodity is null and void.» The adversary may argue: «1 agree on the first premise, but disagree that the sale of non-existent commodities is a sale entailing risk.» He may be answered that this is known through custom (^cadd) in which practice (^curf) decides.⁷⁹
An example of the linguistic is our argument: «Since manumission (^cataq) is carried into effect by equivocal, probable declaration,⁸⁰ and since divorce may be effected by manumission, divorce may be brought about by equivocal, probable declaration.⁸¹ The adversary may agree on the first premise, but disagree on the second; that is, that divorce may be effected by manumission. Then he seeks [to prove his claim] by means of metonymies, figuratives and metaphors.
[An example of] what is established by revealed texts is proving the body-snatcher to be a thief by [citing] ^cAlisha’s statement, may God be pleased with her: «The robber of the dead is just like the robber of the living.» Likewise, we prove //438// that real estate is subject to usurpation by citing the Prophet, peace be upon him: «He who usurps a span of land...» We also prove the homosexual to be an adulterer, by citing the Prophet’s statement: «If a man had a sexual intercourse [with another man] then both men are adulterers.» The mode of reasoning in this case is: ‘Adultery must be punished,’ and ‘He who is involved in homosexuality is an
ritual ablution. The fundamental assumption is that any body of water which has come in contact with other corrupting elements is rendered ritually impure. The reason for impurity is that any change affecting the characteristics of water will nullify its ritual purity. But a body of water which has come in contact with ritually pure substances, such as potash, saffron or musk, remains pure. See Ghazali, al· Wajtz, 2 vols., Cairo, 1317 H., I, 5 ff.; ^cAli b. Abi Bakr al-Marghinani, Hidaya: Sharh Biddyat al-Mubtadi, 4 vols., Cairo, 1980, I, 18 ff.
⁷⁹ Ghazali apparently conceives ^zurf as a component of the existence of that which possesses the characteristic. And the absence of the characteristic indicates the absence of that which possesses the characteristic. Likewise, the existence of the effect indicates the existence of that which produces the effect, and the absence of the effect indicates the absence of that which produces the effect.
An inference concerning a matter from a similar or equivalent matter.^[193] What is established as [a property] of a matter is, by necessity, established as the property of its equivalent or that of a similar matter.
An example of inference on the basis of a characteristic is the argument that the recommended additional prayer (watr) is like the voluntary prayer (nafal) because both can be performed while on a journey. Performing prayer on a journey is coexistent with voluntary prayer and coexclusive to [the minimal] obligatory prayer (JarcPuf). This is so //442// because performing prayer while journeying is characteristic of voluntary prayer; an obligatory prayer (farina) should by no means be performed on a journey.^[194] If the characteristic of a thing exists, it indicates the existence of that thing. The indication [of coexistence] becomes obvious once the. characteristic is accepted [as valid]. The significance of the characteristic is its coexistence with the essence of the thing in a manner which makes them inseparable: The existence of the thing indicates the existence of that which accompanies it, and is inseparable from it.
The adversary may argue: «1 do not agree that the permissibility of performing prayer while on a journey is characteristic of voluntary prayer, because the recommended additional prayer is not like the voluntary prayer, yet voluntary prayer is performed while journeying.» Such an objection is unwarranted because if we first assume that the ruling of recommended additional prayer, which is to be derived through revelatory indications, cannot be established by revelation one way or the other, we will find that recommended additional prayer is drawn towards voluntary prayer due to a characteristic unessential to obligatory minimal prayer. Therefore, in all probability it is [like] voluntary prayer.
We think it conceivable that the adversary may find textual evidence to the effect that recommended additional prayer is like the obligatory minimal prayer. The evidence may thus invalidate this characteristic.^[195] Otherwise, in the absence of textual evidence, the characteristic remains valid, and through it the ruling about recommended additional prayer can be known. If, however, there exists specific textual evidence, the claim for the similarity of characteristics is dropped. But the adversary must produce evidence if he has it.
The adversary may well argue: «The permissibility of performing prayer while journeying is characteristic of non-obligatory prayer; therefore, obligatory prayer must not be performed while journeying. I do agree that recommended additional prayer is not like obligatory minimal prayer (/arits effect; its non-existence signifies the non-existence of proprietorship. We may also argue that if the agent is the proprietor of the entire profit, he would be the proprietor of less than the entire profit. And since the loss [of profit] is incurred by the agent, it indicates that the agent is not the proprietor of the profit.^[199]
All this reverts to the inference that from the absence of a thing its effect is [known to be] absent. Given that there is a conclusion, this inference is demonstrable.^[200]
The characteristic may also be predicated upon the effect. It may be argued that the quest [of the slave] for acquiring livelihood and offspring is the effect and the result of contractual manumission. Its existence signifies the existence of efficient contractual manumission. Likewise, the prohibition against performing prayer on a journey is the effect of mandatory prayer; for if it were made mandatory in its entirety, then it could not be performed partially. So the non-existence of prohibition signifies the non-existence of mandatory prayer. If it were possible to determine the coexistent characteristic, then it would be neither a cause nor an effect, and the two characteristics would be inseparable; thus using one of them to signify the other is acceptable. This inference surpasses the inference based on the effect and the cause.
[3] The third kind [of indicative demonstration] is the inference of a thing from another which resembles it. We may argue that he whose divorce is valid, his temporary separation (zihar) is also valid.^[201] And he who pays tithe and food tax (fitrd) must pay alms tax (zakat). Also, if a small quantity of bodily fluid does not annul ritual ablution (wudu³), a larger quantity of the same fluid can no more annul ritual ablution.^[202] Arguments of this type are numerous.
//446// This [argument] can also be constructed on the basis of resemblance or through the method of coextensiveness and coexclusiveness, as it has been previously expounded. But Shafi^ci, may God be pleased with him, chose the inference of a matter from another similar matter. For his argument is that what is inherent in a matter is inherent in a matter like it.
There is no obscurity concerning this principle. Rather, obscurity lies in the claim for similarity.^[203] The adversary may not agree that divorce resembles temporary separation, or that tithe resembles alms tax, or that the small quantity of bodily fluid resembles a larger quantity. He may argue that a similar matter is indeed that which replaces, substitutes and is equal to, another in qualities which are possible, necessary or impossible. And these qualities can by no means be claimed to exist in these cases. However, resemblance and similarity may be conceived in certain respects, just as equating a male and a female is conceivable with regard to manumission and slavery. Thus male and female slaves are alike concerning the effect of manumission and the mitigation of prescribed penalty (hadd)™³ But this does not preclude differences between male and female slaves with regard to [the requisite of) guardianship in marriage or the right of the master to force the female slave to marry. This also does not preclude differences between men and women in matters of testimony, leading prayer, etc. We can argue for resemblance between males and females only in relation to slavery and manumission.
It may also be argued //447// that the slave is like a freeman who has no means to do penance,^[204] ^{^[205] and that the concubine is like a [free] married woman in that having sexual intercourse with either of them [when such an act is impermissible] necessitates the doing of penance.^[206] These similarities are conceivable provided it is admitted that there are dissimilarities which distinguish marriage from concubinage, although with regard to the nullification of fasting there are no differences. Therefore, it is possible for precepts diverging in characteristics to converge in general issues. We respond that fasting is like prayer as regards intent, and sale is like marriage as regards offer and acceptance. A case is thus inferred from another case on the basis of these properties, and the claim for analogy on the grounds of other matters, which are the characteristics, is precluded.^[207]}
Once this principle is established, Shafi^cI would be justified in arguing that temporary separation is like divorce, that is, temporary separation is like divorce in relation to non-belief (kufr) and Islam, for both [i.e., divorce and temporary separation] are means of prohibiting matrimonial relations with the wife. The way to prevent non-belief is temporary separation, for according to Abu Hanifa, non-believers are not addressed by the revealed law which stipulates impermissibility. Shafi^ci argues that being a revealed law, the prohibition of temporary separation is //448// like the prohibition of divorce, and that resemblance is obvious concerning both the Law and the prevention of non-belief. If the prohibition of divorce were not effected, the prohibition of the likes of divorce must not be effected either.
This is the course of reasoning which the adversary seems to fall back upon, and in which divorce is assumed to contradict temporary separation. According to this, if divorce were not obstructed then temporary separation which resembles it in prohibition is also not obstructed in relation to divine speech and the prohibitions of divine speech.
This evidence [remains] clear until the adversary argues that what is established by temporary separation is a prohibition which is removed by doing penance. Unlike divorce, however, penance is not possible for non-believers. Shafi^ci would answer: «No, penance is possible even though it is agreed that there is no penance. This decisively obstructs prohibition.» And so on concerning the methods of argumentation in this case.^[208]
Shafi^ci, may God be pleased with him, may also argue that if tithe and food tax are obligatory, //449// then all other taxes are obligatory because, being devotional services (^zibaddt), they are both, whether accepted or rejected, equal insofar as childhood is concerned. Thus the claim of similarity is accepted, for the adversary may consider childhood as an obstacle [in the performance of) devotional services [prescribed by] divine speech. Shafi^ci, may God be pleased with him, then argues: If this is obstructed, then tithe and food tax are also obstructed, for they are, insofar as the child is concerned, uniform devotional services just like alms tax and other religious financial obligations. The supposition that the child must pay [the taxes], the designation of a guardian for the performance [of these obligations], and the deferring of [the specifications of] divine speech until the child’s coming of age, are all equally conceivable. This is an argument for similarity on the basis of [a characteristic]. It may be reduced to the supposition of the adversary that childhood is the motive and that tithe annuls it and the knowledge that in this case alms tax resembles tithe.
Over and above this, the adversary must explain why he holds that tithe is a duty incumbent upon all individuals and that food tax is incumbent [only] upon well-to-do individuals.^[209] If this is established then the argument for similarity is annulled. We then argue that tithe and food tax require intent, and that it is lawful to levy land tax from those who do not pay tithe. This will indicate that tithe is not payable by all individuals. However, the reasoner’s evidence is ab initio probable; he thus gives it up //450// and instead draws the attention to differences. We then proceed by invalidating [these] differences. We also argue that if a small amount of venesected blood does not annul [ablution], then a large amount of the same blood does not annul it either, because a little amount is just like a large amount. The likeness between them is known in relation to the argument of the adversary. His argument is that all loci must follow the accustomed substrata, and in the accustomed substrata small and large amounts are alike. So, if a small amount [of blood which comes out] of veins is unlike that [which is emitted] from the accustomed outlet, then it is also unlike a large amount which is in this context like [the small amount of vein’s blood].
If it is asked «What is the reasoning behind the argument of Shafi^cI—may God be pleased with him—that the act of slaughtering which does not render the meat legal does not render the skin ritually pure either?» we answer that the ritual purity of the skin is the result of rendering the meat lawful, for since the meat to be eaten is from heads, trotters and slaughtered sheep^[210] it is determined to be ritually pure and so is the skin on it. If the antecedent (matbu^c) ceases [to yield an effect], the consequent (tdbi^c) ceases as well. The ritual purity of skin must be made either a conclusion or a consequent; the coexistence between the antecedent and the consequent is as known as the coexistence between the effect (natija) and the cause (muntij).
In this discussion, inference through similarity—as we have mentioned in the cases of temporary separation, divorce, tithe and alms tax—is unlikely.
The third type of demonstration is the reductio ad absurdum (burhan al-khulf). It is the setting aside of what one intends to prove, and establishing instead the invalidity of its contrary which stands in diametrical opposition to it. If one of the two opposites is proven invalid, then the other opposite would be proven valid. //451// This essentially reverts to [the method of] division and investigation (taqsim wa-sabr), where some of the parts [in the division] are invalidated so the remaining parts would be proven valid. It also involves another method of demonstration, namely, the enumeration of all the parts of a whole and then they are all invalidated so the whole may be established to be invalid.^[211]
In the first type of the reductio ad absurdum it is argued that «If it were not X it would have been Y; but it is not Y, therefore it is X.»^[212] An example of this is the following: If the sale of a nonexisting object were valid, the sale would be binding. But the sale is not binding; therefore, the sale is invalid. If the contraction is invalidated, the sale is proven null and void. Another example is: If the commenda agent were the proprietor of the profit, then he would be the proprietor of the profit resulting from the profit. But he cannot be the proprietor of the profit because this leads to an imbalance in the agreed upon ratio. Therefore, [the owner’s] proprietorship of the profit is invalid.^[213]
This [demonstration] concerns itself with invalidation and validation, as we have mentioned. It is evincible and effective because it is employed in rational demonstration. However, if it falls short of both invalidation and validation, it is of no use in rational demonstration //452// but only in probable knowledge. We may argue, for instance, that if edibility were not the cause, then the cause would be either nutritiveness, measurability or value. But all these are invalid; thus edibility is validated. Given that the procedure of establishing the cause is implemented, this cause [e.g. edibility] may be stipulated. This type of reasoning may be questioned on two counts: The first is the claim for inclusiveness and the second is the claim for invalidity.
If division (taqsim) involves both invalidation and validation, doubt concerning the claim for the invalidation of one of the parts is removed. This is why division which involves both invalidation and validation is operative in rational demonstration. We argue: If the world were not created, then it would be eternal; but it is impossible that it is eternal because it would be required to be unchangeable; therefore, it is established that it is created. Examples of this sort are many.
To this demonstration reverts what some scholars call inverted inference (qiyas aimaks). To illustrate it they use the following example of Abu Hanifa, may God be pleased with him: «Had fasting been unnecessary in the performance of iStikqf, it would have also been unnecessary in vowing, just like in prayer.»
Other scholars argued that this is invalid because it is an inference from the opposite (istidldl bil-didd). However, this claim is false //453// because the inference reverts to the demonstration of reductio ad absurdum. The course of this demonstration is: Had fasting not been an obligation, it would not have been an obligation when vowing either. But it is an obligation in vowing; therefore, it is obligatory. This is the demonstration of reductio ad absurdum. Its proponent may be asked: Why do you argue that if fasting were not obligatory, then it would not have been obligatory in vowing? And to what extent would vowing be a cause (sabab)? So, we argue with him about this impossibility. He may demonstrate its impossibility and argue: «If fasting were necessary in vowing when performing iHikqf, then prayer would be necessary in vowing.» This reverts to the inference about something from its likeness. He may argue: «Prayer is—insofar as it is rendered obligatory by vowing and relevant in ihikaf—tikt fasting. And if prayer is not rendered obligatory in vowing, how would fasting then be rendered obligatory?»^[214] ^{^[215]}
The crux of the argument is that he who vows to fast while performing i^tikdf must fast while performing iStikaf. This is either because fasting is a prerequisite for iHikaf, or because of the necessary concomitance (iltizam)—though one does not think it to be a prerequisite. It is false, however, to attribute it to necessary concomitance because if concomitance renders it obligatory it would also be obligatory in prayer even though //454// it is not a prerequisite. Prayer is like fasting with regard to vowing. If vowing is nullified in the like of prayer, it would be nullified in fasting as well, and its contrary would be confirmed. We have to show that a difference exists between fasting and prayer, and that fasting is an abstention from the same genus of iHikdf\ it blends with iHikdf_y and, unlike prayer, is influenced by it when it coexists with it.¹¹⁴ This is in accordance with what has been written about this question.
In the second type of reductio ad absurdum demonstration the parts of a whole are enumerated, and then each part is invalidated until all the whole is invalidated. We argue: If ild*^[216] were a kind of divorce (taldq), it would be pronounced either through an unambiguous declaration (yarih) or an allusion (kindya). Since it is invalid to carry it out through unambiguous declaration or allusion, it is not therefore a divorce. This reverts to two premises and a conclusion, namely, ‘No divorce [is effected] except through an unambiguous declaration or an allusion,’ but ‘There is neither an unambiguous declaration nor an allusion’; therefore, ‘there is no divorce.’
All this belongs to the methods of proof, most of which are interconnected. Division (taqsirri) and //455// reductio ad absurdum demonstration have much to do with all arguments. Most analytical knowledge revolves around them.
This is all we wish to mention about the illustration of the methods by which the causes of the original cases (usul) axe known.^[217] As we have already stated [at the outset of the treatise], this is one of the five principles [which lead to] the knowledge of legal inference (qiyds). The four principles remaining to be investigated are the original case, the assimilated case, the judgment and the cause. What we have discussed [in the present chapter] are the methods by which one of the principles—namely, the cause—is identified.}

<< | >>

↑

Source: Hallaq Wael B.. Law and Legal Theory in Classical and Medieval Islam. Routledge,2022. — 344 p.. 2022

More legal literature on Laws.Studio

Ill LOGIC, FORMAL ARGUMENTS AND FORMALIZATION OF ARGUMENTS IN SUNNI JURISPRUDENCE

More on the topic Ill LOGIC, FORMAL ARGUMENTS AND FORMALIZATION OF ARGUMENTS IN SUNNI JURISPRUDENCE: