13 — Reports
|A report is that which can be characterized by truthfulness or lying.||أمّا1 الأخبار فالخبر ما يدخله الصدق والكذب.|
|Reports are of two types: individually transmitted and collectively transmitted.||والأخبار تنقسم إلى2 آحاد ومتواتر،|
|A collectively transmitted report makes knowledge obligatory. It is related, at every stage of its transmission all the way back to the one from whom it is reported, by a group the likes of which could not conspire to lie. It must originate in eye-witnessing or hearing, not in diligent inquiry.||فالمتواتر ما يوجب العلم وهو أن يروي جماعة لا يقع التواطؤ على الكذب من مثلهم، إلى أن ينتهي إلى المخبَر عنه ويكون في الأصل3 عن مشاهدة أو سماع لا عن اجتهاد.|
|Individually transmitted reports make action obligatory, but do not make knowledge obligatory. They are of two types: those that are merely attributed to the Prophet, and those that are actually traced to the Prophet. A traced report has a continuous chain of transmission. If an attributed report is attributed to the Prophet by anyone other than a Companion, it is not an authoritative proof, except for the reports attributed by Saʿīd ibn al-Musayyab, which have been scrutinized and found to be traceable to the Prophet.||والآحاد هو الذي يوجب العمل ولا يوجب العلم، وينقسم إلى مرسَل ومسنَد، فالمسنَد ما اتّصل إسناده والمرسَل4 إن كان من مراسيل غير الصحابة فليس بحجّة إلّا مراسيل سَعيد بن المُسَيَّب فإنّها فُتِّشَت فوُجِدت مسانيد.|
|Relating a report as “from so and so”5 counts as a form of tracing. If a teacher read a report, the one who later relates it from him may say “he told me” or “he informed me.” If the one who relates the report read it to the teacher, he says “he informed me,” but does not say “he told me.” If the teacher authorized him to relate traditions from him, the one who relates from him says “he authorized me,” or “he informed me by authorization.”||والعنعنة تدخل على الإسناد، وإذا قرأ الشيخ يجوز للراوي أن يقول حدّثني و أخبرني، وإن قرأ هو على الشيخ فيقول أخبرني6 ولا يقول حدّثني، وإن أجاز الشيخ7 فيقول الراوي أجازني أو أخبرني إجازةً.|
Reports are merely channels of the Sunna
Like his doctrine of consensus, al‑Juwaynī’s analysis of different types of reports is not a description of a source of law, but a set of tools for interpreting revelation and debating its interpretation.
A report (khabar, by which al‑Juwaynī means specifically a report from the Prophet, often called a ḥadīth) is not a source of law in its own right; it is a channel through which we know the source, the Prophet’s Sunna, which consists of his actual words, deeds, silences, and inactions. All of these were discussed in the first half of the treatise, so we already know how to interpret the Sunna. This section is about how we know the Sunna, and how the way we know it affects its interpretation.
Al‑Juwaynī never explicitly mentions the Sunna or even the Prophet in this section. His analysis of reports could apply to historical reports, reports about the meanings words were coined to convey, reports about the existence of consensus, or anything else. But he shows his real interest when he alludes to the Prophet by mentioning the Companions who “attribute” reports to him; and he alludes to the Sunna itself when he insists that it, and not someone’s interpretation of it, is what legally relevant reports must be about. This is the point of insisting that the first narrator or transmitter of a report should see or hear the Prophet directly, rather than just giving the narrator’s own opinion about what the Prophet wanted the law to be. For example, when the Prophet’s wife ʿĀʾisha said that “if the Prophet had known what the women were doing, he would have barred them from the mosque,” she was reporting her own considered opinion rather than the Prophet’s Sunna itself, and we know that al‑Juwaynī did not consider the opinion of a Companion to be legally binding for others.
Some early Ẓāhirī legal theorists were even stricter than al‑Juwaynī, requiring that the Prophet’s exact words be related verbatim. Al‑Juwaynī does not say that a report must be verbatim; his definition of reports implies rather that he is interested in “truthfulness” as opposed to “lying.” We know that as an Ashʿarī he is more interested in the inner speech-meaning and intent behind the Prophet’s verbal and non-verbal communication, than in his words per se. He does insist that reports should be about the source itself, not someone’s interpretation of the source reflecting his own diligent inquiry. But by defining reports in terms of truthfulness versus lying, rather than in terms of literal or historical accuracy, he shows that his concern is not to recover an exact text of the Prophet’s words or an exact history of his deeds. Rather, the message of this section is that as long as we have reason to believe that a bit of the Sunna is being reported truthfully rather than deceptively, then that report imposes on us an obligation, even if we do not know with certainty that it is historically true.
Authority requires continuous transmission, not certainty
This whole section, in fact, seems to be about the various degrees of certainty with which the Sunna can be known through reports. Al-Juwaynī’s first distinction is between reports that might possibly be a lie, and reports that could not possibly be fabricated because they are related by numerous independent chains of witnesses, from the Prophet’s time down to our own. It turns out, however, that the mere possibility of lying does not affect the practical value or legal authority of a report: obedience to a bit of Sunna is obligatory whether or not we are certain that the report about it is truthful. The legal obligation to act does not depend on the epistemological certainty of one’s knowledge. There were groups, early in Islamic history, who insisted that legal obligation could only be based on indubitable knowledge such as that provided by the Qurʾān (which was considered to have been collectively transmitted from teacher to student by numerous chains of Qurʾān reciters). Some early Muʿtazila and Ẓāhiriyya had taken this stance, and since they generally held that only a very few ḥadīth were this widely transmitted, this severely limited the amount of legal detail they could expect to find in revelation. For some of them, this was exactly the point: they preferred to limit religious law to the few points dictated in the Qurʾān, and leave the rest of life unregulated or open to the decrees of political rulers. Al‑Juwaynī, however, like virtually all the Sunnī legal tradition, adopted al‑Shāfiʿī’s idea that the massive body of individually transmitted reports about the Prophet’s Sunna provided a detailed elaboration of the Qurʾān’s requirements, covering all of life. In this vision of revealed law, all reports about the Sunna are binding, as long as their chain of transmission (isnād) meets a minimum standard.
What, then, is al‑Juwaynī’s purpose in discussing the various levels of certainty provided by different types of reports? We will have to keep this question in mind as we examine his next two ways of classifying reports.
Having already distinguished multilaterally from individually transmitted reports, al‑Juwaynī now further subdivides the latter into those that are “traced” to the Prophet and those that are merely “attributed” to him. The former are those whose text or content (matn) is accompanied by a full chain of transmitters (isnād) that goes all the way back to the Prophet. Such a report has the form: “X related to us, from Y, that Z (a Companion of the Prophet) said he heard the Prophet say….” By contrast, a report that is cited with an incomplete chain of transmitters might have the form: “X told us that Y (a Successor, of the generation following the Companions) related that the Prophet said….” In this case the Successor attributed a saying or action to the Prophet without actually tracing it to him by specifying which Companion related the event. Here al‑Juwaynī’s distinction makes a decisive difference: such an “attributed” report is of no legal value or authority whatsoever. He makes only one exception, for a Successor named Saʿīd ibn al‑Musayyab, of whom al‑Shāfiʿī had said that he always transmitted his reports from specific Companions even though he routinely failed to name them.
Al-Juwaynī’s last distinction, however, has no such decisive consequence: he explains that reports can be classified according to the kinds of transmission that took place between one person and the next in their isnāds. These modes of transmission reflect the various methods of teaching and learning that were employed in mosques and madrasas. Al‑Juwaynī sets up an implicit hierarchy among these modes of teaching. The best method is for your teacher to read you a ḥadīth from his own notes (or recite it from memory). In that case you have the right to authenticate the ḥadīth, when you recite it one day to your own students, by saying “he [my teacher] told me.” You could also say “he informed me,” but then you would not be distinguishing your transmission from the next best kind, which can only be authenticated with the words “he informed me.” In this method, you write out for yourself a transcript of your teacher’s notes or lectures, and then you read them aloud to the teacher, who confirms that you have written them down correctly, and thus gives you the right to transmit them to others. The third method is not as prestigious, and therefore has to be qualified as “he informed by authorizing me,” or simply “he authorized me” to transmit this ḥadīth. This is the best you can say if you never actually heard a formal recitation of all the ḥadīth you learned from your teacher, or read out to him all the ḥadīth you copied, but merely received his authorization to pass on to others all that you learned from him, because the teacher knew you had studied the material well. Finally, the weakest form of authentication, which al‑Juwaynī accepts without qualification (though some did not) was simply to say “I got this ḥadīth “from” my teacher, without specifying how you had learned it.
Al‑Juwaynī sets up this elaborate classification of modes of transmission, but never says what difference it makes. All of them are adequate, so all of them make acting upon the content of the ḥadīth obligatory. What, then, is the purpose of this classification?
Institutional context matters
One way in which I think al‑Juwaynī’s classification of modes of transmission is highly significant is that it establishes in the mind of the person learning, transmitting, or interpreting a ḥadīth a sense of being related in a certain way to the words or events it relates. An interpreter does not just encounter a ḥadīth, analyze its isnād, and then assign it a certain degree of epistemological certainty—a 60% or 70% chance of being truthful. In itself this would not really affect the interpreter’s understanding of the ḥadīth, or his duty to follow it. What is more important is that the isnād places the reader or interpreter into a tradition of teaching, so that he receives the ḥadīth not as an isolated text on a loose piece of paper that he can make what he wants of, but as a piece of a larger heritage into which he has been initiated through his legal training. The ḥadīth is not detached from its author, open to whatever meanings the reader’s situation might lead him to project onto it; it has always been attached, generation after generation, to certain understandings of the ḥadīth and of its legal implications, which have been handed down from teacher to student as part of his education. He can no more reinvent the meaning of the ḥadīth than he can discard the interpretive framework established by the prior consensus of jurists, by the range of their interpretations, and by the moral values and presuppositions his teachers have instilled in him.
Al‑Juwaynī’s analysis of transmission, then, establishes a certain institutional, educational, and social context through which both the texts of revelation and the context for their interpretation are handed down. If he were only after epistemological certainty, he might care more about historical plausibility, or about written records; but in fact he prefers to make legal authority rest on participation in an oral tradition that transmits not only texts but also the proper “horizon” for understanding them. The uninitiated cannot just come to the texts with their own “horizons of understanding.” A written text is in a sense “set loose” from its author, from previous interpretations of it, and from the linguistic and cultural frameworks that previously gave it meaning; but a text with a personal and largely oral isnād is bound (at least in the reader’s imagination) to the earlier scholars who transmitted it, and to the frameworks of interpretation that shaped their understanding of it, and also to the author himself. When scripture is imagined as transmitted in this way, the reader is not imagined to have independent control of its meaning—despite all that I have said before about the flexibility of al‑Juwaynī’s rules for interpretation. He is imagined as bound to an interpretive tradition, and through it to the intent of the author.
This social, institutional, and educational context has been seriously disrupted, however, by colonialism, the introduction into Muslim societies of modern forms of education (which are much less focused on individual interaction between teachers and their students), and by the use of printing and electronic modes of storing and accessing texts. Already in al‑Juwaynī’s day the practice of teacher-student transmission had been interrupted somewhat by the use of popular books – ḥadīth collections such as those of al‑Bukhārī and Muslim—that came to serve as reference works and thus eliminated the need to learn all one’s ḥadīth directly from living transmitters. Already in those collections the ḥadīth were set loose, to some degree, from the tradition that transmitted them. But they were still organized in ways that preserved their compilers’ understanding of the ḥadīth, and more importantly they were not studied in a vacuum; they were read with the help of commentaries, under the guidance of a teacher who had himself been trained in the same interpretive tradition. How different things are today, when anybody can perform an online search and come up with a dozen ḥadīth on any given topic, extracted from their literary contexts by the search engine, and read in the privacy of one’s laptop or phone, in a social and cultural context transformed by modernity and postmodern thought, without the guidance or limitations provided by a teacher or an interpretive tradition. This is a far cry from the interpretive situation that al‑Juwaynī envisioned for readers of ḥadīth. In such a transformed context, his flexible rules of interpretation have the potential to wreak havoc on the meaning of revelation.
We can see, then, some significance in al‑Juwaynī’s epistemological grading of ḥadīth. Even if his analysis of degrees of certainty does not affect the meaning or authority of ḥadīth, it does establish a certain minimum standard of transmission (at least one complete isnād all the way back to the Prophet) and a certain context for their interpretation.
Authority is determined by acceptance, not certainty
Al‑Juwaynī knew lots of other ways of evaluating the authenticity of ḥadīth, which he skips over. He could have presented the whole “science of the men who related traditions” (ʿilm al‑rijāl), which he mentions in chapter 17 as a necessary qualification for diligent inquiry. This science lays out the lives and reputations of transmitters, thus enabling scholars to evaluate the likelihood that a given report is trustworthy by assessing the reliability of the men (and women) who transmitted it. Isnāds could also be evaluated based on their plausibility: did X really move to Baghdād in time to hear this report from Y before Y died? Some legal theorists also proposed that ḥadīth could be evaluated based on the plausibility of their contents: if they contained obvious anachronisms, or contradicted the Qurʾān or well-established norms, then they could be rejected as inauthentic—though generally jurists were more comfortable criticizing an isnād than the content of a ḥadīth.
In this brief handbook, however, al‑Juwaynī was content to keep his tools for evaluating ḥadīth fairly simple. His overall message was that nearly all ḥadīth, regardless of the degree of epistemological certainty they offer, have the authority of legal sources, as though they were themselves direct reflections of the Prophet’s actual Sunna.
This was, in fact, how ḥadīth were generally treated by jurists. For example, Shāfiʿī jurists were nearly unanimous in their view that the monetary compensation (diya) that should be paid for the murder of a woman was only half what should be paid for the murder of a man. They supported this view by claiming (usually without citing reports) that it was held by various Companions—even though legal theorists like al‑Shāfiʿī and al‑Juwaynī did not consider their opinions authoritative. They also cited a report from the Prophet, but the only ḥadīth collection containing this report says explicitly that its isnād is inauthentic. In the end the strongest evidence in favor of their view was simply the fact that it was widely accepted, not only among the Shāfiʿiyya but among other schools as well, to the point that they could say “we know of no disagreement on this issue.”8 The epistemological weakness of the revealed evidence was no obstacle to the scholarly community’s acceptance of a rule.
Even al‑Juwaynī’s rejection of “attributed” ḥadīth did not keep them from being used. Some of the Shāfiʿiyya considered them authoritative, but even those who rejected them in principle made use of them in specific arguments. For example, the sole Prophetic ḥadīth that explicitly supported their rejection of women’s testimony in marriage and divorce cases was “attributed” to the Prophet by a scholar named al‑Zuhrī who lived several generations after the Prophet; but they relied on it nevertheless, and justified this by pointing out that the Ḥanafiyya, against whom they were arguing about women’s testimony, did accept attributed reports.9
Thus the real criterion for the use of ḥadīth was not their degree of epistemological certainty, but their degree of correlation with one’s own legal opinions. In fact, historical research suggests that from the very beginnings of “ḥadīth science” (which invented the categories and criteria used by al-Juwaynī) the legal authority of a ḥadīth was assessed not according to the quality of its isnād, but according to its conformity to accepted norms. Jonathan Brown has argued that early Sunnī scholars treated ḥadīth as authoritative not because they thought their isnāds guaranteed that what they reported was precisely what happened, but because they thought their content reflected the gist of what they felt the Prophet had taught, as reflected in the norms and practice of the early community.10 In a similar vein, Eerik Dickinson has argued that when early ḥadīth critics declared that a certain transmitter was reliable, what they meant was not that he was a trustworthy person with a good memory, but that the ḥadīth he transmitted tended to be in line with what everyone else reported.11 If these historians are right—and I certainly respect their work very much—then the criteria and categories al‑Juwaynī has given us for evaluating the degree of epistemological certainty afforded by ḥadīth were not actually developed for the purpose of figuring out which reports should and should not be used in constructing the law. Prevailing conceptions of the law itself determined which ḥadīth were most important.
Another flexible interpretive tool
This tells us something important about the purpose of al‑Juwaynī’s epistemological analysis of isnāds: it is not about defining the canon of revelation, or recovering the original text of revelation or the actual words of the Prophet (as in Biblical criticism). Al‑Juwaynī’s goal is not to draw a sharp line around some ḥadīth and declare them “in” while declaring others “out.” He emphasizes that nearly all reports, regardless of their mode of transmission and their epistemological certainty, have to be obeyed. In practice even those reports he excludes—ḥadīth about Companions and “attributed” reports—turn out to be useful as evidence of the law. There was in fact a kind of canonization process that singled out some ḥadīth—especially the books of ḥadīth collected by al‑Bukhārī and Muslim—as having a special prestige;12 but al‑Juwaynī’s words here show that inclusion in a certain book is not the criterion for legal authority. Nor does he seem concerned to determine what the Prophet actually said and did, and eliminate every corruption that may have crept into the corpus of ḥadīth. Most ḥadīth exist in many variants; even the Qurʾān itself exists in multiple texts that differ slightly from one another. Al‑Juwaynī’s theory makes all variant texts authoritative, as long as they meet the basic criterion of having at least one single full isnād. In practice jurists tended to take variant reports, and even different versions of Qurʾānic verses, as so many separate pieces of evidence, all of which could be used and reconciled as if they were independent texts. They did not usually attempt to establish a single authentic text before proceeding to interpret it. The modern idea, common among Muslims and non-Muslims, that there is a single undisputed text of the Qurʾān, and even some completely trustworthy collections of ḥadīth that one can treat as undisputed scriptural canons, is not one that al‑Juwaynī would have affirmed. He was well aware of variants and uncertainty, but these apparently posed no obstacle to his project of grounding law in revelation. Western scholars trained in Biblical criticism sometimes disparage the kind of isnād analysis that al‑Juwaynī has introduced us to, but they do so because they misunderstand it. Its goal is not to establish a single authentic text or a high degree of historical certainty.
If this discourse were about establishing the text of revelation, or the truth about the Prophet’s Sunna, it would have been placed in the first half of the treatise where al‑Juwaynī described the raw material of revelation. Here he is talking about the process of interpretation, and his epistemological analysis of degrees of certainty is just one more interpretive tool, like all the other tools he has given us. Most of those tools have involved establishing intertextual relationships in which one text modifies the meaning of another, or displaces it entirely (as in abrogation). This chapter seems designed to let us establish comparative relationships between reports: this ḥadīth provides a greater degree of certainty than that one. Sure enough, toward the end of the treatise, once he has finished laying out all his interpretive methods, in chapter 16 al‑Juwaynī spells out which types of evidence get priority over which others. Among other things he says that those that produce knowledge are given priority over those that produce belief. This means that if two ḥadīth conflict, or if a ḥadīth conflicts with the law and needs to be somehow modified or overruled, all one needs to do is use one or more of al‑Juwaynī’s concepts to show that the problematic ḥadīth is relatively weak, and find some other ḥadīth or other evidence that is stronger in some way. We are then able to discount the problematic ḥadīth. We are not discounting revelation itself—the Prophet’s Sunna; we are only discounting the report by which we know that bit of revelation, by virtue of some characteristic of the report itself. The Sunna remains unimpeachable, yet can be discounted by virtue of the epistemologically fallible form in which it reaches us.
This is obviously a rather crude way of reconciling or modifying texts. As we saw with abrogation, al‑Juwaynī prefers not to discount a text if he can help it; he would rather reinterpret a text by particularization, clarification, or some other linguistic device. But if such reinterpretation is not possible, or not advantageous, one can resort to abrogation; and if one does not have suitable evidence about the date at which a text was revealed, one can appeal to one of the many factors that make it epistemologically uncertain, and argue that it should be discounted in favor of some other evidence.
This kind of intertextual relationship, then, is yet another tool that an interpreter can use to control the legal meaning of revelation. It too contributes to the flexibility of al‑Juwaynī’s hermeneutic. And like the other methods we have studied, it can be used most easily to reconcile the evidence of revelation with the legal status quo. The science of ḥadīth criticism began by ascribing soundness to precisely those ḥadīth that best fit the existing norms and practices of the Muslim community, and subsequent decisions about which ḥadīth to use and which to discount reinforced the correlation between sound ḥadīth and the legal status quo. Thus despite the flexibility inherent in the tools of ḥadīth criticism, it is always much easier to argue for the authenticity of a ḥadīth that the legal tradition has found “sound” because it was compatible with the law.
This does not mean, however, that ḥadīth criticism is any less a two-edged sword than the other devices we have studied. It can be used just as easily by those who want to reform the law; the only difference is that their use of it will face an uphill battle for communal approval, because it will not have the advantage of fitting the tradition’s expectations. Khaled Abou El Fadl provides a brilliant illustration of the reforming potential inherent in al‑Juwaynī’s legal theory. He develops a more sophisticated theory of ḥadīth criticism, which takes account of the selective and interpretive role that each transmitter plays in shaping our knowledge of the Prophet’s Sunna; and he offers several examples of detailed criticism of several reports that he finds demeaning to women. He argues in detail, for instance, that the famous ḥadīth in which the Prophet said that women are “deficient in religion and intellect” reveals as much and perhaps more about the chauvinism of the culture and the individuals that decided to transmit, authenticate, and interpret it, than about the Prophet’s own view of women. He then makes the same kind of epistemological point that al‑Juwaynī’s theory is designed to let us make: this ḥadīth cannot itself bear the weight of the severe legal consequences that it is often used to support, but must be set aside in favor of other revealed evidence that better fits the norms and values of the present Muslim interpretive community. His argument raises hackles, but it is just a continuation and refinement of the very legal theory proposed by al‑Juwaynī.
For an overview of the significance of Prophetic ḥadīth in many fields including legal theory, see Jonathan A. C. Brown, Hadith: Muhammad’s Legacy in the Medieval and Modern World (Oxford: Oneworld, 2009).
On the formation of ḥadīth criticism, see Eerik Dickinson, The Development of Early Sunnite Ḥadīth Criticism: The Taqdima of Ibn Abī Ḥātim al-Rāzī (240/854–327/938) (Leiden: Brill, 2001).
On the process of canonization, see Jonathan Brown, The Canonization of al-Bukhārī and Muslim: The Formation and Function of the Sunnī Ḥadīth Canon (Leiden: Brill, 2007).
For an innovative approach to ḥadīth criticism, see Khaled Abou El Fadl, Speaking in God’s Name: Islamic Law, Authority, and Women (Oxford: Oneworld, 2001).
- Spr. 601 has instead وأما .
- Spr. 601 has instead والخبر ينقسم إلى قسمين .
- Here ends the original text of Spr. 601; the final folio was added later, completing the text in a very different hand. Since it is based on a quite different version of the text, I will not cite its variants.
- The Maḥallī tradition adds ما لم يتصل إسناده ﻓ . This phrase was not in Ibn al‑Firkāḥ’s mutūn, however, for he says explicitly that al‑Juwaynī did not explain what mursal means.
- Al-ʿanʿana, the practice of indicating the chain of transmission by saying “from (ʿan) so and so,” instead of using a more complete form such as “so and so told me.”
- SH has instead أخبرنا . When he repeats this expression in his own comments Ibn al‑Firkāḥ says أخبرني , but that proves nothing because he often changes the pronouns when repeating these expressions.
- The Maḥallī tradition and one of SH’s mss, Riyadh 5878, adds من غير قراءة .
- Scott C. Lucas, “Justifying Gender Inequality in the Shāfiʿī Law School: Two Case Studies of Muslim Legal Reasoning,” Journal of the American Oriental Society 129 (2009): 237-258.
- Jonathan A. C. Brown, “Did the Prophet Say It or Not? The Literal, Historical, and Effective Truth of Ḥadīths in Early Sunnism,” Journal of the American Oriental Society 129.2 (2009): 259–285.
- Eerik Dickinson, The Development of Early Sunnite Ḥadīth Criticism: The Taqdima of Ibn Abī Ḥātim al-Rāzī (240/854–327/938) (Leiden: Brill, 2001), ch. 6.
- See Jonathan Brown, The Canonization of al-Bukhārī and Muslim: The Formation and Function of the Sunnī Ḥadīth Canon (Leiden: Brill, 2007).