2 - Whither Hadith Studies.pdf PDF

Preview

Summary

2 Whither €adÏth Studies?1 Harald Motzki “ We have to adopt a highly critical attitude towards our own theories if we do not wish to argue in circles: the attitude of trying to refute them.” — Karl Popper2 I. Introduction: Juynboll on N¥fi¢, the Mawl¥ of Ibn ¢Umar3 Two problems face historical resea...

Description

Accelerat ing t he world's research.

2 - Whither Hadith Studies.pdf Harald Motzki

Related papers

Download a PDF Pack of t he best relat ed papers 

“Murder in Khaybar: some t hought s on t he origin of t he qasama procedure in Islamic law,” Isla… Ruud Pet ers

Hadit h (

)

Pavel Pavlovit ch 1 Art icle - MOHAMMAD SAEED MIT WALLY IBRAHIM AL-RAHAWAN _ Det ect ing Text ual Addit ions _ 08-12… Mohammad Al-Rahawan

2 Whither €adÏth Studies?1 Harald Motzki “ We have to adopt a highly critical attitude towards our own theories if we do not wish to argue in circles: the attitude of trying to refute them.” — Karl Popper2

I. Introduction: Juynboll on N¥fi¢, the Mawl¥ of Ibn ¢Umar3 Two problems face historical research into the textual sources on early Islam. First, the sources are — apart from a few “remnants”, such as inscriptions, coins, etc. — only “traditions” and most of these are of Muslim provenance.4 Secondly, these traditions are available only in sources originating more than one and a half centuries after the events they purport to relate. Scholars have quite reasonably felt justified in questioning the epistemic value of these traditions as a basis for reconstructing the historical events in early Islam. On the question of the historical value of ^adÏths late nineteenth century Western scholarship provided different answers. Opinions range from a broad acceptance of these traditions as historical sources to complete rejection. €adÏths are rejected because they are thought to have been influenced by later political, religious and legal developments. What’s more, some scholars are convinced that the ir point of origin can be traced to these developments, and that they can plausibly be considered projections of them. As long as the value of the traditions as historical sources is a subject of dispute, any attempt to reconstruct the political, religious and legal developments in early Islam is on shaky ground. There are ways to escape this sterile situation. On the one hand, we can critically examine the previous attempts to assess the historical value of the Muslim traditions and seek to establish whether their premises, methods and conclusions are really satisfactory. On the other, we can examine anew the sources themselves, with the aim of finding criteria which will 1 This is a modified translation of my article “Quo vadis €adÏt-Forschung? Eine kritische Untersuchung von

G.H.A. Juynboll, ‘N¥fi¢, the Mawl¥ of Ibn ¢Umar, and his Position in Muslim €adÏth Literature’” in Der Islam 73 (1996), 40-80; 193-231. The translation was made by Dr. Frank Griffel and revised by Dr. Paul Hardy. I wish to thank them for all their help. 2 Karl Popper, The Logic of Scientific Discovery (London: Hutchinson, 1983). 3 G.H.A. Juynboll, “N¥fi¢, the Mawl¥ of Ibn ¢Umar, and his Position in Muslim €adÏth Literature” in Der Islam 70 (1993), 207-244. 4 On “remnants” (Überreste) and “traditions” (Überlieferungen) as technical terms in the field of the historical evaluation of sources cf. Ahasver von Brandt, Werkzeug des Historikers (Stuttgart: Steiner Verlag, 1973) 51- 64.

2

allow safe judgements on their historical value. In the present chapter, this dual path will be pursued by scrutinising a study which deals with the historical value of traditions allegedly going back to N¥fi¢, a client (mawl¥) of one of the Prophet’s companions, ¢Abd All¥h b. ¢Umar. It is well known that the majority of Muslim traditions about the first century consists of a text (matn) and of corresponding information on how this text was handed down (isn¥d). The chain of transmission or isn¥d allegedly provides detailed information on how the text got to the author of the compilation in which it is to be found. Muslim €adÏth criticism judged the reliability of a given tradition first of all from the standpoint of its isn¥d. In contrast, Western scholarship, with its aim of assessing the historical value of a tradition, has restricted it’s attention mainly to the text. The latter approach was prompted by the conviction that the isn¥d is, in the most cases, purely fictional. This opinion was shared by Ignaz Goldziher, one of the founding fathers of Western €adÏth studies. On the other hand, due to the scarcity of sources, it was extremely difficult, if not impossible, to verify the reliability of the isn¥ds. This could not be done without referring to information given by the early Muslims themselves in the biographical dictionaries on the traditionists. Yet this information was not considered to be very reliable by Western scholars. One of the few scholars in Western €adÏth studies who tried to find methods which include the isn¥d as a criterion to value the sources was Joseph Schacht. Although he relied on the contents and structure of the texts in his attempt to evaluate and date the traditions, he also developed premises and methods to improve the dating of texts by studying the evidence of how they were handed down. Josef van Ess and G.H.A. Juynboll took over Schacht’s attempts and Juynboll in particular has developed and improved them. Michael Cook, however, was highly critical of Schacht’s method of dating the traditions by studying their isn¥ds.5 Juynboll systematically arranged the different chains of transmission into clusters, representing one or a number of similar traditions in perspicuous diagrams to a high degree of accuracy. He coined new terms to characterise particular phenomena within these isn¥d clusters, terms like “single strand,” “partial common link,” “spider,” “dives” and others. He then provided a historical explanation for these conspicuous phenomena in his diagrams of isn¥d bundles. Based on these observations he deduced certain rules for the dating and evaluation of the authenticity of individual isn¥ds. Juynboll’s article “N¥fi¢, the Mawl¥ of Ibn ¢Umar, and his Position in Muslim €adÏth Literature” is a good example of the current stage that isn¥d analysis has reached in Western €adÏth studies. It also reflects the evaluation of Muslim traditions as sources for historical research based on this analysis. One may summarise the results of Juynboll’s studies on N¥fi¢ as follows: 5 For a more detailed summary of Western €adÏth studies and their methods, and for the relevant literature cf.

Chapter One of this volume.

3

1) The Prophetic ^adÏths with the isn¥d, N¥fi¢ Æ Ibn ¢Umar, prized by Muslims and found in the “canonical” collections of €adÏth, do not go back to N¥fi¢. For the most part they go back to M¥lik b. Anas. 2) The Muslim biographical literature claims that M¥lik is a pupil of N¥fi¢, even though this relationship appears to have no historical basis. 3) Those ^adÏths which have the isn¥d, N¥fi¢ Æ Ibn ¢Umar, and which are handed down by other supposed pupils of N¥fi¢ are without exception fabrications either by the authors of the “canonical” collections or their teachers. In this chapter, I shall argue against Juynboll’s results, since all three points are either wrong or, at the very least, lack sufficient justification. This is not to deny that the analysis of isn¥ds is a very useful tool for attempting to date the traditions. Nevertheless, it is necessary to call into question a number of premises and methods recently introduced by Juynboll into €adÏth studies. II. Chain Analysis: Method and Concept A. The Common Link and its Single Strand Several of Juynboll’s publications follow Schacht’s earlier attempts at isn¥d analysis by pointing out that the majority of the isn¥d bundles fail to exhibit the kind of structure one would expect if the Prophetic traditions had been handed down in continuous fashion from generation to generation. It is significant that they do not divide into several branches immediately after the Prophet. In the majority of cases they divide only after a succession of three to four transmitters who form a single strand.6 The transmitter after whom the chains of transmission divide into several branches Schacht named “common link.” Juynboll follows Schacht in this. Such an anomaly clearly requires explanation. According to Juynboll, explanation is to “be sought in the chronology of the birth of the isn¥d.”7 In his view of the origins of the isn¥d, the naming of authorities or authenticators of information about the Prophet and the so-called Companions became a requirement only in the third quarter of the first century after the Hijra (A.H.) during the “second fitna” of 61-73 A.H. / 681-692 C.E.

6 To avoid any complication we will use the terminology established by Juynboll. I also follow Juynboll’s

method of arranging the collectors (the starting point of an isn¥d) at the top of an isn¥d, and the end (e.g. the Prophet or companion) at the bottom. J. Schacht, M. Cook and others do it the other way round. 7 Juynboll, “N¥fi¢, the Mawl¥ of Ibn ¢Umar, and his Position in Muslim €adÏth Literature,” 210.

4

There is something to be said for this theory of the genesis of the isn¥d. The question is, however, whether it adequately explains the phenomenon of the common links with their single strands going back to earlier authorities. Juynboll seems to suggest that the single strands below the common links are the consequence of the late origin of the isn¥d. If it was not usual to name the sources of reports during most of the first century, then the strands with very early transmitters must be later fabrications. Since we owe these early isn¥d elements to the common links, they must be the fabricators. That means that the single strand containing the early transmitters is not reliable. This explanation has several shortcomings. First, it does not explain Juynboll’s observation that common links are not usually found at the level of the “Successors” (t¥bi¢‰n) but one or more generations later. Second, the general conclusion that the common links must be the fabricators of their single strands which are, therefore, historically unreliable seems to be questionable. The “Successors” flourished in last quarter of the first Islamic century and the first twenty years of the second. If the isn¥d came into force at the beginning of this phase, a high incidence of common links should be found amongst the Successors. In defence of Juynboll’s claim, one might suggest that the various centres of learning adopted the isn¥d at varying rates. Indeed, there is some evidence that this is the case.8 But even then the number of Successors among the common links should not be so small. On the other hand, Juynboll’s claim that only few Successors, if any at all, are true common links is perhaps an artefact of his own style of isn¥d analysis. As our investigation unfolds, the reader will find that his analysis is based on doubtful premises so that, in the end, the error in his hypothesis will become clear. I mean his view that N¥fi¢, though a member of the generation of Successors, is only an apparent common link. The failure of Juynboll’s hypothesis probably means that more of the Successors are common links than he would like to suppose. Be that as it may, the custom of naming authorities, introduced in the last quarter of the first century A.H., is certainly a pre-condition for the single strands in the lower end of the isn¥d bundles. Given that people became accustomed to give authorities for their reports only at the end of the first or the beginning of the second century A.H., does that necessarily mean that the authorities they named are fictitious? Is this an explanation of the common link phenomenon, i.e. the fact the transmission lines mostly fan out only in the third or fourth generation after the Prophet? I doubt it and propose another explanation for the common link phenomenon. It is better to look upon the common links as the first great collectors and

8

Cf. H. Motzki, Die Anfänge der islamischen Jurisprudenz, 210, 214-215; The Origins of Islamic Jurisprudence, 235, 240-241.

5

professional teachers of knowledge in general and of traditions about persons living in the first century of Islam in particular. This makes it easier to understand the single strand below the common link as well. It is the isn¥d given by a first systematic collector. He wrote down the traditions of the first century after the Hijra and transmitted them in circles of learning or “lectures.” With his isn¥d such a first collector stated from whom he had received a particular tradition, that is, from which Successor, and from whom the latter in turn, had allegedly received it. It is conceivable, in view of the chronology of the birth of the isn¥d, that the collector asked his informant for the source of his tradition and was told the name or, at least, some name. But it is equally possible that the collector did not make this enquiry of his informant. He simply inserted the name that to him seemed most likely. Single strands are, thus, the consequence of the fact that the early collectors — unlike later ones — usually gave only one source (and thereby only one isn¥d) for a tradition. The reason may have been that they only transmitted those traditions that they considered to be the most reliable and/or that there was as yet no requirement that several authorities and their informants be cited. However, the single strand does not mean that this was the only path of transmission by which the tradition circulated. That is, it does not mean that no one other than those persons named in the isn¥d knew of the tradition. Juynboll holds that the phenomenon of a single strand indicates that this is so. Indeed, he states: “As appears from this schematised bundle, each Companion mentioned at the beginning of an isn¥d strand is allegedly the only one who transmits that particular tradition from the Prophet and, what is more, he allegedly passes it on to only one pupil, a Successor, who himself has allegedly only one pupil, a younger Successor or a member of the generation following that of the Successors, who likewise has allegedly only one pupil etc.”9 Juynboll himself calls this interpretation “sheer unfathomable coincidence.” Probably, no one supports it. But the single strand means simply that the common link in the dissemination of the ^adÏth mentioned only one path of transmission. Other paths which the same matn might have taken have “died out” because they were not passed on by one of these first early collectors. In some instances, later compilers or the pupils of these early collectors sought further strands of transmission of a particular tradition that may have survived in oral tradition or with minor collectors. Where they were successful, one or more strands “dive” below the common link in the isn¥d bundle. This interpretation still allows us to postulate early collectors who mixed in traditions of their own with the genuine ones, adding fictitious isn¥ds, or later collectors who dived below the common link, again with fictitious isn¥ds. However, this possibility does not justify the conclusion that all single strands between the 9 Juynboll, “N¥fi¢, the Mawl¥ of Ibn ¢Umar, and his Position in Muslim €adÏth Literature,” 209-210 (my

emphasis).

6

Prophet and the common links are false and that the texts are the inventions of the common links. Nor should strands diving below the common link be summarily dismissed as fictitious. Yet this is precisely what Juynboll does when he claims that the common links “are to be held responsible not only for the matn of the tradition they brought into circulation, but also, in response to the compulsory authentication device, for the series of names linking their time with that of the first and oldest alleged authority.”10 His further statement that “Companions and mostly also Successors are themselves not responsible for their names cropping up in isn¥d strands” must be taken to mean that they have nothing to do with the tradition. In other words, Juynboll is suggesting that they have been falsely named as transmitters by the common link. For if they had really passed the tradition on to the common link, they would of course be “responsible” for their name appearing in the isn¥d. Juynboll’s analysis of the lower part of the isn¥d amounts to an a priori denial of its historical authenticity. In this he continues to follow Schacht who held this lower part of the isn¥d to be wholly fictitious. 11 This makes the isn¥d an invalid aid for the evaluation of the origin of the text and it’s possible authenticity beyond the common link. But does this conclusion necessarily follow? One cannot exclude the possibility and, indeed, the feasibility that a ^adÏth, transmitted by a common link, was received from the person named by that common link as his authority, whatever one may think of the remainder of the isn¥d.12 The historicity of the source cited by the common link cannot or can only very rarely be determined solely by analysis of an isn¥d bundle. Close examination of the various versions of the text and its claimed sources will prove more productive.13 But greater certainty about the reliability of a common link can only be acquired by scrutinising large bodies of text which allegedly go back to this common link.14 The hypothesis that the significant common links in the isn¥d bundles were the first systematic collectors and professional teachers of traditions explains why single strands are

10 Ibid. 210. 11 J. Schacht, The Origins of Muhammadan Jurisprudence, 171, 175. 12 It seems that Juynboll is willing to accept this, at least, in exceptional cases. Cf. his “Early Islamic Society

as Reflected in its Use of Isnads,” 182. 13 A good example for such an investigation is Josef van Ess’s study on the emergence of the ^adÏth traditions concerning predestination in his Zwischen €adÏt und Theologie (Berlin/New York: W. de Gruyter, 1975) although he does not cover the problem of ^adÏth forgeries sufficiently. On the issue of the history of traditions before the common link, cf. the study on some sÏra traditions of ¢Urwa b. al-Zubayr by Gregor Schoeler, Charakter und Authentie der muslimischen Überlieferung über das Leben Mohammeds; and H. Motzki , “The Prophet and the Cat” and idem “The Murder of Ibn AbÏ l-€uqayq.” 14 My book, Die Anfänge der islamischen Jurisprudenz. (Eng. trans. The Origins of Islamic Jurisprudence) may serve as a model for such an approach. In recent years, Juynboll has devoted much work to the systematic analysis of common links. In the light of his study on N¥fi¢ one cannot help but get the impression that he primarily interprets the isn¥ds and evaluates other information of the sources in the light of his findings taken from isn¥ds.

7

found below the common link and why the majority of common links are not at the level of the Prophet’s Companions, but belong to the three subsequent generations. B. Partial Common Links and Single Strands Looking at the upper part of the isn¥d bundle, Juynboll assumes that the (true) common link of an isn¥d bundle was the originator of the tradition in question. In contrast to the situation below the common link, the transmission fans out above it in a variety of routes that can be reconstructed from the later collections. This allows one to make verifiable statements about the dissemination of a ^adÏth from the common link onwards and possibly about the historical authenticity of the chains of transmission (isn¥ds) as well. In order to distinguish credible chains of transmission from false ones and to establish the genuine common link from which the text and the lower end of the chains of transmission derive, that is, in order to date the tradition, Juynboll has constructed the following methodological rule: The more persons there are who transmit a saying from a master to later generations, or the more persons there are who transmit something to someone, the more easily we can lend credence to that point of transmission as possibly historical. Conversely, this also means that the historicity of any strand peopled solely by Ful¥ns … is in view of this adage questionable. It will be obvious that this rule is also based on simple logic.15 A few lines later he adds: “It is...