Kitāb Sulaym ibn Qays: Report ten (First part)

In report 10 of the recent Anṣārī edition of Kitāb Sulaym, Imam ʿAlī is asked about the differences in transmission and Quranic interpretation found amongst the people.Footnote ¹⁸ In response, he lists the four (and only four) types of transmitter he has identified. These I have named the hypocrite, the presumptuous, the mistaken and the reliable, and in the report each is given a short description of their characteristics. These descriptions are significant for, as we shall see, they presume a certain level of hermeneutic awareness. Only the fourth (the reliable) can be trusted to transmit the reports; only this fourth type has mastery of the hermeneutic skills necessary to ensure accurate transmission. The first part of the report ends with a general summary of the hermeneutic categories within the speech of the Prophet, which are (so Imam ʿAlī claims) identical with those of God's speech (i.e. the Quran). The second part of the tenth report tells of ʿAlī’s relationship with the Prophet, and how he collected knowledge from him; and also affirms Sulaym's probity as a transmitter. The two parts were probably originally separate and have been brought together within the Kitāb Sulaym. The first part of the report constitutes for al-Sharīf al-Raḍī in the Nahj al-Balāgha, a separate integral unit which can be cited on its own merits and does not need the second part found in the Kitāb Sulaym version in order to make a doctrinal contribution.

The version of this report in the Kitāb Sulaym begins:

It could be argued, at the very outset, that the concern for understanding the Quran and the transmission of the Prophetic Sunna must stem from a time when these two had been established as the prime and indubitable pair of sources for legal deduction. The dating of this has been much debated, but central to their establishment as sources was, of course, Muḥammad b. Idrīs al-Shāfiʿī (d. 204/820). Without wishing to enter into that particular debate, the focus on Quran and Sunna within this report would at least push a likely context for its emergence to the latter half of the eighth century ce.Footnote ²²

Turning to the report's text in detail, the listings of category pairs are a common means of presenting the findings of hermeneutic reflection. Most of the pairings given here were taken up within the later hermeneutic tradition and given technical definitions: ṣidq/kidhb, nāsikh/mansūkh, ʿāmm/khāṣṣ and muḥkam/mutashābih.Footnote ²³ They are well known and regularly found located together (often with supplements, such as ẓāhir/bāṭin, ḥaqīqa/majāz and muṭlaq/muqayyad) in later tafsīr and uṣūl works. The muḥkam/mutashābih pairing is, of course Quranic (Q. 3:7);Footnote ²⁴ the notion of naskh is less explicitly (or easily) traced within the Quran;Footnote ²⁵ and while the other terms exist within the Quran, they do not appear as hermeneutic categories, either individually or in pairs. I would argue that the collocation of the categories here, as a list of pairings into which revelatory material can be placed, probably reflects a mature hermeneutic science, rather than any rudimentary exegetical theory of the first century ah. This apparent anachronism hints at the report being considerably later than the period of “Sulaym”. An examination of whether the terms (either individually or in pairs) are used in a manner congruent with later conceptions of (say) abrogation and particularization also indicate a point of formulation sometime after the turn of the second century ah (late eighth century ce).

The claim that the Prophet's message, like the understanding of the Quran, has been misunderstood, warped, or fabricated, is a common motif in Shīʿī argumentation, and (it could be argued) forms the basis of the Shīʿī position concerning the deviant direction taken by the early Muslim community. Even the Prophet himself knew about his message being changed by some of his supposed followers:

Here, the way in which one can avoid misrepresenting the Prophet's message is to be found through distinguishing between the various hermeneutic categories listed above.

ʿAlī lists four ḥadīth “transmitter types”: hypocrite, presumptuous, mistaken and reliable. One could see the whole system of transmitter categorization as fitting in with a developing science of ḥadīth criticism, something which only really became widespread after al-Shāfiʿī.Footnote ²⁶ Early lists of transmitter categories establish abstract qualities of the transmitter combined with an assessment of whether their ḥadīth can ever be trusted. Clearly the categorization scheme in Kitāb Sulaym is part of a Shīʿī polemic against their opponents, foregrounding the notion of a single true transmitter of the Prophet's Sunna. It becomes, then, another expression of the Shīʿī election motif and the total rejection of their opponents’ positions. For this polemic to be maximally effective, however, it presents itself as a twist on the existing categorization scheme, indicating again a ninth-century ce development.Footnote ²⁷

When we focus on the exegetical theory which the report assumes, the obvious qualities of a trustworthy memory are supplemented by other qualities. As will become clear, reliable transmitters should be able to understand and assess the significance of the material they are transmitting. It is in the description of these exegetical skills that the report reveals a hermeneutic science assumed in the background.

Imam ʿAlī begins his description of the various categories with the statement:

The description of the first category – the hypocrite – is almost entirely polemic, and clearly aimed at the opponents of the Shīʿa:

The passage tells us little about hermeneutics, but instead references a host of established Shīʿī themes: the rejection of the Companions of the Prophet as assumed reliable transmitters, the connection between these false witnesses and political power after the Prophet's death, and the notion of “the ones protected by God” (the Imams and perhaps their Shīʿa also) who are set apart from “the people” (al-nās – perhaps a reference to the non-Shīʿa though this is not the only possible interpretation). The target need not necessarily be the Umayyads, though they would appear the most likely immediate candidate: the repudiation could, though, be general and apply to all those who distort the Prophet's message. It is sentiments such as these which, most probably, persuade Modarressi to classify the Kitāb Sulaym as an “anti-Umayyad polemic”.Footnote ³⁰

The second type of transmitter is described thus:

Here, the transmitter fails to memorize the report exactly (ʿalā wajhihi – “as it is” or perhaps “as it should be memorized”), but instead presumes he knows what it means and therefore presents his own version of its meaning (his own wahm). He does this in an unthinking manner, unlike the hypocrite; and if he had realized he was presenting it inaccurately he would have rejected it, as would the Muslims if they had known. The implication here is that there is no blame attached to him, but his report is to be rejected. There is a preference for precise verbal transmission (rather than transmission of the assumed meaning of the Prophet's statement), but it is possible that someone who can accurately understand the meaning of the report (i.e. without inserting his own wahm into the transmission) may be able to transmit it in an acceptable form. It is possible, from the above description, that a meaning-related transmission which compromises the precise wording of the report could still be a transmission “ʿalā wajhihi ”. The presentation prefigures, of course, the pairing of lafẓī/maʿnawī (verbal/meaning-based) transmission, though the terminology is not used. I return to this distinction below.

It is in the discussion of the third transmitter type that abrogation comes to the fore. This transmitter, it appears, does not take note of the occurrence of abrogation (naskh):

Exactly what the transmitter does not know is not clear, but from the context, it seems most likely that he does not know of the Prophet's later order (or prohibition) and hears only the first. It is possible that what is meant here is a more radical misunderstanding of the process of naskh – namely that he hears both order and the prohibition but takes the first because he does not understand the procedure of abrogation, preserving the mansūkh rather than the nāsikh. If he had known it was mansūkh (i.e. if he had known how naskh operates), he would have rejected it. In either case, there is a clear assumption of a developed notion of abrogation: a later ruling abrogates an earlier ruling; the earlier ruling is rejected (rafaḍa). Rejected is, perhaps, stronger than al-Shāfiʿī’s “left/abandoned” (taraka), and may reflect a polemic edge to the rejected ruling.Footnote ³² There is no indication here that the Sunna could abrogate the Quran (or vice versa), and perhaps an implicit rejection of this mode of naskh: the Sunna is portrayed as only abrogating the Sunna. This view (that there is no “inter-source” abrogation), as is well-known, was advocated by al-Shāfiʿī, and was apparently viewed as novel when he proposed it.Footnote ³³ There is no explicit reference to one distinction which became central to later theories of naskh: that between text and ruling. In later writings of legal theory (uṣūl al-fiqh) the abrogation of the ruling but not the text was entertained as a possible mode of naskh. This idea was not fully exploited by al-Shāfiʿī, though there are indications that he accepted the possible naskh of the text, but not of the ruling.Footnote ³⁴ In Kitāb Sulaym the text and ruling appear bound up and, if identified as mansūkh, are to be rejected together. For the Muslims who hear the report, on the other hand, it is not so clear whether it is the ruling they are rejecting or the report containing it. The phrase concerning their possible reaction (“If the Muslims had known it was mansūkh…”) is ambiguous.Footnote ³⁵ The distinction does not appear as a universally assumed element of naskh in the presentations of either al-Shāfiʿī or Kitāb Sulaym, perhaps indicating they are working with the same, proto-hermeneutic assumptions. In sum, then, the discourse of the naskh process described in this report in the Kitāb Sulaym reflects the debate as it stood at the time of al-Shāfiʿī, sharing some (though not all) of the emerging technical terminology of Muslim hermeneutics.

The last transmitter type, whose reports alone are to be trusted, is one who memorizes (or preserves) a saying exactly “as it should be” (ḥafiẓa mā samiʿa ʿalā wajhihi), and does not add to or omit anything from what he has heard:

The lafẓī/maʿnawī distinction is, once again, the most obvious reference here, though the terminology is not used. It is possible that ʿalā wajhihi (“as it is”, “as it should be”) does not mean precise verbatim transmission, but “in a way that perfectly preserves the meaning” or some such locution. However, the rest of the description, including the phrase “he brings it just as he heard it; he does not add anything, nor does he take anything away”, would indicate, as in the description of the second type, a preference for lafẓī transmission. As already mentioned, the requirement to transmit verbatim rather than by meaning seems to have been a minority position at the time of al-Shāfiʿī. Al-Shāfiʿī has some concerns about it, but he does not dwell on it at length. Indeed, the debate expressed throughout his al-Risala around the ability of reports of limited transmission (aḥād) to act as proof is not so much around the wording/meaning debate (there is an implicit acceptance of wording variants being non-fatal to the epistemological chances of a report). Rather it concerns an acceptance that reports of limited transmission may fail to reach one type of certainty, but have sufficient probative force to be used as legal sources.Footnote ³⁹ That absolute certainty as to precise wording was not a requirement for utility in the later ḥadīth sciences is well knownFootnote ⁴⁰ and the lafẓī/maʿnawī distinction appears to have been fully discussed (and occasionally problematized) only later in the development of that science.Footnote ⁴¹ The lafẓī/maʿnawī discussion is not entirely absent from al-Shāfiʿī’s discussion, however, though the discussion is naturally rather rudimentary compared to the sophistication of the later tradition. For example, al-Shāfiʿī expresses some concern that a transmitter summarizing a report might change meaning:

The categories of lafẓ and maʿnā, though, are not fully worked out in his al-Risāla, and in other places al-Shāfiʿī indicates that transmission which retains context, even if it is non-verbatim, is preferable to accurate verbal transmission which ignores context and thereby distorts meaning.Footnote ⁴³

The insistence on lafẓī transmission in the Kitāb Sulaym appears, then, as an implicit criticism of (and reaction to) the opinion of the majority who allow maʿnawī transmission. However, the report does not contain the technical terminology which later became standard, and this may indicate the report predates the wholesale incorporation of the terminology (and perhaps also the concepts) of lafẓ and maʿnā into the analysis of juristic thought. Once again, a late eighth or early ninth century ce context is the most natural one for explaining the hermeneutic assumptions within the report.

This fourth reliable transmitter type also has other characteristics: he knows the abrogating and the abrogated, and acts on the former, discarding the latter. This is the characteristic which is lacking in the second transmitter type (referred to and discussed above). When analysing the ideal type of transmitter being constructed here, the individual is portrayed as knowing not only that a later ruling abrogates an earlier ruling (i.e. he is familiar with the theory of naskh); he also knows how to identify which reports abrogate and which are abrogated, and he preserves only the nāsikh. This skill is not simply a matter of knowing the relative dates of the reports; it requires the transmitter to recognize that the two reports cannot possibly be reconciled and are in direct contradiction. For naskh to be operative, the two orders under examination need to address precisely the same legal subject such that accepting one means the rejection of the other. To carry out this identification process requires a certain level of hermeneutic skill, and it was a lack of this skill (rather than a faulty memory or any mendacity) which made the second transmitter type unacceptable. It is, then, presented as a decisive criterion for the acceptability of a transmitter, and establishes that his transmissions are acceptable in contrast to the second and third types.

Furthermore, the phrase “He preserves the abrogating from [or ‘instead of’, or ‘out of’] the abrogated”, (ḥafiẓa al-nāsikh min al-mansūkh) implies that the nāsikh is preserved and the mansūkh is not.Footnote ⁴⁴ The final phrase of this description maintains that the mansūkh is to be rejected/discarded (wa-rafaḍa al-mansūkh). The idea that there is no need to preserve the abrogated Sunna is not tackled in later uṣūl with quite the same precision as that relating to the abrogated Quran. The abrogated elements of the Quran, of course, have ritual importance as elements of God's revelation, but have no legal relevance. With regard to the Sunna, the discussion was probably not as pressing since very little of the Sunna was (in later terminology) qaṭʿī al-wurūd (“of certain provenance”).Footnote ⁴⁵ The transmitter here does not even bother to remember the abrogated, because the purpose of the transmitter is not to record everything the Prophet said, but only to record that which is legally relevant. The transmitter, with his knowledge of naskh, becomes the gatekeeper to the Sunna in that he knows what to preserve and what to reject. He has the qualities, then, of a jurist rather than a simple transmitter. As with the lafẓī/maʿnawī distinction, the report rejects a mainstream position – in this case, it was the division between jurists and ḥadīth transmitters which was becoming more widely recognized in the late eighth century.Footnote ⁴⁶ The report is clearly participating in a debate which emerged much later then the seventh century ce.

In the closing paragraph of this first section of the report, three scriptural category pairings are said, by ʿAlī, to apply to both elements of revelation, the Quran and the Sunna: nāsikh/mansūkh, ʿāmm/khāṣṣ and muḥkam/mutashābih. There is no engagement with the thorny issue of whether the Quran can abrogate the Sunna referenced above – or the equally widely discussed idea that the two sources can particularize each other (which al-Shāfiʿī discussed in detail in al-Risāla). Nevertheless, the pairings and an established theory (though not fully expressed) underlying them are implied here. Overall, there is a recognition that these categories need to be employed in order to understand what God and the Prophet mean in their statements. The ʿāmm/khāṣṣ distinction receives particular treatment:

From this rather scant description, identifying whether a statement of the Prophet or God is general or particular is central to recognizing the intended meaning. If one is unable to recognize this (and simply “hears” the speech), then the meaning drawn from the statement will not be the same as the intended meaning.Footnote ⁵⁰ The hermeneutic apparatus to make sense of this statement is relatively developed: there is a meaning which a text has in and of itself (its literal meaning, for want of a better term), and this literal meaning may be different from the intended meaning; to gain access to the former (“literal” meaning), one needs a knowledge of the language (this is what those who simply hear understand by the statement); to gain access to the latter one needs an interpretative skill: the ability to distinguish the ʿāmm from the khāṣṣ. The passage ends with a dig at some of the companions (a recurrent Shīʿī theme) who did not always understand what the Prophet was saying:

The first part of the statement is a simple rejection of the innate ability of the companions to understand the Prophet's message, and represents a clear element of Shīʿī polemic in the text. The second part, referencing the visit of the Bedouin or stranger, is possibly a continuation of this polemic tone, as the Companions could not understand the Prophet's words, and perhaps were too embarrassed to ask for clarification. Consequently, they were pleased when someone else asked, and they avoided revealing their ignorance. It could be interpreted as exhibiting the belief amongst the early grammarians that the language of scripture (ḥadīth included) was to be understood as an instantiation of the perfect Arabic language, and hence ambiguities and difficulties can be solved through reference to this linguistic corpus. The companions liked it when a bedouin came to ask the Prophet questions, as they would be able to hear him converse with the Prophet and perhaps later ask him what the Prophet meant. The bedouin and the Prophet shared the same linguistic code. The rather dim-witted companions were not blessed with total competence in this ability. An alternative interpretation might be that the bedouin lacked the timidity of the companions when faced with a prophetic statement they did not understand. Under either interpretation, however, the portrayal of the companions is generally negative. While these assumptions became standard grammatical doctrine in later years, the linguistic self-awareness required to carry out such an analysis was really only developing very late in the Umayyad period, and reached the level of maturity displayed in this passage from Kitāb Sulaym only in the early ʿAbbāsid period.Footnote ⁵²

It is of course possible that the description here is a veiled polemic. Perhaps, only the Imam fulfils the criteria of this fourth transmitter type, and the Imam (according to some Shīʿī doctrines) has complete knowledge of the Sunna. The Imam, then, can discard elements of the Sunna without the fear that his judgement as to them being abrogated is potentially incorrect. I would consider such a reading as possible, but unlikely to have been part of the original formulation, as the text appears to be exhorting Sulaym to be choosy about whose reports he selects. If this meant only those transmitted through ʿAlī, one would have expected a more straightforward way of expressing this notion. Furthermore, the emergence of the Imam as the ultimate arbiter of the correct interpretation of the Prophet's Sunna took some time to develop in Shīʿī (particularly Imāmī) thought, and really only became established as mainstream in the late ninth century.

If the basic message of the report is that only the Imams can transmit the Prophet's Sunna since only they have knowledge of these categories, then ʿAlī is, effectively, putting himself precisely in this role of a companion who always understood what the Prophet meant. ʿAlī could distinguish the general from the particular and the abrogating from the abrogated, and perhaps the implication here is that the other companions could not. It would seem that ʿAlī (and by implication perhaps the Imams after him) are to be viewed as the archetypal perfect transmitters of the Prophet's Sunna. But even if this is so, there is no indication in the report that the ability to transmit perfectly is exclusively held by the Imams. Indeed reliable transmission seems to consist of (first) a technical (and hence learnable) skill (to remember words as they are spoken), and (second) an intellectual ability (knowledge of the Arabic language and being able to distinguish the general from the specific). The strictly behavioural criterion (i.e. moral probity or ʿadāla), which became a central element in ʿilm al-rijāl, is only hinted at here (viz. the transmitter acts on the basis of the abrogating and not the abrogated, and hence he acts in conformity with the law). There is no indication in the report that some sort of divine knowledge, exclusively available to the chosen Imams, is necessary to transmit the Sunna reliably. As I have argued elsewhere, the need to present the Imam's hermeneutical tools as available for the general community of Shīʿī exegetes is a relatively late development, perhaps coinciding with the period of lesser occultation.Footnote ⁵³

In sum, then, the reliable transmitter in this report is more than simply someone who can memorize the words he hears faultlessly: he has exegetical skills, such as his ability to recognize the processes of abrogation and particularization occurring. Indeed, this ability is what distinguishes him from the other transmitter types, who either mendaciously warp the Prophet's words, or simply fail to understand them. In the report, the ideal transmitter is also an accomplished exegete. The establishment of an intellectual justification for community members other than the Imams being authoritative exegetes is traditionally viewed as a ghayba development (i.e. late ninth, early tenth century). If, as I have argued, this element of the report is best dated as emerging in the late eighth/early ninth century (based on its level of hermeneutic sophistication), then the report becomes indirect evidence for the attempted establishment of an alternative religious authority source (what might be called the proto-ʿulamāʾ amongst the Shīʿa) in the period immediately following the Ismaili schism.Footnote ⁵⁴ From the perspective represented by this report, the Imams certainly have these exegetical skills, but they do not necessarily have exclusive claim over them; and if others can acquire them, then they cannot just transmit the Sunna but must also offer its correct interpretation. The tension between scholarly authority and that of the Imam, which became a prevalent theme in later Shīʿī jurisprudence and theology, is nascent (perhaps latent) here in the report in the Kitāb Sulaym b. Qays.

The content of the first section of the tenth report appears, then, as a rather audacious attempt to attribute to ʿAlī knowledge and mastery of exegetical techniques and a level of hermeneutic sophistication which came into existence in the late eighth/early ninth century. Having said that, there are points in the text where the fit between the use of technical terminology and concepts within later Muslim hermeneutic understanding and those found in the report is not perfect. This perhaps indicates that the appropriate context in which to view the report is the early formative period of hermeneutic thinking in the Muslim religious sciences (namely the late eighth and early ninth century ce), rather than the fully flourished theoretical awareness one finds in tenth-century works of tafsīr and uṣūl al-fiqh.

The report's isnād

Isnād analysis is an exacting, and at times a rather inconclusive exercise. Nonetheless, it is worth examining whether anything relevant to the dating of the report can be learned from the isnāds available for this report. If one uses every available isnād from every available source, then one has a total of fourteen isnāds attached to segments of variants of report 10 in the Kitāb Sulaym b. Qays. The various isnāds of the report are shown in Figure 1.Footnote ⁵⁵

Figure 1.

The isnāds of the report in the Kitāb Sulaym b. Qays

Following the methodology of Juynboll (itself a disputed approach to isnād analysis),Footnote ⁵⁶ it would appear that Ibn Udhayna (death date not known, but a companion of either or both Imams al-Ṣādiq (d. 148/765) and Kāẓim (d. 182/798)) is the common link, which would fit with the account of his reception and then distribution of the Kitāb Sulaym from Abān (d. c. 138/755–6) outlined in the introduction. If we are to credit Ibn Udhayna with the report's circulation, this would be some forty to fifty years earlier than the dating, arrived at in the above analysis, based on the report's matn. The two isnāds which branch off from Abān (found in collections by al-Karājikī (d. 449/1057) and al-Ḥaṣkafī (d. 551/1156–57)) should probably be seen as “dives” (pace Juynboll), since Ibn Udhayna is considered “weak”.Footnote ⁵⁷ Ḥammād b. ʿĪsā (d. 209/824–5 or 208/823–4) possibly acts as a secondary common link: he provided al-Faḍl b. Shādhān, al-Kulaynī (d. 329/941) and Ibn Bābawayh (d. 381/991) with the embellishments noted in the footnotes above. The problem with such an analysis is that al-Faḍl does not count Ibn Udhayna as one of his transmitters, preferring to go with the less popular, but rival third link of Ibrāhīm b. ʿUmar al-Yamānī (who, according to many Shīʿī isnāds relates Imams al-Ṣādiq (d. 148/765) and Kāẓim (d. 182/798), as well as from Sulaym directly). Ḥammād b. ʿĪsā could be the actual common link, and Ibn Udhayna is an apparent common link, in that he provides us with the earliest three versions of the report (al-Faḍl b. Shādhān, al-Kulaynī and Ibn Bābawayh) and it is only later that Ibn Udhayna is made to spawn the versions of al-Ḥaṣkafī, al-Nuʿmānī (d. 360/971) and al-Kashshī (d. c. 340/951). A further complication is al-Ṣaffār's (d. 290/903) isnād found in Baṣāʾir al-Darajāt, a work which predates both al-Kulaynī and Ibn Bābawayh. Though here, it could be argued that his citation is of only a few lines from the middle section of the report and not the report as a whole: the citation is partial and does not relate to the hermeneutic first section discussed above, and so can be disregarded in isnād analysis. The real difficulty is the version found in the Nahj al-Balāgha which, as already mentioned, without an isnād, bears a resemblance to the other versions examined here, but shows extensive rewording and textual adjustment in order to ensure it conforms more obviously with later notions of ʿāmm/khāṣṣ and nāsikh/mansūkh.

In the end, though, I am not sure the isnād analysis facilitates even an approximate dating. It could, perhaps, provide some indication of the original circulation of the report by Ḥammād b. ʿĪsā, who was abandoned by a later generation of ḥadīth specialists in favour of ʿUmar ibn Udhayna as the primary source of the report. If Ḥammād is responsible, then his dates (i.e. d. 209/824–5 or 208/823–4) would fit with the matn analysis carried out above. But the result does seem, to me at least, rather speculative.