The Four Books of Shiʿi Hadith: From Inception to Consolidation

Since their compilations in the tenth and eleventh centuries ce, the four hadith books, al-Ka﻿̄fī , al-Faqīh , al-Tahdhīb , and al-Istibṣa﻿̄r , have left an indelible mark on Shiʿi religiosity. The present study takes as its starting point the earliest instance in which these four compilations were collectively referred to as the Four Books ( al-kutub al-arba ʿ a ). I investigate the major developments in the period between the inception of this phrase in the fifteenth century and its consolidation as the demarcator of a unique Imami hadith corpus in the seventeenth century. Following the introduction, each section of the article focuses on a figure whose ideas contributed to this consolidation process. In the conclusion I summarize the findings of the previous sections and reflect on the notion of hadith canonicity within the context of Imami jurisprudence during the period under study.


Introduction
When Shiʿi scholars are asked, what are the most authoritative hadith books of their tradition, nearly all respond, the Four Books (al-kutub al-arbaʿa). In Islamic Law and Society (2021) 1-55 providing this answer they follow a tradition stretching back at least five centuries.1 Compendia of statements attributed to the Prophet and the imams, these four books contain teachings on a variety of subjects, albeit with a primary focus on legal precepts. into three parts. The first part contains 3,786 hadiths pertaining to doctrine (uṣūl, lit. "roots"); the second has 11,819 hadiths on law (furūʿ, lit. "branches"); the third, often referred to as al-rawḍa (lit. "garden"), has 597 hadiths on miscellaneous topics that do not directly pertain to those addressed in the first two parts.6 Al-Kāfī is the earliest extant major compilation of Imami hadith.7 The second of the Four Books, Man lā yaḥḍuruhu l-faqīh (henceforth, al-Faqīh), was compiled by Muḥammad ibn Bābawayh (d. 381/991) and completed around 366/978, thirty-seven years after al-Kulaynī's death.8 Whereas al-Kāfī contains hadiths pertaining to both creed and law, al-Faqīh is exclusively dedicated to law.9 As the book's title and preface suggest, Ibn Bābawayh intended his book of 5,963 hadiths to serve as a self-help guide for those without access to a jurist on whose rulings they could rely.10 The remaining two of the Four Books, Tahdhīb al-aḥkām and al-Istibṣār fī-mā ukhtulifa min al-akhbār, were compiled by Muḥammad ibn Ḥasan al-Ṭūsī (d. 460/1067).11 Born in Khorasan in 385/995, at the age of twenty-three he moved to Baghdad where he would remain for the next forty years;12 it was during this period that al-Tahdhīb and al-Istibṣār were compiled. Like al-Faqīh, both books address only legal topics. Al-Tahdhīb contains 13,590 hadiths and is both longer and more comprehensive than either al-Faqīh or the legal part of al-Kāfī. In al-Istibṣār, al-Ṭūsī selects 5,511 contradictory or conflicting hadiths from his Tahdhīb and attempts to bring them into concord.13 It is to be noted that these four compilations have a great number of hadiths in common.14 These four books -al-Kāfī, al-Faqīh, al-Tahdhīb, and al-Istibṣār -were compiled and widely used long before their consolidation as the Four Books (al-kutub al-arbaʿa). The present study, however, takes as its starting point the earliest instance in which the phrase al-kutub al-arbaʿa enters Imami hadith discourse. It investigates the major developments in the period between the inception of this phrase in the fifteenth century and its consolidation as the demarcator of a unique hadith corpus in the seventeenth century. To this end, I will analyze five thinkers from this period whose views had a decisive impact on the consolidation process. Among the questions I address are: when and in what contexts do the earliest extant references to the Four Books occur? What features distinguish the Four Books from other Imami compilations? What theoretical principles inform each of these five thinkers' methodologies of hadith verification? In what ways are their respective assessments of the reliability of the hadiths in the Four Books informed by each thinker's understanding of the historical trajectory of Imami hadith from its earliest textual manifestation to his time? A key and recurring issue in this article pertains to the contrasting justifications provided for establishing the privileged status of the Four Books and the different methods proposed for verifying their hadiths.

Ibn Abī Jumhūr and the Four Sources
In 888/1483, around the age of fifty, Ibn Abī Jumhūr (fl. 906/1501) completed a concise book on legal reasoning entitled Kāshifat al-ḥāl ʿan aḥwāl al-istidlāl. 15 In the preface, he states that Kāshifat al-ḥāl was written because of a pressing request of the author's friend. Ibn Abī Jumhūr notes that he was asked to compose an epistle outlining the essential elements involved in the process of legal reasoning in regard to one's religious obligations. 16 The purpose of this reasoning, he writes, is to arrive at the knowledge of the precepts that believers must observe.17 According to Ibn Abī Jumhūr, legal reasoning involves knowledge of three kinds of sciences: linguistic, rational, scriptural. The science of language is comprised of the disciplines of lexicography, morphology, and syntax; the rational sciences of logic, theology, and jurisprudence; and the scriptural sciences of exegesis, hadith, and rijāl.18 He next outlines the knowledge required in each of these disciplines. Ibn Abī Jumhūr reassures the readers that in order to engage in legal reasoning they need not possess the expertise of Sibawayh in syntax; 14 On the selection and organization of the hadiths included in each of these four compilations -with a focus on their utilization as material sources of law -see Gleave, "Between Ḥadīth and Fiqh". Ibn Abī Jumhūr conceptualizes istidlāl as the process through which rulings concerning the cases for which no explicit evidence is found in the Qurʾan, nor are they expected to reach, like Avicenna, the summit of logic.19 To ease such concerns, he provides a list of recommended readings for each of the nine above-mentioned disciplines.20 In his view, this suggested reading list provides the uninitiated with a sufficient grounding in the three sciences involved in legal reasoning. Regarding the discipline of hadith, Ibn Abī Jumhūr writes: It is not incumbent upon an investigator to know those hadiths by heart but rather it will suffice him to consult some of the authenticated sources (al-uṣūl al-muṣaḥḥaḥa) and to be capable of accessing those hadiths as they are recorded in one of the esteemed books (al-kutub al-muʿtabara). They say, it is sufficient to refer to only one of the Four Sources, be it In support of his recommended reading list, Ibn Abī Jumhūr invokes the authority of scholars, stating that "they say" it is enough to consult any one of these four books. 22 He does not specify which of his contemporaries or predecessors made this recommendation and whether they also used the phrase the 'four sources' (al-uṣūl al-arbaʿa) to demarcate the sufficient books of Imami hadith from other available works. 23 He does allude to a passage in one of his recommended rijāl books -Khulāṣat al-aqwāl -in which al-Ḥillī (d. 726/1325) refers to three of the four books collectively as the "three books" (al-kutub al-thalātha). Ibn Abī Jumhūr draws attention to the final section of the Khulāṣa where al-Ḥillī documents the transmission path through which Sunna, and consensus are inferred based on the general principles outlined by the Prophet and the imams (see Kāshifat al-ḥāl, 55-57). 17 Ibn Abī Jumhūr, Kāshifat al-ḥāl, 53. 18 The discipline of rijāl is concerned with determining the credibility of hadith narrators. 19 Ibid., 78, 80, 82. 20 See the Appendix for a list of Ibn Abī Jumhūr's recommendations in each of these disciplines. 21 Ibid., 89-92. All English translations in this article are my own. 22 Elsewhere in Kāshifat al-ḥāl (p. 133), Ibn Abī Jumhūr reiterates that it is sufficient for those who seek to derive rulings to turn to these four books as their sources of hadith. 23  The early scholars collected the hadiths from dispersed places, unearthed them from scattered sources and -having explicated their transmission methods, described their transmission chains, and mentioned the names of those who transmitted the hadiths from the imams -recorded them in compilations, such that nothing is excluded from them except for a negli- In his view, as a result of the efforts of early scholars, no tasks remain for their successors other than preserving what they established and studying what they compiled.28 In the context of hadith transmission, the term 'early' scholars (mutaqaddimūn) often refers to those who were contemporary with the imams; they are contrasted with 'later' scholars (mutaʾakhkhirūn) who lived in the period after the twelfth Imam went into Occultation (ghayba) in 260/874.29 Ibn Abī Jumhūr adds a caveat to his assertion regarding the sufficiency of the Four Sources. In his view, the mere presence of a given hadith in any of these four compilations does not automatically establish its authenticity. This is why, in his view, without sufficient knowledge of the science of rijāl, legal reasoning should not be undertaken. He remarks that the science of hadith would not be on a straight course without the science of rijāl, since it is through the latter that authentic hadiths are distinguished from the rest.30 As a result, he writes, those who engage in legal reasoning have the utmost need of this science.31 Hence, he instructs the reader to consult the Four Sources side by side with his recommended list of books on rijāl.32 Although he considers these four compilations as sufficient sources of hadith, no hadith should be accepted as reliable without a verification process.
Ibn Abī Jumhūr outlines this verification process in Kāshifat al-ḥāl and in his hadith compilation Ghawālī al-laʾālī. He divides all hadiths into three 27 Ibn Abī Jumhūr, Kāshifat al-ḥāl, 136-137. 28 Ibid., 137. In the introduction to his own hadith compilation, Ghawālī al-laʾālī, Ibn Abī Jumhūr uses the phrases "dispersed places" and "scattered books" to explain why he set out to compile this book (Ghawālī, 1:3). Whereas in Kāshifat al-ḥāl he refers to "scattered sources" (uṣūl mutafarriqa), in Ghawālī he uses "scattered books" (kutub mutafarriqa), another indication that he uses uṣūl and kutub interchangeably. groups. First, those that yield knowledge in their recipients and hence remove the need for inquiring about the credibility of their narrators. These hadiths are called mutawātir; their widespread and concurrent transmissions eliminate the possibility of forgery and collusion by reporters. Second, those hadiths that have been transmitted in continuous chains by more than three Imami and upright narrators, are renowned and widely quoted in the community, and extensively used by scholars as evidence for their rulings. These hadiths are called prevalent (mashhūr). He explains that although prevalent hadiths do not have the same epistemic status as mutawātir hadiths, they fall only slightly short of yielding knowledge in the recipients, leaving them with a preponderant conjecture (ẓann). The third group of hadiths are called aḥād. These are hadiths transmitted by less than four narrators and leave their recipients with mere conjecture regarding their veracity.33 According to Ibn Abī Jumhūr, aḥād hadiths are themselves divided into three groups. Those with an unbroken chain of transmission beginning from the last narrator all the way back to the Prophet or an imam are called musnad (connected); those with a missing link in the transmission chain are called maqṭūʿ (disconnected); and those in which the first narrator appearing in the transmission chain quotes the Prophet or an imam without having personally heard the hadith and without naming the intermediary person through whom the hadith was initially reported to him are called mursal (conveyed).34 Ibn Abī Jumhūr then introduces a further division according to which -in light of the available rijāl information concerning the narrators present in the transmission chain -any given musnad hadith belongs to one of the following four categories: (1) authentic (ṣaḥīḥ): hadiths transmitted only by upright (ʿadl) Imami narrators;35 (2) sound (ḥasan): hadiths transmitted by Imami narrators whose character has been praised (mamdūḥ) even though their uprightness (ʿadāla) as narrators has not been explicitly recorded; (3) credible (muwaththaq): hadiths transmitted by upright non-Imami narrators, Shiʿi or otherwise; and (4) weak (ḍaʿīf): hadiths that do not fall under any of the above three categories (e.g., when any of the narrators appearing in the chain is accused of forgery appearing in the chain are considered upright and one is recorded as praiseworthy, the hadith would be classified as sound rather than authentic. Ibn Abī Jumhūr notes that the Imami hadith compilers did not classify their hadiths according to this fourfold division. In his view, the compilers' chief concern was to record the hadiths attributed to the imams along with the transmission chains through which they received them. Hence, he argues, the readers of these compilations must identify the narrators mentioned in each transmission chain and examine their status in the rijāl books in order to determine to which of the four categories a given hadith belongs. He notes that such an investigation is difficult and laborious; one can dispense with it, however, by reading several works composed by prominent scholars who have already probed the extant hadiths found in these compilations and evaluated each one's degree of reliability.37 Having presented the fourfold division of hadiths, Ibn Abī Jumhūr points to disagreements among Imami scholars regarding the legitimacy of relying on 44 Non-Imami narrators are not downgraded because they are thought to be incapable of being upright individuals and credible transmitters, but because the imams may have been compelled to practice taqiyya (precautionary dissimulation) in their presence. Due to violent opposition and systematic persecution, the imams at times practiced taqiyya and refrained from expressing their views in public as a means of shielding their followers from persecution. The necessity of dissimulation impelled the imams to modify their statements in ways that would not endanger their followers. Sometimes dissimulation required the imams to express the opposite of their actual views. This was done in order to indicate an apparent agreement between Shiʿi and Sunni positions, when dissenting views would set Shiʿis apart from the rest of Muslims and invite opposition and conflict between them. In this context, if a non-Imami, for example, asked a question and received an answer from the imam, the imam may have practiced taqiyya when answering the question. As a result, in cases in which a non-Imami narrator is found in a hadith's transmission chain, one cannot know for certain whether the hadith expresses the imam's genuine view or whether his statement was uttered under taqiyya. In such cases, a non-Imami transmitter might accurately narrate the imam's statement to others but his hadith will not be treated the same as if an Imami narrator had transmitted it. On taqiyya, see Kohlberg, In Praise of the Few, ch. 15.
As Ibn Abī Jumhūr explains in the preface, Kāshifat al-ḥāl is a sufficient but not a comprehensive guide to legal methodology. Consequently, he does not elaborate on his verification approach to the Four Sources. Systematic discussions on verification methods are often found in works of hadith criticism (dirāyat al-ḥadīth).45 Although no text on hadith criticism is found among his extant writings, a trace of Ibn Abī Jumhūr survives in the earliest extant Imami book on hadith criticism, namely, al-Bidāya fī ʿilm al-dirāya of Zayn al-Dīn al-ʿĀmilī.46 Written seven decades after Kāshifat al-ḥāl, it presents a more extensive discussion of the Four Sources. The significance of Ibn Abī Jumhūr's collective reference to al-Kāfī, al-Faqīh, al-Tahdhīb, and al-Istibṣār as the Four Sources, along with his assertion that they stand as sufficient hadith books to be relied upon for the purposes of legal reasoning, was duly noted by his readers. Al-Shahīd ii intended for al-Bidāya and al-Riʿāya to serve as summary introductions to hadith criticism.53 Al-Bidāya is terse, and the author notes in his autocommentary that the text's succinct length is designed to facilitate its memorization and increase its benefits, adding that, "people's dispositions in this era do not bear the many burdens of knowledge, especially concerning this topic".54 In other writings, al-Shahīd ii faults his contemporaries, Persian scholars in particular, for spending their lives in pursuit of philosophy and logic instead of dedicating themselves to acquiring "religious knowledge", noting that they adhere to the Prophet's religion at the same time that they strive to revive the religion of Aristotle.55 In his view, their neglect of the transmitted sciences has led to the scarcity or extinction of the books of hadith and rijāl even though, he writes, these two are the foundations of the sharīʿa.56

Al-Shahid ii and the Four Books
were from that region. To avoid confusion and clutter, I have dropped al-ʿĀmilī after its first appearance for each of the authors surveyed in the article and instead refer to them by another title or name associated with them. For a biographical dictionary of ʿĀmilī scholars, see al-Ḥurr al-ʿĀmilī's Amal al-āmil. indicate that philosophy and logic were not al-Shahīd ii's forte; at times, a lack of substantive intellectual engagement with those he criticizes gives his criticisms of philosophers a tenor Al-Shahīd ii laments that in his time an "eminent" scholar might reach the end of his life without "having even looked at one of the books of hadith such as al-Kāfī, al-Tahdhīb, al-Faqīh, and others". He remarks that the fault does not rest merely with his contemporaries since their teachers and the teachers of their teachers had the same negligent attitude towards hadith books. According to al-Shahīd ii, because of this generational neglect, even if a person is aware of the problem and is determined to pursue these hadith books, he can hardly do so, since copies of these books have become scarce.57 Al-Shahīd ii's critical appraisal of his contemporaries is accompanied by his effort to remedy what he perceives as their disregard for hadith by writing three works on hadith criticism.58 He defines hadith criticism (dirāyat al-ḥadīth) as the science of reflecting on the transmission and content of hadiths in order to determine which ones are to be accepted and used as a basis for legal action.59 Having emphasized the centrality of hadith and the sciences of rijāl and dirāya, he proceeds to discuss the extant corpus of Imami hadith and what he believes constitutes its representative sources, the Four Books (al-kutub al-arbaʿa).
In addition to being the author of the first extant Imami handbook of hadith criticism, al-Shahīd ii is also the first scholar, thus far identified, to use the phrase 'the four books' (al-kutub al-arbaʿa) to refer to al-Kāfī, al-Faqīh, of disdain and turns his lamentation of their disregard for hadith and rijāl into contempt. In his view, his counterparts lack religion, have little piety, and are of weak resolve (ibid., 49). Elsewhere, he declares that it is religiously prohibited to engage with certain forms of philosophy. See via free access al-Tahdhīb, and al-Istibṣār.60 As noted, Ibn Abī Jumhūr was the first person, thus far identified, to refer to these four compilations collectively as 'the four sources' (al-uṣūl al-arbaʿa) of Shiʿi hadith.61 Al-Shahīd ii had read Ibn Abī Jumhūr's Kāshifat al-ḥāl.62 His elaboration of the genesis and authority of the Four Books, however, is more detailed than that of his precursor.
In al-Riʿāya, al-Shahīd ii prefaces his discussion on Imami hadith sources by cautioning against drawing rigid boundaries for the hadith corpus. In his view, limiting the corpus by declaring that it contains a specific number of hadiths -as he claims Ibn Ḥanbal had done -is untenable since hadiths were initially transmitted orally; this leaves open the possibility for certain hadiths to have been lost or not transmitted to everyone.63 This possibility, he adds, is greater for Shiʿis since numerous individuals narrated hadiths from the imams. He then addresses the transformation of hadiths from oral to textual sources. The hadith sources of the early scholars consisted of 400 compositions (muṣannafāt) by 400 composers. They called these compositions 'sources' (uṣūl) and upon them they relied. Eventually, however, circumstances led to the loss of the majority of these [400] sources, and then a group of scholars gathered them [i.e., the remaining sources] into specialized books (kutub) in order to make them accessible for the seeker.64 From these, the best collections are al-Kāfī by Muḥammad ibn Yaʿqūb al-Kulaynī and al-Tahdhīb by al-Shaykh Abū Jaʿfar al-Ṭūsī. Neither of these two books is sufficient apart from the other: the first covers hadiths related to various subsidiary fields and the second covers hadiths that specifically pertain to legal rulings. Al-Shahīd ii explains that during the formative period,67 Shiʿi scholars had composed "400 sources" (al-uṣūl al-arbaʿumiʾa) containing the imams' statements on various matters.68 He notes that most of these sources were gradually lost. At a later stage, these "400 sources" were incorporated into abridged collections, the most prominent of which are al-Kāfī, al-Tahdhīb, al-Istibṣār 63 Ibn Ḥanbal (d. 241/855) reportedly held that there are slightly less than 700,000 ṣaḥīḥ hadiths. See Jalāl al-Dīn al-Suyūṭī, Tadrīb al-rāwī fī sharḥ taqrīb al-Nawawī, ed. N. al-Fāryabī, 2 vols. (Riyad: Maktabat al-Kawthar, 1415), 1:41 and 106. 64 This sentence is equivocal. Al-Shahīd ii's statement may mean that after the majority of the "400 sources" were lost, a group of scholars gathered the remaining ones into collections to preserve them and make them accessible to future readers. Alternatively, it was only after a group of scholars gathered them into collections that the "400 sources" were lost. It is to be noted that less than twenty of these purported "400 sources" have survived; sixteen of them are collected in al-Uṣūl al-sitta ʿashar, ed. Ḥ. Muṣṭafawī (Qom: n.p., 1405). 65 In the preface to al-Istibṣār, al-Ṭūsī remarks that his aim in this book is to reconcile the seemingly conflicting hadiths found in his previous compilation, al- Tahdhīb and al-Faqīh. He singles out al-Kāfī and al-Tahdhīb as the "best" extant Imami compilations and regards al-Istibṣār and al-Faqīh as complimentary rather than essential sources. He is neither precise in outlining the features that render these books the "best" compilations nor does he mention other available hadith sources that might have been included in this category but, due to one or another shortcoming, are set aside. He notes, however, that although the imams' statements are not limited to what is found in these four compilations, what has been left out from them is no longer readily available. Al-Shahīd ii concludes that al-Kāfī, al-Faqīh, al-Tahdhīb, and al-Istibṣār provide jurists with sufficient hadith sources to issue their rulings; he adds that jurists are not required to search for sources other than these four books to determine whether or not there are any hadiths pertinent to a given ruling.69 Notice that al-Shahīd ii does not mention the phrase "al-kutub al-arbaʿa" in the above passage, although it does appear in some of his transmission licenses (ijāzāt) to his students.70 Nine years prior to al-Riʿāya's completion, in his license to ʿAtāʾ Allāh al-Mūsawī, dated 950/1543, al-Shahīd ii refers to the "Four Books of hadith" (kutub al-ḥadīth al-arbaʿa), calling them the "foundations of religion and the ground of certainty".71 This was sixty years after Ibn Abī Jumhūr's reference to al-Kāfī, al-Faqīh, al-Tahdhīb, and al-Istibṣār as the Four Sources. In another license, dated 957/1550, he characterizes the "Four 67 In this context, "formative" refers to the period in which the imams' followers had direct access to them. This changed in 260/874, when the twelfth and last imam went into concealment, beginning the era known as the Occultation (ghayba). According to Imami eschatology, this era will continue until the appearance of the twelfth Imam at the End of Times. 68 These "400 sources" reportedly were compiled by the companions who directly heard the hadiths from the imams and recorded them in writing. See Modarressi, Tradition and Survival, xiv. On these "400 sources", see E. Kohlberg to be determined whether the unidentified narrators from whom he transmitted were also credible. Hence, al-Shahīd ii does not accept relying on mursal hadiths and argues that, because there are many weak narrators in chains of transmission, any unidentified narrator in the chain may be categorized as weak if his identity becomes known.78 As noted, Ibn Abī Jumhūr allows for certain exceptions in his application of the fourfold method. For example, he justifies relying on hadiths with defective narrators (e.g., non-Imami) or defective transmission chains (e.g., irsāl) if they are accompanied by corroborating indicants external to their transmission chains. Such indicants include cases in which a hadith is widely narrated (mash-hūr) in the community or in which acting on the content of the hadith has been a prevalent practice among jurists.79 Al-Shahīd ii addresses this argument from prevalence (shuhra) in al-Riʿāya. He notes that some Imami scholars sanction acting on a non-ṣaḥīḥ hadith if it is widely narrated (shuhrat riwāya), meaning that it is found in several sources with either the same wording or with different wordings that are close in meaning. Likewise, he remarks, some Imami scholars hold that it is acceptable to rely on a non-ṣaḥīḥ hadith if it is widely relied upon (shuhrat fatwā), meaning that in their legal books jurists have issued rulings based on the content of this hadith.80 He explains that this position is due to the strong conjecture (quwwat al-ẓann), stemming from prevalence, that the narrator of such a hadith must have been credible since many jurists have relied upon the content of his transmitted hadith. Hence, according to this view, it is acceptable to rely on non-ṣaḥīḥ hadiths that satisfy these criteria even though their transmission chains are defective; the prevalent status of these hadiths compensates for their defective transmission path.81 Al-Shahīd ii finds these arguments unconvincing; he rejects the claims to prevalence as a means of compensating for a defective transmission chain and rendering an otherwise unreliable hadith reliable. A critical factor for him is the source and the period of prevalence. He argues that most rulings regarded as 'prevalent' today were not necessarily considered as such in the past. He identifies al-Ṭūsī as the source to which most claims of juristic prevalence are traced. In his view, al-Ṭūsī's immediate successors, due to his great influence, blindly followed his opinions, as did most later scholars. The only exceptions were a minority such as Ibn Idrīs who criticized al-Ṭūsī's reliance on aḥād reports.82 According to al-Shahīd ii, what al-Ṭūsī's successors treat as 'prevalent' legal rulings are in fact al-Ṭūsī's personal rulings and thus are not representative of the earlier legal tradition. In order to distinguish between al-Ṭūsī and his predecessors, he points to al-Murtaḍā who, in contrast to al-Ṭūsī, categorically rejected the authority of aḥād hadiths, much less sanctioning relying on non-ṣaḥīḥ aḥād hadiths. Hence, he concludes, post-Ṭūsī claims to prevalence do not compensate for deficient transmission chains. Al-Shahīd ii's rejection of prevalence, however, is not categorical. He endorses reliance on the prevalent status of a hadith if the hadith is known to have been acknowledged as such by the Imami scholars preceding al-Ṭūsī;83 nevertheless, limiting the authority of prevalence to pre-Ṭūsī scholars restricts its scope as an authenticity-granting mechanism.
Unlike Ibn Abī Jumhūr, for al-Shahīd ii the principle means of verifying a hadith is through its transmission chain; external corroborating indicants do not redeem a hadith's defective transmission chain except in rare cases. This leads him to regard as unreliable most hadiths with defective narrators (e.g., non-Imami) and/or defective transmission chains (e.g., irsāl). These include a large number of the hadiths narrated in the Four Books. Hence, despite their agreement on the centrality of the Four Books and their adherence to the same verification method, Ibn Abī Jumhūr and al-Shahīd ii provide different assessments of the reliability of the hadiths narrated in these four compilations. According to some scholars, the application of al-Shahīd ii's method to the transmission chains of the hadiths narrated in al-Kāfī would 81 Al-Shahīd al-Thānī, al-Riʿāya, 58. He remarks that some jurists prefer a weak but prevalent hadith over an authentic hadith whose content was not widely relied on by early scholars (ibid., 56 yield 5,072 "authentic" hadiths, less than a third of its total hadiths.84 Such an assessment is intriguing since he regards al-Kāfī as a book "whose likeness is not found in the world" and states that "nothing like it has been compiled in Islam".85 Hence, despite considering al-Kāfī and al-Tahdhīb as the "best" Imami hadith compilations -and regarding them, along with al-Faqīh and al-Istibṣār, as "the foundations of religion, the ground of certainty, the mainstay of the [Imami] tradition, the sustainer of faith, and the ground of the pillars of Islam" -in al-Shahīd ii's assessment not every hadith narrated in the Four Books is authentic; in fact, only a minority of their hadiths may reliably be used for determining legal rulings. In the preface, al-Ḥārithī remarks that he was inspired to write the book by the enthusiasm of the people of his adopted home for hadith and law.89 A further motivation for writing the Wuṣūl, al-Ḥārithī notes, was his contemporaries' general neglect of hadith. He regards reflection on God's religion as obligatory for all Muslims; this reflection, al-Ḥārithī maintains, rests upon the transmitted hadiths of the Prophet and the imams.90 Hence, he asserts, those with understanding and zeal must dedicate themselves to examining the transmission of hadiths and their narrators, to become proficient in incorporating hadiths in their argumentations, and to learn the technical hadith terminology used by Shiʿi scholars.91 Like al-Shahīd ii, he laments that the sciences of hadith have declined in his time. By contrast, he writes, earlier generations of scholars assiduously engaged in hadith criticism, and although they did not compose books dedicated to this topic, they articulated their views in numerous books on law, jurisprudence, hadith, and rijāl.92 His Wuṣūl seeks to fill this gap.

Al-Ḥārithī and the Five Hadith Sources
Al-Ḥārithī's remark that no Imami scholar before him had written a text on hadith criticism may indicate that when he wrote the Wuṣūl, he was not aware of al-Shahīd ii's al-Bidāya which, as noted, was the first extant Imami text in this field.93 In any case, Wuṣūl is a more comprehensive book than al-Bidāya and its autocommentary, al-Riʿāya. Al-Ḥārithī begins with a lengthy introduction in which he discusses the significance of hadith as a source of guidance, the biographies of the Prophet and the imams, the status of the Prophet's Companions as transmitters of his Sunna, and the shortcomings of the Sunni hadith corpus, including its two most authoritative sources, the Ṣaḥīḥān of al-Bukhārī and Muslim ibn al-Ḥajjāj.94 This introductory section, together with al-Ḥārithī's discussion of several topics pertaining to hadith criticism, render the Wuṣūl a longer text than al-Shahīd ii's Bidāya, which provides a concise treatment of the topic without engaging in doctrinal discussions.
Al-Ḥārithī describes hadith and its transmission in reverential terms. For instance, in a chapter on the codes of conduct observed by hadith scholars, he quotes three related sayings that highlight the status of hadith: every religion has its knights, and the knights of this religion are the companions of hadith transmission (aṣḥāb al-asānīd); there is no heretic in the world except that he loathes the people of hadith (ahl al-ḥadīth); nothing is more burdensome and  Al-Ḥārithī's passage has a few points in common with al-Shahīd ii's statement regarding the Four Books. Like al-Shahīd, he holds that these sources contain most but not all of the narrated hadiths of the Prophet and the imams. Al-Ḥārithī and al-Shahīd ii also agree that not every hadith quoted in these sources is authentic. He remarks that the compilers of the Five Sources collected hadiths "authentic or otherwise" on various topics. In principle and with some minor modifications,98 al-Ḥārithī subscribes to the fourfold model of hadith verification and ascertains the reliability of hadiths based on the information found in their transmission chains.
Al-Shahīd ii and al-Ḥārithī disagree over the number of compilations they regard as major sources of hadith. Al-Ḥārithī includes Madīnat al-ʿilm as a fifth major source of Imami hadith and refers to them collectively as the Five Sources (al-uṣūl al-khamsa). Two things are noteworthy about al-Ḥārithī's inclusion of Madīnat al-ʿilm. First, this work, attributed to Ibn Bābawayh, is no longer extant;99 even prior to al-Ḥārithī, who seems to have been one of the last individuals with access to it, Madīnat al-ʿilm was not a widely-cited He concludes that either the whereabouts of this book remain unknown or it was not an important work, was not extensively copied, and hence is no longer extant.107 Until further evidence becomes available, these observations remain speculations. In the present context, it suffices to say that al-Ḥārithī is the only figure to add Madīnat al-ʿilm to the Four Books and present them as the five foundational sources of Imami hadith.
Al-Ḥārithī's flight to Iran left an enduring impact on Safavid intellectual developments in the sixteenth century. He continued to promote al-Shahīd ii's teachings, was appointed by royal decrees to important posts,108 and through his writings as well as training a series of influential students, played an instrumental role in the revival of hadith in Iran. Two of these students were pivotal in consolidating the status accorded to the Four Books: al-Shahīd ii's son, Ḥasan and al-Ḥārithī's son, Bahāʾ al-Dīn. 104 The  Al-Shaykh Ḥasan is chiefly remembered for his Maʿālim al-dīn wa-malādh al-mujtahidīn (994/1586).110 Like many books of jurisprudence, Maʿālim contains a chapter on hadith authority, transmission, and authenticity.111 Al-Shaykh Ḥasan also outlines his views on hadith authenticity, with a focus on the Four Books, in another work, Muntaqā al-jumān fī l-aḥādīth al-ṣiḥāḥ wa-l-ḥisān (1006/1597). A collection of hadiths on the obligatory rituals (ʿibādāt), such as the daily prayers and fasting, Muntaqā opens with a lengthy introduction arranged in twelve sections. He remarks that he was prompted to compile the book by witnessing his contemporaries' disregard for hadith. Apparently, the efforts of al-Shahīd ii and al-Ḥārithī, who expressed the same concern, did not fully satisfy al-Shaykh Ḥasan. In his view, the neglect of hadith had led to numerous errors, distortions, alterations, and forgeries of texts.112 This is particularly troubling, he states, since in most cases deriving legal rulings depends on hadith and the jurists' responsa on most issues are traced back to hadiths.113 He finds his contemporaries' negligence of hadith contrary to the diligence of the early scholars.
Al-Shaykh Ḥasan is also troubled by the negligence of some of his contemporaries in matters pertaining to hadith verification. He avers that some scholars, instead of taking heed of the intricate matters pertaining to the hadith 108 Note that al-Ḥārithī had a complicated relationship with the Safavid establishment. See corpus, have widened the scope of reliable hadiths. In his view, they quote in their books hadiths that they regard as "reliable" without taking into consideration the difference between authentic and weak transmission methods (ṭarīqs) and without distinguishing between sound and defective transmission chains (isnāds).114 Rather, he claims, they often rely on what they take to be indicants (qarāʾin) and hints (amārāt) that lead to accepting otherwise "weak" transmissions.115 In Muntaqā, al-Shaykh Ḥasan attempts to counter this approach to hadith verification. Muntaqā's reputation is partly due to al-Shaykh Ḥasan's decision to compile a book comprised only of the authentic and sound hadiths quoted in the Four Books. This is because, he explains, these four books have received "excessive attention" from later scholars due to the excellence they see in them (ziyādat al-iʿtināʾ li-mā raʾaw lahā min al-maziyya). As a result, the Four Books have been privileged (ustuʾthirat) over many other well-known and available hadith books.116 The later scholars he is referring to include Ibn Abī Jumhūr, al-Shahīd ii, al-Ḥārithī, and al-Ardabīlī who, in two of his works -Majmaʿ al-fāʾida and Zubdat al-bayān -mentions the phrase "the Four Books".117 Al-Shaykh Ḥasan, however, was the first scholar to use the phrase al-kutub al-arbaʿa in the context of a hadith compilation and in hadith criticism discussions.118 Unlike Ibn Abī Jumhūr, who considered the Four Books as sufficient hadith sources for legal purposes, and al-Shahīd ii, who described al-Kāfī and al-Tahdhīb as the best Imami hadith compilations and regarded them, together with al-Faqīh and al-Istibṣār, as the "mainstay of the [Imami] tradition", al-Shaykh Ḥasan characterizes them as the later scholars' privileged compilations. In his view, it is the privileged status assigned to these four books that sets them apart from other available hadith works.119 Al-Shaykh Ḥasan provides another explanation for limiting Muntaqā to the hadiths found in the Four Books. He writes that al-Kāfī, al-Faqīh, al-Tahdhīb, and al-Istibṣār are the only available hadith books whose attribution to their respective compilers is generally conclusive (tawātur ijmālī) and for which there are circumstantial indicants that their content is known to have been correctly transmitted.120 For instance, it is known through generally conclusive transmissions that it was al-Kulaynī and not someone else who compiled al-Kāfī, and circumstantial indicants establish that its content, as transmitted, was, in fact, entirely compiled by him. Al-Shaykh Ḥasan's comment, therefore, has no bearing on the reliability of the hadiths quoted in the Four Books and it remains to be ascertained whether a given hadith has been reliably transmitted from the Prophet and the imams, in every link within the chain, until it reached al-Kulaynī, Ibn Bābawayh, and al-Ṭūsī. Regardless of the reliability of the hadiths included in the Four Books, in al-Shaykh Ḥasan's view, nothing has been deleted from or added to these four books and the integrity of their manuscripts has been preserved over the centuries.121 It was for this reason, he explains, that when he composed Muntaqā, he limited himself to the Four Books, noting that other hadith books were available, but that none of them shares the above-mentioned characteristics with these four compendia.122 Al-Shaykh Ḥasan's book, therefore, is comprised of authentic and sound hadiths pertaining to the acts of worship quoted in the Four Books.  For al-Shaykh Ḥasan, a hadith is authentic when all of the narrators appearing in the chain of transmission are Imami, upright (ʿadl), and accurate (ḍābiṭ); these requirements must be met in every link in the chain. Any defect in the chain of transmission, such as irsāl or qaṭʿ, renders the hadith unreliable.129 In al-Shaykh Ḥasan's verification scheme, a narrator's uprightness is established through the testimony of at least two individuals of recognized uprightness. He defines uprightness as the quality present in a person's soul that restrains him from committing major sins, from being likely to commit minor ones, and from things that, although are not technically sins, undermine honorable conduct.130 Presumably, the presence of these qualities in a person prompts him to be truthful in his transmissions. The majority of Imami scholars considered the attestation of one person of recognized uprightness a sufficient criterion for determining a narrator's uprightness.131 For al-Shaykh Ḥasan, however, if a narrator is judged as upright by only one individual, that does not establish that he was in fact upright.132 This disagreement between al-Shaykh Ḥasan and his predecessors regarding the criteria for uprightness has significant implications for the scope of authentic hadiths. Whereas his predecessors, following Ibn Ṭāwūs and al-Ḥillī, divided hadiths into four categories, al-Shaykh Ḥasan, while accepting the fourfold division, added a modification to the category of authentic hadiths, thereby creating a fivefold division. He distinguishes between two standards of hadith authenticity. According to the first standard, a hadith is authentic if the uprightness of every narrator in its chain of transmission is confirmed 127 E.g., having direct access to the imams or hearing the hadiths in person from highly regarded companions of the imams. 128 Ibid., 1:3. 129 Al-Shaykh Ḥasan, Muntaqā, 1:15; idem, Maʿālim, 296. 130 Ibid., 278. Al-Shaykh Ḥasan maintains that just as a plaintiff's claim in court must be corroborated by the testimony of at least two upright witnesses -except in cases such as adultery where four witnesses are required -a narrator's uprightness must be affirmed by at least two individuals of recognized uprightness. In his view, just as the testimony of a single witness does not obligate the judge to issue a verdict, a single testimony affirming the uprightness of a given narrator does not establish that the narrator in question is upright. He acknowledges that two testimonies do not necessarily and conclusively prove a narrator's uprightness. However, because certain precepts in the Qurʾan (e.g., 2:282, 5:106, 65:2) and hadith set the minimum number of court witnesses at two, this number is sufficient.  Compared to others who used the fourfold verification method, al-Shaykh Ḥasan's criteria undermined the reliability of a greater number of hadiths in the Four Books.134 Like al-Shahīd ii, he rejects the authority of mursal hadiths irrespective of the prominence of their transmitters (e.g., Ibn Abī ʿUmayr). Likewise, he rejects the authority of hadiths transmitted by upright non-Imami narrators. As mentioned in § 2, Ibn Abī Jumhūr endorsed acting on mursal and muwaththaq hadiths when they are supported by corroborating indicants. Al-Shaykh Ḥasan's strict adherence to the fourfold method resulted in rejecting the reliability of numerous hadiths.135 Apart from a few exceptions, his contemporaries rejected al-Shaykh Ḥasan's approach to the fourfold verification method.136 Their disagreements with him pertained either to the details of the method or its application. A key representative of a critical stance toward al-Shaykh Ḥasan was his contemporary, Bahāʾ al-Dīn al-ʿĀmilī, better known as al-Shaykh al-Bahāʾī.

6
Al-Bahāʾī and the Four Central Books   All of our hadiths -except in rare cases -are transmitted on the authority of our twelve imams (God's peace be on them all) and they transmit on the authority of the Prophet (peace be on him and his family) since their knowledge is obtained from that niche…our early hadith scholars (God be pleased with them) collected the hadiths of our imams (God's peace be on them all) that reached them in 400 books called the uṣūl. Subsequently, a group of later scholars (may God reward them for their effort) embarked upon collecting and arranging those books, so that they might be conveniently disseminated and made accessible to the seekers of those al-Kāfī, al-Faqīh, al-Tahdhīb, al-Istibṣār as the pivots of Imami hadith, al-Bahāʾī writes that he relies only on the "authentic" hadiths transmitted in the Four Books that pertain to religious precepts. In this way, he states, his book stands as "a criterion (qānūn) for the men of religion of the saved Imami sect and an edict (dustūr) upon which legal experts rely for deriving substantive legal matters."142 He proceeds to elaborate the factors that render a hadith authentic. ∵ Like al-Shahīd ii and al-Shaykh Ḥasan before him, al-Bahāʾī draws attention to the different standards of hadith authenticity prevalent among earlier and later scholars.143 He points out that later scholars replaced the earlier twofold division with a fourfold division. In his view, this new development was due to changing circumstance that deprived later scholars, with damaging effects, of the sources of information that were available to the early scholars. According to al-Bahāʾī, because of the fear of persecution, Shiʿis were often unable to freely disseminate and preserve their textual hadith heritage. This led to the gradual disappearance of many reliable texts; as a result, hadiths from reliable and unreliable sources were mixed together. Furthermore, many of the corroborating indicants that were available to the early scholars and aided them in verifying the authenticity of numerous hadiths were lost to their successors. As a result, later scholars perceived the need to devise a "criterion" for distinguishing unreliable from reliable hadiths. "They", he writes, "devised this new terminology -may God reward them for their effort -and in doing so brought us closer to the distant past". This new terminology is the division of hadiths into ṣaḥīḥ, ḥasan, muwaththaq, and ḍaʿīf. He notes that the first Shiʿi scholar to introduce this new approach was Ibn al-Muṭahhar al-Ḥillī.144 Al-Bahāʾī argues that the introduction of the new verification method did not entirely efface the twofold division. Later scholars, despite the adoption of the fourfold method, would sometimes use the early scholars' criteria and take into consideration factors other than the chain of transmission to determine a hadith's reliability. For instance, later scholars treat the mursal reports of figures such as Ibn Abī ʿUmayr and Ṣafwān ibn Yaḥyā (d. ca. 210/825) as authentic even though their defective transmission chains suggest that these hadiths are weak. As another example, al-Bahāʾī refers to certain non-Imami narrators whose hadiths are regarded as authentic even though the definition of authenticity adopted by later scholars would downgrade the status of such hadiths.145 Hence, he argues, even later scholars such as al-Ḥillī who adhere to the fourfold method, would "at times" allow for exceptions and determine a hadith's reliability through means other than its chain of transmission.146 What is "at times" for al-Ḥillī, however, becomes most of the time for al-Bahāʾī.
Despite his approval of the fourfold method, and his remark that the twofold division is no longer tenable, in practice al-Bahāʾī largely adopts the early scholars' approach to hadith verification.147 The crux of his disagreement with al-Shahīd ii and al-Shaykh Ḥasan revolves around whether it is justified to rely on any indicants other than the chain of transmission to verify a hadith's reliability. Al-Bahāʾī is among those scholars who put a strong emphasis on non-sanad indicants and, as a result, assign reliability to numerous hadiths with defects in their transmission chains.148 A few examples highlight his differences in this regard with al-Shahīd ii and his followers.
known as aṣḥāb al-ijmāʿ.149 This phrase refers to a group of the companions of the fifth to the eight imams who are considered, based on a reported consensus, to be among the most credible hadith narrators. The earliest reference to these companions is found in the Rijāl of al-Kashshī (d. ca. 340/951), who writes that there is a consensus regarding their credibility. Al-Kashshī mentions the names of 18 individuals under the aṣḥāb al-ijmāʿ. He remarks, however, that there are some disagreements between scholars over who should be included in the category of aṣḥāb al-ijmāʿ. While all scholars include 16 out of the 18 individuals mentioned by al-Kashshī, they disagree over six other individuals. As a result, the names of twenty-two individuals are associated with the aṣḥāb al-ijmāʿ.150 There is no precedent for al-Kashshī's remark concerning the community's consensus regarding these companions' credibility in the surviving works of his predecessors and contemporaries. In fact, the phrase "aṣḥāb al-ijmāʿ " was not coined by al-Kashshī, and even after him, there was no explicit mention of it for over two centuries. Discussions concerning aṣḥāb al-ijmāʿ gained traction after Ibn Shahrāshūb's (d. 588/1192) reference to them in Manāqib Āl Abī Ṭālib.151 Ibn al-Muṭahhar al-Ḥillī was the first major scholar to discuss this consensus in some detail. He argued that al-Kashshī's statement indicates the credibility of these companions as hadith narrators even though some of them belonged to "deviant" Shiʿi sects (i.e., non-Twelver) such as the Faṭḥiyya and the Nāwūsiyya.152 A few decades later, Ibn Makkī proposed a different perspective. In his assessment, the consensus regarding these companions, in addition to their credibility as narrators, establishes two further points: the reliability of In Ḥabl al-matīn and Mashriq al-shamsayn, al-Bahāʾī invokes al-Kashshī's statement on these companions and accepts the reported consensus regarding their credibility as a legitimate means of giving credence to their hadiths, irrespective of any defects in their transmission chains.154 Hence, he argues, even if a hadith's transmission chain contains missing links or narrators with questionable status, the hadith would still be acceptable since it has been transmitted by one of aṣḥāb al-ijmāʿ. In al-Bahāʾī's view, the companion's narration indicates that the hadith is reliable, since otherwise he would not have narrated it to others.155 As noted, al-Bahāʾī also maintains that early scholars would regard a hadith as authentic if it appeared in a source with a well-attested attribution to an individual whose veracity was a matter of consensus; likewise, they would consider a hadith authentic if it was affirmed as such by an individual whose veracity was a matter Bābawayh, like those of Ibn Abī ʿUmayr, should be assigned the same status as musnad hadiths.159 In his commentary on al-Faqīh, al-Bahāʾī states that 2,050 of Ibn Bābawayh's 5,963 hadiths are mursal. Nevertheless, he maintains that the mursal hadiths of credible narrators are authoritative and should not be disregarded simply due to their defective transmission chains.160 According to this view, Ibn Bābawayh, who is a credible narrator, would not transmit hadiths from anyone unless he considered the narrator to be credible. The corroborating indicant in this case is the reports that assert Ibn Bābawayh's credibility as a transmitter.161 Although this argument had already been put forward with regard to Ibn Abī ʿUmayr, al-Bahāʾī seems to be the first person who explicitly justifies relying on Ibn Bābawayh's mursal hadiths and treating them as if they were musnad.162 Al-Bahāʾī also accepts a hadith's prevalent status (shuhra) as an indication of its reliability. This approach has significant implications for the number of hadiths regarded as acceptable: a hadith's prevalence is invoked to compensate for defects in its chain of transmission that would otherwise render it unreliable. Al-Bahāʾī's position on this issue is similar to that of Ibn Abī Jumhūr and represents another divergence from al-Shahīd ii and al-Shaykh Ḥasan. As mentioned, according to the latter two, the status of a hadith depends on its transmission chain, not on its widespread narration or the jurists' widespread reliance on its content. For them, a hadith's prevalence does not compensate for its defective transmission chain. Al-Bahāʾī disagrees and considers prevalence as a critical factor in determining a hadith's reliability, irrespective of its transmission chain. 157 Since these twenty-two companions appear in the transmission chains of numerous hadiths in the Four Books, one's position on the consensus claimed for their credibility has consequences on the reliability status of the Four Books. Ḥusayn al-Nūrī al-Ṭabrisī (d. 1320/1902), aware of these consequences, remarks that discussions concerning aṣḥāb al-ijmāʿ is one of the most important topics in the discipline of rijāl. In many instances, he notes, thousands of hadiths that otherwise would not be included in the authentic category either enter this category or are considered as such through their transmission via these companions. Al-Bahāʾī also departs from some major rijāl principles of his predecessors and, as a result, widens the range of what he considers to be reliable hadiths in the Four Books. An example of al-Bahāʾī's rijāl approach is his view regarding the status of Ibrāhīm ibn Hāshim al-Qummī, whose veracity as a narrator is not explicitly mentioned in the early rijāl literature.163 One's position on Ibrāhīm ibn Hāshim has a significant implication for the status of the hadiths quoted in the Four Books since he appears in 6,414 transmission chains.164 How do al-Bahāʾī and others justify Ibrāhīm's credibility? They note that in rijāl works, Ibrāhīm is described as "the first person to promulgate hadiths of the Kufans in Qom". In their view, this description accords the highest status of credibility to Ibrāhīm as a transmitter.165 Al-Shahīd ii, however, challenged Ibrāhīm's hadiths on the grounds that a transmitter should be considered credible once he is described as such in the rijāl literature. In that literature, however, Ibrāhīm is characterized as "praiseworthy" (mamdūḥ) but not explicitly described as "credible" (muwaththaq). Hence, al-Shahīd ii concluded, Ibrāhīm's hadiths must be downgraded from the authentic to the sound category.166 Al-Bahāʾī also disagrees with al-Shaykh Ḥasan's criterion for determining a transmitter's reliability, namely, that a transmitter's uprightness is established once it is corroborated by the testimony of at least two upright witnesses (see § 5). Al-Bahāʾī rejects this position. On the one hand, he argues, al-Shaykh Ḥasan requires the testimony of at least two witnesses before confirming a 163 See You are aware that the rijāl scholars whose books have reached us at this time all transmit the information contained in their books regarding the uprightness of most transmitters from other sources [i.e., they did not personally meet the transmitters concerning whose status they pass judgment]. Hence, if two of them agree about the uprightness of a particular transmitter, this is of no use in determining the authenticity of a hadith unless it is also established that the authors of these rijāl books refrained from considering a transmitter to be reliable on the basis of one upright person. This, however, is very difficult to establish. Indeed, the opposite appears to be true: al-ʿAllāma [al-Ḥillī] in his books on jurisprudence explicitly mentions that the assertion of a single individual is sufficient [for establishing a transmitter's uprightness]. One may also glean from the statements of al-Kashshī, al-Najāshī, al-Shaykh [al-Ṭūsī], Ibn Ṭāwūs, and others that they rely upon the testimony of one individual to assess whether a transmitter is upright or not. This is evident to anyone who is acquainted with their books.167 Hence, al-Bahāʾī argues, if one accepts the two-witness criterion, the rijāl books do not provide a firm basis for determining the uprightness of the narrators who appear in a hadith's transmission chain, even if two or more rijāl scholars agree about the narrators' uprightness. This is because, he argues, there is no guarantee that when the rijāl scholars composed their books, they required the testimony of two truthful individuals before stating that a transmitter is reliable or unreliable. In al-Bahāʾī's view, the reverse is the case. Consensus (ijmāʿ) and prevalence (shuhra) feature prominently among the indicants that early scholars, according to al-Bahāʾī, utilized to evaluate a hadith's reliability. For instance, he maintains that they would consider a hadith authentic if it appeared in a source with a widely-recognized attribution to an individual whose veracity is a matter of consensus among the Imami community; or if there is a consensus that certain individuals affirm it as such; or if there is a consensus that sanctions relying on the hadiths transmitted by certain individuals. He adds that early scholars would accept a hadith as authentic if they found it quoted in any of the books that were widely considered credible and relied upon by their predecessors, regardless of whether or not the compilers of these books were Imamis. Mashriq,[27][28][29] In Ḥabl al-matīn and in his other writings, al-Bahāʾī quotes and relies on muwaththaq hadiths, a departure from al-Shahīd ii and al-Shaykh Ḥasan who, in principle, do not consider such hadiths reliable. As Pakatchi observes, one of al-Bahāʾī's arguments against a categorical rejection of muwaththaq hadiths pertains to the narrators' conversion to Imami Shiʿism from another sect and vice versa. He argues that although certain narrators are described as non-Imamis in the rijāl books, some of the hadiths they transmitted may be from the period before their conversion; in that case, the status of their hadiths should not automatically be downgraded to muwaththaq. Similarly, a narrator may have been a non-Imami at the time of receiving (taḥammul) the hadiths but later became an Imami. According to al-Bahāʾī, one must determine whether the narrator was an Imami at the time of delivering (adāʾ) the hadith and not at the time he first received the hadith. Jumhūr, he discusses the Four Books in relation to the law and avers that when issuing legal rulings, jurists are not obligated to search for any hadith sources besides these four. Al-Shahīd ii's foremost disciple, al-Ḥārithī, agreed albeit with a caveat. In his view, the representative sources of Imami hadith were gathered in five compilations, the Four Books plus a now lost compendium, Madīnat al-ʿilm, attributed to Ibn Bābawayh. Al-Shaykh Ḥasan was the next major figure who played an important role in the consolidation of the Four Books. Whereas Ibn Abī Jumhūr and al-Shahīd ii singled out the Four Books as sufficient sources of legal hadiths, he described them as the preferred compilations of later scholars. For him, these four works were different from the rest due to the widespread recognition they had received from Imami scholars. Al-Shaykh Ḥasan added that the Four Books display another unique feature: they were the only available hadith books whose attribution to their respective compilers was generally conclusive and for which there were circumstantial indicants that their contents were reliably transmitted over the centuries. Hence, nothing had been lost or added to these books that their compilers had not intended. This view played a decisive role in the consolidation of the Four Books as the most authoritative corpus of Imami hadith. After al-Shaykh Ḥasan, the Four Books' manuscript transmission is increasingly presented as uniquely reliable.171 While he found this feature significant, it had no bearing for him with regard to the authenticity of the hadiths quoted in the Four Books. Like Ibn Abī Jumhūr and his successors, al-Bahāʾī accorded a privileged status to the Four Books. Following al-Shahīd ii's historical narrative, he asserted that al-Kulaynī, Ibn Bābawayh, and al-Ṭūsī compiled the Four Books based on the materials available to them from the "400 sources". He often described the Four Books as the sources that constitute the focal point of Imami law. Al-Bahāʾī's enduring impact on the discourse surrounding the Four Books pertains to his delineation of methodologies for accepting and discounting their hadiths.

7.2
Long before scholars referred to al-Kāfī, al-Faqīh, al-Tahdhīb, and al-Istibṣār collectively as the Four Books, they had identified each as a pivotal source of hadith. First among those who assigned a special status to these compilations were the compilers themselves. Al-Kulaynī regarded his al-Kāfī as a sufficient source of religious knowledge;172 Ibn Bābawayh described al-Faqīh as a comprehensive guide for those without access to a jurist;173 and according to al-Ṭūsī, al-Tahdhīb and al-Istibṣār, together with al-Nihāya, eliminated the need to consult any other book of law.174 Many of their successors concurred with these assessments.175 Nearly all major Imami jurists active during the four centuries between al-Ṭūsī and Ibn Abī Jumhūr relied on al-Kāfī, al-Faqīh, al-Tahdhīb, and al-Istibṣār when formulating their legal opinions. These include influential figures such as al-Muḥaqqiq al-Ḥillī, Ibn al-Muṭahhar al-Ḥillī, and Ibn Makkī.176 Although each of these four compilations was widely used prior to the introduction of the term the "Four Books", the collective reference to them as the "Four Books" contributed to their consolidation as the most authoritative sources of Imami hadith. This process is evident in transmission licenses issued by scholars who characterize the Four Books as the "most central" (akbad),177 "most renowned" (ashhar),178 and "chief" (al-ʿumad) 179 hadiths sources, which serve as "reference" (marjiʿ)180 works for Imami jurists.

7.3
One must not readily assume that the Four Books were selected from a large number of hadith compilations. They may be nearly everything that was available to Ibn Abī Jumhūr or al-Shahīd ii. It remains to be further investigated whether the Four Books were, as some of the figures examined in this article contended, the "most renowned" and "chief" hadith sources or perhaps the only major extant compilations of legal hadith.183 Some might argue that al-Kāfī, al-Faqīh, al-Tahdhīb, and al-Istibṣār may in fact have been the "most renowned" or "chief" hadith sources even if many other compilations were in circulation. Did the five jurists examined in this study have access to other major compilations of legal hadith? And, after studying them, did they judge them to be unreliable or not sufficiently comprehensive to be mentioned along with the Four Books as the most authoritative sources of Imami hadith? At present, there is no strong supporting evidence that this was the case. It is clear, however, that the consolidation of the Four Books as the most authoritative sources of hadith was a major factor in stimulating discussions regarding the status of individual hadiths contained in them. In a period of less than sixty years, several theories were developed to explain the special status of these compilations, while either restricting or expanding the probative force (ḥujjiyya) of their individual hadiths.

7.4
Neither the compilers themselves nor successive generations of Shiʿi scholars considered all of the hadiths quoted in the Four Books to be authentic. It is noteworthy that none of the four books in this hadith quartet has the word ṣaḥīḥ (authentic) in its title. By contrast, in the Sunni tradition six hadith compilations came to be referred to collectively as the Six Authentic Books (al-ṣiḥāḥ al-sitta).184 Of these six, the collections of al-Bukhārī and Muslim are explicitly called ṣaḥīḥ.185 In the Shiʿi tradition, the term ṣaḥīḥ was not applied to al-Kāfī, al-Faqīh, al-Tahdhīb, or al-Istibṣār -either before or after the phrase 'the four books' was coined -and, more importantly, there is no consensus on their authenticity.186 Despite their awareness of the Six Sunni Books, prior to Ibn Abī Jumhūr, Shiʿi scholars were not keen to set a specific number for the hadith sources they considered authoritative, although they sometimes specify the compilations they consider significant and compare these with the Six Books. Ibn Makkī (d. 786/1384), for instance, remarks that al-Kāfī alone has more hadiths than these six Sunni compilations combined and that the same is nearly the case for al-Faqīh, al-Tahdhīb, and al-Istibṣār.187 After the phrase 183 In The lack of consensus among Imamis concerning the reliability of the Four Books contrasts with the Sunni tradition, where a consensus eventually was claimed for the 'the four books' was coined, it became common to compare the Six Authentic Books with the Four Books. Al-Bahāʾī, for example, asserted that Shiʿi compilations contain more hadiths than the Six Authentic Books and noted that the Four Books constitute the central sources of Shiʿi hadith.188 Eventually, the consolidation of the Four Books as the central corpus of hadith led a small group of Imami scholars to treat every hadith in the Four Books as either authentic or, at the least, as sufficiently reliable sources of legal knowledge (see § 7.5). Before this pivot to canonization, most Imami scholars primarily discussed methods of hadith authentication rather than authenticating any specific compilation.

7.5
The discourse on the reliability of the Four Books developed chiefly in the context of law and jurisprudence.189 This is due to the juristic contention that the actions of believers must as much as possible be based on the Qurʾan and hadith.

1.
Claiming consensus for the credibility (thiqa) of certain transmitters, whether it be an individual, such as Ibn Abī ʿUmayr, or a group, such as aṣḥāb al-ijmāʿ. Once a hadith was known to have been transmitted by any such transmitter, its authenticity was established and questions regarding its defective transmission chain dissolved. Closely related to this is the necessity of reliance on certain privileged biographical dictionaries, which record the status of narrators. These include the rijāl compilations of al-Kashshī, al-Najāshī, al-Ṭūsī, Ibn Dāwūd al-Ḥillī and Ibn al-Muṭahhar al-Ḥillī. Recall, for instance, that Ibn Abī Jumhūr required that the Four Books must be read in conjunction with the rijāl books.192 2. Demarcating a historical period -often the first three centuries ah -and considering certain texts, verification methods, and scholarly practices from this period as authoritative. Examples include the "400 sources" -reportedly compiled by the imams' disciples -from which the Four Books reportedly were drawn; to give credence to the twofold verification scheme claimed to have been employed by 'early scholars' (qudamāʾ); to accord a privileged status to prevalent (mashhūr) practices of the early jurists, such as their legal opinions or their widespread transmission of a hadith. For instance, if numerous early jurists issued rulings on the basis of a particular hadith, the hadith would be treated as reliable even if its extant version has a defective transmission chain. This verification approach assumes that early scholars who were either contemporaries of the imams or active shortly thereafter would not issue a legal ruling based on a hadith they knew to be unreliable. Those who take this verification approach likewise treat the widely circulating hadiths during the early period as reliable regardless of any defects in their narrators or transmission chains. In both these cases, a hadith's prevalent status compensates for any defects in its transmission chain. 3. Upholding certain methods as the sole legitimate means of hadith verification (e.g., the fourfold method). 4. Designating categorical authenticity to an entire compilation or a textual corpus. For advocates of this approach, the Four Books, for example, are authentic in their entirety. 191 See, for example, al-Shahīd al-Thānī, al-Riʿāya, 37-41. As mentioned in § 2, reports that are not mutawātir are regarded as aḥād. While mutawātir reports leave no questions or doubts in the recipient regarding their authenticity, the authenticity status of non-mutawātir reports remains inconclusive. 192 Ibn Abī Jumhūr, Kāshifat al-ḥāl, 92-93, 137-142.
As discussed throughout this article, these four methods sometimes overlap with each other. Beginning with Ibn Abī Jumhūr, who is the first known scholar to coin the phrase, no major Imami scholar has disputed the privileged status accorded to the Four Books; instead, scholars disagreed over the verification methods used to ascertain the reliability of the hadiths quoted in them. All the figures examined in this study adhered to the fourfold verification method, albeit with disagreements concerning its application. Discussions about the authority of the Four Books, along with the defence, modification, and rejection of the fourfold verification method, dominated the Imami hadith discourse during the sixteenth century. Two contrasting approaches to the fourfold method and the Four Books resulted from these discussions. The key issue was whether, and to what extent, other means of hadith verification may be utilized to determine the reliability of the hadiths transmitted in the Four Books.
While Ibn Abī Jumhūr and al-Bahāʾī were keen to accommodate exceptions to the fourfold method and used other means of hadith verification, al-Shahīd ii and, especially, al-Shaykh Ḥasan were reluctant to give credence to any indicants external to a hadith's transmission chain. In Muntaqā, al-Shaykh Ḥasan criticized those who, while using the fourfold method, maintained that a hadith is authentic if its transmission has no features that would discredit it (even if its transmission chain has some defects). In his view, it is reasonable for those who use the twofold method to argue this point but unreasonable for those who use the fourfold method. Some scholars would suspend the fourfold method on certain occasions, such as giving credence to the hadiths transmitted by Ibn Abī ʿUmayr despite defects in their transmission chains. Al-Shaykh Ḥasan, like al-Shahīd ii, found this position flawed; in his view, to concede to it would render the fourfold division redundant. The fourfold verification method is used in order to distinguish authentic from non-authentic hadiths, based on the reliability of their transmission chains. If in certain cases authenticity can be attributed to hadiths with defective transmission chains, the fourfold division becomes superfluous. According to al-Shaykh Ḥasan, such an approach goes against the very reason for which the fourfold division was designed. Hence, he rejected mursal hadiths even if they were transmitted by the most credible narrators.193 In addition to requiring strict observance of the fourfold method, he stipulated that a hadith is authentic only when the uprightness of the narrators in its transmission chain is confirmed by at least two individuals of recognized uprightness. In this way, while upholding the the four books of shiʿi hadith Islamic Law and Society (2021) 1-55 | 10.1163/15685195-28040002 Downloaded from Brill.com04/26/2022 08:58:47AM via free access privileged status of the Four Books, al-Shaykh Ḥasan's verification method renders the majority of their hadiths unreliable.
Al-Shaykh Ḥasan's proposed standards fueled extensive criticisms and elicited several responses over the course of the next century (ca. 11th/17th).194 For example, disregarding al-Shaykh Ḥasan's call for a vigilant observance of the fourfold framework, al-Bahāʾī expanded the methods of hadith verification beyond sanad analysis. While upholding the fourfold method, he asserted that in addition to the transmission chain, it is legitimate to take into consideration factors that may corroborate a hadith's reliability even if the hadith fails to meet the standards of the fourfold verification method.195 Although this approach was not unprecedented among Imami scholars, al-Bahāʾī expanded its scope and presented it in a more systematic manner. Al-Bahāʾī also disagreed with al-Shaykh Ḥasan's criterion of hadith authenticity and rejected his requirement of (at a minimum) two witnesses of recognized uprightness for establishing a narrator's uprightness as hadith transmitter. Like most Imami scholars before and after al-Shaykh Ḥasan, he considered the testimony of one upright witness to be sufficient.
Despite these differences, al-Bahāʾī did not regard every single hadith in the Four Books as an acceptable source of law. In his view, many hadiths not only had defective transmission chains, but also lacked any corroborating indicant that might justify their reliability. Nevertheless, al-Bahāʾī's criticisms of al-Shahīd ii's and al-Shaykh Ḥasan's verification methods were widely accepted. His promotion of non-sanad corroborating indicants to determine a hadith's reliability undermined their demand for strict observance of the fourfold method as the only viable method of hadith verification; eventually it was al-Bahāʾī's position that found favor among the majority of Imami scholars.
Perhaps unwittingly, al-Bahāʾī's undermining of the methods advocated by al-Shahīd ii and his followers contributed to the rise of their antithesis in the thought of Muḥammad Amīn al-Astarābādī (d. 1036/1626) and the Akhbārī 193 Al-Shaykh Ḥasan, Muntaqā, 1:12-13. 194 At times these criticisms would turn into fierce attacks. See al-Baḥrānī, Luʾluʾat al-baḥrayn, 43-45. 195 In many instances, juristic disagreements concerned not whether but which exceptions can be accommodated in the application of the fourfold verification method. For instance, a scholar might consider it legitimate to rely on indicants that in his view point to the reliability of hadiths transmitted by Ibn Abī ʿUmayr; the same scholar, however, might disagree with his colleagues' arguments in favor of Ibrāhīm al-Qummī's credibility as a transmitter, based on other corroborating indicants. In many cases the issue is not only whether relying on corroborating indicants is a legitimate means of hadith verification but also what counts as a corroborating indicant. movement he inspired.196 Al-Astarābādī and his followers have been the subject of several studies in the past three decades.197 It suffices for our purposes to note that al-Astarābādī took his predecessors' discussions on hadith authenticity in an entirely new direction. In his al-Fawāʾid al-madaniyya (1031/1622), al-Astarābādī characterizes the Four Books not just as the "chief" or "most renowned" compilations but rather as "authentic" sources of Imami hadith.198 He rejected the fourfold verification method, which had remained unchallenged for three centuries, characterizing it as a heretical innovation adopted from Sunni hadith scholarship.199 This sharp break from his predecessors represented a decisive attempt to canonize the Four Books. Writing 140 years after Abī Jumhūr recommended the Four Books as sufficient hadith sources for aspiring jurists, al-Astarābādī argued in favor of their categorical authenticity. In contrast to his predecessors, al-Astarābādī focused on the status of the Four Books as compilations rather than on articulating a method for ascertaining the reliability of specific hadiths. In the 17th and 18th centuries, al-Astarābādī's followers elaborated on his ideas, making waves that rippled widely across the Shiʿi world. These waves ultimately failed to bring the Four Books to the shore of indisputability. Contestations over the authenticity and the interpretation of the hadiths transmitted in the Four Books have persisted down to the present. Unlike the Qurʾan, the Four Books are not considered a sealed scriptural canon and efforts to restrict their authority or to expand the hadith corpus beyond them are still ongoing.200 Nonetheless, the reception history of the Four Books indicates that they have rarely been perceived as a mere corpus. These four compilations have been widely disseminated, subjected to numerous commentaries and glosses, translated to other languages, and held in high esteem among scholars regardless of theological and jurisprudential orientations. Indeed, there is hardly a major book of law that does not contain numerous citations from the Four Books. The consolidation of al-Kāfī, al-Faqīh, al-Tahdhīb, and al-Istibṣār into the Four Books did not place them beyond reproach; neither did it resolve debate regarding their status. While the overwhelming majority of Imami jurists regard the Four Books as indispensable sources of legal hadiths, they do not consider the reliability of the hadiths transmitted in them to be indisputable.

Acknowledgment
This article is a revised and slightly abridged version of a chapter from my doctoral dissertation completed in 2019 at the University of California, Berkeley. The revisions were chiefly carried out during my postdoctoral fellowship under the auspices of the European Research Council Advanced Award: "Law, Authority and Learning in Imami Shiʿite Islam" (LAWALISI: no. 695245), based at the University of Exeter. Further revisions were undertaken during my postdoctoral fellowship at the Humboldt University of Berlin. I thank Asad Q. Ahmed and Robert Gleave for their constructive comments on the earlier drafts of this article. I am grateful to Kumail Rajani for his detailed and productive suggestions on the final draft. I also wish to thank Hassan Rezakhany, who read parts of this study and offered valuable feedback. Two anonymous reviewers provided helpful comments; my thanks go to them as well. Finally, I am grateful to David S. Powers for his judicious editorial suggestions. Any shortcoming is solely my responsibility. Dr Mahmoud Tabatabaei (1967-2020), a dear relative of mine, died suddenly in Tehran at the age of 53 due to complications from the coronavirus. He was a kind and generous man. This article is dedicated to his memory.