Jarh wa-Ta’dil: the Science of Narrator Criticism

In the previous article we saw how Imam Malik ﵀ brought forth the Muwatta’ in Medina, the first publicly published hadith book, and how the strength of that book ultimately rested on a single golden chain: Malik ﵀ from Nafi’ ﵀ from Ibn Umar ﵁. A book is worth no more than the men who carry it. We saw that the reliability of the Muwatta’ depended on what the community knew about each of those transmitters, on their memory, their integrity and the question whether teacher and pupil had really heard one another. But who established that, and how? Behind every chain of a published book stands a judgement about men, and that judgement had to come from somewhere and be preserved somewhere.

That judgement began in the field, with the travelling collectors. Whoever trekked for weeks through the desert to hear a chain of transmission with his own ears returned home not only with text but with an impression of the man who carried the text. What the travellers did in the field, however, had to be stored somewhere. A man like Yahya ibn Ma’in ﵀ did not walk thousands of miles only to let the collected names evaporate again. Every meeting in the field yielded a date, a face, a memory, a verdict. Those verdicts had to be preserved. They had to be classified. They had to be compared. They had to be available to the next generation, which in its turn would be able to judge the judges. Between roughly 150 AH and 350 AH, that which the rihla yielded in the field was transformed, in the studies of Basra, Kufa, Baghdad, Wasit, Naysabur and Bukhara, into a science with its own vocabulary, its own methodology and its own literature. That science is called Jarh wa-Ta’dil, literally “wounding and vindicating”. We know it in modern language as the science of narrator criticism, and it is, after the text of the Quran itself, the second great apparatus of preservation that the Islamic civilisation has brought forth.

In this article we open the workshop where that apparatus was built. We follow the three generations that designed it, we look into the biographical dictionaries it left behind, we draw out one concrete chain as a third-century student would have done, and we show why this infrastructure refutes the twentieth-century orientalist theory of the “back-projected” hadith even before its pronouncement. The article closes with a look ahead to a post-canonical scholar from Nishapur whose Sahih, and its surviving folio, shows that this apparatus kept working for generations after the canon.

What the Rihla Yielded Had to Be Preserved

Whoever lays the accounts of the early hadith collectors side by side sees one pattern always returning. The traveller returns home with more than text. He returns home with a dossier. He has looked the man he spoke with in the eyes. He has heard his breathing as he recited the names of his teachers. He has seen whether the scholar hesitated, whether he consulted his book, whether he quoted confidently from memory. He has asked where he studied, when he was born, who his pupils were. He has, if he was thorough, heard the same hadith from two different teachers to compare the wording. And he has, if he was patient, waited to see whether the man, at a second meeting, still told the same account in precisely the same way.

That dossier was the raw material of what would later become Jarh wa-Ta’dil. Every scholar carried thousands of such dossiers in his head. Yahya ibn Sa’id al-Qattan ﵀ in Basra, who died in 198 AH, reportedly had a detailed judgement ready on hundreds of living transmitters. Abd ar-Rahman ibn Mahdi ﵀, likewise from Basra, who died in the same year 198 AH, could recite from memory the names, years of study and weaknesses of hundreds of Kufans, Medinans and Hijazis. But as long as those dossiers sat in heads and not on parchment, the community was vulnerable. One shipwreck on the Tigris, one plague epidemic in Wasit, and half a century of collected memory could be lost.

The solution was obvious. What sat in the heads had to go into books. But what kind of books? No longer only collections of hadith. Collections of men. Books in which not the transmission but the transmitter stood central. Books in which you could look up a name as a Roman jurist consulted a code of law, and find out: when did this man live, in which city, with which teachers did he learn, to which pupils did he pass on, and, most importantly, what did the masters of his time judge about his memory, his integrity and his reliability?

That book was invented in the third century AH. It carries names such as Kitab at-Tarikh (book of history), Kitab al-Jarh wa-t-Ta’dil (book of wounding and vindicating), Tabaqat (layers of generations), or simply Mu’jam (alphabetical register). The foundation was laid by three masters whose names return to this day on every page of every serious hadith study.

The Founders: Shu’ba, Yahya ibn Sa’id, Abd ar-Rahman ibn Mahdi

The first of these three, and the one to whom the byname amir al-mu’minin fi’l-hadith was given, was Shu’ba ibn al-Hajjaj ﵀, born in Wasit around 82 AH and died in Basra in 160 AH. Amir al-mu’minin fi’l-hadith, leader of the believers in hadith, was no title handed out lightly. In later centuries it went also to Sufyan ath-Thawri ﵀, to Ahmad ibn Hanbal ﵀, to al-Bukhari ﵀, a handful of men per century. For Shu’ba ﵀ it was coined because he was the first to begin making a systematic and methodical distinction between who was allowed to pass on a hadith and who was not.

Shu’ba ﵀ had a disconcerting manner of working. When he heard of an unknown scholar for the first time, he did not immediately attend his lessons. He followed him first in the street. He watched how the man cared for his animals, how he dealt with his neighbours, whether he traded honestly in the market, whether he performed his prayers on time. Only when the human credibility of the man had been established did Shu’ba ﵀ enter his teaching circle. His famous dictum runs: “No man has told me hadiths without my having looked at him, and if he did not perform his prayer well, I wrote nothing down from him.” So simple, so implacable. Whoever was not faithful to his Creator could not be reliable in the transmission of the words of His Prophet ﷺ.

Sufyan ath-Thawri ﵀ from Kufa, who died in 161 AH, worked according to a comparable principle but with an accent of his own. In the early practice, three things were as a rule kept in view with a transmitter: his capacity of memory (hifz), his accuracy in repetition (itqan), and his freedom from sectarian preferences that could cloud the faculty of judgement. When a transmitter satisfied the first two conditions but showed under the third a noticeable inclination, for example a strong Murji’i, Qadari or Shi’i conviction, this was noted alongside, without necessarily disqualifying the man, for known leanings were in their eyes less dangerous than hidden ones.

The second founder, Yahya ibn Sa’id al-Qattan ﵀, a Basran, who died in 198 AH, brought the manner of working of Shu’ba ﵀ and Sufyan ﵀ to a system. His byname al-Qattan comes from qutn, cotton; he traded in cotton, and in his shop he received students in between the work. His methodological contribution lay in the strictness with which he examined every chain from beginning to end, instead of trusting blindly in a renowned name at the end. He handled, moreover, with great precision the distinguishing vocabulary that the classical tradition would fix: thiqa (reliable), saduq (truthful but with small weaknesses), layyin al-hadith (soft in his transmission), da’if (weak), matruk (abandoned), kadhdhab (liar). These terms received their canonical place only in the later codification, notably in the introduction to al-Jarh wa-t-Ta’dil of Ibn Abi Hatim ﵀, but the practice from which they arose was honed in the workshops of men like Yahya ﵀.

His best friend and kindred spirit in Basra was Abd ar-Rahman ibn Mahdi ﵀, who died in the same year 198 AH. Yahya ﵀ and Abd ar-Rahman ﵀ were regarded as the two balances on which the hadith of the first part of the second century was weighed. Their agreement about a man carried such weight that later generations treated it as almost binding. Their difference of opinion was noted and passed on, for even their disagreements were methodologically instructive.

Between these four men, Shu’ba ﵀, Sufyan ﵀, Yahya ﵀ and Abd ar-Rahman ﵀, lay the foundation. They did not yet have biographical dictionaries at their disposal, for their pupils would write those. They had at their disposal memory, methodology, and the readiness to judge. To pronounce a judgement about a fellow scholar was no light matter in a culture in which reputation was everything. That they did it, and that they did it in the service of something they held greater than friendship, namely the purity of the transmission of the Prophet ﷺ, gave their work the moral authority that the whole later science would carry.

The anecdotal character of such exchanges must not conceal what they meant methodologically. A hadith scholar was now judged as a witness in a sharia court was judged. When the community distrusted a man with the safekeeping of a borrowed dirham, it could not trust him with the safekeeping of a saying of the Prophet ﷺ. The standard was not learning alone, but credibility in the whole of his life.

The Next Generation: Yahya ibn Ma’in, Ali ibn al-Madini, Ahmad ibn Hanbal

What the first generation had designed as a method, the second generation codified as a science. Three names pressed forward in the first decades of the third century AH as the absolute authorities in Jarh wa-Ta’dil: Yahya ibn Ma’in ﵀, Ali ibn al-Madini ﵀ and Ahmad ibn Hanbal ﵀. All three worked in Baghdad and its surroundings, they knew one another, they debated with one another, they consulted one another, and they held themselves jointly responsible for what they passed on.

Yahya ibn Ma’in ﵀ was born in 158 AH in Qadisiyya near Baghdad and died in Medina in 233 AH on the way back from the hajj. He was the son of a rich administrator in the service of the Abbasid caliph. His father left him a fortune of about a million dirhams, wealth with which he could have settled in luxury until the end of his life. Yahya ﵀ chose otherwise. He spent everything on books, on journeys to hear hadith, on visiting teachers in Mecca, Medina, Syria, Egypt, Wasit, Kufa and Basra. When he died, his fortune was spent and he reportedly did not even own a shroud.

What that money had brought him was a memory that was already regarded as legendary in his own time. Yahya ﵀ knew, according to his pupils, about a million hadiths with their complete chains. That is the figure given in the classical sources and sometimes contested or qualified in the modern literature, but even if we cut the figure amply in half it remains overwhelming. He was, as his younger contemporary Ahmad ibn Hanbal ﵀ would say, the man who had given his life to wash the lie out of the hadith.

To him is attributed the saying: “When I see a hadith that I have never seen before, my heart trembles, and I do not pass it on until I have compared it with what I have heard from others.” That principle of internal verification, the cross-referencing of an unknown transmission with the complete corpus he carried in his head, was his trademark. He did not even trust himself. What came from others had to fit into the pattern he had already established in tens of thousands of other transmissions.

Ali ibn al-Madini ﵀, born in 161 AH in Basra from a family of Medinan descent, from which he derives his nisba al-Madini, died in 234 AH, was his complementary counterpart. Where Yahya ibn Ma’in ﵀ was the master of narrator criticism, Ali ﵀ was the master of what one would now call ilm al-ilal, the science of the subtle defects in a chain. A hadith could at first sight seem completely in order, all transmitters thiqa, no interruptions in the chain, and yet contain a hidden flaw that only someone with Ali’s ﵀ knowledge could see. A teacher who died in year X and a pupil who was born in year Y where Y was greater than X, a meeting that was geographically impossible because the two men had never been in the same city, a deviation in the choice of words that pointed to a confusion with another hadith: these were the defects that Ali ﵀ could track down. He wrote a well-known work, Kitab al-Ilal, which was recorded by his pupils and which is to this day a reference work in the discipline.

Between them stood Ahmad ibn Hanbal ﵀, born in 164 AH in Baghdad and died in 241 AH in the same city. Everyone knows him as the imam of one of the four Sunni schools of law and as the man who, during the mihna, the inquisition of the Abbasid caliph al-Ma’mun and his successors, was flogged for years because he refused to declare that the Quran was created. What is less often emphasised in the history of the hadith science is that Ahmad ﵀ was also one of the strictest critics of transmitters that the third century has known. His Musnad, with about thirty thousand hadiths arranged by Companion ﵃ of the Prophet ﷺ, was no uncritical collection, but a work in which every chain had passed through his own judging eye.

The three men worked together as one college. When Yahya ﵀ passed a judgement, he asked Ahmad ﵀ for confirmation. When Ali ﵀ discovered an unusual defect, he shared it with Yahya ﵀. When Ahmad ﵀ hesitated about a man, he consulted Ali ﵀. A student in Baghdad who had had the favour of studying with all three later described them as three mirrors in which every man who passed on hadith saw his true face. The terminology that had been proposed in Basra was now refined and refined again by these three, until the classical tradition distinguished more than a dozen degrees of reliability, from thabt hujja (firm and probative) at the top to kadhdhab waddaa’ (notorious forger) at the bottom.

It was this generation that first began to write real biographical dictionaries. Yahya ibn Ma’in ﵀ left behind several recensions in which he had recorded his judgement on thousands of scholars: a version through his pupil ad-Duri, one through Ibn Muhriz, and still others through ad-Darimi and Ibn al-Junayd. Ali ibn al-Madini ﵀ wrote dozens of monographs, of which unfortunately only fragments have been preserved. Ahmad ibn Hanbal ﵀ had his judgements on thousands of men taken up in a work that his son Abdullah ﵀ ordered after his death, the famous Kitab al-Ilal wa-Ma’rifat ar-Rijal (Book of the Defects and the Knowledge of Men).

The Canonical Six

What the second generation had given as system, the third generation delivered as application. Between roughly 200 AH and 280 AH, on the shoulders of Shu’ba ﵀, Sufyan ﵀, Yahya ﵀, Abd ar-Rahman ﵀, Yahya ibn Ma’in ﵀, Ali ibn al-Madini ﵀ and Ahmad ﵀, the canonical hadith literature was built up that has remained normative for the Sunni community to this day. Six collections, al-kutub as-sitta, compiled by six men who largely knew one another, sometimes studied together, sometimes were one another’s students.

Muhammad ibn Isma’il al-Bukhari ﵀ was born in Bukhara in 194 AH and died in Khartank, a village near Samarkand, in 256 AH. He was of Persian descent, grandson of a newly converted Mughira al-Ju’fi. At the age of ten he had reportedly learned the Quran by heart. At the age of sixteen he departed on rihla to Mecca, and in the course of a working life of nearly half a century he visited Egypt, Syria, Hijaz, Iraq, Khurasan and Transoxania. He collected, according to his own later statement, about 600,000 hadiths in the course of those journeys. From these he chose, after a strict selection process that cost him sixteen years, 7563 for his al-Jami’ as-Sahih, the book that is now simply called Sahih al-Bukhari. When the same hadith occurs several times under different chapters to illustrate different legal points, about 2602 unique hadiths remain according to later counting.

The selection process of al-Bukhari ﵀ was not only quantitative. He set for himself conditions stricter than any predecessor. A hadith was included only when it was completely connected (no missing links), the transmitters were all uncontested thiqa, two successive transmitters had demonstrably met one another in person (the so-called liqaa’ condition, which was stricter than the condition of Muslim ﵀, who required only contemporaneity), and the transmission did not conflict in content with a stronger transmission or with the Quran. Every hadith he included, he later assured, he had approached beforehand with two rak’at of prayer and istikhara, asking for Divine guidance in the selection.

Muslim ibn al-Hajjaj an-Naysaburi ﵀ was born in 206 AH in Naysabur and died in the same city in 261 AH, five years after his master al-Bukhari ﵀. They had studied together in Naysabur during al-Bukhari’s stay there, and the relationship between them was that of two masters who recognised one another. Muslim ﵀ wrote his own Sahih, in short Sahih Muslim, which in the usual count contains about 7275 hadiths with repetitions, or about 3033 when the doubly occurring transmissions are subtracted. His conditions were somewhat less strict than those of al-Bukhari ﵀ as regards the liqaa’, but his organisation of the material was considered pedagogically superior by many later scholars. The combination of the two works, in Arabic as-Sahihain (the two Sahihs), forms the heart of the Sunni hadith canon.

Abu Dawud as-Sijistani ﵀, born in 202 AH in Sijistan (the present Iranian-Afghan border region) and died in Basra in 275 AH, wrote Kitab as-Sunan, about 4800 hadiths with the emphasis on legal subject matter. Abu Dawud ﵀ had studied with Ahmad ibn Hanbal ﵀, and his work is sometimes regarded as the hadith compilation that most directly serves the Hanbali school of law.

Muhammad ibn Isa at-Tirmidhi ﵀, born in 209 AH in Tirmidh (Termez, in present-day Uzbekistan, on the Amu Darya) and died there in 279 AH, was a pupil of both al-Bukhari ﵀ and Abu Dawud ﵀. His al-Jami’, usually called Sunan at-Tirmidhi, counts about 3956 hadiths, and his methodological contribution was unique: with each hadith he mentions not only the chain, but also a judgement of his own on the status (sahih, hasan, gharib, da’if) and references to variant transmissions of the same content from other Companions ﵃. The work is, in that respect, the first real handbook for the student who wants to learn not only what has been transmitted, but also how the transmitted is weighed.

Ahmad ibn Shu’ayb an-Nasa’i ﵀, born in 215 AH in Nasa (Khurasan, present-day Turkmenistan) and died in 303 AH in Mecca, wrote as-Sunan as-Sughra (the Small Sunan), with about 5761 hadiths. An-Nasa’i ﵀ was known within the six as the strictest after al-Bukhari ﵀ and Muslim ﵀; some of his conditions were even stricter than those of Abu Dawud ﵀.

Muhammad ibn Yazid Ibn Maja ﵀, born in 209 AH in Qazwin (present-day Iran) and died there in 273 AH, completed the canon with his Sunan Ibn Maja, about 4341 hadiths, with the broadest acceptance criteria of the six. Some of his hadiths are to be found only with him, and these are treated by the later science with more caution than the material that also occurs in the Sahihain.

What all six have in common is that they made their selection on the basis of the apparatus of Jarh wa-Ta’dil. None of them worked by intuition. Every chain in each of the six works can be traced back, and for every man in every chain a judgement is available in the biographical literature. It is no faith in the six; it is a faith in the method that produced them.

The Biographical Dictionaries

Beside the hadith collections there arose, hand in hand with them, the second great genre of the third and fourth centuries: the biographical dictionaries of the transmitters themselves. Here lies the library on which the hadith canon stands.

Al-Bukhari ﵀ himself wrote, beside his Sahih, a biographical work called Kitab at-Tarikh al-Kabir, the Great Book of History. It contains short biographical notices on many thousands of transmitters, alphabetically ordered. For each he gives name, kunya, place of origin, place of settlement, principal teachers, principal pupils, year of birth so far as known, year of death, and the critical judgement that he himself or his predecessors had pronounced on the man. He wrote in addition a Tarikh al-Awsat and a Tarikh as-Saghir, a medium-length and a short version. According to his own statement he began these works at the age of seventeen, sitting by the grave of the Prophet ﷺ in Medina under the full moon.

Abd ar-Rahman Ibn Abi Hatim ar-Razi ﵀, who died in 327 AH, wrote al-Jarh wa-t-Ta’dil, a work in nine parts that took up from his father Abu Hatim ﵀ and the masters of his generation the judgements on several thousands of transmitters. He ordered the material by generation and within that alphabetically, and he cited for every man the verbatim pronouncements of his father, of Yahya ibn Ma’in ﵀, of Ahmad ibn Hanbal ﵀, of Abu Zur’a ar-Razi ﵀, and of others. When the masters disagreed, he noted both judgements. In the introduction to this same work he fixed the degrees of jarh and ta’dil systematically, and it is that introduction which made the distinguishing vocabulary canonical for the later science. The work is, for the student who wants to trace the sources for one man, indispensable.

Muhammad ibn Hibban al-Busti ﵀, who died in 354 AH, wrote two works that mirrored one another: Kitab ath-Thiqat (Book of the Reliable), with several thousands of positively judged transmitters, and Kitab al-Majruhin (Book of the Wounded), with the names of those who had been rejected for various reasons. Ibn Hibban ﵀ was, according to later criticism, sometimes somewhat too generous with tawthiq (declaration of reliability), but his works are rich in biographical detail.

The high point of this literature came only centuries later. Yusuf ibn Abd ar-Rahman al-Mizzi ﵀, who died in 742 AH in Damascus, completed Tahdhib al-Kamal fi Asma’ ar-Rijal, a monumental fusion of earlier works in thirty-five parts. It contains the biographies of about 8000 transmitters whose names occur in the six canonical works, plus the Muwatta’ of Imam Malik ﵀ and the Musnad of Ahmad ﵀. His pupil Shams ad-Din adh-Dhahabi ﵀, who died in 748 AH in the same city, wrote an abridgement of al-Mizzi’s ﵀ work, entitled al-Kashif, and in addition a second adaptation, Tadhhib at-Tahdhib. World-famous, however, he became for two other works: Siyar A’lam an-Nubala’ (Lives of the Noble Notables), a monumental biographical history of the Islamic civilisation with many thousands of biographies, and Mizan al-I’tidal (The Balance of Equilibrium), a critical dictionary that concentrates specifically on transmitters who have been attacked by someone, regardless of whether that attack was justified. Mizan contains about 11,000 entries.

Ibn Hajar al-Asqalani ﵀, who died in Cairo in 852 AH, completed the edifice with three masterworks: Tahdhib at-Tahdhib, a more extensive adaptation of al-Mizzi ﵀ in twelve parts with about 12,460 biographies of the transmitters from the six and the supplementary works; Taqrib at-Tahdhib, a compact summary with the same names and a one-word verdict for each; and Lisan al-Mizan, a supplement to adh-Dhahabi’s ﵀ Mizan, with once again several thousands of transmitters who do not occur in the main collections but do in other sources. His Lisan counts about 6500 entries. It is this Tahdhib at-Tahdhib, the work of Ibn Hajar ﵀ and not of adh-Dhahabi ﵀, that is to this day the standard reference work on the transmitters of the six.

Whoever adds up these figures gets an idea of what the Islamic civilisation had built: a biographical archive of several hundreds of thousands of named individuals, with a still much greater number of cross-references between teachers and pupils, spun over six to eight centuries. For no other ancient or medieval culture does a comparable prosopographical corpus exist. The Roman Empire has its Prosopographia Imperii Romani, but that work is a twentieth-century modern reconstruction from scattered inscriptions; it does not contain the self-documented teacher-pupil relationships that the Islamic literature recorded in its own time. The Talmudic literature knows hundreds of rabbinic biographies, but no alphabetically ordered corpus, edited in its own time, that can be compared with this scope and methodology.

How It Works in Practice: One Chain Dissected

Every hadith is like a footnote, in the sense that it refers to a chain of men each of whom can be looked up separately in the biographical literature. Let us not leave this abstract and instead draw out one concrete chain as a third-century student would have done. Let us take a hadith from Sahih al-Bukhari, with the chain as al-Bukhari ﵀ gives it:

Haddathana Abdullah ibn Yusuf, qala: akhbarana Malik, an Nafi’, an Ibn Umar, qala: Qala Rasul Allah ﷺ: …

Four men between al-Bukhari ﵀ and the Prophet ﷺ. Four dossiers to open, four series of questions to ask.

First link: Abdullah ibn Yusuf at-Tinnisi, the direct teacher of al-Bukhari ﵀. He was of Egyptian origin (Tinnis, a port city in the Nile delta), studied extensively with Malik ﵀ in Medina, and afterwards returned to Egypt, where he died around 218 AH. When we look him up in Tahdhib at-Tahdhib, we find: reliable (thiqa), of the generation of pupils who visited Malik ﵀ in his last years, a specialist in the transmission of the Muwatta’. Yahya ibn Ma’in ﵀ calls him thiqa, Abu Hatim ar-Razi ﵀ calls him thiqa thabt, an-Nasa’i ﵀ likewise calls him thiqa. He told hadith to al-Bukhari ﵀ in person in Egypt during al-Bukhari’s journey there around 215 AH, when al-Bukhari ﵀ was about twenty-one years old. Dating: confirmed. Character: confirmed. Meeting with pupil: confirmed.

Second link: Malik ibn Anas ﵀, born in 93 AH in Medina, died there in 179 AH. The biographical literature on him fills whole volumes by itself, for he is the eponymous founder of the Maliki school of law. He was a pupil of Nafi’, of az-Zuhri ﵀, of Hisham ibn Urwa ﵀, and hundreds of others. He was a teacher of ash-Shafi’i ﵀, of Abdullah ibn al-Mubarak ﵀, of al-Layth ibn Sa’d ﵀, and of thousands of others. His thiqa status has been confirmed by every generation. His meeting with Abdullah ibn Yusuf is documented. His method of hadith transmission, oriented on the people of Medina and the practice there, is described in detail.

Third link: Nafi’ mawla Ibn Umar, died in 117 AH in Medina, was the freed slave and companion of Abdullah ibn Umar ﵁. He is extensively described in Tahdhib at-Tahdhib: Berber in origin, reportedly taken captive during a campaign of conquest in or around the time of Umar ibn al-Khattab ﵁, and then ransomed by Ibn Umar ﵁, who kept him in his household for about thirty years. Imam Malik ﵀ wrote of him: “When Nafi’ told me a hadith of Ibn Umar ﵁, I asked no one else any more.” Yahya ibn Ma’in ﵀: thiqa. Ahmad ﵀: thiqa. Adh-Dhahabi ﵀ later: thiqa thabt hujja. The chain Malik > Nafi’ > Ibn Umar is known among the hadith scholars as as-silsilat adh-dhahabiyya, the golden chain, and counts as one of the most reliable chains to be found in the whole literature. Imam al-Bukhari ﵀ uses this chain in his work many times.

Fourth link: Abdullah ibn Umar ﵁, born about ten years before the Hijra in Mecca, died in 73 AH in Mecca at about 84 years of age, was the son of the second caliph Umar ibn al-Khattab ﵁ and one of the most productive transmitters among the Companions ﵃. About 2630 transmissions are ascribed to him. He was known for his extraordinary caution in transmission; when he doubted a single word, he refused to pass the hadith on. He knew the Prophet ﷺ personally throughout his whole youth and young adult life. As a boy he was still found too young to fight in the first battles, and only at al-Khandaq was participation permitted him. He is one of the central witnesses for the practice of the prayer, the hajj, and the daily life of the Prophet ﷺ.

Four men, four biographies, four sets of data. When we draw out this chain on a timeline, we get:

• Ibn Umar ﵁: ca. 10 BH - 73 AH (Mecca/Medina)
• Nafi’ ﵀: ca. 30 AH - 117 AH (Medina)
• Malik ﵀: 93 AH - 179 AH (Medina)
• Abdullah ibn Yusuf ﵀: ca. 130 AH - 218 AH (Egypt, with study in Medina)
• Al-Bukhari ﵀: 194 AH - 256 AH (Bukhara/world traveller)

Every transfer is chronologically possible. Every transfer is geographically verifiable. Every transmitter has been independently verified by his contemporaries and later masters. Every interruption or conflict would appear in the biographical dictionaries as a red light. If a later orientalist were to suggest that this chain was in reality a later fabrication, he would have to explain four separate biographies as fabricated, with all their details about teachers, pupils, birthplaces, dates of death, anecdotes and cross-references to hundreds of other transmissions in which the same men occur.

Multiply this one example by the thousands of hadiths in each of the six collections. Multiply it by the tens of thousands of hadiths outside the six. Multiply it by the hundreds of thousands in the broader literature. Then you see what the apparatus rests on. Not on the goodwill of one chain, but on the coherence of a corpus that checks itself in every generation.

The Projection Theory of Schacht

Against this background we must consider the orientalist theory that exercised the most influence in the twentieth century on the Western academic perception of the hadith. It was formulated by Joseph Schacht, a German-British professor of Islamic law, in his work The Origins of Muhammadan Jurisprudence of 1950, followed by An Introduction to Islamic Law of 1964.

Schacht’s central thesis can be summed up in one sentence: the legal hadiths that the Sunni schools of law ascribe to the Prophet ﷺ arose, in the majority of cases, only between the second and third centuries AH, and were then projected back via fabricated chains to the Prophet ﷺ or his Companions ﵃. In Schacht’s words: “every legal tradition from the Prophet, until the contrary is proved, must be taken as not authentic, as a fictitious expression of a legal doctrine formulated at a later date”. The basic evidence would lie in what he called the “projecting back of isnads”: an early legal doctrine that circulated anonymously or under the name of a Tabi’i was given, in a later phase, an added chain that made it end with a Companion ﵁ or the Prophet ﷺ, to raise its authority.

The attraction of Schacht’s theory in the West lay in its elegance. With one hypothesis he could call the whole legal hadith literature into doubt, and, if accepted, it yielded him a legal-historical narrative in which Islamic law was in essence a second- and third-century Iraqi invention, instead of a direct continuation of the practice of the Prophet ﷺ and his Companions ﵃.

What Schacht could not explain, however, or deliberately passed by, is precisely the apparatus we have described in this article. A forger who in the third century AH wished to produce a hadith with a chain did not stand before an empty field. He stood before a field already filled with the biographical dossiers of hundreds of thousands of scholars, drawn up by men whose moral and scholarly authority stood above every suspicion. Imagine what such a forger would have to do.

He would have to invent a chain that not only was chronologically correct (that is still the easiest part), but that also closed geographically (the two men had to have been in the same city at some point in their lives), that agreed with the documented teacher-pupil relationships of every link in Tahdhib at-Tahdhib and related works, that did not conflict with the known sectarian preferences of the transmitters (a state servant of the Umayyads would not have studied with a fervent Shi’i, and vice versa), and that moreover did not go against, in content, other transmissions on which the scholars in six generations did agree.

Moreover, the forger would not only have to invent the men he used in the chain, but also weave them through the whole rest of the literature. When he claims that Nafi’ ﵀ passed on a hadith of Ibn Umar ﵁ in 100 AH in Medina to a certain Hisham, then that Hisham must also surface somewhere in other chain data, otherwise he stands as an isolated figure in the open air and is immediately conspicuous. The hadith scholars were not naive. An unknown name in an isolated chain was the first red flag they signalled. It was precisely this anchoring that led Azami ﵀ to write that there is no reason whatsoever to distrust the chain Malik > Nafi’ > Ibn Umar: every link is confirmed by dozens of independent transfers in the literature.

Suppose that our hypothetical forger nonetheless overcame these obstacles, then he would next have to ensure that his fabrication remained unnoticed by every master of Jarh wa-Ta’dil. Yahya ibn Ma’in ﵀, who for the distinguishing of errors laid the notes of eighteen pupils of Hammad side by side and whose heart trembled at every new hadith, would have to be silenced. Ali ibn al-Madini ﵀, who had developed ilm al-ilal into a precision instrument, would have to be misled. Ahmad ibn Hanbal ﵀, who was described by his students as someone who smelled a fabricated transmission as a money-changer smells a false coin, would have to be outplayed. And all these masters would have to be misled simultaneously and unanimously for one hadith, and then again for the next, and the next, and the next, thousands of times, in six parallel compilations of six different men.

Moreover, there is a numerical problem. We have seen that al-Bukhari ﵀ himself says he examined 600,000 hadiths. Other masters give comparable or greater numbers. The projection theory requires that the vast majority of these chains is fiction. But every fictitious chain requires a fictitious set of contemporaries, and those contemporaries would in their turn need fictitious biographies, and those fictitious biographies would, to be credible, in their turn have to occur in other sources, and so forth. The required quantity of coherently fabricated data exceeds every realistic estimate of what a conspiracy of scholars in six generations could have produced, without a single confession or early refutation.

The Islamic hadith science knows, for that matter, its own disasters, and it has named them specifically. There were in the second century indeed movements of forgery, organised by political factions, by sectarian groups (some extreme Shi’i directions, some extreme Khariji factions), by zealots who with good intentions wished to see the Prophet ﷺ exalted. These forgeries the hadith scholars have not denied; they have precisely catalogued them. Al-Azami refers in this connection to the counts that the later critics kept of the invented material per subject: dozens of forgeries about the Prophet ﷺ, dozens about the first caliphs, and so forth, each time with the forger named. Works such as Ibn al-Jawzi’s ﵀ Kitab al-Mawdu’at and al-Suyuti’s ﵀ al-La’ali al-Masnu’a contain thousands of suspected or proven forgeries, with the names of the forgers and the way in which the deceit was discovered. If Schacht’s theory were true, one would have to find the six canonical works in these forgery books as well. That does not happen. The forgery books concentrate on material outside the canon, material that had been explicitly rejected by the masters of Jarh wa-Ta’dil. The apparatus worked.

Wensinck and the Hadith of the Five Pillars

A second twentieth-century orientalist who occupied himself with the hadith and likewise raised a specific criticism was Arent Jan Wensinck, a Dutch scholar who in Leiden compiled the first real concordance to the hadith literature. His Concordance et indices de la tradition musulmane, whose publication began in 1936 and which was completed by others after his death, is to this day a useful reference work that even modern hadith scholars use.

Wensinck’s own theoretical contribution was less careful than his instrumental work. His dissertation Mohammed en de Joden te Medina of 1908 dealt with the Jewish communities of Medina and the relations of the Prophet ﷺ with them; it is a different subject from hadith criticism. His argument about the creeds he formulated later, in his English-language work The Muslim Creed of 1932. In it he argued that the hadith about the five pillars of Islam (the well-known transmission in al-Bukhari ﵀ and Muslim ﵀ that runs: Buniya al-Islam 'ala khams, “Islam is built on five [pillars]”) could not be authentic. His reasoning: the formal enumeration of five pillars as a structured list would be a later theological codification, arisen in the same milieus where the first creeds were also formulated.

Al-Azami has given a specific refutation in his Studies in Early Hadith Literature and in his later work On Schacht’s Origins of Muhammadan Jurisprudence (1985). He traced the chain of the hadith back, identified the transmitters (the hadith comes among others through Ibn Umar ﵁, through Hanzala ﵀, through Ikrima ibn Khalid ﵀, and is included by al-Bukhari ﵀, Muslim ﵀, at-Tirmidhi ﵀ and an-Nasa’i ﵀), showed the chronology, and pointed out that elements of the five pillars were already anchored in the earliest practice and that the same hadith with the same structural build-up is transmitted from several independent Companions ﵃. The independent variation between the transmissions, in details such as the order of the pillars, rules out a common late fabrication. Wensinck’s hypothesis is thereby refuted not by dogma, but by the working of the apparatus itself, the same apparatus that has been described in this article.

It is remarkable that both Schacht and Wensinck, both familiar with the Arabic sources, largely left the apparatus of Jarh wa-Ta’dil out of their argumentation. They consulted the biographical dictionaries, that cannot be denied, but they did not acknowledge its cumulative force for their own hypotheses. It has been, in retrospect, a weakness of the Western study of the hadith in the first half of the twentieth century that it did not take the Islamic self-criticism in its own seriousness. The community had itself spent centuries separating what was strong from what was weak; a later academic who dismissed the whole harvest in one sweep was not more careful than his predecessors, but less.

The Apparatus Kept Working

All this brings us to a threshold. One might think that the science of narrator criticism had completed its work with the canonical six and thereafter came to a standstill, and that what was still written after al-Bukhari ﵀ and Muslim ﵀ was no more than pious repetition of a once-fixed judgement. That thought is incorrect. The apparatus we have described in this article did not close itself off with the canon. It remained, generation after generation, in the service of scholars who did not copy the canon but tested it anew with the same means.

Half a century after al-Bukhari ﵀ there worked in Nishapur a scholar to whom his contemporaries awarded the title Imam al-A’imma, imam of the imams. He composed a Sahih of his own, not by copying out the canonical two but by leading every chain anew along his own critical apparatus, and on points where he deviated he noted why. Of that work more than a judgement has been preserved: there exists a physical folio, a written page from an early manuscript, on which stand the names of a chain of transmission that can be traced in precisely the same biographical dictionaries that we have gone through in this article.

Who that scholar was, how his Sahih confirmed the canonical project with independent means, and why the surviving folio from his work shows that narrator criticism remained a living and documented instrument instead of a congealed inheritance, is the subject of the next article. We shall see how the science that had been built up in the workshops of Shu’ba ﵀, Yahya ibn Ma’in ﵀ and Ahmad ﵀ still proved, in a Nishapuri study at the beginning of the fourth century AH, to be what it had always been: an implacable instrument for separating true from false.