Untangling the “Unwritten Documents” of the Prophet Muḥammad. An Isnād-cum-Matn Analysis of Interwoven Traditions

Abstract

Since the earliest studies of Islam by non-Muslims were carried out, variant traditions (aḥādīth) have been regarded as a proof of forgery or editing within the ḥadīth material. Early studies have shown that variances are the result of different processes, some intentionally and others mistakenly; some caused by editing processes, while others through the process of transmission across the first centuries of Islam. During the transmission process, or the genesis of a tradition, accounts are constantly shaped and adjusted. The use of topoi forms a part of this process as well as the inclusion of motifs in different accounts. The present article will explore one of these motifs, specifically, the instruction of the Prophet Muḥammad, on his deathbed, to bring him writing materials so that he could prepare a document for his community. This motif appears in a number of accounts with different settings, characters and details on the nature of the document itself. This article examines whether there exists a direct relationship between the different accounts and, if so, what does this mean. Through this study, we will see that additional motifs have been added to this tradition during its transmission process and that some of these motifs can be attributed to regionalisation or specific transmitters.

1. Introduction

Contradictory stories or discrepancies within traditions sit at the heart of discussions about the usefulness of the Islamic tradition (ḥadīth pl. aḥādīth) material as a source of information on the early centuries of Islam. The discrepancies have led scholars to develop the science of ḥadīth criticism and new methodologies and theories to study the ḥadīth material. This was especially important for those traditions in which a sunna (custom) of the Prophet Muḥammad was described, because his behavior and sayings form instructive examples of the ideal Islamic way of life for Muslims. Contradictions also appear in biographical accounts of his life that deal with specific events. One of those events, about which a number of conflicting narratives exist, takes place towards the end of the Prophet’s life and is framed as his final illness.

The narratives are centred around Muḥammad’s command to bring him writing materials in order to pen a document (kitāb) for his community. The wider details of the tradition, such as who was present and other minor details, differ. Nevertheless, the final result is the same: Muḥammad foregoes his original instruction and the document remains unwritten. The setting and almost identical wording of the instruction indicate that the narratives are derived from a common source. Although the exact nature of the document remains unknown, some narratives do allude to its content.

It is precisely this lack of clarity about the content that has led to heated Sunni-Shīʿī contestations over the centuries. Examples of those discussions can be found within the works of classical ḥadīth scholars, such as Ibn Kathīr (d. 774/1373) and Ibn Ḥajar (852/1448).1 Even today, lively debates continue on the internet.2 The discussions focus on the purpose and content of the document, but often also discuss the variant narratives in which Muḥammad’s request is embedded. While the narratives typically state the purpose of the document, which is to serve as guidance for Muḥammad’s community, its content remains unknown because the Prophet eventually refrains from writing it. This ambiguity allows for diverse interpretation of the document’s content and import. For example, that it would be instructive with regards to the succession to the Prophet, provide specific guidelines for his fellowship, or serve as a test for his community.3 Ibn Kathīr, therefore, complains:

“[t]his ḥadīth has served to feed the imaginations of certain foolish persons, who advocate improper innovative practices. These adherents of the shīʿa and others, all claimed that the Messenger of God4 wished to write in the document […] what they purpose in their own statements. […] Whatever the Messenger of God wished to write came previously in those aḥādīth that lend themselves to clear and unambiguous interpretation.”5

Ibn Kathīr here refers to the different interpretations, both by the ambiguity in this tradition and its variant narratives.

The present article will not go into the various discussions that were triggered by this ambiguity, but instead focus on the development of the narratives in the ḥadīth material. The aḥādīth will be studied using the isnād-cum-matn (ICM) analysis developed by Harald Motzki and Gregor Schoeler.6 This method is based on the fact that a ḥadīth usually consists of a text (matn pl. mutūn) and a chain of transmitters (isnād pl. asānīd), which claim to represent the transmission path between the first narrator of the story and the collection into which this tradition culminated. The number of people in a chain varies from source to source and can range from four to five people in the earliest collections of the third Islamic century to fourteen or more people in the later collections of the eighth century.7

Two problems are here highlighted within the Islamic tradition material: First, that the available collections date from a period that is at least two hundred years later than the time they describe. The question then is whether the event described took place in the way described or at all? The second problem is that the chain of transmitters represents a transmission process in which changes inherently take place. What we have at our disposal is the end result, being texts that the author of the collection claims to have received via the persons in the accompanying chains. The conflicting information in the narratives on Muḥammad’s unwritten document is indicative of the fact that the material has been edited. To answer the question of who is responsible for each part of the account, we need to look beyond the versions in the collections. On the basis of these collections, we can only study the conscious and unconscious choices collectors have made in the material they present. An example of this type of analysis is the discussion of Gurdofarid Miskinzoda. She shows that the narratives were linked to the discussions about the succession of the Prophet and the status of other writings with guidelines for the community in addition to the Qurʾān, and how the position of the collectors developed.8

The ICM analysis is aimed at the period before the ḥadīth collections. It compares textual variations between traditions with overlapping asānīd, i.e., the traditions have a number of people in common. If the asānīd represent the actual transmission path of a tradition, in case of overlap, part of the text should be the same. If there is textual overlap without the chains, which have a similar overlap, then one of the asānīd is (partly) forged or erroneous.9 However, if traditions report the same topic, but do not overlap in the chains nor are they comparable in terms of word usage and structure of the text, then we are dealing with two separate tradition complexes, i.e., two separate stories describing the same event. The following ICM analysis of the traditions of Muḥammad’s unwritten document will show to whom the earliest version of a narrative can be attributed and which parts that basic narrative consisted of. Furthermore, it will show how the narrative evolved and how, over time and across specific regions, other textual motifs have been added and omitted, creating new narratives.

2. Isnād-cum-Matn Analysis Applied

Based on the asānīd, the traditions in which the motif appears can be divided in five groups. The first group contains traditions ascribed to the famous Qur’ān scholar and Companion of the Prophet, ʿAbd Allāh ibn ʿAbbās (d. 67/686–7). The majority of the traditions (35) belong to this group. The second group10 is ascribed to the Companion Jābir b. ʿAbd Allāh (d. 78/697) and contains seven traditions. The third group, consisting of two traditions, is traced back to the Companion and second caliph ʿUmar b. al-Khaṭṭāb (d. 23/644). The fourth group, with five traditions, is ascribed to Muḥammad’s nephew and son-in-law, ʿAlī b. Abī Ṭālib (d. 40/661), and the last group with the second largest number of traditions (15)11 is traced back to ʿĀʾisha (d. 58/678), said to have been Muḥammad’s favourite wife and daughter of the first caliph Abū Bakr. All of the alleged first transmitters of these traditions belonged to the circle of Muḥammad’s intimates and the sīra material describes their frequent interaction with the Prophet.

The ICM analysis begins with a brief biography of the individuals to whom the traditions are attributed per group. To obtain an overview of which people have handed down the tradition according to the asānīd and to identify common transmitters per group, the asānīd have been drawn in a figure. The earliest common transmitter, the so-called common link, is by way of hypothesis assumed to be the distributor of the tradition in question. Subsequently, the mutūn of the traditions within a given group are first compared with each other and then with the groups discussed earlier.

2.1. Group 1: The Ibn Abbās Narrative

The first group of the traditions is traced back to ʿAbd Allāh b. ʿAbbās. Ibn ʿAbbās was related to the Prophet through his father al-ʿAbbās, the brother of Muḥammad’s father ʿAbd Allāh, and his mother, who was the sister of the Prophet’s wife Maymūna.12 At the time when the story is set, the death of the Prophet, he was still young, between ten and fifteen years old.13 Ibn ʿAbbās is one of the most controversial Companions of Muḥammad within scholarly debates. In Muslim scholarship, he is revered as one of the greatest Qurʾān exegetes and his name appears in the asānīd of countless traditions on the Prophet Muḥammad.14 In non-Muslim scholarship, the authenticity of the ascription of the majority of these traditions is criticized. Herbert Berg and Claude Gilliot argue that Ibn ʿAbbās should be regarded as a symbolic figure to authenticate the information in the tradition.15

Figure 1 is a simplified representation of the chains showing only the earliest generations of transmitters.16 The common link according to Figure 1 would be Ibn ʿAbbās.17 In order to determine whether these traditions are indeed from Ibn ʿAbbās, the texts (mutūn) of the traditions will be compared. The textual analysis revealed six different versions of the tradition about the unwritten document, which correspond to the numbers indicated in Figure 1. Three of them go back to Saʿīd b. Jubayr (d. 94/714), a Kufan scholar of the Qurʾān, jurisprudence and Ḥadīth, who, according to Muslim biographies, was a student of Ibn ʿAbbās.18

2.1.1. Ibn ʿAbbās version 1—Sufyān b. ʿUyayna

The Ibn ʿAbbās version 1 of Sufyān b. ʿUyayna (d. 198/814), the first transmitter common to all asānīd in this subgroup, from Sulaymān b. Abī Muslim (n.d.), appears most often in the Islamic ḥadīth collections. A reconstruction of Sufyān’s text19 based on the thirteen20 traditions I found is:

[…] on the authority of Ibn ʿAbbās, he said:21 “Thursday, what a Thursday!” (Then Ibn ʿAbbās wept so hard that his tears wet the pebbles. Saʿīd or someone else asked Ibn ʿAbbās, “What about Thursday?”)22 The pain of the Prophet became severe, so he said, “If you bring me (iʾtūnī) [something]23, then I will write a document for you (lakum), after which you will never go astray.” People began to argue with each other, although a dispute in front of a prophet is improper. They said, “What is the matter with him? Is he talking deliriously? Ask him for an explanation.” (So they went back to him, repeating those remarks to him,)24 and the Prophet replied, “Leave me alone! The state I am in is better than that for which you are calling me.” He instructed (for them)25 three things, “Expel the polytheists from the Arabian Peninsula, give the quantity of water sufficient to pass therewith from one watering-place to another to the delegations, as I used to do.” [Sulaymān remarked], “Either Saʿīd said nothing about the third one intentionally, or he said it and I have forgotten it.”

Ibn ʿAbbās’ exclamation on “Thursday”, at the beginning of the tradition, places the subsequent event on this day. Muḥammad is in a room with several people. He is very sick at the time and wants to prepare a document to guide the people (“for you”: lakum). In fact, he appears to be so sick that people think he is delirious. The debate that arises after his request becomes too much for him and he asks everyone to leave without having written the document. At the end, it is said that Muhammad commanded three things, only two of which are mentioned. This implies that he wanted to record these three commands in the document.26

2.1.2. Ibn ʿAbbās Version 2—Mālik b. Mighwal

The second Ibn ʿAbbās version that is traced back to Saʿīd b. Jubayr, according to the asānīd, is from Mālik b. Mighwal (d. 159/776) via Ṭalḥa b. Muṣarrif (d. 113/731), both of whom were from Kufa. This version is shorter than Sufyān’s, but similarly begins with Ibn ʿAbbās’ exclamation on Thursday.

[…] on the authority of Ibn ʿAbbās, he said27: “Thursday, what a Thursday!” Then I [ Saʿīd b. Jubayr] looked at his tears running down his cheeks as if they were a string of pearls. He said: The Messenger of God said: “If you bring me (iʾtūnī) a shoulder blade and an inkpot (or a tablet and an inkpot)28, then I will write for you a document, after which you will never go astray.” They said, “The Messenger of God was talking deliriously.”29

The cry motif is present in both Ibn ʿAbbās versions, but has been formulated differently. The comparison of the tears with pearls can only be found in version 2 of Mālik b. Mighwal from Ṭalḥa b. Muṣarrif. Muḥammad’s instruction to bring writing materials and the reason for that are the same. The only reference to the discussion that then takes place in Ibn ʿAbbās version 1 of Sulaymān al-Aḥwal is that they (several persons) say that Muḥammad was delirious. This is also the end of the story in this second Ibn ʿAbbās version; the threefold command motif of version 1 is completely missing. The comment about the delirium is the only indication that this story is set during the Prophet’s illness.

2.1.3. Ibn ʿAbbās Version 3—Sulaymān al-Aʿmash

The third version of Saʿīd b. Jubayr, according to the asānīd, comes from the Kufan scholar Sulaymān al-Aʿmash (d. 148/765) via his fellow townsman ʿAbd Allāh b. ʿAbd Allāh [al-Rāzī] (n.d.). Two slightly different accounts have been preserved by Ibn Saʿd and al-Ṭabarānī. Ibn Saʿd’s account is30:

[…] on the authority of Ibn ʿAbbās, he said: The Prophet became sick on Thursday, so he, i.e., Ibn ʿAbbās, began to cry and say, “Thursday, what a Thursday!” The pain of the Prophet became severe, so he said, “If you bring me an inkpot and a piece of paper31, then I will write for you a document after which you will never go astray.” He said: Some of those who were with him said that the Prophet is certainly talking deliriously. He said: It was said to him (=Prophet), “Shall we not bring you what you asked for?” He replied, “Or after what?” He said: So he did not summon it.

Al-Ṭabarānī’s tradition is a shortened version of this:32

[…] on the authority of Ibn ʿAbbās, may God be pleased with them: When it was Thursday, what a Thursday! Then he cried and said: The Messenger of God said, “If you bring me (iʾtūnī) a piece of paper and an inkpot, then I will write for you a document after which you will never go astray.” They said, “The Messenger of God is talking deliriously.” Then they said nothing and he (=Prophet) said nothing. They said, “Shall we not bring [it] to you later?” He replied, “After what?”

Although both traditions share similarities with the first two discussed Ibn ʿAbbās versions, both attributed to Saʿīd b. Jubayr, they are distinct from one another. The two traditions contain the same deviant motifs, which is why they are considered one version.33 Like Ibn ʿAbbās versions 1 and 2, version 3 begins with the exclamation motif and the cry motif. However, no description is given of the crying and this version is clear about the writing material, a piece of paper and an inkpot. Again, Muḥammad is thought to be delirious. The question they then ask the Prophet and his answer are unique to version 3, although they seem to be a vague echo of version 1. As in version 2, the threefold command is missing.

The unique motifs of each version can be provisionally attributed to the partial common links from Figure 1, Sufyān b. ʿUyayna for Ibn ʿAbbās version 1, Mālik b. Mighwal for version 2, and Sulaymān al-Aʿmash for version 3.

The motifs the three Ibn ʿAbbās versions have in common can be tentatively attributed to the first transmitter common to these versions, Saʿīd b. Jubayr (d. 95/714):

• The exclamation of Ibn ʿAbbās “Thursday, what a Thursday!”;
• The crying of Ibn ʿAbbās (even if the three versions differ in the details);
• Muḥammad’s instruction for writing materials (even though the three versions differ in the materials listed);
• The reason for his instruction: to write a document for them after which they will not go astray.
• People who wonder if Muḥammad is delirious.

Only Ibn ʿAbbās version 1 explicitly states that Muḥammad is seriously ill, as well as one tradition of version 3, while version 2 and the other tradition from version 3 state this implicitly through the question of whether he is delirious. The use of a plural form in the conjugation of the verb (“iʾtūnī”) indicates that several people are present during this event. The presence of several people and the motif of Muḥammad’s illness can therefore also be attributed to Saʿīd b. Jubayr.

2.1.4. Ibn ʿAbbās Version 4—Ibn Shihāb al-Zuhrī

Ibn ʿAbbās version number 4, according to the asānīd, does not come from Saʿīd b. Jubayr but from the famous Medinan ḥadīth scholar and jurist Ibn Shihāb al-Zuhrī (d. 124/742) via ʿUbayd Allāh b. ʿAbd Allāh b. ʿUtba (d. appr. 102/720), who also lived in Medina. The vast majority of this group of traditions comes from the Yemenite scholar ʿAbd al-Razzāq b. Hammām al-Ṣanʿānī (d. 211/826) (see Appendix B). His nine traditions are very similar, making it possible to make a reconstruction of ʿAbd al-Razzāq’s text:34

ʿAbd al-Razzāq—Maʿmar on the authority of al-Zuhrī on the authority of ʿUbayd Allāh b. ʿAbd Allāh (b. ʿUtba) on the authority of Ibn ʿAbbās, he said: When the Messenger of God reached the point of death, during which there were men [present]in the house, among whom was ʿUmar b. al-Khaṭṭāb, the Prophet said, “Now then! I shall write for you a document after which you will not (never)35 go astray.” ʿUmar said, “The Messenger of God became overpowered by the pain and you have the Qurʾān. The book of God is enough for us.” The people in the house disputed and quarrelled. Some of them said, “Make them let him (/the Messenger of God)36 write a document for you after which you will not go astray.” Some of them said what ʿUmar said. When the nonsense (/noise)37 and the disagreement intensified in front of the Messenger of God, the Messenger of God said, “Go away.”

ʿUbayd Allāh (/ʿAbd Allāh)38 said: Ibn ʿAbbās used to say, “The most terrible disaster is that their disagreement and their noise came between the Prophet and him writing that document.”

In addition to ʿAbd al-Razzāq’s version from Maʿmar, there is a second Maʿmar version from his fellow townsman Hishām b. Yūsuf (d. 197/813).39 The biggest differences from ʿAbd al-Razzāq’s text are al-nabī (the Prophet) instead of rasūl Allāh (the Messenger of God), the omission of the preposition ghalabahu al-wajʿ (the pain overpowered him) instead of ghalaba ʿalayhi al-wajʿ and the Prophet saying “Go away from me” (qūmū ʿannī) instead of “Go away” (qūmū). In addition, there are three traditions from Yūnus b. Yazīd from al-Zuhrī, which are very similar to ʿAbd al-Razzāq’s text, but, like Hishām’s version, nevertheless contain unique formulations40, making it possible to ascribe Ibn ʿAbbās version 4 to al-Zuhrī and to date it to the first quarter of the second Islamic century.

In version 4 of al-Zuhrī, the beginning with Ibn ʿAbbās’ tear motif and Thursday’s exclamation is missing, but a similar emotional statement returns at the end, in which Ibn ʿAbbās speaks of a disaster. Like the first three versions, the event takes place during Muḥammad’s disease. Although the Prophet expresses the same desire to write a document for his community, there is no mention of writing materials. For the first time, one of the people present is mentioned by name, ʿUmar, the second caliph of the Islamic empire and the one who was involved in appointing Abū Bakr as the first caliph after the death of the Prophet.41 He is the one who makes the call not to obey the Prophet’s wish. His argument is that no second document is needed besides the Qurʾān.42 The discussion in Ibn ʿAbbās version 4 is more drawn out by mentioning ʿUmar’s counter argument. Ibn ʿAbbās’ statement at the end on the disaster makes it clear that the Prophet did not write the document. Also in version 4, the only reference to the content of the document is given in the Prophet’s request, to not let his community go astray. No further information on its content is provided.

The following motifs can be tentatively attributed to al-Zuhrī, as they appear only in his accounts: the disaster motif, the omission of writing materials, ʿUmar’s presence and his argument regarding the Qur’ān. That would also mean that the motifs that al-Zuhrī’s version has in common with Saʿīd b. Jubayr’s other three versions could possibly come from Ibn ʿAbbās, as he is the only transmitter common to all traditions. These are: the setting of the story during the Prophet’s illness, his desire to write a document for the people (“you”) to not let them go astray, the quarrelling over the Prophet’s request, the idea that Muḥammad’s illness is the cause of his request, Muḥammad not writing the document, and finally, the emotional outburst of Ibn ʿAbbās (in the versions of Saʿīd b. Jubayr expressed with tears and the Thursday exclamation, in that of al-Zuhrī with the disaster motif). However, there are two other versions attributed to Ibn ʿAbbās that we must include in the comparison before any more definitive statements can be made about the attribution of the motifs to individuals.

2.1.5. Ibn ʿAbbās Version 5—Layth b. Abī Sulaym

Ibn ʿAbbās versions 5 and 6 are not as widely preserved in the ḥadīth collections as the first four versions. Version 5 comes in three traditions, which can be found in the collections of Ibn Ḥanbal and al-Ṭabarānī. The common link according to the asānīd is the Kufan traditionist Layth b. Abī Sulaym (d. 138/755 or 143/761) (see Figure 2). All three traditions are much shorter than the versions discussed above and differ from each other.

Version T1 Ibn Ḥanbal:43

When the Messenger of God reached the point of death, he said, “If you bring me a shoulder blade, then I will write for you on it a document [so that] two men from among you do not disagree after me.” He said: The people (qawm) began to shout. The woman said, “Woe unto you! [It is] the command of the Messenger of God!” (wayḥakum ʿahd rasūl Allāh).

Version T2 al-Ṭabarānī:44

The Messenger of God called for a shoulder blade. He said, “If you bring me a shoulder blade, then I will write for you a document [so that] after me you never disagree.” There was a stir (laghaṭ) among the people who were with him. A woman among those in attendance said, “Woe to you! [It is] the command of the Messenger of God!” One of the people said, “Be quiet! You have no knowledge!” (lā ʿaql laki). The Prophet said, “You have no understanding!” (antum lā aḥlām lakum).”

Version T3 al-Ṭabarānī:45

The Messenger of God said, “If you bring me a shoulder blade and an inkpot, then I will write for you a document on which two men will not disagree.” He said: They delayed (fa-abṭiʾū) the shoulder blade and the inkpot and God took him (fa-qabaḍahu Allāh).

Although the three traditions differ from each other, they still have a few characteristics in common, which makes them classifiable under the same version. All three mention a shoulder blade as writing material, which is unique for this Ibn ʿAbbās version 5. Furthermore, unlike the Ibn ʿAbbās versions 1–4, the purpose of the document is to avoid disagreement between two men. Since this is mentioned in two traditions (T1 and T3), I also consider this a peculiarity of the Ibn ʿAbbās version from Layth. The same applies to the noise or shouting after the request of the Prophet and the correction of those present by the (unknown) woman (both present in T1 and T2). Two traditions (T1 and T3) indicate that this event took place just before the death of the Prophet.46 Of the other elements in the text that only appear in one tradition, it is not clear whether they come from the transmitters above Layth in Figure 2 or if they come from variations in the transmission by Layth. Only the textual elements appearing in two or all three of the traditions can be attributed to the common link of this group, Layth b. Abī Sulaymān.

2.1.6. Ibn ʿAbbās Version 6—Ibn Saʿd

The last Ibn ʿAbbās version can be found in one tradition in Ibn Saʿd’s al-Ṭabaqāt al-Kubrā with the Medinan isnād Muḥammad b. ʿUmar [al-Wāqidī] (d. 207/823) -> Ibrāhīm b. Ismāʿīl b. Abī Ḥabība (d. 165/781–2) -> Dāwud b. al-Ḥuṣayn [al-Qurashī al-Umawī] (d. 135/752–3) -> ʿIkrima [al-Qurashī al-Hāshimī], a mawlā of Ibn ʿAbbās who died in 105/723–4 -> Ibn ʿAbbās (d. 67/686–7).47 The text of the traditions is:

[…] on the authority of Ibn ʿAbbās, that the Prophet said during his illness what he died of, “If you bring me an inkpot and a piece of paper, then I will write for you a document after which you will never go astray.” ʿUmar said, “Who belongs to so-and-so and so-and-so of the cities of the Byzantines? The Messenger of God is not dead until we conquer them and if he dies, we will wait for him as the Banū Isrāʾīl waited for Moses!” Zaynab, the wife of the Prophet, said, “Do you [people] not listen to the Prophet charging you?” They shouted (laghaṭū) and so he [the Prophet] said: “Get out!”. When they left, the Prophet was taken on the spot (qubiḍa al-nabī makānahu).

As in the other Ibn ʿAbbās versions, the event takes place during the Prophet’s illness, but in this version, it happens even during the last moments of his life. In the last sentence, we are told that the Prophet passed away while those present are still leaving the place. The same time and the same verb (although active form qabaḍa instead of the passive form qubiḍa in Ibn Saʿd’s tradition) was mentioned in tradition T3 of Ibn ʿAbbās version 5. The writing materials mentioned, inkpot and a piece of paper, previously appeared only in both traditions of Ibn ʿAbbās version 3 of Sulaymān al-Aʿmash, one of them from the work of Ibn Saʿd (tradition SA1) and in one tradition (S2) of Ibn ʿAbbās version 1, also part of Ibn Saʿd’s work. Tradition S2 is the only one among the traditions of Ibn ʿAbbās version 1 that contains the combination of inkpot and a piece of paper. If we combine this departure from the other version 1 traditions with the fact that all three Ibn ʿAbbās traditions that mention inkwell and a piece of paper as writing materials are included within the same section of Ibn Saʿd’s work, he seems to be the one responsible for the unity of writing materials within these three Ibn ʿAbbās traditions. Yet he also mentions two other traditions from Ibn ʿAbbās, one with different writing materials and one without writing materials. It is therefore not possible to determine with certainty to whom the deviating formulations belong.48

As in Ibn ʿAbbās version 4 of al-Zuhrī, it is ʿUmar who responds to the Prophet’s request. His arguments for not responding to the request, however, have nothing to do with objecting the writing of the document as in version 4, but with rejecting the death of the Prophet. The first reference to the cities of the Byzantines is anachronistic in the context of this story, as the conquests of Byzantine cities did not take place until the caliphate of ʿUmar (r. 13-23 AH/634-644 CE).49 ʿUmar’s rejection of the death of Muḥammad in the second part of the sentence comes from another tradition in which ʿUmar makes a similar comparison with Moses during a speech after the Prophet’s death.50 Since none of the other Ibn ʿAbbās traditions contain a similar rejection by ʿUmar, the Moses motif is taken from another tradition and placed in the unwritten document narrative rather than vice versa.

The story continues with the Zaynab motif. She seems to be responding to ʿUmar’s statement, but the use of the plural form in the verb (tasmaʿūniī) and the suffix (ilaykum) makes it clear that she is addressing several persons. Since the sentence about the arguing is placed after her comment, the arguing appears to be due to her comment rather than the other way around, as we saw in tradition T1 of version 5. The last part of this tradition about the quarrelling and the Prophet who sends them away is familiar again and appears in different Ibn ʿAbbās versions. This is where the story ends, and while the purpose of the document is apparent from the Prophet’s request, further information on its content is lacking.

2.1.7. Conclusion Ibn ʿAbbās Traditions

The textual analysis of the traditions on the unwritten document attributed to Ibn ʿAbbās showed that there are six different versions that differ to a greater or lesser degree from each other. This confirms the branches that the isnād bundle shows in Figure 1. Each version contains one or more motifs that only appear in that particular version. These characteristic motifs can thus be attributed to the common link of that particular version (see Table 2).

In addition, however, the six versions also contain a number of common motifs that are explicitly or implicitly mentioned in the texts. The first is the time of the event: The Prophet is very sick. It is unclear on which day the event takes place. According to the first three versions, this is on Thursday, as appears from the exclamation of Ibn ʿAbbās. Versions 4, 5 and 6 do not mention a specific day, but state that Muḥammad is about to die, with versions 5 and 6 even indicating that he dies on the same day.51 Miskinzoda’s observation that most traditions agree on Thursday and some on Monday,52 has to be adjusted. The dating of this event on Thursday in versions 1–3 must be attributed to Saʿīd b. Jubayr, the common link of versions 1–3. While Saʿīd b. Jubayr’s traditions are indeed the most common in the collections and consequently, his version seems to be the most accepted, this should be counted as one version, as they all come from the same transmitter. The importance of the difference in dating is the degree of drama in which the event is placed. When placed during the Prophet’s illness, there is still hope for a second opportunity for the Prophet to write the document. However, when placed on his day of death, then the chance of another opportunity is lost. The internal dating of the event of Ibn ʿAbbās versions 5 and 6 to the day of death enhances the drama of the event.

The second motif is the presence of several people, although this is not explicitly stated in most versions, but is indicated by means of the plural form of person suffixes and verb conjugations (f.e. iʾtūnī and lakum).

The third motif is the Prophet’s instruction to bring writing materials so that he can write a document for them after which they will not go astray. This is the only sentence that appears almost verbatim in all traditions, except for the type of writing materials. The materials vary per version and sometimes even within the traditions of one version.53 Only version 5 is slightly different in that the purpose of the document is to avoid disagreement. This deviation must therefore be attributed to Layth b. Abī Sulaymān, the common link of this subgroup of traditions. The last common “motif” is that none of the versions refers to a written document as the end result nor to the content of the document that the Prophet intended to prepare. The threefold command of the Prophet in version 1 of Saʿīd b. Jubayr, which seems to refer to the content of the document, is not confirmed by his other two versions (Ibn ʿAbbās versions 2 and 3) and must therefore be attributed to Sufyān b. ʿUyayna and dated to the last quarter of the second Islamic century, as he died in 198 AH. Ibn ʿAbbās version 4 is the only one that explicitly states that the document is not written. Al-Zuhrī indicates with an additional isnād for this statement that ʿUbayd Allāh heard this from Ibn ʿAbbās and that the latter made this statement more often (kāna Ibn ʿAbbās yaqūlu). The explicit mention that the document was not written must therefore be attributed to al-Zuhrī.

The only transmitter common to all versions is Ibn ʿAbbās. The common motifs should therefore be attributed to him. Given the discussion in non-Muslim scholarship about the authenticity of attribution of traditions to Ibn ʿAbbās, we must ask whether the common motifs may not originate from Ibn ʿAbbās, but perhaps from Saʿīd b. Jubayr? Further, are versions 4, 5 and 6 attributed to another informant to give more authority to the attribution to Ibn ʿAbbās by a spread of asānīd? Both options do not seem likely here. Versions 4, 5 and 6 contain its own characteristics. They differ more from each other and from versions 1, 2 and 3, while the latter are more similar in content, sometimes almost verbatim. The greater degree of similarity corresponds to the information from the chains of transmission, as the chains of versions 1–3 have one transmitter more in common, i.e., Saʿīd b. Jubayr, than the other three versions. The larger deviations in versions 4, 5 and 6 indicate that they are derived from other transmitters, which is visible in their asānīd. Yet, all six versions contain a common core. If the discrepancies are explained by a difference in the transmitters mentioned in the asānīd, then a common transmitter—in this case the only transmitter all asānīd have in commong (Ibn ʿAbbās)—should also provide similar information in all versions: the above discussed common narrative motifs.

Beside the motifs that appear in all six versions of Ibn ʿAbbās, there are also a few that only appear in several Ibn ʿAbbās versions. Can they also be ascribed to Ibn ʿAbbās? Despite the aforementioned variation in writing material, the inkpot occurs in four Ibn ʿAbbās versions (versions 1, 2, 4 and 6) and in one of the three traditions of version 5. Versions 1, 4, 5 and 6 mention discord (tanāzuʿ or ikhtilāf) or uproar (laghaṭ) after the Prophet’s request. In several versions, the Prophet orders those present to leave (versions 1, 4 and 6). In versions 1–3 and 4 it is suggested that Muḥammad’s illness is the cause of his request (formulated in versions 1–3 as delirious and in version 4 as overcome by pain). These motifs are very likely also from Ibn ʿAbbās, as they are supported by different versions.

ʿUmar is mentioned in two versions (4 and 6). Since version 6 consists of only one tradition, it is difficult to determine from whom or from what time the ʿUmar motif originates. A striking similarity between the asānīd of versions 4 and 6 is that it circulated in Medina in the earliest generations of transmitters. Since the other Ibn ʿAbbās versions (1–3 and 5), of which the earliest transmitters were not from Medina (except Ibn ʿAbbās), do not mention ʿUmar, the ʿUmar motif cannot be ascribed to Ibn ʿAbbās. ʿUmar may have been added to the Medinan versions as well as suppressed in the (mainly) Kufan versions.54 Another interesting motif is that in version 1 the Prophet rebukes those present for having the idea that he is delirious, while in versions 5 and 6 a woman (identified as Zaynab, the wife of the Prophet, in version 6) rebukes them for not obeying the Prophet’s command. It is also impossible to say whether this comes from Ibn ʿAbbās.

The aforementioned similarities between the different Ibn ʿAbbās versions are all based on substantive similarities. Except for the sentence containing the Prophet’s request and a few single words, these motifs are discussed and worded differently in each version. This points to an oral tradition in the first few generations. Above the common links of the different versions (Sufyān b. ʿUyayna, Mālik b. Mighwal, Sulaymān al-Aʿmash, al-Zuhrī and Layth b. Abī Sulaym) there is greater similarity between the structure and wording of these narrative versions, suggesting a transition to written transmission or transmission through dictation sessions. If we look at the first few generations of transmitters, it is striking that the Ibn ʿAbbās versions mainly circulated in Iraq and the Hijaz (see

Table 1. The unique motifs of the Saʿīd b. Jubayr versions.

Version 1 Sufyān b. ʿUyayna (d. 198/814)	Version 2 Mālik b. Mighwal (d. 159/776)	Version 3 Sulaymān al-Aʿmash (d. 148/765)
The tears wet the pebbles	The comparison of the tears with pearls	No description of the crying
Different writing materials: a piece of paper, shoulder blade or document	Two different writing materials: shoulder blade or tablet	One kind of writing material: a piece of paper
Description of the disagreement and dispute among the people		The question to Muḥammad whether they should bring it later
Muḥammad’s command to leave him alone and his remark on his state.		Muḥammad’s counter question: After what?
Threefold command

). The Ibn ʿAbbās versions do not discuss the contents of the unwritten document (except in additions of later transmitters) and only inform us on its purpose: to provide guidance to the followers of Muḥammad.

Finally, what is striking about all these traditions is that none of them places Ibn ʿAbbās explicitly in the space where the Prophet and the group of unknown persons are located. According to the asānīd, he tells about this event and, according to some versions, seems emotionally affected by it, but in each version, Ibn ʿAbbās relates the story in a third person objective point of view. The contrast of the almost detached description of the main event with Ibn ʿAbbās’ emotional outburst in the introduction is enhanced in some traditions of version 155 and almost all traditions of version 2 by the first-person point of view of Saʿīd b. Jubayr. The first and second person are only used when somebody speaks. The mainly third-person point of view separates Ibn ʿAbbās from the quarelling. In the next part, we will look at the extent to which the traditions about the unwritten document ascribed to Jābir b. ʿAbd Allāh match the Ibn ʿAbbās versions.

2.2. Group 2: The Jābir b. ʿAbd Allāh Narrative

The following group of traditions that reflect the motive of Muhammad’s request for writing materials for a document he intends to prepare for his community have been attributed to Ibn ʿAbbās’ contemporary Jābir b. ʿAbd Allāh. Jābir, like Ibn ʿAbbās, was one of the Prophet’s Companions. He died in 78/697 at the age of 94, meaning that he was older than Ibn ʿAbbās when Muḥammad died. Jābir belonged to the tribe of the Khazraj, one of the two Arab tribes who lived in Yathrib (later called Medina) before the arrival of the Prophet. Together with his father, he would have attended the second ʿAqaba meeting with Muḥammad, shortly before his hijra, where Jābir swore allegiance to Muḥammad and converted to Islam together with his father. Although he was not present at the first two famous battles of Badr and Uḥud, he participated in numerous other battles of the Prophet. Various reports describe regular contact between Muḥammad and Jābir’s family. A critical note from Kister on Jābir’s tradition material is that traditions were attributed to him that did not always adhere to the correct rules of ḥadīth transmission. For example, the famous scholar al-Ḥasan al-Baṣrī is said to have reported directly from Jābir without being a student of his.56

Group 2 includes seven traditions, all of which go back to Jābir b. ʿAbd Allāh through the Meccan traditionist Abū l-Zubayr Muḥammad b. Muslim b. Tadrus al-Qurashī (d. 128/746). While, according to the asānīd, the traditions of Ibn ʿAbbās circulated mainly in Kufa and Medina in the first generations, the traditions of Jābir seem to have circulated mainly in Basra and Mecca during this period (see Figure 3).

The first thing that strikes one about these seven traditions is that they are all short: the focus is on the Prophet’s instruction to get writing materials. Five of the seven traditions look very much alike and are, according to the asānīd, from Qurra b. Khālid, who lived in Basra. The text of these traditions is:57

[…] on the authority of Abū l-Zubayr on the authority of Jābir, he said (/that)58 at his death, the Messenger (/Prophet)59 of God called for a piece of paper to write on it a document for his community (li-ummatihi) [so that] they will not go astray nor be led astray. In the house was noise and ʿUmar b. al-Khaṭṭāb spoke, so he (the Prophet)60 relinquished it (the document or the piece of paper).

The other two Abū l-Zubayr traditions, J1 and J7,61 differ slightly from Qurra’s version.62 In J1 the reference to ʿUmar is missing, i.e., “they shouted in front of him” (laghaṭū ʿindahu) instead of “there was noise and ʿUmar spoke” (laghaṭ wa-takallama ʿUmar), but since ʿUmar’s presence is confirmed in J7, i.e., “ʿUmar b. al-Khaṭṭāb was opposed to it” (fa-khālafa ʿalayhā ʿUmar b. al-Khaṭṭāb), we can most probably attribute the omission or suppression of ʿUmar’s name to Ibrāhīm b. Yazīd.63 The noise motif is completely missing in J7 and has been replaced by the aforementioned phrase that ʿUmar was opposed to it. Mūsā b. Dāwud or ʿAbd Allāh b. Lahīʿa must have been responsible for this.64

The common motifs in all traditions can be attributed to Abū l-Zubayr (d. 128/746), the common link of the asānīd, and dated to the first quarter of the second Islamic century. Abū l-Zubayr’s narrative is very similar to the Ibn ʿAbbās versions. The event takes place towards the end of the Prophet’s life. Muḥammad orders writing materials to write a document, but due to noise around him, he abandons it. Moreover, from the version of Abū l-Zubayr we do not learn anything about the content of the document and only the purpose is described. However, there are also clear differences from the Ibn ʿAbbās versions.

Unlike the Ibn ʿAbbās versions, Abū l-Zubayr’s version is lacking any mention of the Prophet’s illness. The Prophet’s command is formulated differently: “the Prophet/Messenger of God called for” (daʿā al-nabī/rasūl Allāh bi-) instead of the characteristic “bring me” (iʾtūnī bi-). Although the piece of paper (ṣaḥīfa) also appears as a variant in some Ibn ʿAbbās versions, for example in version 6 and some traditions of version 1, in the Abū l-Zubayr version the inkpot is missing. The difference between the direct speech of the Prophet’s command in the Ibn ʿAbbās versions and the indirect speech in the Abū l-Zubayr version extends to the group for whom the document is intended, i.e., “you” (lakum) in the former and “his community” (li-ummatihi) in the latter. The Abū l-Zubayr version identifies the location of the event, while this remains unknown in the Ibn ʿAbbās versions, except that of al-Zuhrī, who mentions “the people of/in the house disagreed” (fa-khtalafa ahl al-bayt). Finally, the Abū l-Zubayr version explicitly states that the Prophet gave up the idea of writing a document “he abandoned it” (rafaḍahu/-ha). Only in one tradition of Ibn ʿAbbās version 3 and version 4 of al-Zuhrī does a similar motif occur, respectively, “So he did not summon it” (fa-lam yadʿu bihi) and “Ibn ʿAbbās used to say, ‘The most terrible disaster is that their disagreement and their noise came between the Prophet and him writing that document’” (kāna Ibn ʿAbbās yaqūlu: inna al-raziyya kull al-razāyā mā ḥāla bayna rasūl Allāh wa-bayna an yaktuba lahum dhālika l-kitāb min ikhtilāfihim wa-laghaṭihim).

It is striking that these two characteristic motifs of al-Zuhrī’s Ibn ʿAbbās version (see

Table 2. The unique motifs of Ibn ʿAbbās versions 1–6.

Version 1 Sufyān b. ʿUyayna (d. 198/814)	The tears wet the pebbles; different writing materials (a piece of paper, shoulder blade or document); description of the disagreement and dispute among the people; Muḥammad’s command to leave him alone and his remark on his state; threefold command
Version 2 Mālik b. Mighwal (d. 159/776)	The comparison of the tears with pearls; two different writing materials (shoulder blade or tablet)
Version 3 Sulaymān al-Aʿmash (d. 148/765)	No description of the crying; one kind of writing material (a piece of paper); the question to Muḥammad whether they should bring it later; Muḥammad’s counter question: After what?
Versions 1–3 Saʿīd b. Jubayr (d. 95/714)	The exclamation of Ibn ʿAbbās “Thursday, what a Thursday!”; the crying of Ibn ʿAbbās (even if the three versions differ in the details); people who wonder if Muḥammad is delirious
Version 4 Ibn Shihāb al-Zuhrī (d. 124/742)	Ibn ʿAbbās speaks of a disaster; no writing materials mentioned; ʿUmar’s presence; ʿUmar makes the call not to obey the Prophet’s wish; ʿUmar’s argument that no second document is needed besides the Qurʾān; counter argument; the people of/in the house (ahl al-bayt)
Version 5 Layth b. Abī Sulaymān	Shoulder blade as writing material; purpose document is to avoid disagreement (between two men); noise or shouting after the request of the Prophet; the correction of those present by a woman
Version 6 al-Wāqidī—ʿIkrima	ʿUmar’s rejection of the death of Muḥammad; reference to the cities of the Byzantines; Zaynab, the Prophet’s wife, corrects those present

) are far part of Abū l-Zubayr’s version, even though the wording is different. A third characteristic can be added to this, since another similarity that al-Zuhrī’s version shares with Abū l-Zubayr’s version is that ʿUmar is mentioned (as does Ibn ʿAbbās version 6). Ibn ʿAbbās version 4 (and 6) circulated in Medina and this regional proximity is probably the cause of the similarities. Abū l-Zubayr (d. 128/746) and al-Zuhrī (d. 127/742) were contemporaries, lived in the same area, the Hijaz, and both transmitted traditions from each other.65 Yet the list of differences also shows that the version of Abū l-Zubayr is a separate version and does not show the characteristics of the Ibn ʿAbbās traditions. This seems to indicate that the attribution to two different informants is correct. Whether those informants are indeed Ibn ʿAbbās and Jābir b. ʿAbd Allāh cannot be established definitively. What speaks for the reliability of the attribution is that both were in close contact with the Prophet, although they do not emerge as active participants in the story. Both Ibn ʿAbbās’ and Abū l-Zubayr’ s versions provide an outsider’s view of the event. While the Ibn ʿAbbās versions show a change of perspective, Abū l-Zubayr’s version is told entirely from a third person point of view, suggesting a greater distance between the narrator and the event. The common core of the stories indicates a common source: either the actual event or a well-known story on the unwritten document circulating in the Hijaz and Iraq in the second half of the first Islamic century, with the ʿUmar motif possibly being part of the Hijazi stories.

2.3. Group 3: The ʿUmar b. al-Khaṭṭāb Narrative

The Hijazi occurrence of the ʿUmar motif is also evident in the traditions of group 3 attributed to the Companion and second caliph of the Muslim empire ʿUmar b. al-Khaṭṭāb (d. 23/644). ʿUmar had close ties to the Prophet, which were confirmed by Muḥammad’s marriage to ʿUmar’s daughter Ḥafṣa. According to Islamic tradition, ʿUmar played a major role in the appointment of Abū Bakr as leader of the Muslim community after the death of the Prophet Muḥammad. Immediately following the death of the Prophet, the religious community Muḥammad had established a decade earlier started to fall apart. Separately, several groups of Muslims in Medina gathered to discuss their future course. During one of these debates, ʿUmar b. al-Khaṭṭāb took the hand of Abū Bakr and swore allegiance to him, shortly followed by the other men present in the hall. The next day, Abū Bakr’s leadership was announced in Medina.66 Abū Bakr later appointed ʿUmar as his successor, and ʿUmar became caliph of the Muslim empire after Abū Bakr’s death.

Group 3 contains only two traditions, preserved in the collections of Ibn Saʿd and al-Ṭabarānī,67 and the common link according to the chains of transmission is Hishām b. Saʿd (d. 160/776–7) from Medina, who belonged to the tribe of the Quraysh. Ḥadīth authorities such as Ibn Ḥanbal (d. 241/855) and Yaḥyā b. Maʿīn (d. 158/775) labelled him a weak transmitter (laysa bi-muḥkim li-l-ḥadīth/laysa bi-dhāk al-qawī/ḍaʿīf).68 This may also be the reason why few traditions have been preserved in the ḥadīth collections. Al-Ṭabarānī remarks that only Hishām relates these traditions from his informant Zayd, and only Mūsā, who received the tradition from Hishām according to his isnād, from Hishām, and that the same applies to Muḥammad b. ʿAlī b. Khalaf (see Figure 4). Al-Ṭabarānī’s comment indicates that he had no variants of this tradition in his day. He was apparently unfamiliar with al-Wāqidī’s tradition.69

Ibn Saʿd’s tradition goes back to Hishām through al-Wāqidī, while that of al-Ṭabarānī goes back to him through Mūsā b. Jaʿfar b. Ibrāhīm b. Muḥammad b. ʿAlī b. ʿAbd Allāh b. Jaʿfar b. Abī Ṭālib (al-Jaʿfarī)70. According to both asānīd, the tradition is handed down through a chain of members of the same family.71 Hishām b. Saʿd received the tradition from his guardian Zayd b. Aslam (d. 136/754), who got it from his father Aslam al-Qurashī al-ʿAdawī who was a client (mawlā) from ʿUmar b. al-Khaṭṭāb. All are from Medina. The text of Ibn Saʿd’s tradition from Hishām b. Saʿd is:

[…] on the authority of ʿUmar b. al-Khaṭṭāb, he said: We were with the Prophet, and between us and between the women was a curtain (ḥijāb). The Messenger of God said, “If you cleanse me with seven water skins and bring me a piece of paper and an inkpot then I will write for you a document after which you will never go astray.” The women said: “Bring the Messenger of God what he needs.” ʿUmar said: I said, “Be quiet. You are his companions. When he (i.e., Muḥammad) is sick, you squeeze your eyes and when he is healthy you take his neck!” The Messenger of God said, “They are better than you (minkum)!”

The bold phrases are identical to al-Ṭabarānī’s tradition and his tradition is largely similar in content. The main difference from Ibn Saʿd’s tradition is that the event is explicitly placed at the time of the Prophet’s illness, “when the Prophet was ill” (lammā maraḍa al-nabī), which in Ibn Saʿd’s tradition can be implicitly inferred from the penultimate sentence. The seven water skins are missing.72 The Prophet repeats his command one more time, because the persons present (“we”) preferred not to do so (fa-karihanā dhālika ashadd al-karāha). In both traditions the women are behind a partition, which is described as ḥijāb by Ibn Saʿd and as sitr73 by al-Ṭabarānī. In al-Ṭabarānī’s tradition, the women ask those present if they have not heard what the Prophet asked (a-lā tasmaʿūna mā yaqūlu rasūl Allāh?).74 The similarities in formulation and content can be attributed to the common link Hishām b. Saʿd, and therefore dated to the middle or the second quarter of the second Islamic century. Since there are only two traditions, it is not clear which of the differences in the two traditions are from Hishām and which are from later transmitters.

The event in the traditions of Hishām b. Saʿd, like the Ibn ʿAbbās versions, is linked to the Prophet’s illness (explicitly in al-Ṭabarānī’s tradition U2 and implicit in Ibn Saʿd’s tradition U1). The Prophet’s command is formulated differently in the two traditions: “bring me” (iʾtūnī bi-) in U1 is similar to the Ibn ʿAbbās versions and “summon for me” (udʿū lī bi-) in U2 resembles Abū l-Zubayr’s version from Jābir, i.e., “the Prophet/Messenger of God called for” (daʿā al-nabī/rasūl Allāh bi-). The writing material Muḥammad asks for, a piece of paper and an inkpot (bi-ṣaḥīfa wa-dawāh), and the stated purpose of the document, “after which you will never go astray”, are more like the Ibn ʿAbbās versions by mentioning the inkpot (which is missing in the version of Abū l-Zubayr) and the use of the indirect speech “for you” (“for your umma” in the version of Abū l-Zubayr). It is even identical to Ibn ʿAbbās version 3 of Sulaymān b. al-Aʿmash and Ibn ʿAbbās version 6 of Ibn Saʿd through ʿIkrima. The presence of several persons is not explicitly stated, but can be derived from the plural form of person suffixes and verb conjugations, all second person plural masculine, i.e., the aforementioned “bring me” or “summon for me”, “for you” (lakum), “you will not go astray” (lā taḍillū), “than you” (minkum).

Contrary to the versions of Ibn ʿAbbās and Abū l-Zubayr, there is no uproar or discord among the persons present, except for the discussion between ʿUmar and the women. Neither the location of the event nor the content of the document is discussed. In fact, the document is no longer mentioned in the story at all. Hishām’s version focuses on ʿUmar’s derogatory statement about the women behind the curtain and Muḥammad’s rebuke by. According to this version, ʿUmar is angry with the women for urging those present to carry out the Prophet’s command, yet he does not appear to give them time to comply with his (the women’s question immediately follows the Prophet’s request). Muḥammad’s reprimand would then have to be directed against ʿUmar. However, it is also addressed to other people, which would only make sense if a discussion took place amongst them. This shows that Hishām’s version is a secondary version. In Hishām’s version a switch is made to a first-person perspective: ʿUmar, the alleged narrator of this tradition, relates the event from his point of view (f.e., “we were”, “between us”, “I said”). In contrast with the versions of Ibn ʿAbbās and Abū l-Zubayr, he is an active participant in the event.

The noise or discord motif occurs in both Ibn ʿAbbās’ and Abū l-Zubayr’s versions, but the ʿUmar motif only in those of Hijazi origin, versions 4 and 6 of Ibn ʿAbbās and the version of Abū l-Zubayr. The Medinan origin of the two traditions of Hishām, who is from Medina himself and provides a Medinan isnād, is consistent with the regional occurrence of the ʿUmar motif. Is Hishām’s version derived from Ibn ʿAbbās’ or Abū l-Zubayr’s? So far, Hishām’s version has more similarities to that of Ibn ʿAbbās than that of Abu l-Zubayr, such as that no location is mentioned, nor that Muḥammad is about to die, and the wording of the Prophet’s command. One part has not yet been discussed: the women motif.

The women motif occurs in both of Hishām b. Saʿd’s traditions from ʿUmar and can therefore be traced back to him and dated to the second quarter or the middle of the second Islamic century, since Hishām died in 160/776–7. We have already encountered the (one) woman motif in two other versions: Ibn ʿAbbās versions 5 and 6. In two of the three traditions of Ibn ʿAbbās version 5 of Layth b. Abī Sulaym, a woman rebukes the people present for shouting instead of following the command of the Prophet. Although what she says is different from Hishām’s version, it conveys the same sentiment: a rebuke for not following the Prophet’s command (“Woe unto you! [It is] the command of the Messenger of God!” (wayḥakum ʿahd rasūl Allāh)).

The woman motif in Layth’s traditions can be dated to the second quarter of the second Islamic century, since Layth died in 138/755–6 or 143/760–1. In Ibn ʿAbbās version 6, the woman is identified as Zaynab, one of the Prophet’s wives. Although her reprimand follows a statement by ʿUmar, she addresses several people, revealing that part of the story is missing from this tradition. The woman motif in this tradition is more difficult to date, because there is only one tradition of it. It is part of Ibn Saʿd’s work in which he collected nine traditions about the unwritten document.75 Of these nine traditions, two contain the woman motif, for which Ibn Saʿd both times lists his teacher al-Wāqidī as informant. However, of these nine, Ibn Saʿd traces four traditions back to al-Wāqidī, two of which do not contain the woman motif. It is therefore unlikely that Ibn Saʿd or al-Wāqidī included the woman motif in these two traditions. That would put the dating of the women motif in Ibn ʿAbbās version 6 with al-Wāqidī’s informant, Ibrāhīm b. Ismāʿīl, who died in 165/781–2, i.e., to the same period as Hishām’s. The woman motif (either one or more women) thus seems to have been introduced into the story of the unwritten document in the second quarter of the second Islamic century. Unlike the ʿUmar motif, the woman motif is not restricted to any particular region (Layth and his informant Ṭāwūs are not from Medina, but from Kufa and Janad (Yemen), respectively), but occurs only in traditions that were not widespread, indicating that they were not widely accepted.

2.4. Group 4: The ʿAlī b. Abī Ṭālib Narrative

The traditions of the penultimate group are all traced back to Muḥammad’s nephew ʿAlī b. Abī Ṭālib (d. 40/661), the fourth caliph of the Muslim empire. The Islamic tradition material describes a strong bond between Muḥammad and his nephew ʿAlī that began before Muḥammad’s prophethood and continued until the latter’s death. ʿAlī was one of the first converts and married Fāṭima, the Prophet’s daughter. When Muḥammad died, ʿAlī led the preparation of his body for burial.76 Because of this, he did not take part in the discussions about succession of the Prophet. The debates on the day the Prophet died, and the way in which Abū Bakr became the leader of the community, show that Muḥammad had not or—at that time—yet arranged his succession, or that any indication Muḥammad might have given during his life was not followed up. The succession of Muḥammad and the leadership of the Islamic community would become one of the main factors causing a schism in the religious community, leading to the distinction between Sunnī and Shīʿī77 Islam. The majority of the Shiites (Twelver and Ismāʿīlī Shiites) do not recognize the legitimacy of Abū Bakr’s reign and claim that Muḥammad had appointed his nephew ʿAlī as his successor and that only descendants of ʿAlī and (according to the majority of the Shiites) his wife Fāṭima had the right and the qualifications to lead the Islamic community.

The five78 traditions ascribed to ʿAlī are all from ʿUmar b. al-Faḍl al-Sulamī or al-Ḥarashī, who lived in Basra (see Figure 5). The biographical works do not mention his date of death. According to al-Mizzī, he narrates of four persons and eight persons of him,79 which seems to indicate that he was not a prolific transmitter. Inferred from the dates of death of the two persons in the chain who transmit from him, he likely died in the second half of the second Islamic century. Even less is known about his informant, Nuʿaym b. al-Yazīd. According to al-Mizzī and Ibn Ḥajar, he only narrates from ʿAlī and only ʿUmar b. al-Faḍl transmits from him. Ibn Ḥajar considers him majhūl (an unknown transmitter).80

Two slightly different versions of the narrative can be distinguished in the five traditions. The text of the first version in traditions A1 and A2 from Ibn Saʿd and al-Bukhārī is:

[…] ʿAlī b. Abī Ṭālib informed us that the Prophet (A1)/Messenger of God (A2) said when he became heavy [in sickness], “ʿAlī, if you bring me a plate (ṭabaq) then I will write on it what would prevent my community after me81 from going astray.” He (ʿAlī) said: “I was afraid that his soul would depart (before I return), so I said, ‘I can memorise better from my forearm than from a piece of paper’.” He (ʿAlī) said: “His head was between my forearm and my upper arm. He (Muḥammad) started to bequeath the prayer, the zakāh (almsgiving) and what your right hands own (=slaves).” He (ʿAlī) said: “[Muḥammad continued] thus until his soul departed. He ordered the shahāda (creed) that there is no god but God and that Muḥammad is his servant and his Messenger (in A1: until his soul departed). Whoever witnesses them is forbidden to the fire.”

The second variant from Ibn Ḥanbal82 is similar in content but ends with Muḥammad’s first three commands, that is, up to and including “what your right hands own”. Other differences include the indirect speech of the Prophet’s request, the omission of the explanation on how to memorize better from the forearm and the use of synonyms in certain places.83 Since there are only two variants of the tradition from ʿUmar b. al-Faḍl, it is unclear whether ʿUmar himself related the tradition in two ways or who is responsible for adding or leaving out the last parts. However, for comparison with the previously discussed versions of this story, this does not matter, because the last part of the story is unique to ʿUmar b. al-Faḍl’s version from ʿAlī.

The version of ʿUmar b. al-Faḍl is the most different version up till now. Although the setting is the same with Muḥammad being critically ill and asking for writing material to prevent his community from going astray, for the first time only one person is present, Muḥammad’s nephew ʿAlī, and Muḥammad directs his request only to him. Despite the family relationship to Muḥammad and the close connection they are said to have, none of the other versions lists ʿAlī as one of those in attendance. The persons generally remain unknown except for ʿUmar b. al-Khaṭṭāb, who appears only in Hijazi versions, and Zaynab in one tradition.84

The material that Muḥammad wants to write on, ṭabaq (a thin plate or bone),85 is also a new element in this story. For the first time, we learn what Muḥammad intended to write, namely two (A3, A4 and A5) or three (A1 and A2) of the five pillars of Islam, and slaves. We have encountered the motif of the threefold command earlier in Ibn ʿAbbās version 1 of Sufyān b. ʿUyayna (d. 198/814), i.e., expulsion of the polytheists, an instruction for dealing with delegations, and a third, forgotten, command. It was implied that the threefold command was the content of the document. In the version of ʿUmar b. al-Faḍl, the link is made explicit by ʿAlī telling Muḥammad that he can better memorize what Muḥammad wants to write down. After all, ʿAlī is afraid that Muḥammad will be dead before he returns with the requested writing material. Since ʿUmar b. al-Faḍl is a contemporary of Sufyān b. ʿUyayna, the addition of the threefold command motif to the story of the Prophet’s unwritten document—even though the command itself is different—can be dated to the second half of the second Islamic century. Both lived in Iraq, Sufyān in Kufa and ʿUmar in Basra, which means that the threefold command can also be linked to a certain region.

However, this does not necessarily mean that this motif was put into circulation by ʿUmar b. al-Faḍl. The threefold command in ʿUmar b. al-Faḍl’s version from ʿAlī is very similar to a tradition whose common link appears to be Qatāda b. Diʿāma, a legal scholar from Basra who died in 117/735.86 In this tradition, which takes place just before the death of the Prophet, the Companion Anas b. Mālik or Muḥammad’s wife Umm Salama87 says:

“The general testamentary statement made by the Messenger of God, when his death approached was, ‘(Uphold) prayer; and (care for) what you right hands possess’, until his chest began to gurgle as he spoke, and his tongue could scarcely express it.”88

While an isnād-cum-matn analysis of these traditions is interesting and may show whether Qatāda b. Diʿāma is the common link of the traditions and thus the earlier source for this motif, it goes beyond the purpose of this article. Suffice it to conclude that there is an interdependence between the traditions from Qatāda b. Diʿāma and the ones from ʿUmar b. al-Faḍl. Ibn Kathīr noticed the same similarity, for he placed Ibn Ḥanbal’s tradition from ʿUmar b. al-Faḍl in the midst of the traditions from Qatāda. Ibn Kathīr notes that Ibn Ḥanbal is the only one who gives it like this (tafarrada bihi Aḥmad [b. Ḥanbal] min hādhā l-wajh).89 ʿUmar b. al-Faḍl’s version from ʿAlī is thus most probably a mix of a number of motifs from different traditions. Just as he seems to have adopted and edited the section on the testamentary statement, so did he adopt and edit the tradition about the unwritten document of the Prophet. A third clue to his adaptation of this tradition is that the position in which Muḥammad dies—with his head between ʿAlī’s forearm and upper arm—is very similar to the tradition attributed to Muḥammad’s favorite wife ʿĀʾisha. In it she tells that the Prophet died with his head between her chest and her chin.90

The reason for creating an ʿAlī version of this tradition may have to do with the time period and region in which ʿUmar b. al-Faḍl grew up. Although we are not sure when ʿUmar was born and died, he likely witnessed the change of power in the Islamic empire in the middle of the second Islamic century. The Umayyad dynasty, who had ruled the empire after the death of caliph ʿAlī, were overthrown by the ʿAbbāsids with the help of Shīʿī Muslims on the promise that a descendant of ʿAlī would be proclaimed caliph. Their disappointment with the ʿAbbāsid-appointed caliph resulted in a hostile attitude of the Shīʿī’s towards the ʿAbbāsid caliphs.91 Many followers of ʿAlī could be found in the area where ʿUmar b. al-Faḍl lived. Although Basra was not completely pro-ʿAlid like Kufa –it was largely Sunnī- there were also supporters of ʿAlī.92 While the version of ʿUmar does not address the issue of succession from the Prophet, it does underscore ʿAlī’s closeness to the Prophet and in this sense displays a pro-ʿAlid tendency.

Besides these differences from the previously discussed versions of Ibn ʿAbbās, Abū l-Zubayr and Hishām b. Saʿd, there is another striking difference. There is no opposition to obeying the Prophet’s instruction for writing material. Although ʿAlī does not follow the instruction to the letter (he memorizes them), he ensures that Muḥammad’s words are preserved. The mention of ʿAlī’s good memory assures the reliability of the commands and a reason is provided for not obeying the order: ʿAlī is afraid that Muḥammad will die before his return. We will also encounter this lack of opposition in the discussion of the next narrative.

2.5. Group 5: The ʿĀʾisha bt. Abī Bakr Narrative

The last group of traditions, revolving around the Prophet’s instruction to get writing materials, is ascribed to ʿĀʾisha (d. 58/687), the favourite wife of the Prophet Muḥammad and the daughter of Abū Bakr, the first caliph of the Islamic empire after the death of Muḥammad. According to the Islamic ḥadīth material, the Prophet married her a few years before the hijra after the death of his first wife Khadīja. At the time, ʿĀʾisha was still very young. Although ʿĀʾisha was not his only wife and their marriage to the Prophet may also have been concluded from a political point of view to strengthen ties with Abū Bakr, she continued to hold a special position among Muhammad’s wives. Contrary to ʿUmar b. al-Faḍl’s version from ʿAlī of discussed above, the common belief is that the Prophet died in her arms, after which he was buried in her apartment.93

All the traditions in this group are attributed to the Ibn Abī Mulayka, who, like ʿĀʾisha and Abū Bakr, belonged to the Taym clan within the tribe of the Quraysh. Ibn Saʿd considers Ibn Abī Mulayka a reliable (thiqa) transmitter, belonging to the second generation in Mecca, who narrated many aḥādīth (kathīr al-ḥadīth), including traditions from ʿĀʾisha and Ibn ʿAbbās. He died in Mecca in 117/735.94

Two slightly different versions can be distinguished within the ʿĀʾisha traditions.

The reconstruction of ʿĀʾisha version 1 is:95

[…] on the authority of ʿĀʾisha, she said: The Messenger of God said to me during his illness of which he died, “If you call (udʿī/udʿū)96ʿAbd al-Raḥmān b. Abī Bakr for me, then I will write a document for Abū Bakr on which no one after me disagrees.” Then he said, “Leave it/him (daʿīhi)97. God forbid, that the believers disagree about Abū Bakr.”

The common link of version 2 is the Kufan scholar Abū Muʿāwiya al-Ḍarīr (d. 194/810). His reconstructed text is:98

[…] on the authority of ʿĀʾisha, she said: When the Messenger of God became heavy in sickness, he said to ʿAbd al-Raḥmān b. Abī Bakr, “Bring me a shoulder blade so that I can write a document for Abū Bakr on which no one disagrees.” When ʿAbd al-Raḥmān started to get up, he said, “God and the believers forbade disagreeing about Abū Bakr.”

The most striking differences between the two versions are the role of ʿĀʾisha, the writing material and the wording of the last sentence. In version 1, the Prophet’s request is addressed to ʿĀʾisha (AA2, AA4, AA7)(udʿī) or to a group of unknown persons (udʿū)99, who must get her brother ʿAbd al-Raḥmān, while in version 2 from Abū Muʿāwiya al-Ḍarīr, her brother is already present and Muḥammad directly instructs him to get writing materials. In version 1, ʿĀʾisha takes an active role and the story is told from a first-person perspective, while in version 2 she is (only) the source of the account and she relates the tradition from the third-person point of view.

In version 1, the Prophet’s request does not include any reference to writing material, while in version 2 he asks for a shoulder blade. After the request, Muḥammad tells ʿĀʾisha in version 1, “Leave it/him” (daʿīhi)’, which is missing in version 2. In version 2 it is said that ʿAbd al-Raḥmān stands up (dhahaba ʿAbd al-Raḥmān li-yaqūma or qāma ʿAbd al-Raḥmān)100 and this is again missing in version 1. The last sentence looks similar in the translation, but is worded differently. Version 1 has maʿādha Allāh an yakhtalifa al-muʾminūna (God forbids that the believers disagree), while version 2 has abā Allāh wa-l-muʾminuna an yukhtalafa (God and the believers forbade disagreeing).

Both ʿĀʾisha versions take place during the last days of the Prophet’s illness. In version 1, this is referred to as “during the disease he died from” (fī maraḍihi alladhī māta fīhi) and in version 2 as “when he became heavy in sickness” (lammā thaqula). The former expression is identical to that of Ibn ʿAbbās version 6 from Ibn Saʿd, but it is such a general wording that it cannot be seen as evidence of interdependent transmission. The latter is unique to Abū Muʿāwiya al-Ḍarīr’s version 2 from ʿĀʾisha. In both ʿĀʾisha versions, the Prophet addresses one person, similar to the ʿAlī narrative discussed above. Muḥammad wants ʿAbd al-Raḥmān, the son of Abū Bakr and a full brother of ʿĀʾisha101, to write a document for Abū Bakr. Both ʿAbd al-Raḥmān and Abū Bakr have not been mentioned as participants in the story about the unwritten document of the Prophet in any of the discussed versions, and their appearance is therefore unique to the ʿĀʾisha versions. Similar to the ʿAlī version of ʿUmar b. al-Faḍl, the people present, ʿĀʾisha or ʿAbd al-Raḥmān b. Abī Bakr, do not dispute the Prophet’s instruction.

For the first time ever, the document is not intended for the community—mentioned explicitly in some versions, “li-ummatī” (for my community), and implicit in others, “lakum” (for you)—but for one person. The purpose of the document is to avoid disagreement. We encountered this before in Ibn ʿAbbās version 5 of the Kufan transmitter Layth b. Abī Sulaymān (d. 138/755–6 or 143/760–1). Since both Layth and the common link of ʿĀʾisha version 2, Abū Muʿāwiya al-Ḍarīr, and the possible common links of ʿĀʾisha version 1, Muḥammad b. Abān al-Jaʿfī (d. 175/792) or ʿAbd al-ʿAzīz b. Rufayʿ (d. 130/787–8),102 (ʿAbd Allāh b. ʿAdī 1997) lived in Kufa, the disagreement motif appears to have regional origins. The disagreement, in both ʿĀʾisha versions, concerns a dispute over Abū Bakr, as shown in the last sentence, “God and the believers forbade disagreeing about Abū Bakr”. Not only God, but also the “the believers” reject it, thereby implying that disagreement equals non-belief. Given the time at which this story takes place, just before the death of the Prophet, this seems to refer to the disagreement that arises over Abū Bakr’s succession of Muḥammad as leader of the Muslim community.

A second similarity to Ibn ʿAbbās version 5 of Layth b. Abī Sulaymān is the shoulder blade mentioned in ʿĀʾisha version 2 of Abū Muʿāwiya al-Ḍarīr, which shows the interdependency of the traditions of Abū Muʿāwiya al-Ḍarīr and Layth b. Abī Sulaymān. According to al-Mizzī, Abū Muʿāwiya al-Ḍarīr transmits from Layth, which corresponds to the findings of the isnād-cum-matn analysis.103 The shoulder blade motif in Abū Muʿāwiya al-Ḍarīr’s version from ʿĀʾisha is therefore most likely from Layth. Something else seems to have happened with the disagreement motif. Except for Ibn ʿAbbās version 5 of Layth, we only encountered the disagreement motif with Ibn Abī Mulayka, where it is part of the basic narrative. According to al-Mizzī, one of Layth’s informants is Ibn Abī Mulayka104 and therefore the latter is very likely the source of Layth’s disagreement motif.

However, there is even more to the ʿĀʾisha narrative. The similarities of both ʿĀʾisha versions indicate that they are from the same source, Ibn Abī Mulayka (d. 117/735), according to Figure 6. The common motifs of the two ʿĀʾisha versions thus date to the end of the first or the beginning of the second Islamic century. The ḥadīth material contains several traditions that are attributed to Ibn Abī Mulayka and that take place during the Prophet’s illness. From these traditions a clear picture emerges in which the Prophet prefers Abū Bakr over others in different settings. Various phrases from the ʿĀʾisha narrative are also present in these traditions, such as “during his illness of which he died” (fī maraḍihi alladhī māta fīhi), “call Abū Bakr for me” (udʿūh lī (Abī Bakr))105, “God and the believers forbid that” (yaʾbā Allāh dhālika wa-l-muʾminūna).106 Ibn Abī Mulayka narrates from Ibn ʿAbbās107 and thus appears to be the one who combined the motif of the unwritten document with the motif of the disagreement over Abū Bakr, just as he may have done with other motifs.

3. Conclusions

The ICM analysis of the unwritten document narratives has shown that there are roughly five different narratives in the narrative material about the unwritten document that Muhammad intended to write during his illness. By attributing those narratives to companions of the Prophet, they appear to be separate narratives, but the similarity in setting and in the Prophet’s request suggests that there might be a connection between the traditions. The ICM analysis helped to entangle the interwoven tradition complexes. By comparing the chains of transmission with the texts, it was possible to date the different narratives, identify the oldest kernel of each narrative and to determine who is resposible for certain parts in the account. By separating the earlierst core and later motifs, it is possible to make much more precise statements about possible historical elements of these sīra stories.

One of the earliest versions is ascribed to Ibn ʿAbbās. The event takes place when the Prophet is very sick. Several persons are present when he asks for an inkpot and something to write on to prepare a document for his people after which they will not go astray. What he wants to write remains unknown, because after his request a commotion arises among those present, since they think the request was caused by his illness. At one point the Prophet orders them to leave. The earliest transmission of this basic narrative was probably oral and circulated in the first few generations mainly in Iraq and the Hijaz. A second, distinctive narrative, attributed to Jābir b. ʿAbd Allāh, was circulating in the same region. The common core of these two stories points to a sommon source that can be dated at least to the second half of the first Islamic century—the actual event or a well-known story on the unwritten document. Although the attribution to Ibn ʿAbbās and Jābir b. ʿAbd Allāh could not be confirmed, the characteristic features of both versions indicate that it cannot be excluded. Ibn ʿAbbās and Jābir were Companions of the Prophet and had access to him.

Unlike later versions of the story, their texts relate the story itself in an almost detached way (except for the emotional context in which four Ibn ʿAbbās versions are placed). Further remarkable elements of these earliest versions are the opposition to the writing of the document, even accusing the Prophet of being delirious, the lack of clarity about its content and the Prophet’s abandonment of writing the document. These ambiguous elements, which can be dated to at least half a century after the death of the Prophet, speak for an actual event rather than a story. The later additions revealed by the ICM analysis fill in the gaps in this narrative or explain ambiguities. For example, later transmitters of the Ibn ʿAbbās narrative added their own details. Some of these can be traced back to certain narrators, such as the dating on Thursday, while others are of regional origin. An example of the latter is the identification of ʿUmar b. al-Khaṭṭāb as one of those present. The ʿUmar motif is part of traditions of Hijazi origin and the earliest dateable occurrence is the first quarter of the second Islamic century, based on Ibn ʿAbbās version 4 from al-Zuhrī (d. 124/742) and the Jābir traditions from Abū l-Zubayr (d. 128/746). Because al-Zuhrī and Abū l-Zubayr transmit from each other and their versions contain similarities, there seems to be interdependency in the transmission, which speaks for inclusion of the ʿUmar motif in Hijazi traditions instead of suppression of the ʿUmar motif in Iraqi traditions. Another later addition is the woman (or women) motif. It seems to have been introduced into the story of the unwritten document in the second quarter of the second Islamic century. Unlike the ʿUmar motif, the woman motif is not restricted to any particular region but occurs only in traditions that were not widespread, indicating that they were not widely accepted. Half a century later, in the second half of the second Islamic century, the threefold command motif is introduced into the story. The introduction of the motif also originated in a specific region, in Iraq in the vicinity of Basra and Kufa.

The two most deviating narratives are those of ʿAlī b. Abī Ṭālib and ʿĀʾisha. Both narratives consist of combinations of motifs from different stories, including that of the unwritten document. The document story is used to spotlight one person, ʿAlī and Abū Bakr, respectively. Both narratives are related to the discussion of the succession of the Prophet, which may be why both narratives not only refer to the content of the document but also contain no opposition to Muḥammad’s wish to write it. Although the ʿAlī narrative does not explicitly address the issue of succession like the ʿĀʾisha narrative, it does underscore ʿAlī’s closeness to the Prophet and in this sense displays a pro-ʿAlid tendency. The narrative probably originated in a period and region of pro-ʿAlid support. The ʿĀʾisha narrative emerged earlier, towards the end of the first or the beginning of the second Islamic century. After the Ibn ʿAbbās narrative it is the most common narrative in the ḥadīth collections. While the storyline of the earliest versions of the unwritten document is still visible in the ʿAlī narrative, in the ʿĀʾisha narrative it has been snowed under by motifs from other traditions. However, that is an interwoven tradition complex that still has to be untangled.