The exchange of letters in early Sufism: A preliminary study

Abstract This paper introduces the exchange of letters in early Sufism, analyses the significance of these exchanges, and examines these documents not for their general literary qualities or for theoretical discussion of appropriate conduct but, rather, for actual data relating to personal and interpersonal relationships. Furthermore, this paper emphasizes the crucial need for creating a corpus of Sufi letters and pieces of correspondence. The discussion is divided into methodological and conceptual-historical perspectives. The methodological perspective includes a survey of sources, the question of transmission, letter fragment usage by later authors, and a reconstruction attempt of the actual circumstances of these documents. The conceptual-historical perspective analyses content, rhetoric, argumentation forms, and self-representation.


Introduction
Exchanges of letters (rasāʾil) and correspondence (mukātabāt) during the early phase of Sufism, whether between Sufi figures or occasionally between Sufis and non-Sufis, have rarely attracted scholarly interest. Sufi works of the late third/ninth and fourth/tenth centuries provide us with an impressive number of fragments of early Sufi letters and pieces of correspondence. However, no corpus of this material has yet been generated. Sufis addressed letters to their counterparts over a wide variety of circumstances and purposes. An impressive number of letter fragments appear to have been taken from original and longer versions of correspondence that certain Sufis, prior to the tenth century, had addressed to each other; this period witnessed the composition of the great Sufi manuals. These letters were not written primarily for the purpose of recording mystical theories, but rather served particular purposes in their respective historical period. The letters are generally undated and in many cases do not even include the addressees' names. The purposes of these letters were wideranging: they included instructing Sufi novices, providing pieces of counsel, resolving some enquiries and controversial issues, and even criticizing and debating with addressees. While, broadly speaking, the exchange of letters helped early Sufis establish their networks and cultivate their collective religious identity, the different circumstances that surrounded the letters' authorship need to be thoroughly investigated. Unlike earlier Sufi letters and correspondence, letters produced after the fourth/tenth century were limited to a few scholarly endeavours. The correspondence of Sharaf al-Dīn Muḥammad al-Balkhī with his master Majd al-Dīn al-Baghdādī (d. 616/1219) was the focus of Fritz Meier's paper in which he relies on the Köprülü manuscript (1589), translates the text into German and comments upon it. 2 From the late sixth/twelfth century, the short letters of the Sufi sheikh of Baghdad ʿUmar al-Suhrawardī (d. 632/1234) to some of his contemporaries including the renowned theologian Fakhr al-Dīn al-Rāzī (d. 606/ 1210) gained a brief reference in my book on the life and teachings of al-Suhrawardī; 3 the latter is an example of a Sufi letter addressed to a non-Sufi contemporary and is in itself an additional interesting aspect of this research. My reference to other of al-Suhrawardī's letters relies on different manuscripts from the Biblioteka Jagiellońska (Poland) and the Yehuda collection at The National Library, Jerusalem. Al-Suhrawardī's letter to al-Rāzī has been published by Pourjavadi. 4 Another Sufi master, Ibn ʿArabī (d. 638/ 1240), used his letter to al-Rāzī to convert the latter to Sufism. 5 William Chittick relied on a manuscript that he discovered in the Süleymaniye Library in Istanbul to draw attention to the Arabic correspondence of Qāḍī Burhān al-Dīn Aḥmad ibn Muḥammad (d. 800/1398) with one member of the School of Ibn ʿArabī. 6 In another work, Chittick examines Ṣadr al-Dīn al-Qūnawī's (d. 673/1274) correspondence with Naṣīr al-Dīn al-Ṭūsī (d. 673/ 1274). 7 All these letters revolve around arguments of mystical theories and metaphysics.
Risāla (plural rasāʾil) was a very common genre during the first centuries of Islam. As Albert Arazi and Haggai Ben-Shammay indicate, the Arabic term risāla appears in very ancient inscriptions of Arabia. 8 While the term originally meant the oral transmission of a message, it took on the meaning of written text by the mid-third/eighth century. Works of adab, historiographies, and biographies provide us with an enormous number of letters addressed by men of authority to their opponents to debate political issues; examples of these are the letters of ʿAlī ibn Abī Ṭālib to Muʿāwiya ibn Abī Sufyān and other contemporaries. 9 Other instances may be seen in correspondence from scholars and ascetics (zuhhād, singular zāhid) to their contemporaries or rulers who they aspired to counsel: examples may be found in the letters of al-Ḥasan al-Baṣrī (d. 110/ 728), the renowned ascetic of Baṣra, to one of his companions, 10 or to the Umayyad khalīfa ʿUmar ibn ʿAbd al-ʿAzīz, as well as the exchanges between al-Ḥasan ibn ʿAbd Allāh ibn Saʿīd al-ʿAskarī, the famous lexicologist (d. 382/ 992), and al-Ṣāḥib ibn ʿAbbād the Buyid statesman. 11 Risāla developed during the Umayyad period and denoted a "monograph", "treatise" or a "literary epistle" on particular religious, theological or philosophical issues. Ḥājjī Khalīfa (d. 1067/1657), while referring to the Rasāʾil of Abū al-ʿAlāʾ al-Maʿarrī (d. 449/1057), distinguishes between long rasāʾil that serve as monographs, and short rasāʾil that serve as correspondence (mukātaba). 12 Adrian Gully, in his work on the culture of letter-writing in pre-modern Islam, focuses primarily on letter-writing as part of artistic prose composition (inshāʾ). Gully proposes dividing what he calls "the epistolary genre" into two parts: official or formal letters, and informal letters (ikhwāniyyāt). As regards ikhwāniyyāt, Gully criticizes Arazi and Ben Shammay's definition of this genre which is seen as purely literary letters whose exclusive subject is deep affection for one's friends. According to Arazi and Ben Shammay, although these letters were usually prompted by a particular event such as the birth of a son or the death of a relative, the authors do not seek to discuss any particular event, but only to formulate and celebrate the motif of brotherly friendship. 13 Gully argues that informal letters were not exclusively reserved for correspondence between friends; 14 and insofar as my studies of Sufi letters are concerned, I agree with Gully. In these letters, while abstract literary contents concerning Sufi conduct, general manners and beliefs still occupy a prominent place, many of the letters go beyond displays of Sufi manners and include references to particular circumstances and networks that are not necessarily of a friendly nature. These letters can, therefore, be considered neither purely literary nor restricted to the theme of friendship. This paper has nothing to do with risāla as pure monograph, and it is limited to pieces of correspondence that were circulated among early Sufis. Nonetheless, separating risāla as a literary epistle from risāla as an actual letter is not easy. In the Sufi domain, this separation is especially difficult since structural constraints imposed by the literary genre very often influence the actual content, which is the practical message flowing from a particular historic moment and giving rise to writing the letter. The following discussion will further elaborate this idea.
Klaus Hachmeier's discussion of non-Sufi private letters and official correspondence written in Arabic under Buyid rule focuses on the value of these documents as historical sources and can shed light on our topic. For example, Hachmeier makes use of the term überreste, which was introduced earlier by A.V. Brandt to relate to "everything that has remained immediately and directly of the historical facts of events". Hachmeier places private and official letters produced during Buyid rule under a sub-category called "written überreste". This subcategory includes poetry as well as both private and official exchanges in the form of letters, messages, deeds, etc.: these have survived in three forms: (i) as original documents; (ii) as exchanges cited in other works; and (iii) in letter collections. 15 In his most recent study, Hachmeier provides a more comprehensive survey of a large collection of letters written by Abū Isḥāq Ibrāhīm al-Ṣābiʾ (d. 384/994), the Buyid secretary, based on manuscripts and other sources. 16 In a Sufi context, the verb kataba often appears before quoted letters leaving the impression that the term risāla denotes a written letter. Evidence of oral correspondence is documented but the terms risāla or rasāʾil are not used in these instances. A reference to such oral correspondence is documented concerning Dhū al-Nūn al-Miṣrī (d. c. 245/859-860) and Fāṭima of Nishapur. 17 Having said this, the current paper seeks to sketch the major outlines of the study of Sufi letters and is a preliminary attempt to prepare the ground for a comprehensive study of exchanges of letters in early Sufism. This preliminary research will relate to two perspectives: the first methodological and the second labelled "conceptual-historical" since it relates to contexts, contents and rhetoric.
From a methodological perspective, this paper seeks to examine the following questions: what are the types of sources in which these letters were preserved, and how should these sources be treated? What methods should be used to research Sufi letters? Did all references to such correspondence indicate written forms of correspondence or were some acts of oral transmission? In other words, do terms like "he wrote to" (kataba ilā) always indicate written letters or might they also refer to oral contact? How are terms like "he wrote to" or "he addressed a letter to" in Sufi contexts understood when used in non-Sufi texts? Do Sufi letters differ in structure and style from non-Sufi letters? Is the use of particular Sufi terminology the only dividing feature? To what extent should traces of Sufi letters be a source with social or historical values? This last question has to do with both methodology and content. From a conceptual-historical perspective, this paper aims to analyse the different forms of available letters, the particular theoretical discussions implied therein, and the identities and relationship between both senders and addressees. The paper will also examine certain philological issues of structure and semantics as they relate to this type of research. Finally, the paper will discuss the ways in which the study of Sufi letters should be able to enrich our understanding of early Sufi piety by taking a broad approach and incorporating an analysis of social frameworks as well as other thematic features. This will help establish the basis through which one may reconstruct the complexity of the dynamic networks in early Sufism. This type of research is intended to promote further studies of Sufism, in relatively new territory away from most studies of Sufi works which presently deal with abstract Sufi theories or the development of particular forms of piety and Sufi rituals such as samāʿ and dhikr. It will also help us reconstruct the development of early Sufi piety as a process motivated and dominated by people whose personal concerns, tensions and aspirations could not, and should not, be overlooked.

Methodological perspective
Methodologically speaking, the first step in this research project is to create a corpus of Sufi letters and pieces of correspondence that were written between Sufis and, on certain occasions, between Sufis and non-Sufis, between the late ninth and the thirteenth centuries. These letters should be distinguished from Sufi monographs or treatises holding the title risāla. This corpus would include complete texts of correspondence, fragments of letters and response letters in Sufi published works as well as unpublished manuscripts and non-Sufi biographies, historiographies and works of adab. This raises the following ques   19 This interest in al-Junayd's letters was a very significant contribution to the study of early Sufism; however, a further process of creating and analysing a larger corpus of correspondence that goes beyond al-Junayd, and includes more key Sufi personalities, is still sorely needed. One of the major methodological challenges in treating al-Junayd's letters is the tendency towards a dominant literary nature which then turns the letter into a sort of a treatise on abstract Sufi matters. I argue that in spite of such a dominant literary nature and structure, Sufi letters can be a sufficient source in reconstructing certain elements of Sufis' social lives and communal engagements. Al-Ḥakīm al-Tirmidhī (d. c. 295/908) is another source for the exchange of letters in early Sufism. Three letters that he wrote in response to two of his contemporaries are known to us. Two of these letters were addressed to Muḥammad ibn al-Faḍl (d. 319/931) and are found in Sara Sviri's unpublished critical edition entitled Masāʾil wa-rasāʾil, which is based on the Leipzig manuscript (No. 212). 20 The third letter, published by Bernd Radtke, was addressed to Abū ʿUthmān al-Ḥīrī (d. 297/910), the renowned master of the malāmatiyya group of Nishapur in al-Tirmidhī's days. 21 In her chapter on al-Ḥakīm al-Tirmidhī and the malāmatī movement, Sara Sviri briefly refers to al-Tirmidhī's criticism of the concept of self-blame. This is the most fundamental doctrine of malāmatiyya, as implied in one of the letters to Ibn al-Faḍl, and it hints at the significance of Sufi correspondence as a primary source in reconstructing the dynamic and complex networks between the mystical circles in Khurāsān and Baghdad. 22 Kitāb al-Lumaʿ of Abū Naṣr al-Sarrāj is an important source of Sufi correspondence. Written in the course of the fourth/tenth century, it is the earliest source to include fragments of letters from a large group of figures; as well as point out the importance of correspondence between Sufi figures as a source of veiled, secret Sufi teachings, the concealment is an attempt to hide the material from non-Sufi eyes. On one occasion in the section devoted to Sufi terminology, particularly under his definition of the term ramz (symbol), al-Sarrāj refers very briefly to Sufi exchanges of correspondence: "The one who seeks to have an understanding of our masters' symbols, he should look for them in their exchanges of correspondence and letters; since their symbols are to be sought therein not in their compilations" (man arāda an yaqifa ʿalā rumūz mashāyikhinā fa-l-yanẓur fī mukātabātihim wa-murāsalātihim fa-inna rumūzahum fīhā lā fī muṣannafātihim). 23 Al-Sarrāj compiles a separate chapter in which he gathers a large number of Sufi letters in addition to the opening sections of letters whose originals are no longer available to us. This chapter appears as part of a long section that al-Sarrāj devotes to Sufis' mukātabāt, ṣudūr (introductions of letters), poems, prayers (daʿawāt), and pieces of counsel (waṣāyā) that Sufis used to send to each other. These topics demonstrate different aspects of the Sufi art of writing which al-Sarrāj chooses to locate directly after his thorough treatment of Sufi manners and before his discussion of practical Sufi piety; this practical piety includes both samāʿ (audition, the act of listening to a recitation of poetry or a song) and wajd (ecstatic state). Scholars of Arabic belles-lettres have noticed that, in the early history of Arabic literature, written messages were very often rhymed poems. In the Sufi realm, this notion helps explain why al-Sarrāj combines Sufi correspondence with poetry under the same section; by virtue of its structure and content, the implication is that the Sufi tradition of writing is a fundamental component of Sufi piety in general. Al-Sarrāj's belief in the role of Sufi writing is best manifested in his life as well. 24 Pieces of correspondence in al-Lumaʿ are not just included in the above-mentioned section but are scattered throughout the work. One example appears in the chapter on the concept of intimacy (uns), where a correspondence between an anonymous man and the Egyptian mystic Dhū al-Nūn al-Miṣrī is mentioned. None of these occasions have as yet attracted much research.
Another Sufi author of the late fourth/tenth century, ʿAbd al-Malik al-Khargūshī had a similar chapter on Sufi correspondence in his Tahdhīb al-asrār where he quotes short fragments of letters attributed to both Sufi and non-Sufi figures of early Islam. 25 Unlike al-Sarrāj, al-Khargūshī's chapter mixes references to Sufi letters with references to non-Sufi letters, probably as a means of emphasizing the integral position of Sufism within early medieval Muslim societies and culture. The data is, remarkably, dominated by a literary form of presentation with a clear tendency on the part of the author to convey proper Sufi ethics and morality. While this is also the general tendency in al-Sarrāj's work, we are still able to glean some features of the historical-social backgrounds of certain quoted fragments of letters.
In addition to the above-mentioned sources, there is an impressive body of correspondence and fragments of correspondence in the works of al-Iṣfahānī's Ḥilyat al-awliyāʾ, ʿAbd Allāh Anṣārī Haravī's Ṭabaqāt al-ṣūfiyya, Abū al-Qāsim al-Qushayrī's al-Risāla fī ʿilm al-taṣawwuf, and Abū Khalaf al-Ṭabarī's Salwat al-ʿārifīn wa-uns al-mushtāqīn. In order to create a broader corpus, however, we need to search further for letters in manuscript archives and libraries. The collection of Şehit Ali Paşa (No. 1374), for instance, includes several letters ascribed to al-Junayd al-Baghdādī. One of the major difficulties is that most of the references to the term risāla in the bio-bibliographical indexes of Ḥājjī Khalīfa (kashf al-ẓunūn) and Fuat Sezgin (Geschichte des Arabischen Schrifttums) relate, in fact, to long monographs. Titles that include risāla as private letters, kitāb ilā (a letter to), or jawāb ilā (a response letter to) should be searched for and added to the corpus. After creating the corpus, the process of identifying the bulk of letters pursuant to different criteria of both context and content as discussed under "conceptual-historical perspective" is needed.
Relevant textual material included in non-Sufi sources, in works of adab (belles-lettres), biographies and historiographies in particular, should be included. In these sources, short references to Sufi letters are occasionally provided. As well as being an occasional source for Sufi letters, non-Sufi works can also shed light on the historical contexts suggested in these letters. In al-Khaṭīb al-Baghdādī's (d. 463/1071) Taʾrīkh Baghdād, for instance, I found the following reference to Sumnūn ibn Ḥamza (d. c. 297/910): "A man wrote to Sumnūn asking him about his, that is Sumnūn's, situation after leaving his company. Sumnūn wrote to that man in response [. to ʿAmr ibn ʿUthmān al-Makkī (d. 291/903-904 or 296/909), for instance, we must look for further data in non-Sufi sources: this will be discussed in detail below. These sources very frequently refer to letters in the form of "so and so addressed many letters to so and so" without quoting from the letter. Another interesting example is found in Taʾrīkh Baghdād. Al-Ḥallāj's (executed 310/922) son Aḥmad ibn al-Ḥusayn ibn Manṣūr is quoted as having described both the hostility of ʿAmr al-Makkī to his father and the efforts made by al-Makkī to send many denunciatory letters ( yaktubu al-kutub) to the people of Khūzistān, where al-Ḥallāj lived at that time. 28 Based on these instances, we can see how letters are a good source for ideas not included in the famous large-scaled Sufi manuals and magna opera.
If the famous Sufi manuals are designed to underline Sufi morals and to address the general concerns of the Sufi community, Sufi letters are sometimes meant to express more personal emotions and interests. Even when typical expressions recur in letters composed by the same Sufi figure, it is still worth examining particular characteristics that distinguish each letter regarding both the context and dynamics of interpersonal connections. While there are expressions common to all of al-Junayd's letters, for instance, each letter relates to one aspect of his diverse network of relationships. The section of this paper entitled "Conceptual-historical perspective" will examine this network through diverse structural discourses and rhetoric.
As with other Sufi materials, including anecdotes, sayings, and poetry, the authenticity of fragments of letters is not easily ascertained. More interesting than the question of authenticity are the reasons that led later Sufi authors to preserve these fragments. Al-Sarrāj, as I have already indicated, was the first to perceive the importance of these fragments in describing the nature of Sufism and the aspirations of its followers. On two occasions, he uses the expression "wajadtu fī kitāb" (lit. "I found in one kitāb") before quoting an anecdote about al-Junayd through the authority of Jaʿfar al-Khuldī (d. 348/959), and before quoting one saying that appears in a kitāb of Abū Turāb al-Nakhshabī (d. 245/859). 29 The phrase "wajadtu fī kitāb bi-khaṭṭī. . ." appears very frequently in ḥadīth literature, especially when the transmitter is eager to emphasize that he has consulted actual manuscripts in which the particular tradition and its chain of transmission are preserved. The word kitāb in such cases refers to a piece of paper, a notebook or a booklet but not to a regular book. 30 If the reference is to a regular book, then the title of the book usually appears after the term kitāb. 31 Al-Sarrāj's use of the phrase "I found in a kitāb" most likely relates to a piece of paper or a notebook; where he refers to the term kitāb alone, such as in his reference to one kitāb sent by al-Junayd to Mimshādh al-Dīnawarī (d. 299/ 911-2), the reference is to an original form of a letter. In such a case, al-Sarrāj relies on the authority of others who transmitted both the reference to the letter and its circumstances to him; al-Sarrāj indicates that he "heard Aḥmad ibn ʿAlī al-Karajī" telling him that "al-Junayd addressed a letter to . . .". 32 On another occasion, al-Sarrāj indicates that "it was narrated about al-Shiblī that he addressed a letter to al-Junayd". 33 In cases where Sufi authors did not have such original versions in hand, they probably referred to consensual, orally transmitted versions. While referring to the response letter of Yūsuf ibn al-Ḥusayn on al-Junayd's letter, Abū Nuʿaym al-Iṣfahānī indicates that the transmitter "read the response letter". 34 On another occasion, in his Ḥilyat al-awliyāʾ, al-Iṣfahānī indicates that al-Junayd addressed a critical letter to his close companion, Ibrāhīm ibn Aḥmad al-Māristānī, concerning al-Māristānī's agreement with one of the mutaʾawwilīn, which most likely refers to Muʿtazila and their exegesis tradition. 35 Before presenting the text of the letter, al-Iṣfahānī indicates that he had heard it from someone (akhbaranā bihā) who transmitted it from another who transmitted it from someone who lived during al-Junayd's time. 36 While it seems likely that some letters were transmitted orally, it could have been possible for some of these letters to have been handed down in writing among the close circle of the recipient's friends. It is likely that many of these letters were commonly conceived as personal or even secret and that they were not originally intended to be part of any public compilation. We know, for instance, that one of al-Junayd's secret letters fell into the hands of a group of people who could use it to threaten al-Junayd, as he himself indicates in another letter that he addressed to Abū Bakr al-Kisāʾī al-Dīnawarī. 37 On many occasions when recipients shared letters with their coteries, some of their companions copied parts of these letters, or at least the introductory parts, thereby creating their own "ṣūrat kitāb" (lit. an image or copy of a letter). This term is very common in Classical Arabic works of adab, biographies, and historiographies. Physically, these copies were handwritten on paper or skin patches (sing. ruqʿa) and passed among acquaintances, authors, and scholars. 38 It would be possible to assume, furthermore, that there were also occasions on which certain short messages were orally transmitted. This assumption fits with the idea that mutual inquiries (masāʾil) and answers (ajwiba) on Sufi terms and conditions played a fundamental role in early Sufi tradition. Many of these inquiries and answers were a form of orally transmitted letters. 39 What began as an oral message, however, was later documented in writing through different forms of transmission when acquaintances of both the sender or the recipient had interest in the content. While this process of transmission succeeded in preserving a huge number of letters, we should bear in mind that this occurred to a greater extent with administrative letters and letters with typical literary values. Personal letters, on the other hand, were easily missed and ignored due to the lack of interest in what was seen as private "business" between relatives or companions. On one occasion in ʿAbd al-Karīm ibn Muḥammad al-Rāfiʿī al-Qazwīnī's (d. 623/1226) historiography, he states that he himself saw a handwritten document of the qāḍī ʿAbd al-Malik ibn al-Muʿāfā in which he, the qāḍī, quotes a verse taken from the introductory part of a letter (ṣadr kitāb) addressed by Abū Ṭāhir al-Jaʿfarī, the honourable personality of Qazwīn, to his grandfather Muḥammad. 40 On another occasion, al-Rāfiʿī indicates that he saw another person in a high administrative position in Qazwīn writing some verses at the top of a letter addressed to one of his friends. After quoting these verses, al-Rāfiʿī tells a story about Aḥmad al-Ghazālī (d. 517/1123 or 520/1126), the young brother of Abū Ḥāmid al-Ghazālī (d. 505/1111), which appeared in the main body of that letter (khilāl al-kitāb). 41 It is important to remember that Sufi letters are rarely just personal or private documents. They usually take the form of public literary epistles due to the large space they devote to discussions of Sufi conduct. Personal and interpersonal features, however, can still be gleaned; but these are not always easy to see and a cross-check with other types of sources is required. Comparing Sufi letters with private and official letters that belong to general medieval Islamic epistolography 42  überreste, and this is useful for two major reasons. It helps situate Sufi letters into their broader religious, social and historical contexts as well as delineate specific structural attributes that distinguish Sufi letters as a unique category of Islamic epistolographical literature. In Sufi letters that I have been able to access, abstract literary contents concerning Sufi conduct and general manners and beliefs still occupy a prominent place; however, many of the letters go beyond the displays of Sufi manners to include references to particular circumstances and events. Having said this, we are still unable to know with any confidence whether senders of early Sufi letters intended exclusively to address particular real-life situations or whether they were taking advantage of such situations to express appropriate conduct for future generations of Sufis. I would argue that, occasionally, a letter could have been a response to a real set of circumstances, which was then amended to reflect general Sufi conduct, thereby giving the letter more timeless value. The following example relates to Bishr ibn al-Ḥārith (d. 227/841) and sheds light on two correlated issues: the first is the nature of relationships between historical context and theoretical presentation of morals in one Sufi letter; and the second is the problematic issue of authenticity, which questions the circumstances that surrounded the composition and address for each letter. This question is raised very often when the given letter is shaped in a way that combines a common structure of letterwriting in general with the actual historical event that motivated the writing of the letter. Some Sufi letters are likely to be shaped differently without any connection to a particular event. These follow the general format of "so-and-so wrote a letter to so-and-so . . .". pure devotional life in the search for worldly profits. The passage quoted by al-Sarrāj is: "I have heard that you have become satisfied with working in spinning wheels for your subsistence. With whom will you be able to seek refuge if God deprives you of your senses of hearing and seeing?" We do not know if this was a sort of a completed telegram-style letter or just a single passage taken from a longer original letter. What we do know based on what al-Sarrāj narrates is that the letter left a deep impact on Bishr and caused him to abandon completely the profession he was pursuing with his sister. 44 On the assumption that this short fragment was taken from a longer authentic letter, we might suggest that the original letter included further passages of a general literary nature celebrating the high Sufi ethos of absolute dependence on God (tawakkul) and a full devotional life; however, we should keep in mind that this does not change the main focus of the original letter which is to castigate a specific Sufi figure for a specific controversial behaviour at one specific point in time. It is still interesting to note that the way this letter is introduced to Kitāb al-Lumaʿ presupposes that al-Sarrāj probably intended to show Bishr's approach towards the dilemma of working for one's subsistence or choosing to commit oneself totally to tawakkul. It is also possible that al-Sarrāj sought to emphasize the idea that Bishr changed his approach on this subject and finally abandoned working for his subsistence. The brevity of this letter does not allow further investigation into the questions of how and to what extent moral discussions were able to set the tone of the letter originally written in response to a certain event or under the influence of a specific situation. General moral discussions are usually common in literary letters, that is, in rasāʾil as treatises or monographs. This investigation is possible with a longer example letter and will be discussed in the following section.

Conceptual-historical perspective
In this section, I will discuss questions of content, as well as different expressive forms and styles relating to the changing conditions and historical events in certain relevant letters. This examination will also involve analysing rhetoric and forms of argumentation, self-justification and self-representation of the senders and their use of frequent structural features, metaphors and symbols. It will focus on structures used by either different senders or by the same sender whether writing to different addressees or for different pragmatic purposes.
Based on content, conditions, and historical events, the bulk of Sufi letters will be identified in different ways. One method will be to distinguish between general letters addressed either to an anonymous individual or to a group from letters whose addressees are specific. This is, for instance, the case in one of Abū Ḥafṣ al-Suhrawardī's letters to "one of the Sufis", which takes the form of a piece of counsel (waṣiyya ilā aḥad al-fuqarāʾ) and revolves around general codes of behaviour and good manners each Sufi aspirant needs to adopt. 45 One interesting example of letters addressed to a general group is the surviving part of the letter of ʿAmr ibn ʿUthmān al-Makkī to the group of the Sufis of Baghdad who are known in the early history of Sufism as ṣūfiyya. Al-Sarrāj tells us that when this letter arrived, it was read in the presence of al-Junayd, al-Shiblī (d. 946/334), and Abū Muḥammad al-Jurayrī (d. 312/924), who were among the most prominent members of the ṣūfiyya group. 46 If the rhetoric of the letters is considered, then we should differentiate between letters whose style is purely poetic, letters which are written metaphorically, and letters which quote the sayings of other Sufis. This method of classification is not always clear-cut, and crossovers do occur. In his famous letter to ʿAmr ibn ʿUtmān al-Makkī, al-Junayd adopts a metaphorical, narrative style by relaying the textual dialogue between a learned man (ʿālim) and a wise man (ḥakīm). In this dialogue, the wise man strives to convince the learned man to replace all the earthly considerations, such as greed, social fame, or leadership, that underpin the learned man's religious science with an exalted type of inner science. 47 The learned man throughout this long dialogue is portrayed as submissive and slavish. He repeatedly confesses that lower earthly benefits motivate his search for religious knowledge. The voice of the wise man, on the contrary, is dominant and very assertive. In one passage the learned man is quoted saying: ʿAmr ibn ʿUthmān al-Makkī was one of the honourable masters of jurisprudence, and when he was appointed to the position of qāḍī of Jeddah, al-Junayd no longer associated with him. 49 On another occasion in his Tārīkh al-Islām, al-Dhahabī indicates that when ʿAmr al-Makkī came to Baghdad, undertook the position of qāḍī in Jeddah and became ill, al-Junayd did not visit him during his sickness. 50 If we add this notion to the structural and thematic components of the letter itself, we come to the conclusion that al-Junayd was critical of the jurists of his day, and that this attitude took form in a poignant letter whose actual effects on the addressee are unknown. This is one example of the rich array of data that Sufi letters can present particularly once they are cross-checked with textual evidence provided by non-Sufi sources.
I have already noted that al-Sarrāj was the first author to dedicate a separate chapter in his work to Sufi letters and correspondence. Only a few Sufi authors who lived after him do that as well. In the majority of cases, Sufi authors choose to include references to letters and to quote fragments from earlier letters in different discussions throughout their work. The choice by a later author to mention a particular letter was generally based on one key concept emphasized in that letter which helps the author prove the correctness of his worldview. 51 By its very nature, this method of using letters contributes to strengthening the literary aspects of these texts over any historical-actual aspects; it disconnects the letters from their original contexts and makes it very difficult for us as researchers to reconstruct the circumstances of writing and generating the original letters.
Al-Sarrāj was unique in the ways he chose both to preserve Sufi letters and to use them in his work. Our survey of the relevant material in Kitāb al-Lumaʿ leaves no doubt that al-Sarrāj was conscious of the fundamental importance of these texts as one of the pillars of Sufis' communal life both from early Sufi history and continuing until his own time. It is surprising, however, that authors of Sufi manuals who lived after him and frequently cited his work did not share his interest in Sufi exchanges of letters; this needs to be the subject of further research.
One of the early Sufi letters that deserves special mention is al-Junayd's letter to Yūsuf ibn al-Ḥusayn al-Rāzī, a distinguished mystic of Rayy of that period. Different sources state that this was one of a series of letters between these two figures. Abdel-Kader, although he refers in some detail to the relationship between al-Junayd and Yūsuf ibn al-Ḥusayn, does not relate to the content of their correspondence, its rhetoric, or the way in which the structure might suggest some aspects of the relationships between them. It is highly likely that it was al-Junayd who initiated this correspondence, as evidenced by the statement "badaʾtuka bi-kitābī" (I began my letter to you) at the end of his letter, 52 and that Yūsuf ibn al-Ḥusayn's letter, or even several letters to al-Junayd, came in response. Al-Junayd opens his letter with a long paragraph whose rhetoric relies on the intensive use of verbal sentences in the past tense. This writing style was very common in Classical Arabic letters and was used by the author to invoke divine favour on behalf of their addressees. At first sight, al-Junayd's opening leaves the impression that he praises Ibn al-Ḥusayn's high spiritual rank as one whom God chooses to grant walāya, an elevated position of closeness and friendship. 53 However, on deeper analysis, what seems like praise presented in the form of positive invocations, turns into a description of how far Ibn al-Ḥusayn needs to travel spiritually in order to attain a high spiritual rank. The first sentences in the opening reads: kashafa al-Ḥaqqu laka ʿan ḥaqīqat anbāʾihi, wa-tawallāka bi-ʿaẓīmi minanihi wa-ālāʾihi, wa-taḍammanaka fī ḍammihi iyyāka ilā sawābigh naʿmāʾihi, wa-jarat ʿalayka bi-rafʿihi laka ilayhi wa-iʿlāʾihi, fa-kunta bi-ḥaythu lā takūnu al-aghyāru laka ilayhi sababan, bal takūnu bi-mā yūjidu bihi minka muntasiban.
May God reveal to you the true nature of His revelation, and grant you the greatness of His favour and graciousness. May He contain you by embracing you yourself in the fullness of His beneficence which, when they reach you, are the grace of His raising you and exalting you. Then will you be where no other is a mediator between you and Him, but you will be in a relationship with God based on that which God has given you. 54 Following this graceful and muted opening, al-Junayd's tone becomes more direct, negative, and obvious. He describes very unfavourably the reality that he has seen with the existence of many pretenders who lack sincerity and truthfulness while looking for earthly benefits in their learning and teaching involvement in circles of ʿilm (religious science). Afterwards, al-Junayd turns to address Ibn al-Ḥusayn, without stating his name or title, and while acknowledging his role as a guide to people on the path of truthfulness, he urges him to strengthen his authority over his disciples: "Turn to your disciples and give them your full attention, face them and concentrate on them [. . .] be with them both by night and by day and give them that special cognisance of your experience". 55 The latter part of the letter implies that al-Junayd felt that his addressee is expected to feel insulted by the powerful and authoritative tone of this letter, and that is why he writes: "Now I began my letter to you with a view to establishing our closer contact, seeking your attention and your good graces and hoping that it would cause you to write back to me [. explicitly urges Ibn al-Ḥusayn to be cautious with his contemporaries and advises him to speak only after he is sure that he knows his listeners. 56 Before examining the response letters of ibn al-Ḥusayn, a brief remark concerning the historical and religious contexts of this exchange of letters should be noted. Recent scholarship on the formative period of Sufism has shown that al-Junayd and his close circle of Baghdadi Sufis succeeded during the course of the third/ninth century in establishing a strong and authoritative institution of Sufi piety based on a group of doctrines and renunciatory practices. 57 Representatives of this group began to spread this type of piety to the eastern parts of the Muslim world in an attempt to ensure the supremacy of their own image of Sufism over all other images and forms that then existed. Baghdadi Sufis, with al-Junayd as their most influential leader, aspired to prove that their perception of taṣawwuf was the best one. Analysing al-Junayd's letters, and more specifically his letter to Ibn al-Ḥusayn of Rayy, through the prism of the debate between the malāmatī trend and the Baghdādī trend, one can see that al-Junayd intended to maintain and strengthen the religious authority of Iraqi piety over the Sufis of the East; the above-mentioned quotations and the arrogant tone implied in its rhetoric and phrasing leave a strong impression that this was the case. In one part of the letter, al-Junayd writes: "My brother, it was not my desire to draw your attention to a privilege and a duty which you neglect, nor to any suspected deficiency [. . .] but God says: Teach for instruction is of avail to those that believe". 58 As for Ibn al-Ḥusayn's response letter, or set of letters, Abū Nuʿaym al-Iṣfahānī preserves a very short passage of what was commonly identified by al-Iṣfahānī and later authors as Ibn al-Ḥusayn's jawāb (response) to al-Junayd. It is likely that this passage was taken from a long-missing letter in which Ibn al-Ḥusayn defends himself against accusations of heresy addressed to him by some jurists of Rayy, along with a critical approach of the Baghdadi Sufis whom al-Junayd leads and represents. Al-Qushayrī indicates that in this letter to al-Junayd, Ibn al-Ḥusayn prays to God that al-Junayd keep himself from tasting his "lower soul's flavour", since, by tasting his lower soul's flavour, he "will not obtain any goodness forever": "lā adhāqaka Allāhu ṭaʿma nafsika, fa-innaka in dhuqtahā lam tadhuq baʿdahā khayan Abadan". 59 This statement, which implies a response to the arrogant tone in al-Junayd's original letter, appears in what seems to be a missing response letter of Yūsuf ibn al-Ḥusayn in the Șehid Ali Pașa manuscript. 60 This is one of two response letters that were, allegedly and most probably, addressed by Ibn al-Ḥusayn. They are preserved in the Süleymaniye Yazma Eser Kütüphanesi in Istanbul. Though there is no reference to the recipient's name in the main body of both letters, a careful reading strengthens the assumption that the sender addressed the letters to one charismatic master whose initial critical letter was hurtful. As I have already mentioned, the first modern scholar to refer to these two letters was Muḥammad Muṣṭafā in his 1978 edition. Muṣṭafā, like Abdel-Kader before him, does not provide any discussion of the texts that he edits. More recently, Jamāl Rajab Sīdbī republished al-Junayd's Rasāʾil, including both his letters and short treatises on tawḥīd and other Sufi issues based on a manuscript of Maʿhad al-Makhṭūṭāt al-ʿArabiyya in Cairo. Appended to the above-mentioned letter of al-Junayd to Ibn al-Ḥusayn, Sīdbī published an additional part that does not appear in Abdel-Kader's edition. 61 This part deals with how different types of Sufis view intoxication (sukr) and it has little relevance to the topic and rhetoric of al-Junayd's letter; this leads us to doubt its authenticity. Later in his edition, Sīdbī published another letter which he maintains was addressed by al-Junayd to someone called Yūsuf ibn Yaḥyā, thus causing confusion between two separate figures of Rayy who had both exchanged letters with al-Junayd: Yaḥyā ibn Muʿādh and Yūsuf ibn al-Ḥusayn. 62 This was the response letter addressed by Ibn al-Ḥusayn to al-Junayd 63 and this view is shared by Muṣṭafā. In Sufism and Early Islamic Piety (pp. 129-34), I relied on the Şehit Ali Paşa manuscript to examine thoroughly the structural and thematic features of these two letters while providing English translations for some parts of them. The opening paragraph is of particular interest. The author, after praising his addressee, clearly indicates that the latter has given him a blow and caused him pain (ḍarabta fa-awjaʿta). He compares his addressee to the prophets who chose leniency while guiding their people down the path of truth and belief; nonetheless, he asserts that, like all prophets, his addressee is not able to guide anyone alone since no one other than God is capable of guiding people. The different rhetorical strategies used by the author to establish this message, in plain terms on some occasions and metaphorically on others, help him refrain from turning his letter into an explicit trigger for personal dispute. At one point in this letter, for instance, the author indicates that he has benefitted from the addressee's previous letters to him in choosing the style and language for his own letter. 64 This letter provides many contextual-historical values. By the end of this letter, the author relates that some renunciants (zuhhād), who had visited the addressee from the author's region, returned home and asked the author to "scale up" (istazādūnā); this could be interpreted as their asking him to commit himself to more devotional rituals as is now their commitment. The second response letter is shorter and its style is similar to the first. Some of its passages were quoted by both Abū ʿAbd al-Raḥmān al-Sulamī (not as a letter but as a general saying) and al-Iṣfahānī. 65 The general tone suggests a sort of a controversy that the author had with the addressee in relation to the passionate language that he, the author himself, chose to express his love and intoxication: "You asked me to avoid writing to you using irrational language (hadhayān) as well as to avoid the talk style of the insane (kalām al-maʿtūh)". 66 I have already referred to the question of why the majority of personal letters in Sufi domains did not come to us in their original handwritten forms, and how some of these letters were known to other contemporaries who helped transmit passages of them at a later date to authors of collections and Sufi compendia. According to one reference that appears in al-Iṣfahānī's Ḥilyat al-awliyāʾ, a man of Rayy, who was a well-known munshid (reciter of poetry), 67 indicates that, when the letter of Ibn al-Ḥusayn reached al-Junayd, he, this man, wished that he could see it due to its charm. It is not clear how this man knew about the charms of Ibn al-Ḥusayn's writing style and what exactly is intended by charm. This man goes on to tell that he left Baghdad for Rayy in order to meet the author of the letter. When this man began to recite poetry, as Ibn al-Ḥusayn requested, the latter began to cry and admitted that the people of Rayy correctly accused him of heresy since reciting the Quran did not cause him similarly to cry! 68 Immediately after this reference, al-Iṣfahānī quotes one of al-Junayd's companions, Abū al-Ḥasan ʿAlī ibn Hārūn, as having said: "I read in the response letter of Yūsuf ibn al-Ḥusayn to al-Junayd [. . .]". What follows is a passage taken from the original second response letter of Ibn al-Ḥusayn 69 and is additional proof for the argument that Sufi letters were accessible to the close circles of both senders and addressees. The way many of these letters were conceived as functional texts might have caused the loss of the originals; however, the close circles of senders and addressees helped preserve certain traces of contents in which these circles were interested so that they succeeded in remembering and transmitting.
In addition to Ibn al-Ḥusayn, many other personalities corresponded with al-Junayd. What we have as evidence is restricted to brief references to their names and the general nature of their letters, without quoted texts. On one occasion in Ibn al-Jawzī's biographical work Ṣifat al-ṣafwa, the author indicates that ʿAlī ibn Sahl was one of the best men at using the symbolic style (min aḥsan al-nās ishāratan), and that he used to exchange letters with al-Junayd while al-Junayd praised him and likened his language to that of angels. 70 Ibn al-Ḥusayn's letters provide excellent examples of how Sufi letters enable us to reconstruct possible conflicts and tense encounters in the daily lives of early Sufis. Sometimes, such encounters are implied in plain brotherly letters, for example the letter of Abū al-Khayr al-Tīnātī 71 to Jaʿfar al-Khuldī, in which al-Tīnātī blames al-Khuldī for not taking responsibility for the ignorance of his followers. 72 Again, the letter was addressed in response to a set of real circumstances, while any existing additional parts dealing with general Sufi conduct have been appended to give the letter more timeless value. In other cases, however, such general and timeless content serve as a fundamental part of the Sufi letter, while the real historical content is hidden. This is the case in al-Junayd's letter, or probably several letters, to the honourable preacher and mystic of Rayy, Yaḥyā ibn Muʿādh al-Rāzī (d. 258/872). Fragments of this letter are preserved in the Şehit Ali Paşa manuscript, 73 while others appear in Kitāb al-Lumaʿ. 74 The reader notices al-Junayd's clear attempt to show his generous knowledge of Sufi terminology and his exceptional writing talent. The literary nature which is dominant in this letter motivates us to ask why al-Junayd chose this general, less historical, rhetoric for one letter while relying on some historical-contextual features in another. Unfortunately, this question remains unanswered. In both cases, however, it is clear that the direct trigger for creating the letters was a real set of events in a certain context. Even when the risāla served as a small book sent by its author to deliver certain teachings and ideas, this was motivated by actual situations and to fulfil specific needs in the living reality of both sender and addressee.
Al-Junayd was not the only Sufi leader who sent letters that contain rebuke and criticism. Hasty and brief references to many other figures who addressed short messages to others also appear in the early sources. These messages usually take the form of telegram-messages that include very few sentences and whose focus is on the sender's critique of certain behaviours of the addressee; this is expressed through a generally moralistic tone without any direct reference to the addressee's identity. One example is the aforementioned letter addressed by Abū al-Khayr al-Tīnātī to Jaʿfar al-Khuldī. In addition to the reference to al-Sarrāj, the letter appears in al-Qushayrī Risāla as follows: The responsibility for Sufis' [lit. poor's!] lack of knowledge is on you; since you were so wrapped up in yourselves and forgot to guide them; they remained uneducated. 75 On certain occasions, the texts of letters, which were primarily dominated by moralistic issues and discussions, leave the impression that those letters were delivered orally. Al-Qushayrī indicates that Yaḥyā ibn Muʿādh of Rayy wrote (kataba) to Abū Yazīd al-Basṭāmī telling him that "here is someone who has 71 Ibn Kathīr states that al-Tīnātī was an originally Arab renunciant who lived in a village called Tīnāt in the region of Anṭākia and was known by the title al-aqṭaʿ, "the amputee", since his arm was cut after he broke a word that he undertook with God and, later on, was arrested by mistake with a group of thieves in the desert where he used to spend his devotional retreats. drunk from the goblet of God's eternal love so that he can never be thirsty again ever". It was said that al-Basṭāmī wrote to Yaḥyā in response: "I was astonished by the weakness of your state! Here is someone who drinks the seas of the universe and he still opens his mouth seeking much more [water!]". 76 It seems probable that the verb kataba, here as on other similar occasions, was used in the sense of "he sent an oral message". On another occasion of al-Risāla al-qushayriyya, a reference was made to Abū ʿUthmān al-Ḥīrī, who "wrote to Muḥammad ibn al-Faḍl asking him: what is the sign of misery?" 77 Sufi literature is fraught with notions of this kind. References to different forms of contact between early Sufi personalities (exchanges of letters, travelling, companionship, ṣuḥba, etc.) were very often designed to emphasize particular religious aspects of these personalities or to help authors defend certain earlier personalities who were targets of non-Sufis' criticism. Assuming that the contact between al-Ḥīrī and Muḥammad ibn al-Faḍl was real, the reference to the letter might preferably be understood as an oral inquiry that al-Ḥīrī addressed to Ibn al-Faḍl by means of a direct encounter or through a messenger. 78 From another perspective, it is possible that the entire story was fabricated by later authors to establish the alleged close relationship between these two figures and to use this relationship to defend Muḥammad ibn al- Shams al-Dīn al-Dhahabī, using al-Sulamī as his authority, indicates that al-Ḥīrī is alleged to have said: "Would I have enough power, I would travel to my brother Muḥammad ibn al-Faḍl so that I can enjoy the delight of seeing him". 80 This implies that if the story of the exchange of letters between these two personalities was fabricated, this fabrication was grounded on the fact that geographically they lived in two distant locations.

Conclusion
This paper introduced exchanges of letters and correspondence between Sufis during Sufism's formative period as documents with social, interpersonal and historical values. By social, interpersonal and historical values I refer to contexts and backgrounds that pushed forward the addressment of letters, as well as the textual nuances that, when cross-checked with relevant data provided in other sources, could demonstrate certain aspects of Sufis' living realities. A deep analysis of the exchange of letters between al-Junayd and Yūsuf ibn al-Ḥusayn, for instance, could provide an extraordinary portrait of some aspects of the relationships between Baghdadi Sufis (the ṣūfiyya) and Sufis of the East during the third/ninth century. The distant relationship between the two personalities as implied in their letters provides an interesting facet of the debate between the Sufi piety of Iraq and that of other parts of Muslim territories, especially the malāmatiyya trend of Khurāsān.
Separating rasāʾil (letters) as literary epistles from rasāʾil as actual letters is even more difficult in Sufi contexts than in non-Sufi ones. The literary genre imposes structural constraints which then influence the content; this content contains practical messages flowing from particular historic moments and giving rise to actual letter writing. Though Sufi letters are rarely just personal or private documents, owing to the large amount of space they devote to discussing Sufi conduct, personal and interpersonal features can still be gleaned. This paper highlights the significance of these documents from two perspectives: the first methodological and the second conceptual-historical.
Using the methodological perspective, this paper emphasized the crucial need for creating a corpus of Sufi letters and pieces of correspondence written in the period between the late ninth and thirteenth centuries between Sufis and, on certain occasions, between Sufis and non-Sufis. A detailed survey of the sources for such material was presented. Questions of transmission and usage of letters by later authors as well as the real circumstances that surrounded the composition of letters or the response letters were also discussed. Very frequently, abstract literary contents concerning Sufi conduct and general manners and beliefs still occupy a prominent place in these letters; this gives these documents a timeless nature. In order to glean interpersonal features, a crosscheck with other types of sources is required. Some examples of this procedure were presented here.
Using the conceptual-historical perspective, questions of content and different expressive forms and styles relating to the changing conditions and historical events in certain relevant letters were discussed. Furthermore, rhetoric, forms 80 al-Dhahabī, Siyar aʿlām al-nubalāʾ, 14: 524. of argumentation, self-justification and self-representation of the senders and their use of frequent structural features, metaphors and symbols were also analysed. Based on contents, conditions and historical events, the bulk of Sufi letters were categorized and identified differently. Throughout this part of the paper, many sample letters and response letters were analysed to demonstrate the methodological method of treatment as well as the real-life situations reflected in what appear to be literary documents.
T H E E X C H A N G E O F L E T T E R S I N E A R L Y S U F I S M