EḴWĀN AL-ṢAFĀʾ, a self-professed brotherhood of piously ascetic scholars. In order to advertise and propagate their special mix of philosophy and religion, the Eḵwān al-Ṣafāʾ wrote a lengthy account of all the known sciences and how the study of each in turn contributes to help liberate the soul and set it on a course toward a future angelic existence once detached from its earthly prison upon the death of the body.

Their work, called simply the “Treatises of the Pious Brethren” (RasāʾelEḵwān al-Ṣafāʾ), is often taken as an early form of the encyclopaedia. Massive and comprehensive in its depiction of the sciences, it encompasses the gamut from arithmetic to magic, including geometry, geology, music, astronomy, biology, psychology, physiology, politics, religion, medicine, education, meteorology, ethics, and more between these extremes. However, while impressive for the amount of learning these treatises contain, they were all compiled for a single, basic purpose—one that their anonymous authors stress over and over again to the point of tedium: learning and studying are training for the soul, a preparation for its eventual life as an independent being once freed from the body; this future life, which is truly authentic of its real disposition and substance, is intellectual and is neither worldly nor based on sensation; the pleasures of physical existence are, by contrast, a hinderance, a trap blocking and restraining the soul from its proper course; knowing the soul, the self, therefore, and how it is or has come to be connected to the body is the primary task of any education (e.g., see Rasāʾel, Beirut , I. p. 76, IV, p. 161).

Realization of the dual nature of humankind as both soul and body allows those who thoroughly understand this connection to value as little as readily tolerable the demands and requirements of bodily existence. The body is nothing but a tool of the traveling soul much like a ship on the ocean for those who ride in it. The soul is like the captain that controls or the wind that moves it; and, like that ship, the body is necessary only until the terrestrial passage of the soul is complete (Rasāʾel I, p. 29, IV, p. 256).

As a consequence of this view of the soul, the Eḵwān regard as genuinely beneficial and basically in harmony with all forms of learning expressing a comparable doctrine any knowledge from any source that supports the soul’s improvement and amelioration. For them, philosophy is one of the major, fundamental categories of this knowledge, as is even more profoundly revelation and prophetic scripture. An attachment to the doctrine of the essential concord between philosophy and religious law characterizes their thought above all else. Both cover divine matters and aim at the same root principle (aṣl), even though differing in methods (forūʿ; Rasāʾel III, p. 30). This point is abundantly obvious and easily recognized by anyone who peruses the Rasāʾel— one already duly noted in the 4th/10th century by Abū Ḥayyān Tawḥīdī (q.v.), who is perhaps our earliest outside witness to these writings. This means also that, for the Eḵwān, philosophy in its proper definition builds upon methods for the purification of the human soul. This was especially true for the philosophy of Pythagoras and Socrates—both of whom belong to the pantheon of the Eḵwān (for Pythagoras, see esp. Rasāʾel III, pp. 199-204).

The inspiration of the Eḵwān al-Ṣafāʾ comes thus directly from philosophy in conjunction with various prophetic sources, the latter ultimately resting on the final revelation given to Moḥammad which produced Islam, the religious law of the Eḵwān themselves and all Muslims. The stated list of their authorities does, however, tend toward the eclectic: Abraham, Joseph, Jesus, and Moḥammad, among the Semitic prophets; the corpus of Muslim Hadith, from post-prophetic, Islamic tradition; Socrates, Pytha-goras, from Greek thought; and Balawhar and Būḏāsaf, from the East (Rasāʾel IV, pp. 58, 175; see BARLAAM AND IOSAPH). To these may be added material taken from and credited explicitly to, for example, Nicomachus (arithmetic), Euclid (geometry), Ptolemy and his Almajest (geography, astronomy) among the Greeks. The Eḵwān depend on many more borrowed sources including stories, information and parables featuring the king or the sages of Persia as well as a king or sages of India. Among the other Iranian materials, the Rasāʾel contain interesting examples of parables featuring animal personification in the manner (quite possibly in direct imitation) of Kalīla wa Demna. For India, the citations and numerous allusions to the Buddha (Būd¨āsaf) and his ascetic instructor, Balawhar, confirm the broad vision of the Eḵwān, who in general reject narrow partisanship in favor of universalism. They claim repeatedly that their Brotherhood does not accept any one religion (mad¨hab) to the exclusion of another but rather supports and promotes the essential, single core teaching of all religions (maḏāheb) and of all the sciences combined (Rasāʾel IV, pp. 42, 52, 167).

The fifty-two separate treatises in the Rasāʾel (including the extended summary of them in the final al-Resāla al-jāmeʿa) follow a preset plan that encompasses four categories or courses of the sciences arranged in ascending order of importance and difficulty (Rasāʾel I, pp. 21-43). The uneducated initiate begins with a course on the introductory tools of instruction: numbers, figures, astronomy, music, proportion, various outlines of the theoretical and practical arts, basic ethics, and the first five parts of logic (Rasāʾel I). Next, in the second course, the adept takes up such subjects as physics, meteorology, biology, physiology, and ultimately several sections specifically on the human being (as a microcosm, as having various aptitudes, as subject to mortality, pleasure and pain, and as divided by differences of language; Rasāʾel II and III, pp. 5-177). The third division of the sciences concerns intellect and soul commencing with an account of intellectual principles according to Pythagoras followed by those of the Eḵwān themselves. This section stresses the concept of the universe having a single soul that behaves as a whole as if it is a macrocosm of the single, individual human. It responds obediently and with love (šawq) for its cause—the Creator—in the same way, for example. Thus its motion, its internal causes, and its order and arrangement lead to its salvation just as they do for each human being (Rasāʾel III, pp. 178-400). Finally, section four turns to divine law and religious legislation (ibid., pp. 401-538, IV). More than the previous three, this part delineates the most intimate beliefs of the Eḵwān, although again always recognizing a curiously eclectic, philosophical component alongside prophetic religion. There is, for example, here an account of philosophical religion as practiced by Greek philosphers, including a discussion of their festivals (ibid., IV, pp. 263-68). The practices of the Eḵwān are not exactly the same but are analogous, just as they are similar to but not the same as standard Islamic rites. The final two divisions of this fourth course cover, first, political governance (sīāsa) and then magic, alchemy, astrology, and spells. Parts of their political doctrine closely parallel the ideas of Abū Naṣr Fārābī from whom they may be derived. Inclusion of a section on the occult sciences is indicative of the comprehensive nature of the whole enterprise: the Eḵwān literally seek the betterment of the soul through the acquisition of the universal elements in all forms of knowledge.

Ultimately, however, the statements in this fourth portion of the Rasāʾel yield few clear details about the doctrines of the Eḵwān, offering instead cryptic allusions that may be interpreted as pointing in several contradictory directions. Hence, the Eḵwān appear to be, on the one hand, strongly Shiʿite but, on the other, not zealously so; they are philosophical, but again not adamantly or exclusively so vis-à-vis prophetic revelation which they continue to insist is superior; they appear Sufi and ascetically otherwordly but nevertheless concerned with the basic material well-being and preservation of the individual and of terrestrial, human associations and society.

From a strictly Islamic perspective, the Eḵwān most consistently present themselves in terms and concepts that reflect the position of the earliest Ismaʿilis. In addition to many reverential tributes to the Ahl al-bayt, to ʿAlī as waṣīy (executor of the prophetic legacy), and to vaguely recognized imams and the various tragedies that have befallen them in the past, the general tone of the hierarchical metaphysics underlying their cosmology is characteristic of other, near contemporary, Ismaʿili authors. A telling example is their use of the term al-sābeq for universal intellect and al-tālī for universal soul (Rasāʾel IV, pp. 200, 209). Thus, although their Neoplatonism—so evident in the account they offer of intellect, and most especially, of soul—does not by itself link them to the Ismaʿilis, added evidence of such a connection does exist in various details of these same doctrines.

Yet, despite these generally acknowledged similarities, a number of scholars have questioned this Ismaʿili affiliation (see particularly Netton). One difficulty, for example, is the propensity of the Eḵwān to counterbalance a pro-Shiʿite reference with another generously approving a virtuous act of Abū Bakr or ʿOmar or ʿOṯmān, thus obscuring, or possibly deliberately confounding, the Shiʿite interpretation of their message. Or, in another context, they boldly declare that, if the forty-six traits of the perfect ruler are not present in a single person at a given period, then forty-six individuals each possessing but one of these traits may substitute as imam in his place (Rasāʾel IV, p. 125). This concept, like that of Fārābī in a similar case, does not seem to be Ismaʿili, or at the least accords poorly with common (Fatimid) Ismaʿili thinking which would necessarily recognize the inherent perfection of the single imam (present in the world at all times). This doctrine seems to preclude a direct link to the Ismaʿilis except possibly for a few of the dissentient factions at an early date who continued to uphold the imamate of Moḥammad b. Esmāʿīl, the grandson of Jaʿfar al-Ṣādeq. In the absence (ḡayba) of this figure, some of the Eastern (Carmathian) groups advocated governance of the Ismaʿili daʿwa (see DĀʿĪ) by members of their hierarchy who were not themselves imams but rather formed a collective of lesser ranks known by the titles Ḵolafāʾ and Lawāḥeq (for a summary of various views about Moḥammad b. Esmāʿīl b. Jaʿfar, see F. Daftary, The Ismāʿīlīs. Their History and Doctrines, Cambridge, etc., 1990, pp. 102-06, 177-78; on the Lawāḥeq, see pp. 236-38). These dāʿīs did not, so far as we can tell, recognize the Fatimids as true imams even as late as the latter half of the 4th/10th century. Therefore, strictly on the basis of mere appearance of the Rasāʾel themselves, one can state that this form of early Ismaʿilism, which also conforms most closely to the doctrines of the Eḵwān, places them among the eastern, non-Fatimid Ismaʿilis. If this is true of them, it does not, as was the case with some other writers of that period, automatically preclude later Ismaʿili acceptance of their works by the mainstream. A number of the eastern authors of similar doubtful affiliation were later brought back into the fold, particularly in the post-Fatimid, Yemeni phase of the movement (Daftary, pp. 246-49).

Needless to say, modern investigators of the Rasāʾel have faced great difficulty in locating the actual authors of the work and, partly as a consequence, in dating the text itself. Ismaʿili tradition has long claimed that the writings of the Eḵwān, in fact, reflect in a rather direct manner the teachings of the pre-Fatimid imams. Credence for this position and for an early, possibly 3rd/9th century, dating for the material in the Rasāʾel, is contained in the modern studies of ʿAbbās Ḥamdānī, seconded by those of Yves Marquet. This interpretation would make the writings, now known under the name Eḵwān-al-Ṣafāʾ, only a later and final recension of what must, accordingly, be taken as official (pre-Fatimid) Ismaʿili teachings. Against this is the lack of a clear, unambiguous citation of them in surviving Fatimid literature or in any other Ismaʿili work of that period or of the 4th/10th century which is when Abū Ḥayyān Tawḥīdī first came in contact with them.

Tawḥīdī’s testimony concerning the authors and the Rasāʾel is, moreover, quite indisputable as far as it goes. In his al-Emtāʿ wa’l-moʾānasa (ed. A. Amīn and A. Zayn, Cairo, n. d.; repr., Beirut, n.d., pt. 2, pp. 4-23), he provides a list of those members of the Eḵwān familiar to him and records in addition portions of several discussions about them or by one of them that he knows of personally. According to the research by S. M. Stern, who was primarily responsible for bringing this information to bear on the problem of dating the Rasāʾel, Tawḥīdī’s evidence indicates that the Eḵwān emanated from Baṣra and that their writings were already in circulation by 373/983-84, but only recently so because the authors were still living.

The figures named by Tawḥīdī (pp. 4-5) are the following: Abū Solaymān Moḥammad b. Maʿšar Bostī, known as Maqdesī, Qāżī Abu’l Ḥasan ʿAlī b. Hārūn Zanjānī, Abū Aḥmad Nahrajūrī, Zayd b. Refāʿa, and ʿAwfī. Zayd b. Refāʿa was a personal friend of Tawḥīdī and was presumably the latter’s source of information about the Eḵwān. In the passages that survive from the anonymous Ṣewān al-ḥekma, this same Maqdesī is, moreover, given credit for composing the Rasāʾel as a whole (text published by Stern, 1946, p. 371). In addition to this the well- known Muʿtazilite chief judge of Ray, Qāżī ʿAbd-al-Jabbār Hamadānī (Taṯbīt dalāʾel al-nobowwa, ed. A.-K. ʿOṯmān, Beirut, n.d., pp. 610 ff.), also mentions in one of his works a similar list and notes particularly that Qāżī Zanjānī seems to be the leader of the group itself, the members of which were still active at that time.

Except for the later Ismaʿilis, who certainly came to revere the Rasāʾel highly, the later influence of the Eḵwān has not yet received sufficient study though there are likely to be traceable effects of their writings especially among those inclined to Neo-pythagorean and Neoplatonic philosophy and mysticism.



There are at least three editions of Rasāʾel, of which those of Cairo (4 vols., 1928) and Beirut (1957) are most frequently cited. The first complete edition, however, appeared in Bombay (4 vols., 1888). Large extracts of the Arabic text were edited by F. Dietrici along with a nearly complete German translation of the whole in his Die Philosophie der Araber im IX. und X. Jahrhundert, vols. 4-8, 10-11, Leipzig, 1865-86. The final “Summary” resāla (al-Resālat al-jāmeʿa) was edited first by J. Ṣalībā (2 vols., Damascus, 1948), then later re-edited by M. Ḡāleb (Beirut, 1974). More recently Susanne Diwald has translated again into German the third section on intellect and soul: Arabische Philosophie und Wissenschaft in der Enzyklopädie … Die Lehre von Seele und Intellekt, Wiesbaden, 1975. Additional shorter portions of the Rasāʾel have appeared in other languages as a part of various studies of them.

Studies. J. ʿAbd-al-Nūr, Eḵwān al-Ṣafāʾ, Cairo, 1954.

Ā. Awa, L’Esprit critique des Frères de la Pureté: Encyclopédistes arabes du IVe/Xe siècle, Beirut, 1948.

A. Bausani, L’Enciclopedia dei Fratelli della Purita, Naples, 1978.

Idem, “Scientific Elements in Ismāʿīlī Thought. The Epistles of the Brethren of Purity,” in S. Nasr, ed., Ismāʿīlī Contributions to Islamic Culture, Tehran, 1977, pp. 123-40.

H. Corbin, “Rituel sabéen et exégèse ismaélienne du rituel,” Eranos Jahrbuch 19, 1950, pp. 181-246.

Idem, Histoire de la philosophie islamique, Paris, 1964.

F. Dieterici, Die Abhandlungen der Ichwān aṣ-ṣafāʾ in Auswahl zum ersten Mal aus arabischen Handschriften herausgegeben, Leipzig, 1886.

B. Goldstein, “A Treatise on Number Theory from a Tenth-Century Arabic Source,” Centaurus 10, 1964, pp. 129-60.

A. Hamdani, “Abū Ḥayyān al-Tawḥīdī and the Brethren of Purity,” IJMES 9, 1978, pp. 345-53.

Idem, “An Early Fatimid Source on the Time and Authorship of the Rasāʾil Iẖwān al-Ṣafāʾ,” Arabica 26, 1979, pp. 62-75.

H. Hamdani, “Rasāʾil Iḵhwān ạs-Ṣafāʾ in the Literature of the Ismāʿīlī Ṭayyibī Daʿwat,” Der Islam 20, 1932, pp. 281-300.

Y. Marquet, “Ikhwān al-Ṣafāʾ” EI2 III, pp. 1071-76.

Idem, La Philosophie des Iẖwān al-Ṣafāʾ, Algiers, 1975.

L. S. H. Nasr, An Introduction to Islamic Cosmological Doctrines, Cambridge, Mass., 1964, pp. 23-104.

I. R. Netton, Muslim Neoplatonists. An Introduction to the Thought of the Brethren of Purity, London, 1982.

Ḏ. Ṣafā, Eḵwān al-Ṣafāʾ, Tehran, 1330 Š./1951.

S. M. Stern, “The Authorship of the Epistles of the Ekhwān-aṣ-ṣafā,” Islamic Culture 20, 1946, pp. 367-72.

Idem, “New Information about the Authors of the Epistles of the Sincere Brethren,” Islamic Studies 3, 1964, pp. 405-28; repr. in Studies in Early Ismāʿīlism, Jerusalem, 1983, pp. 155-176.

A. L. Ṭebāwī, “Ikhwān as-Ṣafā and their Rasāʾil. A Critical Review of a Century and a Half of Research,” Islamic Quarterly 2, 1955, pp. 28-46.

(Paul E. Walker)

Originally Published: December 15, 1998

Last Updated: December 13, 2011

This article is available in print.
Vol. VIII, Fasc. 3, pp. 294-297