BEḤĀR AL-ANWĀR (Oceans of light) by Mollā Moḥammad-Bāqer b. Moḥammad-Taqī Majlesī (d. 1110 or 1111/1699 or 1700), an encyclopedic compilation in Arabic of Imamite traditions (references are to the lithograph edition unless otherwise indicated). This project, which occupied Majlesī for most of his adult life, grew from a modest work known as Fehrest beḥār al-anwār or Fehres moṣannafāt al-asḥāb, which is essen­tially an early version of the Beḥār; this work was published in 1070/1659, when Majlesī was thirty-two. It consists of a table of contents (corresponding roughly to that of the final version) and traditions taken from ten Imamite works, six of them by Ebn Bābūya (editor’s note to the printed edition of the Beḥār CV, pp. 31f.; cf. al-Ḏarīʿa XVI, p. 377, no. 1752). A few years later Majlesī embarked on a great expansion, involving practically the entire corpus of Imamite ḥadīṯ. In his introduction to the Beḥār, the author emphasizes the crucial religious significance of this corpus. He notes that much of it is little known, having been largely forgotten, either as a result of suppression by Sunni rulers or because Imamite scholars preferred later compilations to the early works. Majlesī (I, p. 5 = I, p. 6 of the printed ed.) describes his aim as providing the believers with the means to “humble the despicable heretics” (i.e., the Sunnis). This is to be achieved by collecting the available literature—including texts thought to be lost—and arranging its contents by subject matter.

Majlesī spared no effort in his attempt to realize this aim. He was able to draw on the help of such luminaries as Neʿmat-Allāh Jazāʾerī (d. 1112/1701) and Mīrzā ʿAbd-Allāh b. ʿĪsā Efendī (d. between 1130/1717 and 1140/1727), both of whom were among his pupils (Karl­-Heinz Pampus, Die theologische Enzyklopädie Biḥār al-­Anwār des Muḥammad Bāqir al-Mağlisī [1037-1110 A.H. = 1627-1699 A.D.], Bonn, 1970, p. 138); on oc­casion he also obtained financial assistance from the Safavid court, as when attempting (unsuccessfully) to obtain a manuscript of Ebn Bābūya’s Madīnat al-ʿelm from the Yemen (Ṭabresī/Ṭabarsī, al-Fayż al-qodsī, prefixed to Beḥār I, p. 6 [my pagination] = CV, p. 34 in the printed edition; Aʿyān al-šīʿa XLIV, pp. 97f., 100; al-Ḏarīʿa XX, p. 252). He was a good organizer: once he had decided to incorporate a particular passage, he often had scribes do the actual copying, though he also wrote a considerable portion of the text in his own hand (see editor’s note to the printed edition, CV, pp. 30f.).

Work on the Beḥār did not proceed quite according to plan. Vols. 1-5, 9-13, 22 were completed between 1 Rabīʿ II 1077/1 October 1666 and 27 Rajab 1081/10 December 1670; vols. 6-8 followed on 20 Ramażān 1084/29 December 1673, in Ḏu’l-ḥejja, 1086/February-March, 1676, and in Ḏu’l-ḥejja, 1091/January, 1681, respectively (the information in Pampus, p. 139, is to be corrected accordingly). Next came the Ketāb al-ṭahāra (completed on 14 Ṣafar 1094/12 February 1683) and the Ketāb al-ṣalāt (completed on 21 Šaʿbān 1097/13 July 1686), which together comprise vol. 18; they were followed (in Jomādā II, 1104/February, 1693) by vol. 14. While working on vol. 15, Majlesī realized that the chapters on companionship (ʿešra), originally intended to form part of it, were becoming longer than expected; he therefore announced (at the beginning of vol. 15) his intention to devote a separate volume to them. This volume was subsequently given the number 16. Since, however, the original plan had called for this number to be allotted to the Ketāb al-ādāb wa’l-sonan, there ensued two volumes, each bearing the number 16. The Beḥār is therefore alternatively referred to as consisting of 25 or 26 volumes. The problem was later resolved by incorporating both the Ketāb al-ʿešra and the Ketāb al-ādāb wa’l-sonan within one volume, labeled vol. 16.

Majlesī did not live to see the Beḥār through to completion; at his death he left only preliminary drafts of the later volumes (the two volumes known as 16, as well as vols. 17, 19-21, 23-25; there is some doubt as to whether he completed vol. 15). Particularly noticeable is the absence from these drafts of the explicatory comments (bayān, tawżīḥ) which Majlesī (following the practice of his teacher Mollā Moḥsen Fayż Kāšānī in his Wāfī) had provided for many traditions; instead, there are occasional brief notes, probably written by Majlesī’s pupil and grandson Moḥammad-Ḥosayn Ḵātūnābādī (d. 1151/1739), whose Ḥadāʾeq al-moqarrabīn is one of the earliest sources on the history of the Beḥār. Majlesī’s pupil ʿAbd-Allāh Efendī made fair copies of these drafts, but for some reason refused to make either these copies or the originals available to other scholars. After ʿAbd-Allāh Efendī’s death, Naṣr-Allāh b. Ḥosayn Ḥāʾerī (d. after 1166/1752-53, cf. al-Ḏarīʿa XI, P. 281) was able to copy the fair copies, but even so, not all of this material passed into general circulation. Thus Yūsof Baḥrānī (d. 1186/1772) mentions having seen only three of the final ten volumes (Loʾloʾat al-baḥrayn, Najaf, 1388/1966, p. 57), and even so late an author as Moḥammad-Bāqer Ḵᵛānsārī (d. 1313/1895) says in his Rawżāt (completed in Ḏu’l-ḥejja, 1286/March, 1870) that he has only come across six of them (II, p. 80). Kantūrī (d. 1286/1870) saw neither vol. 14 nor the first part of vol. 19 (Kašf, pp. 80f.). In contrast, the earlier volumes of the Beḥār gained relatively wide recognition and were repeatedly copied by Majlesī’s pupils and admirers, who often added or omitted material. One of the first scholars to take note of the Beḥār was Majlesī’s contemporary Moḥammad b. Ḥasan Ḥorr ʿĀmelī (d. 1104/1693) (the two gave each other ejāzas to transmit traditions which each of them had collected, Rawżāt II, p. 84, VII, p. 103): he cites from vol. 13 in his Eṯbāt al-­hodāt, completed in 1096/1685 (ed. Hāšem Rasūlī ­Maḥallātī, Qom, 1378-79/1958-59, I, p. 60, VII, pp. 162-­73, 366-77), and notes in his Amal al-āmel (completed the following year) that the Beḥār contains 25 volumes (Najaf, 1385/1965, II, p. 248)—an apparent reference to the projected number of volumes as announced by Majlesī.

Even in the late 13th/19th century, not all scholars had a correct idea of the Beḥār: a shaikh from Qom whom E. G. Browne met in June, 1888, not only believed that this work (to which he referred as one of the two “most authentic and esteemed collections of Shiʿite traditions”) was “in 15 or 16 volumes,” but also misidentified its author as the ʿAllāma Ḥellī (see Browne’s A Year Amongst the Persians [1887-1888]3, London, 1950, p. 487). The likelihood of such errors recurring gradually diminished with the increasing availability of the work. Lithographs of individual volumes first appeared at the beginning of the 13th/19th century, most of them in Tabrīz. The first complete edition, dedicated to Nāṣer-al-Dīn Shah and to his son and successor Moẓaffar-al-Dīn, was lithographed in Tehran and Tabrīz between 1303/1885 and 1315/1897­-98. It was edited mainly by Mīrzā Moḥammad-Ḵalīl b. (Moḥammad) Ḥosayn Mūsāwī Eṣfahānī (d. 1315/1897-98) and was published by a wealthy merchant, Ḥājj Moḥammad-Ḥasan Kompānī Eṣfahānī known as Amīn-e Dār al-Żarb, and later, upon his withdrawal in 1314/1896, by his son Ḥājj Moḥammad-Ḥosayn. A useful index entitled Safīnat al-beḥār wa ma­dīnat al-ḥekam wa’l-āṯār was compiled by ʿAbbās b. Moḥammad-Reżā Qomī (d. 1359/1940), who also included in it traditions not found in the Beḥār itself. A supplement (mostadrak) to the Safīna was written by ʿAlī b. Moḥammad b. Esmāʿīl Šāhrūdī. A new edition of the Beḥār was published in Tehran between 1376/1956 and 1394/1974 (later reprinted in Beirut with slight changes). Of its projected 110 volumes, three (54-56) are taken up by an index; a further six (28-34 = most of vol. 8) were not published at the time (see below), although several of these have recently come out. Though this is not a critical edition, it is based on several manuscripts and lithographs and is consider­ably easier to use than its predecessor.

Various volumes of the Beḥār were translated into Persian and some into Urdu; there are also numerous excerpts, abridgements, and supplements (for details see Pampus, pp. 164-70; in the meantime several new Persian translations of individual volumes have appeared). Particularly noteworthy is the use which Majlesī himself made of the Beḥār in his Persian works, some of which are abridged or extracted from volumes of his magnum opus. In his introduction to the Ḥayāt al-­qolūb, for example, Majlesī speaks of the need to make the material in the Beḥār accessible to those with no knowledge of Arabic. And indeed, the three volumes of the Ḥayāt are abridged Persian translations of such material: vol. 1 (completed in mid-Šawwāl, 1085/January, 1675) is taken from Beḥār vol. 5; vol. 2 (completed on 25 Ḏu’l-ḥejja 1087/28 February 1677, see the abridged translation of James A. Merrick entitled The Life and Religion of Mohammed as Contained in the Sheeʿah Tradition of the Hyât-ul-kuloob, Boston, 1850) corresponds to Beḥār vol. 6; and vol. 3 (which remains incomplete) covers part of Beḥār vol. 7. Traditions from Beḥār vols. 6, 9-13 were incorporated by Mailesī in his Jalāʾ al-ʿoyūn; his Meškāt al-anwār (composed, like the Jalāʾ, during the reign of Shah Solaymān) is based on Beḥār vol. 19/1, while Beḥār vol. 19/2 forms the basis for the Zād al-maʿād (completed in Ramażān, 1107/April, 1696).

The importance of the Beḥār for the study of Imamite literature can hardly be exaggerated. It reflects the accumulated knowledge of a millennium of Imamite ḥadīṯ scholarship, and covers most aspects of Imamite religious thought: the concept of knowledge (ʿelm) (vol. 1); tawḥīd and the divine attributes (2); free will and predestination, death and the after-life (3); arguments (eḥtejājāt), mainly of the imams, in defense of Imamite beliefs (4); stories of the prophets (5); biographies of the Prophet and his forebears (6); the special position of the imams (7); the injustices perpetrated against ʿAlī and the ahl al-bayt by the first three caliphs and Moʿāwīa (8); biographies of Fāṭema and the first eleven imams (9-12); the twelfth imam and issues bearing on his occultation (13); cosmology and natural history (14); the concepts of belief and unbelief (15); practical ethics and correct social and religious behavior (16); exhortations (17); the position of the Koran (19/1); the various kinds of supererogatory prayers (doʿāʾ) (19/2-4); the “pillars of Islam” and legal traditions (feqh) relating thereunto (18, 20-21); pilgrimages to the graves of the Prophet, the imams, and other holy men (22); positive law (23-24); the Ketāb al-ejāzāt (25). While most of the works on which Majlesī drew are available today in printed editions, others are still known chiefly through quotations in the Beḥār. Majlesī always identifies the source of each passage in his work, whether it was written by himself or copied from an earlier authority. As a rule, the sources used are quoted accurately; occasional slips (examples of which are given, e.g., in Rawżāt IV, pp. 25, 291f., 338) are due to the haste with which the work was carried out (Aʿyān al-Šīʿa XLIV, p. 97). Although Majlesī aimed at comprehensiveness, there are some texts which he was able to use only partially or not at all, since he only obtained them while the work was already in progress. He intended to incorporate the missing material in a supplement (to be entitled Mostadrak al-­beḥār), but he did not live to write it. The most important of these texts are listed by Ṭabarsī (Fayż apud Beḥār I, pp. 11f. = CV, pp. 60-75 in the printed edition; idem, Mostadrak al-wasāʾel, Tehran, 1382-84/1962-64, III, p. 295; cf. idem, Maʿālem al-ʿebar, appended to Beḥār XVII, p. 250). Furthermore Majlesī, in conform­ity with his stated views, refrained from using Imamite texts with a Sufi or a philosophical bias, thus excluding from the Beḥār a major component of the Imamite literary tradition. He had no qualms, however, about citing Sunni sources, whether for lexicographical or for polemical purposes (see the introduction to the Beḥār I, pp. 10f. = I, pp. 24f. in the printed edition).

Historically, the texts incorporated in the Beḥār may be divided into several major groups. The earliest go back to the pre-ḡayba period (before 260/874). Some are believed by the Imamites to have been composed by the imams themselves; cases in point are the celebrated Nahj al-balāḡa, popularly ascribed to ʿAlī b. Abī Ṭāleb, and al-Ṣaḥīfa al-sajjādīya, containing prayers of the fourth imam ʿAlī Zayn-al-ʿĀbedīn. Others, known as oṣūl, were compiled by disciples of various imams, in partic­ular Imam Jaʿfar al-Ṣādeq; it was thanks to Majlesī that some of these oṣūl were rediscovered and brought back into the mainstream of Imamite literature (E. Kohlberg “al-Uṣūl al-arbaʿumiʾa,” forthcoming in Jerusalem Studies in Arabic and Islam 10). The second group consists of works written during the Lesser Occultation (260-329/974-941) and the Buyid period. Given the central importance in the history of Imamite literature of texts written during those years, it is only natural to find them repeatedly quoted. Particularly prominent are works by Ebn Bābūya (d. 381/991), Shaikh Moḥammad Mofīd (d. 413/1022), ʿAlam-al-Hodā Šarīf Mortażā Rāzī (d. 436/1044), and Shaikh Abū Jaʿfar Moḥammad Ṭūsī (d. 460/1067). The later ʿAbbasid period is represented by works of Ebn Šahrāšūb (d. 588/1192), ʿAlī b. Ṭāwūs (d. 664/1266), and others; while the fourth group, covering the period between the fall of Baghdad (656/1258) and the beginning of the 10th/end of the 15th century, includes compositions by such masters as the ʿAllāma Jamāl-al-Dīn Ḥasan b. Yūsof Ḥellī (d. 726/1325) and al-Šahīd al-­Awwal Šams-al-Dīn Moḥammad b. Makkī (d. 786/1384). Finally, Majlesī also incorporated works written in the Safavid period by (among others) al-­Šahīd al-Ṯānī Zayn-al-Dīn b. ʿAlī (d. 965/1557-58) and Shaikh Bahāʾ-al-Dīn ʿĀmelī (Shaikh Bahāʾī, d. 1030/1621 or 1031/1622).

This enormous body of material comprises a number of different literary forms: commentaries on the Koran, such as the tafsīrs of Forāt b. Ebrāhīm (d. ca. 300/912), ʿAlī b. Ebrāhīm Qomī (d. ca. 307/919), Moḥammad b. Masʿūd ʿAyyāšī (d. 320/932), and the Majmaʿ al-bayān of Fażl b. Ḥasan Ṭabarsī (d. 548/1153); biographies of the imams, such as the Eršād of Shaikh Mofīd and the Eʿlām al-warā of Ṭabarsī; doctrinal and theological works expounding the major tenets of Imamite Shiʿism; legal texts; polemical writings, directed mainly against opponents in the Sunni camp; supererogatory prayers; collections of sayings and anecdotes. Most of this material consists of traditions from the Prophet and the imams, the significance of which was greatly enhanced by the growing influence of the Aḵbārīs (q.v.) in the 11th/17th century; and Majlesī, though not a declared Aḵbārī, was in sympathy with their belief that ḥadīṯ is the repository of all religious knowledge.

Being a compilation, the Beḥār does not present a uniform view on all issues; it contains both moderate and radical traditions. Majlesī was aware of various contradictions and inconsistencies, and on occasion attempted to resolve them. The radical traditions in the Beḥār are concerned with three major issues: (1) The integrity of the Koran (XIX, pp. 11-21 = XCII, pp. 40-77 in the printed edition): certain phrases in the Koran referring to ʿAlī’s rights are said to have been deliberately excised by ʿOṯmān and his accomplices. (2) The status of the Prophet’s companions (ṣaḥāba) (VIII, passim): the great majority of the companions, including in particular the first three caliphs, are portrayed as grave sinners (or even unbelievers) because they usurped ʿAlī’s rights or acquiesced in this usurpation. These two items were offensive to Sunni sensibilities, and at least since the advent of the Saljuqs they had been watered down or suppressed. (Similar considerations may have been behind the decision of the Iranian government of the day not to permit publication in the printed edition of most of the material in vol. 8, cf. Mahdawī, Zendagī-nāma II, p. 272.) These items were also rejected in the middle ages by a number of leading Imamite scholars as incompatible with true Imamite doctrine. Yet with the advent of the Safavids, the often forcible conversion of Sunnis (with which Majlesī was associated), and the bitter Safavid-Ottoman rivalry, anti-Sunni sentiments returned to center stage. The inclusion of such radical traditions was thus in accordance with both the political aims of the rulers and the religious convictions of Majlesī and his circle. (3) The position of the imams (VII, XIX-XIII, passim): they are said to have possessed knowledge of the ḡayb and to have performed miracles. This in­sistence on the imams’ superhuman nature, with the attendant emphasis on the beneficial effects of pilgrim­age to their graves (XXII, passim), was consistent with Majlesī’s endeavors to popularize Shiʿism among the masses.

See also majlesǰ, moḥammad-bāqer.



See also Eʿjāz Ḥosayn Kantūrī, Kašf al-ḥojob wa’l-astār ʿan aḥwāl al-kotob wa’l-asfār, Calcutta, 1330/1912, pp. 76-81.

Moḥammad-Bāqer Ḵᵛānsārī, Rawżāt al-jannāt, ed. ʿA. Esmāʿīlīān, Qom, 1390-92/1970-72, II, pp. 79-84.

Ḥājj Mīrzā Ḥosayn Nūrī Ṭabresī, Fayż al-qodsī (prefixed to Beḥār I; publ. in vol. CV of the printed edition, pp. 2-165).

Browne, Lit. Hist. Persia IV, pp. 359, 409, 417.

Brockelmann, GAL, S. II, p. 573.

ʿA. Qomī, Fawāʾed al-rażawīya, Tehran, 1367/1947-48, p. 412.

Aʿyān al-Šīʿa XLIV, pp. 96-101.

ʿA. Šarīʿatī, Tašayyoʿ-e ʿalawī wa tašayyoʿ-e ṣafawī, Tehran, 1350 Š./1971, pp. 189-203 (includes an attack on Majlesī for incorporating in the Beḥār traditions which ostensibly portray the imams as acquiescing in the rule of the Omayyad and ʿAbbasid caliphs), Moṣleḥ-al-Dīn Mahdawī, Zendagī-nāma-ye ʿAllāma Majlesī II, Isfahan 1401/1980-81, pp. 246-316, 428-30.

ʿAlī Davānī’s introduction to his Mahdī-e mawʿūd (a translation of Beḥār, vol. 13), 2nd ed., 1344 Š./1965, pp. 52-73.

Abdul-Hadi Hairi “Madjlisī,” EI2.

Search terms:

 بحارالانوار behar al anwar bahar al anwaar behar al anvar

(Etan Kohlberg)

Originally Published: December 15, 1989

Last Updated: December 15, 1989

This article is available in print.
Vol. IV, Fasc. 1, pp. 90-93