LITT.D., LL.D., F.B.A.


Sir Thomas Adams’s Professor of Arabic and sometime

Lecturer in Persian in the University of Cambridge

There are complete translations of the Mathnawí in Turkish[1], Arabic[2], and Hindustani[3], but only the first two of the six Books of the poem have hitherto been made accessible in their entirety to European readers, though a number of extracts from Books III–VI are translated in E. H. Whinfield’s useful abridgment. While it may seem surprising that a work so celebrated, and one which reflects (however darkly at times) so much of the highest as well as the lowest in the life and thought of the Mohammedan world in the later Middle Ages, should still remain imperfectly known to Western students, I think that this gap in our knowledge can at least be excused. Judged by modern standards, the Mathnawí is a very long poem: it contains almost as many verses as the Iliad and Odyssey together and about twice as many as the Divina Commedia; and these comparisons make it appear shorter than it actually is, since every verse of the Mathnawí has twenty-two syllables, whereas the hexameter may vary from thirteen to seventeen, and the terza rima, like the Spenserian stanza, admits only ten or eleven in each verse, so that the Mathnawí with 25,700 verses is in reality a far more extensive work than the Faerie Queene with 33,500. On the other hand, it is easily surpassed in length by several Persian poems; and the fact that the Sháhnáma has been translated from beginning to end into English, French, and Italian answers the question asked by Georg Rosen—“Who would care to devote a considerable part of his lifetime to translating thirty or forty thousand Persian distichs of unequal poetical worth?” The size of the Mathnawí is not the chief or the worst obstacle by which its translator is confronted. He at once finds himself involved in the fundamental difficulty, from which there is no escape, that if his translation is faithful, it must be to a large extent unintelligible, and that if he tries to make it intelligible throughout he must often substitute for the exact rendering a free and copious paraphrase embodying matter which properly belongs to a commentary, though such a method cannot satisfy any one who wants to understand the text and know what sense or senses it is capable of bearing. Therefore a complete version of the Mathnawí means, for scientific purposes, a faithful translation supplemented by a full commentary; and considering the scarcity of competent Persian scholars in Europe, no one need wonder that the double task has not yet been accomplished. The most important European translations are enumerated in the following list, which shows incidentally that the greater part of the work already done stands to the credit of this country.

  1. Mesnewi oder Doppelverse des Scheich Mewlânâ Dschelâl-ed-dîn Rûmî, aus dem Persischen übertragen von Georg Rosen. (Leipzig, 1849.)

Being written in rhymed verse, this excellent version of about a third of Book 1 (vv. 1– 1371 in my edition) does not preserve the literal form of the original, but as a rule the meaning is given correctly even where misunderstanding would have been pardonable, while the explanatory notes keep the reader in touch with the mystical background of the poem. The translator has left out a good deal—and in verse-translations of Oriental poetry this is a merit rather than a fault. His book, which was reprinted in 1913 with an introduction by his son, Dr F. Rosen, should help to quicken the growing interest of Germany in Persian literature.

  1. The Mesnevi of Mevlānā Jelālu”d-dīn Muhammed er-Rūmī. Book the First, together with some account of the life and acts of the Author, of his ancestors, and of his descendants, illustratedby a selection of characteristic anecdotes, as collected by their historian, Mevlānā Shemsu”d dīn   Ahmed el-Eflākī el-“Ārifī. Translated and the poetry versified by James W. Redhouse.  (London, 1881.)

Sir James Redhouse’s translation of Book I is much less accurate than Rosen’s. Its peculiarities cause us to speculate why this eminent Turkish scholar, who was not quite at home in Persian mysticism, should have embarked upon a task so formidable; or how, with the sagacity to perceive and the candour to confess his lack of skill in versifying, he allowed himself to be misled by the idea that any kind of verse is superior to prose as a medium for the translation of poetry. The excerpts from Aflákí’s Manáqibu “l-“Árifín, though legendary in character, supply valuable information concerning the poet and the circle of Súfís in which he lived.

  1. Masnaví-i Ma”naví, the Spiritual Couplets of Mauláná Jalálu”d-dín Muhammad Rúmí, translated and abridged by E. H. Whinfield. (London, 1887; 2nd ed., 1898.)

All students of the Mathnawí owe gratitude to Whinfield, who was the first to analyse its contents and illustrate their rich quality by his prose translation of selected passages from the six Books, amounting to something like 3500 verses altogether. His wide and sympathetic knowledge of Oriental mysticism, already exhibited in the notes to his edition and translation of the Gulshan-i Ráz (1880), makes him an admirable guide through the mazes of the Mathnawí, and in general his work deserves the high esteem which it enjoys. I do not wish to criticise it in detail and will only remark that the apparent simplicity of the Persian language is a snare for translators:

  1. The Masnavī by Jalālu”d-dīn Rūmī, Book II translated for the first time from the Persian intoprose, with a Commentary, by C. E. Wilson. (London, 1910.)

This is “a plain literal prose translation,” based on sound principles and carefully executed. Comparing it with my own version of the Second Book, I found that as similar methods produce similar results the two versions often agreed almost word for word, and that where they differed, the point at issue was usually one for discussion rather than correction. My obligations to Professor Wilson are not confined to the turns of phrase which I have borrowed from him now and then: every translator, and particularly the translator of such a poem as the Mathnawí, must feel the advantage of being able to consult the work of a trustworthy predecessor who has gone step by step over the same ground.
The present translation, in which the numeration of the verses corresponds with that of the text of my edition, is intended primarily as an aid to students of Persian; it is therefore as exact and faithful as I can make it, but it does not attempt to convey the inner as distinguished from the outer meaning: that is to say, it gives the literal sense of the words translated without explaining either their metaphorical or their mystical sense[4]. While these latter senses have sometimes been indicated by words in brackets[5], I have on the whole adhered to the principle that translation is one thing, interpretation another, and that correct interpretation depends on correct translation, just as the most fertile source of misinterpretation is inability or neglect to translate correctly. It follows that a translation thus limited in scope will contain a great number of passages which do not explain themselves and cannot be fully understood without a commentary. I should have preferred, as a matter of practical convenience, to include the commentary in the same volume as the translation, but on the other hand I saw grave objections to annotating part of the poem before the whole had been studied and translated. “The Mathnawí,” it has been said, “is easier than easy to the ignorant, but harder than hard to the wise”[6]; and I confess that for me there are still many difficulties, which may perhaps be removed by further study of the poem itself, of works historically connected with it, and of relevant Persian and Arabic literature. The Oriental commentaries, with all their shortcomings, give much help. Amongst those used in preparing this translation I have profited most by the Fátihu “l-abyát (Turkish) of Ismá”íl Anqiraví and the Sharh-i Mathnawí-yi Mawlánáyi Rúmí (Persian) of Walí Muhammad Akbarábádí; I have also consulted the Mukáshafáti Radawí (Persian) of Muhammad Ridá, the Sharh-i Mathnawí (Persian) of Muhammad “Abdu “l-“Alí, who is better known by his title of Bahru “l-“Ulúm, al-Manhaj al-qawí (Arabic) of Yúsuf b. Ahmad al-Mawlawí, and for Book I the Sharh-i Mathnawí-yi Sharíf (Turkish) of “Ábidín Páshá.
As stated in the Introduction to the first volume, no finality is claimed for this edition. Where the text is uncertain, the translation can only be provisional; but even where we feel confidence in the text, cases occur in which every translator of the Mathnawí can but offer the rendering that seems to him possible or probable, and take comfort in the reflection that est quadam prodire tenus si non datur ultra. Some passages, I believe, will always remain mysterious, since the key to them has been lost: one knows that words uttered by a great spiritual teacher may be almost meaningless outside the group of his intimate friends and disciples, or may become so by lapse of time. The loose and rambling structure of the poem leads to other perplexities. When our author gives no sign whether he is speaking in his own person or by the voice of one of his innumerable puppets—celestial, infernal, human, or animal—who talk just like himself; when he mingles his comments with their discourse and glides imperceptibly from the narrative into the exposition; when he leaves us in doubt as to whom he is addressing or what he is describing—the translator is driven to conjecture, and on occasion must leap in the dark. Hence a translation of the Mathnawí, however careful it may be, is necessarily tentative in some respects and capable of being improved, though the process takes time. The corrections which I look forward to publishing at a later stage, when the commentary on this volume appears, are likely to be fewer, but also more important, than those contained in the long list of textual corrections (vol. I, pp. 21–28), three-fourths of which any reader could have made for himself.
Although the question of literary form does not enter very largely into a version so literal as this, I have attempted to preserve the idiomatic flavour of the original[7]—which can be more firmly caught and retained in a prose translation—and also its variety of style, ranging from a plain semi-colloquial manner of expression to a noble and elevated diction like that employed by the author in his mystical odes. On certain topics he is too outspoken for our taste and many pages are disfigured by anecdotes worthy of an Apuleius or Petronius but scarcely fit to be translated into the language of these writers. To omit them, however, would defeat the object I have in view, namely, to provide a complete version of the work which not withstanding the author’s passion for self-effacement, reveals the breadth and depth of his genius most adequately. It is important, for our comprehension of him, to know that he could tell ribald stories in the easy tone of a man of the world, and that the contrast often drawn between him and Sa”dí takes no account of some marked features which the authors of the Mathnawí and the Gulistán possess in common.
This is a translation for students of the text, but I venture to hope that it may attract others neither acquainted with Persian nor specially concerned with Súfism. To those interested in the history of religion, morals, and culture, in fables and folklore, in divinity, philosophy, medicine, astrology and other branches of mediaeval learning, in Eastern poetry and life and manners and human nature, the Mathnawí should not be a sealed book, even if it cannot always be an open one.
The prose headings inserted at short intervals throughout the poem, transliterated words with the exception of proper names, and all direct quotations from the QurӇn except such as occur in the headings are printed in italics. A few foot-notes have been added, some of them for the benefit of the general reader.



December 1925

It would not be fitting that this volume, the first to appear in the “E. J. W. Gibb Memorial” Series since the death of Professor E. G. Browne, should leave my hands without giving some expression to the great sorrow felt by the Trustees at the loss of the Scholar who presided over the foundation of the Trust, took the chief part in organizing and administering it, and for more than twenty years so fully enjoyed the confidence and affection of his surviving colleagues. Amongst the works published by the Trustees or now in course of publication there are few that were not inspired, suggested, supervised, or in some way influenced by him; and his colleagues can never forget how much they have owed to his energy, enthusiasm, and experience—energy sustained by patience, enthusiasm controlled by judgement, experience as ready to acknowledge any mistake of his own as to excuse it in others. The Trustees hope eventually to include in the Series yet another book from his pen, the Catalogue, which he has left almost complete, of his fine collection of Arabic, Persian, and Turkish manuscripts.

R. A. N.

 [1] In prose, by Ismá”il Anqiravi in his commentary entitled Fátihu “l-abyát(Búláq, A.H. 1251 and Constantinople, A.H. 1289). A Turkish verse-translation by Sulaymán Nahifi accompanies the Persian text in the Búláq edition of the Mathnawi (A.H. 1268). For Nahifi (ob. A.H. 1151) see E. J. Gibb, History of Ottoman Poetry, vol. IV, pp. 78-85, where an account is given of the circumstances which led him to compose his version of the Mathnawi. Gibb”s description of it as “a literal and line for line rendering” seems to me a require much qualification.
[2] In prose, by Yúsuf b. Ahmad al-Mawlawi in his commentary entitled al-Manhaj al-qawi (Cairo, A.H. 1289).
[3] I do not think that a complete prose translation in Hindustani has yet appeared, but there is one in verse, entitled Piráhan-ı Yúsufi, by Muhammad Yúsuf “Ali Sháh (Lucknow, 1889; Cawnpore, 1897).
[4] Some day I hope to try in a volume of selected passages whether a translator of the Mathnawi may not merit the praise which Jerome bestowed on Hilary: “quasi captives sensus in suam linguam victoris jure transposuit.”
[5]  Frequently too the terseness of the original demands expansion in order to bring out even the literal sense. The brackets in this version mark off what belongs to a strict rendering of the original text from what has been added for the purpose of explanation. I have not, however, been so pedantic as always to indicate the insertion of certain auxiliary parts of speech which an English translator would naturally use, though they are omitted in the Persian text.

[7]  It may be said that this aim is inconsistent with the translator”s duty to write his own language idiomatically. That is true, and no compromise will unite the contraries, but I have done my best to combine them.
Download PDF