The Musician in the Garden: on Translating Ferdowsi's Shahnameh

Dick Davis

The Ohio State University
Norûz Lecture March 24, 2004


During an idyllic episode in the reign of the Sasanian monarch Khosrow Parviz, the musician Barbad hides himself among the trees of a royal garden, and from his unseen vantage point sings songs of Persia’s ancient glory to his king.

At the moment when the yellow sun set and purple night came on Barbad took his lute and sang the heroic song he had prepared. Hidden in the tree, he sang his beautiful lay, the one we now call “Dad-afarid”, and the king was astonished at the sweetness of his voice. The whole company was amazed, and everyone expressed a different opinion as to what was happening. The king ordered the company to search the area thoroughly, and they looked high and low, but came back empty handed . . .

A beautiful serving girl brought a goblet, and as the king took it from her Barbad suddenly struck up another song, the one called “The Heroes’ Battle”. The wise singer sang and Khosrow listened, drinking his wine as the song progressed. Then he ordered that the singer be found, and that if need be the garden be turned upside down in the search. They searched everywhere in the garden, taking flaming torches beneath the trees, but they saw nothing but willows and cypresses, and pheasants strutting among the flowers. The king asked for another goblet of wine, and leaned his head forward to listen. Again a song began, accompanied by the lute’s sound; it was the one that is called “Green on Green” nowadays, and which is used for magical incantations.


The hidden Barbad sings of justice, warfare, and magic; Khosrow Parviz is charmed and delighted by the songs he hears, but he has no idea of the identity or whereabouts of the musician whose voice conveys them to him. A translator should, ideally, aspire to be like Barbad; he should use whatever talent he has to pass on the ancient stories to which he is privileged to have access, and he should do this with as much skill and sweetness as he can muster, but he should remain hidden, concealed by the foliage of the historic garden whose beauties he celebrates.


A translator passes on what is not his own, and this is why he must remain in the background, concealed among the leaves of a garden at sunset, but his task is such that it cannot be one of simple, unmediated transmission. In order to speak at all, the translator is forced to add his own voice to the material, as Barbad did, and in doing so he must make many choices, some small and some large, but all affecting the presentation and reception of the artifact with which he has been entrusted. In this talk I will discuss some of the choices I made while engaged on translating the Shahnameh.


The first choice a translator of a narrative poem must make is that of form – should he translate into verse or into prose? My instinctive preference has always been that verse should be translated as verse. For a narrative poem this means, practically, that the translator should choose between blank verse and heroic couplets, the two traditional narrative verse forms of English. Some years ago I translated one episode of the Shahnameh, the Legend of Seyavash, into blank verse, and while I do not disown my version I now regret that I used blank verse rather than couplets, for the simple reason that Ferdowsi’s Shahnameh is in couplets, and so, if a translator wishes to bring over any of the formal qualities of the original, couplets are the obvious choice. But to translate the whole of the Shahnameh into couplets would be a massive and absurdly quixotic undertaking. In the early 1980’s Afkham Darbandi and I translated Attar’s Manteq al-Tayr into heroic couplets: its approximately 4,500 lines took us two years to translate, and resulted in a book of 234 pages. The Shahnameh is ten times this length: could I seriously consider giving twenty years to produce a manuscript of almost two and half thousand pages, which no sane publisher would look at? And supposing I could find a publisher, who would read it? Even in the heyday of narrative verse in English, the 16th and 17th centuries, no English poem ever ran to over two thousand pages of couplets or, come to that, of my fall-back choice, blank verse, either; and it is certain that such a poem would find virtually no readers today.


It seemed then that I had no choice but to resort to prose, to reduce the poetic glories of medieval Persian to mundane, prosy modern English. And then I remembered the Naqqali tradition of Iran. The Shahnameh has always had at least two existences within Persian culture – as a text copied at court and revered by the literati, and as a set of stories endlessly reproduced and elaborated by itinerant naqqals in oral performances. These performances are largely in prose, but the high points of the text are rendered as verse. As Kumiko Yamamoto has put it in her fine book The Oral Background of Persian Epics (Leiden, 2003) “prose is used to tell a story, and verse to mark the internal divisions of a performance . . . In terms of narrative structure too, verses which appear intermittently . . . are used, for example, to enhance dramatic effects, to express the internal feelings of characters, or to sum up the story. Hence, verse functions as an attention-getter in the narrative, introducing different rhythms into the prose narration” (p.28). Sometimes these passages in verse are Ferdowsi’s unchanged, sometimes they have been enhanced and polished by generations of performers and are quite far from anything the bard of Tus might have recognized or cared to own. If therefore one wrote down a naqqal’s performance, sticking to his formal choices, one would finish up with a prosimetrum, a text which is largely prose but which contains passages of verse at significant moments of the narrative. The prosimetrum is a common medieval Persian form – Sa’di’s Golestan is perhaps the most famous example – but it was also a not uncommon medieval European form; perhaps the best known European examples are the De Consolatione Philosophae by Boethius and La Vita Nuova by Dante. To translate a medieval Persian poem into what had been a medieval Persian and European form seemed to be an undertaking that would not wholly traduce the formal qualities of the original.


There was a further reason for my choosing the form favored by the naqqals. The naqqali diffusion of the narratives of the Shahnameh has ensured, and been the most obvious manifestation of, these narratives’ popular life within the culture. My aim in translating the Shahnameh was not to produce a text for scholars, but to make it available to a wide non-specialist audience. I hesitate to say a popular audience: perhaps no medieval literary artifact, from any culture, can have a truly popular existence now. We prefer our medievalism to be kitschy and ersatz; The Lord of the Rings rather than Beowulf, Camelot rather than Malory. Nevertheless there is still a world of readers, especially relatively young readers, who are not scholars, who might assay Beowulf or Malory, and it was them I aimed to reach with my translation. I translated not for scholars, who after all have access to the original text, now in relatively good editions, but for that radically endangered species, the general reader. The naqqal’s choice of form, the form of the popular diffusion of the stories of the Shahnameh in Persian, thus seemed all the more appropriate.


But one cannot pretend that there is no loss in choosing this form. A prosimetrum makes an obvious distinction between passages in prose and those in verse, and the implication is certainly that the verse passages are the more interesting. A work which is all in verse is like a 19th century through composed opera, with every word of the text set to an orchestral accompaniment, and so given approximately equal weight. A prosimetrum is like an 18th century opera, with some passages (the arias, corresponding to the verse passages of the prosimetrum) supported by the orchestra, and others (the secco recitatives coming between the arias, corresponding to the prose) supported only by a few chords here and there on the harpsichord. To render a work in one form as a work in the other form is to make distinctions that are not there in the original. Specifically, the translator must decide which parts deserve verse and which parts can make do with prose.


Sometimes the choices were not difficult to make. For example, in stories that contained a significant amount of pointed dialogue, I tended to translate the narrative itself as prose, and the characters’ speeches as verse. Here is an example from the reign of Bahram Gur. Bahram is traveling incognito, and he comes on a miserly house-owner whom he asks for shelter. Because of their pithy, aggressive, epigrammatic quality, I rendered the house-owner’s speeches as verse, and the rest of the tale as prose:

Bahram’s limbs were weak with fatigue; he dismounted and looked about the house. The floors were covered in sheep droppings, but it was a fine building, large and spacious. Bahram said, “You seem a hospitable man; bring me something to rest on.”

The man replied, “Why are you mocking me? There are no carpets here, as you can see, And there’s no food here either – you had best Look for another place to eat and rest.”

Bahram said, “At least bring me a cushion, so that I can sit for a while.”

The man replied, “A cushion! That’s absurd, You might as well expect milk from a bird!”

“Well,” Bahram said, “just bring me some warm milk then; and, if you can find it, some bread that’s not stale.”

“Pretend I gave it you,” the man replied, “And that you ate it and felt satisfied: If there was bread to eat, then I assure you You’d see a healthier host stood here before you”.

Bahram asked, “If you have no sheep, how is it your house is full of their droppings?”

“It’s dark already, look!” the owner said, “This talk of yours has muddled up my head. Go choose a house where they can entertain you And where the owner will be happy to detain you: Why stay with some unhappy wretch who grieves For his bad luck, and has to sleep on leaves? You have a gold sword, and gold stirrups too, You don’t want thieves to come and frighten you. A house like this attracts such dangerous men, And lions might well choose it as their den.”

Bahram said, “If thieves take my sword, I won’t hold you responsible”.

The man replied, “Enough of all this chatter; No one stays in this house; that ends the matter!”

But Bahram persisted, “Why should a wise old man like you be so troubled by my presence? But I think you will have the magnanimity at least to give me a drink of cold water.”

He said, “Go back two bow shots and you’ll see A pond – drink all you want. Stop bothering me, It’s obvious I’m old, worn out, and poor, Too weak to earn my living anymore.”

Bahram replied, “If you’ve no nobility about you, at least don’t quarrel with people. What’s your name?”

He said, “My name is Farshidvard, and I’ve No clothes or food to keep myself alive.”

Bahram asked, “ Why don’t you make some effort to earn your daily bread?”

He said, “I pray for my release; I pray You’ll leave my ruin and be on your way. What’s brought you to this empty house? It’s clear There’s never any wealth or welcome here!”

And as soon as he’d finished talking he began to weep and wail so loudly that the king fled from the noise. Laughing at the old man, he left the town, and his entourage rejoined him.


In a similar story, Bahram stays at the house of a jeweler who has a daughter who is a musician, and here I translated the main narrative as prose, and the girl’s songs as verse. In these cases, formal distinctions – between narrative and speech, between speaking and song – are implied in the original and I used these hints for my own formal distinction between prose and verse.


But sometimes no such distinction is implied in the original, and the decision as to when to use verse is necessarily less clear-cut. A particular problem for the translator of the Shahnameh into English is inherent in the subject matter of the poem. It is easiest for a translator of verse if the target language has a traditional rhetoric that deals with the subject matter of the work he is translating: this rhetoric will then lie to hand for him, he will have a vocabulary and set of conventions available for the relatively easy transfer of material. The Shahnameh is usually characterized as an epic, and epics are about war. Given Britain’s very bellicose history, it may seem strange that there is no traditional verse rhetoric for the depiction of warfare in English, but this is the case. There are no true epics in English: the closest English can claim to an epic is Beowulf, which is in Anglo-Saxon, not English, and which is an epic fragment rather than a full blown example of the genre. In English itself the epic poem par excellence is Paradise Lost, which is not a secular epic, but a religious work, albeit one containing scenes of angelic warfare imitated from Homer and Virgil. Many years ago I reviewed an anthology called The Oxford Book of War Poetry, and I pointed out that virtually all the poems in it were not actually war poems at all, but anti-war poems. Not only does English not have a rhetoric for the depiction of warfare, the contemporary sensibility finds the celebration of warfare, as we traditionally find it in Homer, Virgil, the Icelandic Sagas, and Ferdowsi, distinctly inimical. English itself, as a poetic language, is not hospitable to a rhetoric of war, and, just as significantly for a translator, neither are contemporary educated Western mores, which look on war as a disaster, and at best a necessary evil. We are suspicious of rhetoric that celebrates warfare, and for us such rhetoric is hard to articulate convincingly. When we are faced with it in poetry we otherwise esteem, we don’t want to take it at face value, and instead tend to look for evidence that the poet is using the rhetoric subversively; that is, we want permission to believe that he shares our prejudices.


I’m aware of this, and as a translator of Ferdowsi I deplore it, but being aware of it and deploring it doesn’t stop me from being a child of my time. I find it hard to present warfare in exhilarating verse, and I must admit that I rarely tried to do so in my versions of the Shahnameh’s narratives. There is quite a lot of spirited fighting in the Shahnameh, and I rendered very little of it in verse. Given this, you may wonder why I chose to translate the Shahnameh at all: the 17th century poet Wentworth Dillon, in his Essay on Translated Verse, advises the would-be translator to “Choose an author as you choose a friend”, that is, to translate someone with whose work, and sensibility as it is visible in the work, you feel a deep sympathy. If I don’t feel comfortable translating verse about warfare, what am I doing translating Ferdowsi? The answer is, that although of course verse that actually celebrates war is present in the Shahnameh there is, proportionally, relatively little of it, and what there is tends to be I believe fairly conventional; that is, it does not seem to bear the stamp of personal commitment and conviction on the part of the author. The passages celebrating warfare are balanced, and for me far outweighed in their intensity and nobility, by passages on related themes that derive from warfare – elegy, the perception of loss, the tragic understanding of the inevitability of human conflict, the repeated evocation of past glories, the ubi sunt / ou sont les neiges d’antan theme dear to European and Persian poets alike. For these topics there is a rhetoric in English, and in these passages I do indeed feel a kinship with Ferdowsi’s sensibility, the kind of kinship that Wentworth Dillon recommended, and it is these passages that I chose, whenever I could, to translate as verse.


I said at the beginning of this talk that Ferdowsi’s text is approximately ten times as long as Attar’s Manteq al-Tayr, and yet my translation of the Shahnameh is not nearly ten times as long as Afkham Darbandi’s and my translation of Attar’s poem. Passages have been omitted, and others are presented in summary form. The decisions as to which episodes to pass over and which to foreground were made according to similar criteria to those which helped me to decide whether a given passage should be rendered in verse or prose. In general, I did not omit passages within a given episode, and when I did it was usually because the poem had become highly repetitious at this point. For example, in the later reigns there is an enormous amount of ethical advice handed out by kings either at their coronations or on their death-beds. Much of this advice is extremely repetitious, and much of it I omitted. This of course changes the structure of Ferdowsi’s poem, and ideally a reader should be aware of this. But our appetite for moral sententiae is considerable smaller than that of a medieval audience, and I did not feel I could try the patience of the general reader – who I again emphasize is my intended audience – too high. Conversely, in what I take to be the greatest stories of the poem, for example those of Seyavash and Esfandyar, virtually nothing is omitted. In particular I was scrupulous in translating everything that related to the interiority of the characters and the ethical dilemmas they face, simply because these seem to me to be among the most distinctive and aesthetically admirable sections of the poem.


And as you might suspect, given my previous remarks, I did omit some passages that deal primarily with warfare. The most substantial of these is the episode of the Twelve Champions, the Davazdah Rokh, which occurs during Kay Khosrow’s war against Turan. I believe it was Shamlu who said that this episode is the heart of the poem, which only goes to show that two devoted readers of the Shahnameh can, in all good faith, hold utterly differing views of it. The episode is highly repetitious (this is of course deliberate, but that does not make the repetitions any more palatable for a modern reader), and it also strikes me as otiose in its descriptions, and embarrassingly ethnocentric in its triumphalism. It is an extreme example of the kind of episode that can be found in ethnocentric epics the world over: but no other epic that I am aware of has an equivalent to the story of Seyavash, or to the story of Esfandyar, and since I had to make choices as to what could stay and what could go I felt I had much rather lose the Davazdah Rokh than either Seyavash or Esfandyar


A moment ago I used the phrase “ethnocentric epics”, and you may well feel that the phrase is redundant, that all epics are ipso facto ethnocentric. There is some truth to this, but as I have pointed out in a number of articles this is only partially true of the Shahnameh. Virtually all the major heroes of the legendary section have non-Persian mothers: the non-Persian bestows life in the poem’s legendary narratives, and as these characters often die at the hands or instigation of their Persian kin we can also say that, just as the non-Persian gives life to the poem’s heroes, so the Persian takes it from them. This is hardly an ethnocentric concept. Similarly, epic is traditionally associated with patriarchy and the celebration of masculinity, but I have elsewhere drawn attention to the fact that a remarkable number of women in the Shahnameh, especially in its legendary sections, are celebrated for their determination to shape their own lives. Interiority, and the celebration of the foreign and the feminine, are hardly typical epic concerns, and in my opinion the Shahnameh’s most interesting tales are precisely those that do not correspond to our stereotypes of the nature of epic narrative. Such complications of epic structure seem to me to be essential to the greatness of the Shahnameh, and in my translation I have deliberately lavished care upon the moments when Ferdowsi foregrounds them, and this has meant that I have had to relegate the more stereotypically epic moments, such as the episode of the Davazdah Rokh, to the shadows.


But the fact that the poem is not as long in my version as it is in the original is not a negligible one. There are some works, and I believe that the Shahnameh is one of them, for which their immense length is an intrinsic part of their identity; The Tale of Genjithe Mahabharata, Proust’s A la recherche du temps perdu, could not be short and remain what they are. But our paradigm for poetry tends to be the lyric, that is the relatively brief expression of relatively focused feeling, and epic verse, as I have already indicated, is something outside of our usual literary experience and expectation. In his great essay on translation, Schleiermacher says that the best translators take the reader to the author, they do not force the author to come to the reader. Because of market forces, and my perception of how I might most usefully spend my time, I have, in the matter of length at least, forced Ferdowsi to come to us, instead of insisting that we go to him. The reader should be aware that something is inevitably sacrificed, not least the sense of the huge meandering structure of the poem, which adds immeasurably to the power of many episodes. The tragic consequences of Rostam’s dual with Esfandiyar for example owe much of their strength to the reader’s mind being steeped in Rostam’s innumerable battles on behalf of the Persian throne; the closing pages of the poem have a greatness and finality that draws substantially on the reader’s sense of the multifariousness of what has gone before, and this is unavoidably diminished by summary and compression. A summary translation of The Shahnameh, and especially one that is largely in prose, must inevitably be a little like transcribing Wagner’s Ring for piano and then cutting it to an hour’s length. Much of the transcription may be brilliant, and the choice of cuts felicitous, but one cannot pretend that nothing will be lost.


All this may have left some of you with the feeling that I have written my own Shahnameh, and that what passes under this name in my English version is not Ferdowsi’s work at all. This is not the case. Because, for mundane and practical reasons, I have omitted certain passages and summarized others, because I have chosen to translate some passages as verse and others as prose, this does not at all mean that I have written my own poem. Everything in my version is in Ferdowsi, nothing has been added. My modifications are slight, and are only of emphasis, not of substance. Nevertheless they are there, and it is best to acknowledge them. In truth, no translator can reproduce a work from another language without adding his own emphases and lending his own voice to the artifact in question, and it is better for him to be aware of this, and so consciously compensate for it where possible, than to forge ahead thinking he is somehow absent from the work he is attempting to transmit. I said at the opening of this talk that, ideally, a translator should stay hidden like Barbad concealed among the leaves; but for Barbad to sing at all, for the stories to be communicated, he must employ his own voice.


When I began graduate work my future advisor asked me why I wanted to study Persian medieval literature. “I wish”, I answered, “to be the Arthur Waley of Persian”. Some of you will know the name, to others it will be unfamiliar, so I shall say a brief word about Waley, who was the finest English translator of Chinese and Japanese medieval literature in the first half of the 20th century. His versions of the great Japanese medieval novel, The Tale of Genji, of Classical Chinese lyric poetry, and of innumerable other masterpieces, made these works available to an Anglophone audience in beautiful and highly accessible versions which immediately established their literary presence in the west. His work has subsequently been attacked by some scholars, who consider that his versions contain too many concessions to Western taste. But he won a wide and admiring audience for these works, and it was his literary skill in English that enabled him to do this. It is such an audience that I wish to reach in my versions of Persian poetry. I have written a book and numerous papers on the Shahnameh, and when I write in such forms I do so as a scholar, and wish to be judged as such; but in my translations I write as a lover of literature, of the beauty of others’ artifacts, and – in so far as I can – as a poet.


I began with Barbad, and I should like to end with him. I have said that I consider that some of the finest passages of the Shahnameh are its elegies. In these passages, we feel that the lament for an individual becomes by implication a lament for a civilization, for the inevitable passing of human glory, for the tragic frailty of all our lives. I will end with my verse translation of Barbad’s elegy for his master, Khosrow Parviz:


O great Khosrow, great in your majesty,

In warlike pride, in magnanimity –

Where is that greatness now, your high renown,

Your glory and your throne and royal crown?

Where is your chivalry, your power, my king

Who sheltered all the world beneath your wing?

Where are your wives now, your musicians, all

The nobles who once thronged your audience hall?

Where now is Kaveh’s banner, and the lords

Who flourished in the air their glittering swords?

Where are your valiant warriors and your priests,

Where are your hunting parties and your feasts?

Where is that warlike mien, and where are those

Great armies that destroyed our country’s foes?

Where is your armor, wondrous to behold,

Studded with jewels, fashioned from shining gold?

Where is Shabdiz, your fiery stallion who

Galloped with such impatience under you?

Where are your glorious gold-shod cavalry,

Whose swords sought out one sheath – the enemy?

Where is that generous, never-failing store

Of your largesse? Your eagerness in war?

Where are the camels, horses, elephants,

Their howdahs, trappings and magnificence?

Where is that wise nobility of mind,

That eloquence, at once adroit and kind?

Why are you left deprived, in this sad state,

Who read you this page from the Book of Fate?

Once you desired a son – to be your friend,

A prop to your old age – and in the end

Grief came to you from him: a king’s son gives

His father strength and shelter while he lives,

But to the king of kings this son brought pain

And his success destroyed his father’s reign.

Count Persia as a ruin, as the lair

Of lions and leopards. Look now and despair,

You were the best of Sasan’s line, no one

Will ever reign again as you have done,

And as your seed degenerates, this land

Degenerates beneath an alien hand.

It is the shepherd’s fault when wolves descend

And ravage all the sheep he should defend.

God keep your soul, and may each enemy

Of yours die gibbeted in agony!

I swear by God above, my noble king,

By Mehregan, by New Year and the spring,

That if my hand plays any song again

I should be stricken from the roll of men.

I’ll burn my instruments and never face

The enemies who’ve dealt you this disgrace.”


Then he cut off his four fingers and went back to his house with his hand mutilated; there he built a fire and burned all his musical instruments.