Rumi and the Reed Flute

The Masnavi begins with an instrument complaining that it was cut from where it grew.

Roughly twenty-five thousand couplets follow, and a long tradition of commentary reads the whole work as an unfolding of that first image. What follows is a workbench rather than an essay: two lines of Persian on the table, a gloss, a scansion, and four English drafts stacked on top of each other like tracing paper, each one showing through the last. The broader trouble Rumi's vocabulary makes for translators is laid out image by image in Rumi's Metaphors Do Not Survive Translation. Here the subject is two lines, and a reed.

On the table#

The text, as given on ganjoor.net:

بشنو این نی چون شکایت می‌کند از جدایی‌ها حکایت می‌کند

bishnav īn nay chūn shikāyat mī-kunad az judāyī-hā hikāyat mī-kunad

The romanization follows a classical pronunciation, because that is what the meter needs. A modern Tehran reading sounds closer to beshno in ney chon shekāyat mikonad. In the script, a zero-width non-joiner sits inside می‌کند and جدایی‌ها, holding the prefix mī- and the plural -hā a hair's breadth away from their words without a space. Even the typography is about being almost, but not quite, joined.

A reader without Persian sees a sentence about a reed. The work is in seeing what each word is doing.

Three things on the table will matter for every draft.

چون can be "how" (listen to how it complains) or "since" (listen, because it complains). Most readers take "how." The line does not force it.

جدایی‌ها is plural. The reed tells of separations, many partings, not one abstract Separation.

The two lines are built in parallel: X-āyat mī-kunad / Y-āyat mī-kunad. Complaint and tale rhyme, so the reed's grievance and its story are heard as one act. To complain is to narrate. To narrate is to complain.

The meter, by hand#

The Masnavi keeps one meter for all six books, a form of ramal named by the mnemonic fā'ilātun fā'ilātun fā'ilun. Persian meter counts weight, not stress. A short syllable (‿) is a consonant and a short vowel. A long syllable (–) has a long vowel or closes on a consonant. Word boundaries dissolve: bishnav īn is sung bish-na-vīn.

         fā'ilātun            fā'ilātun            fā'ilun
      –    ‿    –    –  |  –    ‿    –    –  |  –    ‿    –
     bish  na   vīn  nay | chūn  shi  kā   yat | mī   ku   nad
      az   ju   dā   yī  | hā    hi   kā   yat | mī   ku   nad

Eleven syllables, three feet, the last one clipped short. The clipped foot gives each line a falling close, and across thousands of couplets it becomes a walking pace: steady enough to carry a story, a sermon, a joke about a grocer's parrot, and then the story again.

That pace is also why the Masnavi is a masnavi. A ghazal keeps one rhyme for the whole poem and stays short; a masnavi changes rhyme with every couplet and can run as long as breath allows.

So the first couplet rhymes on shikāyat / hikāyat, the second on bubrīda-and / nālīda-and, the third on firāq / ishtiyāq. Each couplet closes its own sound. The sense keeps walking.

Four drafts, laid one over another#

Each draft fixes what bothered the one beneath it, and pays for the fix somewhere else.

First draft: every word, in Persian order.

Listen this reed how complaint does from separations tale does

Faithful to every token, plural included, and both lines end on the same verb the way the Persian does. It is not English. In English, "does" at the end of a line is noise, where in Persian mī-kunad is a drumbeat.

Gives up: the language it is written in.

Second draft: make it a sentence.

Listen to this reed, how it complains, telling a tale of separations.

Readable and close. This is roughly where scholarly prose translation lives. The plural survives.

Gives up: the rhyme of complaint and tale, the repeated verb, any sense of meter. And nay has quietly become "reed" only; the flute is implied by "listen" but never named.

Third draft: find a music that does the work of shikāyat / hikāyat.

Hear this reed, how it complains, how of partings it explains.

Seven syllables a line, trochaic, rhymed, with a doubled "how" echoing the Persian parallel. Complaint and account rhyme again, which is the Persian claim.

Gives up: "explains" is weaker than "tells the tale"; the plural hangs on by "partings"; and the line has caught the nursery lightness English trochees tend to catch. Twenty-five thousand couplets of this would be unbearable.

Fourth draft: let go of form and translate what the image means.

Listen: a flute is a reed someone cut, and every note it plays is the story of the cut.

It reads as a poem. It turns the many separations into one act, the cutting, and makes the backstory explicit.

Gives up: being the first couplet. It contains an interpretation the original keeps for the next line.

Laid side by side, the drafts form a ledger in which no column is full.

No draft keeps everything, and no two keep the same things.

What the table cannot hold#

Some losses do not fit in a ledger cell.

Nay is one word for the plant and the instrument. English must pick, gloss, or footnote, and all three are a kind of cutting. Some readers also hear in nay an echo of the negative, the emptied self that lets breath through; no English word can carry that as a whisper rather than a lecture.

The third couplet does something worse to translators:

سینه خواهم شرحه‌شرحه از فراق تا بگویم شرح درد اشتیاق

sīna ḳhvāham sharha-sharha az firāq tā bigūyam sharh-i dard-i ishtiyāq

I want a chest torn strip from strip by parting, so I can give the full account of longing's pain.

Sharha-sharha is "in strips, lacerated." Sharh is "explanation, commentary." The same root in both lines: the chest must be cut open before it can explain. The wound and the commentary are one word. English has no pair that does it.

Two published traditions sit at opposite ends of the ledger. R. A. Nicholson, in the 1920s and 1930s, produced the complete English Masnavi alongside a critical edition of the Persian and a commentary: the second draft, carried out with great care across six books, explanation pushed into the notes. Coleman Barks, from the 1970s onward, produced free-verse versions worked from existing English translations rather than from the Persian: the fourth draft and beyond, the Sufi frame thinned, the feeling foregrounded. Neither is wrong about what it is doing. The mistake is citing the fourth draft as though it were the first.

The reed bed#

Every draft above leans toward an image the first couplet does not yet show. The second couplet supplies it:

کز نیستان تا مرا ببریده‌اند در نفیرم مرد و زن نالیده‌اند

kaz nayistān tā marā bubrīda-and dar nafīram mard u zan nālīda-and

Ever since they cut me from the reed bed, men and women have wept in my wailing.

Nayistān is the reed bed, the marsh where reeds grow close together, roots in one water. A flute is made by cutting a single stem, drying it, hollowing it, and boring holes along its side. Its voice comes out of what was taken away.

The commentary tradition reads the reed as the soul, the reed bed as its origin, music as the longing to go back. The poem develops that reading itself a few lines on. But the image does not need the doctrine to work. Anyone who has left a place can hear it.

And it is, a little, an image of translation. Every English version is a reed cut from its bed: hollowed, bored with holes it did not have before, made to sound. What comes through it is real music. It is also, in every note, the story of the cut.