Repository path: workshop/translations/hafez-daasht/R58-v1/translation.md · rendered 2026-09-09

«حافظ» غزل ۷۷ rendered whole under R58, in a declared plain register

Source: Ganjoor's Hafez (Qazvini–Ghani), /hafez/ghazal/sh77, fetched 2026-09-01; the Persian, the nine rhyming positions and the frozen selection rule are in ../source.md.

Regime: R58 — the radif carried by inversion, and the inversion's reach recorded.

Register declared before the first English line: plain modern English. No thou/thee/thy, no -eth/-est, no hath/doth, no auxiliary do inserted for metre, no lexis a present-day reader would place before 1900.

Cost: $0 — the lead's own hand, not ledgered (charter §3, A4).

Contamination: none, measured — ../dependence-note.md. Payne's ode 69, which renders this same ghazal, was not opened before this file was frozen and committed.

This is not an independent second opinion of anything (CLAUDE.md, the contamination rule). It is one hand's record of where a word-order constraint bites, made so that the study limb has a first-hand claim to test.

The rendering

A nightingale a bright rose petal in its beak had, and in that leaf and tune sweet cries, full of pain, had.

I said to it: in the very hour of union, what is this moan, this cry? It said: the Beloved's display us at this work had.

If the Friend did not sit with us, there is no ground to complain: he was a king in luck, and for beggary disdain had.

Our pleading and our airs take no hold upon the Friend's beauty; happy the man who from the lovely ones a fortune's gain had.

Rise, and let us scatter our lives upon that Painter's reed, for in his compass' turning all these strange designs he had.

If you are a disciple of love's road, give no thought to a bad name: Sheikh San'an his cloak in pawn at the wine-seller's house had.

Sweet the hour of that sweet wandering dervish, who along the stages of his way the angels' beads of praise inside the pagan girdle had.

Hafez's eye, beneath the eaves of that houri-natured one's hall, the way of gardens with rivers running under them had.

Translator's log

Frozen 2026-09-01, before the study limb of this session was designed and before Payne's ode 69 was opened. Decisions and alternatives, not quality claims. Any evaluative sentence carries internal-judgment-only, and the lead does not judge this rendering.

1. The radif: the options enumerated before any line was written

داشت is the third-person singular preterite of داشتن. It has to stand last, identical, at nine positions, and the nine complements it governs are: a rose petal in a beak (1a), laments (1b), us, at this business (2b), disdain-for-beggary (3b), a fortune with a share in it (4b), all these strange figures (5b), a cloak in pawn (6b), the angels' rosary (7b), the manner of a Qur'anic phrase (8b).

Recorded before writing: unlike کرد at sh130, where English splits one Persian verb into do and make and neither collocates across the poem, داشت has one English collocate that serves all nine. The lexicon is not the obstacle here. Word order is the whole of the obstacle, which is why this poem was selected by a rule that required a finite transitive radif.

2. What the obstacle actually is

English fixes verb before object. To end a clause on had with an overt complement, the complement must get in front of the verb, and English has exactly two ways to do that: constructions that front it for their own reasons (relative and wh-clauses, comparatives, clefts — "all that he had", "more than he had"), and bare topicalisation, which is the marked object–verb inversion framework/v0.2 §7.42 prices.

Both were tried at every position. The construction route was rejected as a systematic solution because using it nine times means nine relative or wh-frames, and the poem has no relative clauses in it: the frames would be a fact added to the English at every bayt, which R51 §1 forbids. It survives only where the Persian already has a relative (4b, آن کز…, he who…).

So the rendering takes the bare inversion, and the log's job is to record how far each one had to travel — R58 §4.

3. The reach of the inversion, position by position

Graded after each line was written, before Payne was opened. LOCAL = only the verb is displaced, the rest of the hemistich stands in ordinary modern order. SPREAD = a second constituent had to move too. REBUILT = the predicate itself was re-shaped (nominalisation, light-verb periphrasis, change of voice). REFUSED = no English tail can stand there.

Tally (computed by runs/RS-20260901-inversion-habit/tally_log.py, which parses this table; R58 §7, note (bey)): radif HELD 9 of 9. Reach: LOCAL 6, SPREAD 3, REBUILT 0, REFUSED 0.

The claim this hands to the study limb. In a declared plain modern register, with no archaism and no reshaping of any predicate, the object–verb inversion that carries this radif was a local move at six of nine positions and never once required the nominalise-and-front rebuild that RS-20260831b-radif-hands §6 caught Payne using at sh130. If that generalises, the inversion is something a translator buys where it is needed rather than a grammar the whole poem has to be written in — and a hand who inverts at a radif position should not, on that account, be inverting anywhere else.

4. The chime

Attempted in the -ain family, the widest English family reaching more than one of the nine senses.

Chime HELD 3 of 9, SHORT 0, REFUSED 6. Persian's -ār family carries nine senses without strain; the English rhyme families that reach two of these senses do not reach a third.

5. Other decisions

• برگ و نوا (1b) is a pun — leaf and tune, and idiomatically means and provision. Rendered as leaf and tune, which keeps the literal pair the bird is holding and singing and loses the idiom. No English pair does both; the alternative (means and measure) keeps the idiom and drops the petal, which the couplet needs because 1a put it in the beak.
• قلندر (7a) rendered wandering dervish rather than transliterated. A transliteration would be the more exact word and would need a gloss the poem has not got; Sheikh San'an is kept untranslated in 6b because it is a proper name and the story is the couplet's whole content.
• جنات تجری تحتها الانهار (8b) is a Qur'anic tag in Arabic inside a Persian line. Rendered in English and italicised, so the reader sees a quotation without being told what it is quoted from — the Persian reader is not told either.
• شیوه (8b) as the way of, not the manner of: manner would ask the line to end on a noun the inversion then has to lift over.
• Line length is not a declared constraint of R58. The English lines run 12–16 syllables against the Persian's 15, and the a-lines run longer than the b-lines because the b-lines end on a monosyllable that the a-lines do not have to reach.