Found on 2026-10-10 by a search the owner asked for: documented cases of a change of cover, for CLAIM-201, whose own illustrations are invented (tu and vous in the abstract) or re-used (CASE-010).
The case
Persian has one third-person pronoun, u, and its possessive clitic -ash has no gender either. The source, Hafez, Divan, ghazal 3, verse 1 (Ganjoor, https://ganjoor.net/hafez/ghazal/sh3, fetched 2026-10-10):
اگر آن ترک شیرازی به دست آرد دل ما را / به خال هندویش بخشم سمرقند و بخارا را
Gloss: “If that Shirazi Turk takes our heart in hand, for his/her Hindu [= dark] mole I’d give Samarkand and Bukhara”. hendu-yash, “his/her Hindu”, is ungendered, and so is “Turk”. Ganjoor’s own prose paraphrase keeps the open pronoun (خال سیاه او, “his/her black mole”). Verse 6 turns to the second person, still ungendered: اگر دشنام فرمایی, “if you deign to curse [me]”. Whether the beloved is a boy, a woman or God (the “true Beloved” of the Sufi reading) is disputed in the scholarship.
English must choose at the first possessive, “his” or “her”, and a capital (“His”) can choose a divine addressee as well. Four renderings choose three ways.
- Jones (1771): a woman. William Jones, A Grammar of the Persian
Language (London: W. and J. Richardson, 1771), read in the archive.org
scan
agrammarpersian00jonegoog, whose title page is the 1771 first edition. His literal prose, near the end of the grammar (about p. 135): “If that lovely maid of Shiraz would accept my heart, I would give for the mole on her cheek the cities of Samarcand and Bokhara.” His verse “A Persian Song” (about p. 137): “Sweet maid, if thou wouldst charm my sight, / And bid these arms thy neck infold; / That rosy cheek, that lily hand / Would give thy poet more delight / Than all Bocara’s vaunted gold, / Than all the gems of Samarcand.” The decision is made even in his grammatical example on verbs: “If that fair damsel of Shiraz would accept my heart”, although he transliterates the line on the same scan as “Egher … turki Shirazi”. The verse also turns the third person into direct address. - Clarke (1891): God. H. Wilberforce Clarke, The Dīvān-i-Hāfiz,
vol. 1 (Calcutta, 1891), p. 40, ode 8 in his numbering (archive.org
thedivan01hafiuoft): “If that Bold One (the true Beloved) of Shiraz gain our heart, / For His dark mole, I will give Samarkand and Bukhara (both worlds).” His gloss on the page lists the alternatives: “Turk-i-Shirazi signifies:— (a) Hāfiz’s beloved (Shakh-i-nabat). (b) The true Beloved (God).” Shakh-i-nabat is glossed on the same page as “A lovely woman”. On p. 43 Clarke reprints Jones’s “Sweet maid” version, saying it gives “neither the metre, nor the rhyme, nor the sense”. One book prints the divine reading and the female one side by side. - Bell (1897): a woman, addressed. Gertrude Lowthian Bell, Poems from
the Divan of Hafiz (London: Heinemann, 1897), pp. 70–71 (archive.org
poemsfromdivanof00hafiiala): “Oh Turkish maid of Shiraz! in thy hand / If thou’lt take my heart, for the mole on thy cheek / I would barter Bokhara and Samarkand.” (The OCR reads “On”; “Oh” is the evident reading.) Like Jones, she also turns the third person into the second. Her notes (pp. 127–129) reject the alternative in print: “The whole poem has received a mystical interpretation which seems to me to add but little to its value or to its intelligibility”. They tell the Timur anecdote with “the black mole on the cheek of thy mistress”. - Davis (2012): a boy. Dick Davis, Faces of Love: Hafez and the Poets
of Shiraz (Mage, 2012; Penguin Classics, 2013), pp. 128–129 by the
book’s own notes, read in archive.org
facesoflovehafezandthepoetsofshiraz: “If that Shirazi Turk would take / My heart within his loving hand / I’d give for his dark mole the towns / Of Bokhara and Samarqand;”. His note on the line tells the same Timur anecdote as “some pretty boy’s mole”.
Davis states the forced choice in his Introduction (page numbers not legible in the scan): “Persian pronouns have no gender distinctions, so that the same word may be translated as ‘he,’ ‘she,’ or ‘it.’ … there is usually no way of telling whether a boy or a girl is being talked about.” And: “Unfortunately a translator of Persian lyric poetry into English usually has to make a decision as to the gender of the person being addressed in a given poem … the reader should be aware that this is in most cases an arbitrary decision, so that he or she is perfectly welcome to change most instances of ‘he’ to ‘she’ and ‘she’ to ‘he’”. He names the history too: “some Victorian commentators and translators tended to bowdlerize these poems by making them always about girls, a blanket insistence that they are always about boys seems to me to be equally tendentious.”
What is held fixed: the mole, the two cities, the wine and the cup-bearer. The renderings differ on which questions they decide, not on those.
What it can show
It is a documented instance of the half of a change of cover that CLAIM-201 calls “the target decides what the source left open”. It shows it on the question the claim needs: who is addressed, and in what relationship. Take the question “Is the beloved a woman, a boy, or God?”. Expert readers of the Persian disagree on it, as Clarke’s two-way gloss, Davis’s “no way of telling” and the scholarship’s dispute attest. A reader of Jones or Bell answers “a woman”, a reader of Davis “a boy” and a reader of Clarke “God”, each on the text’s word. The answer settles what kind of act the poem is: a gallant courtship lyric, a love lyric in the pederastic convention, or devotion to God. That is a change in which contexts exist for which readers, with no other question’s answers needing to move, which is the claim’s form. Davis’s printed licence to swap “he” and “she” shows a translator treating the question as undecidable in the source and decided only by the rendering.
Davis’s licence also bears, as a dispute and not a resolution, on CLAIM-229’s search for renderings that keep the act while changing a proposition. Davis holds that the gender swap keeps the poem’s act (“the real subject is longing and desire, polymorphous and overwhelming”). Bell and Clarke treat the parallel choice, human or divine, as changing it. The proposition is also one the source never asserted, so strictly it is added, not changed.
What it cannot show
- It meets only one half of the definition. CLAIM-201 defines a change of cover as a question determinate for source readers becoming indeterminate for target readers, or the reverse. This case shows only the second of those, a question open in the source that the rendering decides. It shows no question the source decides that a rendering makes undecidable, the claim’s “the source decides what the target cannot”. A candidate in this poem is the honorific farmāyī (“deign”) in verse 6, which marks the lover’s deference; no critic was found saying a rendering loses it, so it is not claimed.
- It measures no reader agreement. The claim’s operational test is agreement among readers above a threshold fixed in advance. The evidence here is a handful of translators and scholars, who are readers but not a sample. Their disagreement shows that experts differ, not that ordinary readers of the Persian answer below a threshold. That the English readers answer determinately is inferred from the explicit pronouns, not measured.
- The strongest reply. A defender of the view that this is no change of cover would say the Persian is not open for its competent readers. The genre convention makes the beloved male by default (Davis: “scholars have generally assumed” the pederastic relationship, and when gender is shown “it’s usually clear that a boy rather than a girl is being referred to”). The Sufi reading is a second layer of interpretation, not an indeterminacy in the first. On that reply, a Persian reader answers “a boy” determinately, and Jones, Bell and Clarke change the answer to a question both sides can decide. That would be distortion within a shared context, which CLAIM-257 measures apart from a change of cover. It would not be the cover changing. Against the reply stand Davis’s own “no way of telling”, Clarke’s two-way gloss, and Lewisohn’s view (known here only through Wikipedia, unverified) that the beloved is “nearly always female”. The reply cannot be settled from the texts.
- What would close the gap. The reader test the search report sketches could serve as items for the discriminating study (CASE-055). Put the question “Is the beloved a woman, a boy, God, or does the poem not say?” to readers of the Persian. Put it to matched English readers of Davis’s version as printed and with “his” changed to “her”, and of Clarke’s capitalised version. Compare agreement against a threshold fixed in advance. The case supports the claim if the Persian readers fall below the threshold and each English group is above it on a different answer. It is the strongest reply’s case if the Persian readers agree on “a boy”.
- It is not contextuality, as CLAIM-201 itself says of a change of cover. Nor does it show the change at small local distortion in any measured sense; it shows that such changes are made.
Leads for the reverse half, unverified. The half this case lacks needs a source that marks what the target cannot. The search’s companion report on compensation names Pushkin’s «Ты и вы» (1828), a poem whose event is a woman’s slip from вы into ты. Its English versions handle it differently: Givental and Wilson-Egolf use “thou” (https://math.berkeley.edu/~giventh/verse/thou.pdf), and Bullock moves the pronouns into French (https://oxfordsong.org/song/tu-et-vous). Whether any English version leaves the slip undecidable for its readers, and whether any critic says so, has not been checked. Dolly’s ты in CASE-066 is the same contrast in prose, filed for compensation, where the translators work to keep the question decidable. Failing a documented case, the search report proposes a stipulated one: a T/V-marked source dialogue whose switch of pronoun is the event, rendered into English, with “Did the relationship change at line n?” put to both groups of readers.
Sources
None of these works is held in the record as a LIT. They are cited here by identifier.
- Hafez, Divan, ghazal 3, Persian text and paraphrase: Ganjoor, https://ganjoor.net/hafez/ghazal/sh3. Read directly.
- Jones, William, A Grammar of the Persian Language (London, 1771): https://archive.org/details/agrammarpersian00jonegoog (Google Books scan). Read directly in the OCR text. The search report took this scan for the 1828 edition edited by Samuel Lee, but its title page reads 1771; the page numbers given are inferred from the OCR’s page heads and should be checked against the page images.
- Clarke, H. Wilberforce, The Dīvān-i-Hāfiz, vol. 1 (Calcutta, 1891): https://archive.org/details/thedivan01hafiuoft. Read directly, pp. 40 and 43.
- Bell, Gertrude Lowthian, Poems from the Divan of Hafiz (London: Heinemann, 1897): https://archive.org/details/poemsfromdivanof00hafiiala. Read directly in the OCR text. Page numbers are inferred from the OCR’s page heads.
- Davis, Dick, Faces of Love: Hafez and the Poets of Shiraz (Mage, 2012; Penguin Classics, 2013): https://archive.org/details/facesoflovehafezandthepoetsofshiraz, whose copyright page gives ISBN 978-1-101-62717-4 (an e-book ISBN; the print ISBNs are unchecked). Read directly in the OCR text.
- Wikipedia, “Shirazi Turk”, section “Gender of the Turk”: https://en.wikipedia.org/wiki/Shirazi_Turk. Read only through the search report. It cites Bicknell, Herman, Háfiz of Shíráz (1875), p. 20 (“within his hand”), and Lewisohn, Leonard (ed.), Hafiz and the Religion of Love in Classical Persian Poetry (2010), p. 48. Neither was opened; both are unverified.
- Anushiravani, Alireza and Atashi, Laleh (2012), “Cultural Translation: A Critical Analysis of William Jones’s Translation of Hafez”, Persian Literary Studies Journal 1(1): 41–58, doi:10.22099/jps.2013.601. Known only from its abstract as returned by search, which argues that Jones turns the poem’s mystical core into secular love. Not read.
- Arberry, A. J. (1946), “Orient Pearls at Random Strung”, Bulletin of the School of Oriental and African Studies 11(4), JSTOR 608587. Collects English versions of this ghazal. Not read.
- For the reverse-half lead: Pushkin, «Ты и вы» (1828), with Givental and Wilson-Egolf and Bullock at the URLs above, known here only through the companion search report. Unverified.
Note of 2026-10-10: the reverse half has a candidate
The reverse half this case lacks, a question the source decides that a rendering makes undecidable, now has a documented candidate in CASE-059: Tatiana’s letter in Eugene Onegin III, whose вы→ты→вы is lost in Kline’s “you” throughout and half kept in Spalding’s “thou” with no return. The lead named above, Pushkin’s «Ты и вы», was checked and fails: the poem names its pronouns, so every rendering keeps the switch, and what English endangers is the value of “thou”, not the question. The new case supplies attested texts and a visible loss. It does not supply what this case also lacks, measured reader agreement on either side, and its native-reader attestation is thinner than this case’s scholarly one.