The first work in French to mention Hafez was Les six voyages written by Jean-Baptiste Tavernier (q.v.; 1605-89) in 1676, in which Tavernier described Hafez’s tomb (see HAFEZ xiv). The first translation of one of his poems into French appeared in the Voyages (1686, repr., 1711) of Jean Chardin (q.v.; 1643-1713). However, the first complete translation of all of Hafez’s ḡ azals (q.v.) was published in 2006 by Charles-Henri de Fouchécour, well over three centuries after the first reference to the poet in a French work, whereas the complete translation of the Divān in German by Joseph von Hammer (1812) and in English by Henry Wilberforce Clarke (1891) had already appeared in the 19th century (see HAFEZ x and HAFEZ xi).

Between 1686 (Chardin) and 2006 (Fouchécour), there were several incomplete translations, ranging from only a few ḡ azals to about a hundred. Paradoxically, the first extensive translation in French was by the Anglo-Welsh Sir William Jones (q.v.; 1746-94). In 1770, Jones translated verses of three ḡ azals in his Histoire de Nader Chah and ten complete ḡ azals in Un traité sur la poésie orientale, where he provided two translations for each ḡ azal. In his account of Nader Shah, Jones translated the three ḡ azals into French verse (repr., V, pp. 46, 107, 115-16) and also offered a literal translation for each at the end (repr., V, pp. 229-30). In Un traité sur la poésie orientale, he did the opposite: In his initial description of the poetic genre (repr., V, pp. 463-71), Jones gave literal translations of ten ḡ azals, followed at the end of the book by verse translations of those and others (repr., V, pp. 484-503). The first translators to translate more ḡ azals into French were Charles Devillers and Arthur Guy (1874-1945). Devillers translated 124 ḡ azals in 1922, and Guy provided a translation of 175 ḡ azals in 1927, despite his intention to translate 573 as stated in his introduction (p. ix).

It is also paradoxical that for the most famous Persian poet there are only about 20 translations in French, in comparison to Khayyam’s (q.v.) quatrains, which were translated at least 119 times. The reason for this apparent imbalance is most probably due to the degree of difficulty in translating Hafez, but also to the need for a deep understanding of his poetry, whether regarded as lyrical or mystical. The first translators, such as Auguste Herbin (1783-1806) in 1806 and Jean-Baptiste-André Grangeret de Lagrange (1790-1859) in 1813 and 1814, pointed out that Hafez could be compared to the Greek lyric poet Anacreon (570-488 BCE) or to the Roman lyric poet Horace (65-8 BCE). On the other hand, A. L. M. Nicolas (1864-1939), in his explanatory notes on some verses, interpreted Hafez’s poetry as mystical (1898). Nevertheless, all three translators chose to translate the “beloved” Hafez celebrates in his poems without the initial capital letter and usually as a feminine word: “bien-aimée” (Grangeret, 1813 in all the three translated poems; Nicolas, p. 5, for example), and even “notre mie” in Nicolas’ translation, adding a footnote explaining its meaning as “la divinité, objet de notre unique amour” (p. 59). Some other translators, however, almost systematically translated “beloved” with an initial capital letter and as a masculine word: “Aimé,” for example, is used systematically in translations by Guy and Fouchécour. The understanding of Hafez’s poetry is therefore indispensable and decisive for its translation into French.

Translators themselves were and still are aware of the possibility of ambiguities that the Persian language allows and with which Hafez plays (see, e.g., the introduction of Lazard, 2010, pp. 31-32). Some of the most frequently used words and notions by Hafez, including the highly problematic rend and rendi (see HAFEZ viii. HAFEZ AND RENDI) are also the most complex to translate. This point was further highlighted by Gilbert Lazard (1920-2018), who explained that he translated the frequently occurring notion of rendi in Hafez’s ḡ azals by using several different words in French, depending on the context: “libertin,” “hors-la-loi,” “sans foi ni loi,” “vaurien,” “errant,” “homme libre” (2010, p. 32). Fouchécour on the contrary, always translates rendi using the French word: “libertinage” (pp. 89-90). Some poetic images are similarly almost impossible to translate as they are foreign to French culture and sensibility. In such cases, translators almost always chose to change them or leave them untranslated. This is the case with sel ‘salt’ (Pers. namak) and r ôti ‘roasted meat’ (Pers. kabāb) in the twelfth ḡ azal translated by Nicolas (see his explanation of the choices he made in his translation, pp. v-vii; the use of “roasted meat” only appears in the preface).

Hafez’s prosody is another contentious issue for the translators; whether to opt for prose or verse. If they chose the latter, they then had to decide which French poetic meter to use. This question applies to all poetic translations, including of course to Hafez. Most of the translations are in prose but translate each bayt (couplet) separately, with the exception of Charles-François Defrémery (q.v.; 1858) and Henri Massé (1950), who presented the ḡ azals in a paragraph form. Vincent-Mansour Monteil (1913-2005) and Lazard’s translations are in verse and use heptasyllables or octosyllables, which are closer to the Persian metrics of Hafez (see the explanations in Monteil, 1989, pp. 16-17 and Lazard, 2010, p. 30). Sir William Jones preferred to translate the three ḡ azals mentioned in Histoire de Nader Chah with a 12-syllable meter corresponding to the French alexandrine. In Un traité sur la poésie orientale, he chose to translate the ten ḡ azals with octosyllables. Édouard Servan de Sugny (1799-1860) used quatrains, with the French alexandrine for the first three verses and with octosyllables for the fourth one. Guy tried to imitate the Persian meter by alternating short and long syllables and explaining that this arrangement of syllables can be employed in French under certain conditions (pp. xxxvii-xxxviii). Verse translations often diverge from a literal reproduction of the original. For example, the “hair of the beloved” became the “eyes” in Jones’ translation (quoted by Shams-Yadolahi, p. 74) or a “parrot” became a “hummingbird” in Lazard (2010, p. 31).

Rhyme is yet another point to consider when translating poetry into verse. Even though Arthur Guy’s translation is in prose, he endeavored to keep the rhyme close to the original Persian, that is, a single rhyme pattern throughout the poem. Monteil (1983 and 1989) used both rhymes and assonances in his translation with the latter as his preferred form and therefore the most frequently used. For almost half of the 101 ḡ azals translated, Lazard (2010) used rhyme; 21 poems with the single rhyme pattern just as in Persian and 26 with the scheme “aabb” (i.e., rimes plates or rimes suivies in French poetry). The translation of Sir William Jones presents rhyme patterns as well. In the Histoire de Nader Chah, the rhymes are either “aabb” or “abab” (rimes croisées) or “abba” (rimes embrassées). In Un t raité sur la poésie orientale, all the poems are built with the scheme “abba.” Servan de Sugny used the same rhyme scheme, “abba.”

Given the fact that almost all the French translations are selections from Hafez, the choices made by the translator and the form of the presentation are significant. The choice could be attributed to the preoccupations of the translator, but this supposition can be refuted on the grounds that translators usually tried to offer some poems bearing a lyrical inspiration and some others with the mystical one. Although Herbin translated only four ḡ azals, he nevertheless strove to demonstrate the variety of Hafez’s writing styles by presenting the first poem in lyric style, the second one on wine, and the third one bearing a mystical tone. Concerning the fourth one, it is highly probable that it was not written by Hafez, as suggested by Zahra Shams-Yadolahi (p. 76). About his own translation, Monteil attests that the tone of the first four poems is lyrical and mystical for the last five (1983, p. 34). Regarding the order of the translated ḡ azals, most of the translators did not provide any clear cut insight into their method. As mentioned previously, Monteil organized the poems in accordance to their lyric or mystic tonality. Lazard explained that he could have relied on the ḡazals’ themes as a basis for arranging his presentation. However, that might have given the misleading impression that the arrangement reflected an evolution in Hafez’s thought. His presentation therefore imitated the Persian order, that is, it followed the alphabetical order of the rhyme (2010, p. 32). Some translators gave titles to their translated poems (e.g., Herbin; Guy; Massé, 1950; and Monteil, 1983 and 1989). Some translations provide the text in Persian, in Arabic script (Monteil) or in Latin characters. In the case of Latin characters, the Persian poems are sometimes entirely transliterated (Herbin and Monteil, 1954). Sometimes it happens that this occurs only for the first verse (Guy, integrating the name of the Persian meter used in each translated ḡ azal) in order to help the readers, especially the scholars, to find the original poem in the Divān. Some translators indicated the edition used for their translations. For instance, Defrémery (p. 419) based his translation on Hermann Brockhaus’ (1806-1877) edition. Guy (p. XXXV) used Aḥmad Sudi’s (d. ca.1598) edition. Lazard (2010, p. 32) used Parviz Nātel Khānlari’s (q.v.) edition, and Fouchécour (pp. 1255-57) based his on several editions but adhered to Khānlari’s order of numbering the ḡazals.

These translations made it possible for the cultured readership in France, in particular writers and poets, to become acquainted with the work of Hafez. However, unlike most other Persian poets, Hafez was sometimes already known to writers in France thanks to English or German translations. For example, the first French poet to quote Hafez was André Chénier (1762-1794). He discovered Hafez in 1787 during a stay in London where he read Jones’ translation in English (Ḥadidi, 1977, p. 655).

The understanding of Hafez by French writers evolved depending on the interpretation of the translators. Victor Hugo (1802-85) quoted Hafez’s bayts “Écoutez: je vais vous dire des choses du cœur”  in the epigraph for the edition of Les odes et poésies diverses in 1822 (which became Odes et ballades in 1828, without this epigraph). Naturally, the French Romantic poets shared this lyric perception of Hafez. Along the same line of thinking, after quoting Hafez’s bayts “Oh! permets, charmante fille que j’enveloppe mon cou avec tes bras” in the  epigraph of the poem “Sultan Achmet” (Poem XXIX in Les orientales), Hugo has the sultan say the verses “Je donnerais sans retour / Mon royaume pour Médine, / Médine pour ton amour,” which echoes the famous verses of Hafez about Samarkand and Bukhara given in exchange for the mole of the beloved (agar ān tork-e širāzi be dast ārad del-e mā rā / be ḵāl-e hendu-aš baḵšam Samarqand o Boḵārā rā). Hafez was subsequently read as a mystic poet, which was basically the understanding of the French Parnassian poets. This is the case with poets such as Théophile Gautier (1811-72), Armand Renaud (1836-95), and Jean Lahor (1840-1909). Gautier quoted Hafez in the preface of Émaux et camées (1852). When Renaud wrote Les nuits persanes (1870), he was influenced by famous Persian poets such as Saʿdi (q.v.) and Jalāl-al-Din Rumi, although mainly by Hafez (Ḥadidi, 1977, pp. 662-64). He called him, in his preface, “Hafiz, le voluptueux aux profondeurs mystérieuses” (Les nuits persanes, p. 8). This influence is particularly noticeable in the first part of the volume, “Gul & Bulbul – La rose et le rossignol.” In L’illusion (1888), Lahor entitled one of his poems “Hafiz,” saying that he was “un rossignol fou de toutes les roses” (p. 141).

André Gide (1869-1951) also began the first chapter of Les nourritures terrestres (1897) with a verse of Hafez in the epigraph “Mon paresseux bonheur qui longtemps sommeilla / S’éveille…” and quoted him several times in the same book. He admitted and recognized the influence Persian poetry, and also Hafez, had on his work in the Journal Parse in 1921 (see also Honarmandi, pp. 27 sqq). Tristan Klingsor (1874-1966) wrote Schéhérazade (1903), presenting Hafez as his master and himself as his bad student. In Les éblouissements (1907), Anna de Noailles (1876-1933) wrote “Le jardin-qui-séduit-le-cœur,” in which she talked about a garden in Shiraz in reference to Hafez. In “Le parfum entré par une fenêtre” (Henry de Montherlant [1895-1972], Encore un instant de Bonheur, 1934), the love for jasmine is an allusion to the love for the rose in Hafez. Jérôme Tharaud (1874-1953) and Jean Tharaud (1877-1952) conjured up in Vers d’almanach (1946) a discussion between Hafez and Timur concerning the verses about Samarkand and Bukhara (Ḥadidi, 1977, pp. 658-59). These examples clearly demonstrate the manner in which French writers found new themes and images in Hafez’s poetry to renew their own inspiration.

Nevertheless, one may wonder why there are so few translations in French of the work of Hafez compared to other Persian poets. This is perhaps linked to the fact that, as noticed by Fouchécour, Saʿdi corresponds better than Hafez to the French way of thinking (reported in Shams-Yadolahi, p. 118). Another reason could simply be the difficulty in translating Hafez, as mentioned previously; the rhythm of his poems; the complexity of his images; and the different interpretations, lyric and mystic, of his work. As aptly put by Henri Massé (Khanlari, p. 168), translating Hafez is like trying to lock up moonlight in a vase.