Translations of One Thousand and One Nights have been made into most of the world's major languages. They include the French translation by Antoine Galland (titled Les mille et une nuits, finished in 1717). Galland's translation was essentially based on a medieval Arabic manuscript of Syrian origin, supplemented by oral tales recorded by him in Paris from Hanna Diyab, a Maronite Arab from Aleppo.
The first English translation appeared in 1706 and was made from Galland's version; being anonymous, it is known as the Grub Street edition. There are two extant copies, one kept in the Bodleian Library and one in Princeton University Library. After this, several English reissues appeared simultaneously in 1708. As early as the end of the 18th century the English translation based on Galland was brought to Halifax, Nova Scotia, Montreal, Philadelphia, New York and Sydney. Generally, translations starting from Galland were censored due to lewd content.
Meanwhile, the original scattered Arabic texts were collected and printed in four corpuses: the so-called Calcutta I or Shirwanee Edition (1814–18, 2 volumes) the Bulaq or Cairo Edition (1835, 2 volumes) the Breslau Edition (1825–38, 8 volumes) the Calcutta II or W.H. Macnaghten Edition (1839–42, 4 volumes)
Galland-based English translations were superseded by that made by Edward William Lane in 1839–41. In the 1880s an unexpurgated and complete English translation, The Book of the Thousand Nights and a Night, was made by Richard Francis Burton.
French translations
Galland
Galland based his translation on what is now known as the Galland Manuscript, a three-volume (or perhaps once four-volume) manuscript from the fourteenth or fifteenth century now kept in the National Library of France. He published his translation in a twelve-volume set between 1704 and 1717. Galland's translation altered the style, tone and content of the Arabic text. Designed to appeal, it omitted sophisticated or dark elements while enhancing exotic and magical elements and became the basis of most children's versions of One Thousand and One Nights.
Mardrus
In 1926–1932 a lavishly decorated 12-volume edition of J. C. Mardrus' translation, titled Le livre des mille nuits et une nuit, appeared. Soviet and Russian scholar Isaak Filshtinsky, however, considered Mardrus' translation inferior to others due to presence of chunks of text, which Mardrus conceived himself to satisfy the tastes of his time. According to Robert Irwin, "Mardrus took elements which were there in the original Arabic and worked them up, exaggerating and inventing, reshaping the Nights in such a manner that the stories appear at times to have been written by Oscar Wilde or Stéphane Mallarmé". In response to criticism of his translation by academic Arabists, Mardrus promised to produce a tome of learned commentary and justificatory pieces which he, however, failed to do.
Russian translations
The first Russian translation of One Thousand and One Nights, in Russian: (Týsjača i odná nočʹ), was made by Alexey Filatyev in 1763–1774. It was based on Galland's translation and consisted of 12 volumes. Later Russian translations were also based on European translations. For instance, a three-volume translation by Yulia Doppelmayr (1889–1890) was based on Galland, the six-volume translation by Lyudmila Shelgunova (1894) was based on that by Edward William Lane and an anonymous translation (1902–1903) was based on Mardrus.
The first Russian translation directly from the Arabic source (from Calcutta II) was made by Mikhail Salye, published in eight volumes by Academia in 1929–1939. Salye also translated seven tales not contained in Calcutta II (from the manuscript in the National Library of Russia).
English translations
Departing from the Grub Street version, Jonathan Scott made the first literal translation of Galland. This rendition, entitled The Arabian Nights Entertainments, appeared in 1811.
Henry Torrens would later translate the first 50 nights from Calcutta II, published in 1838. Having heard that Edward William Lane had begun his own translation, Torrens abandoned his work. Lane translated from the Bulaq corpus. In his opinion, "Galland [had] excessively perverted the work". According to Lane, Galland's "acquaintance with Arab manners and customs was insufficient to preserve him always from errors of the grossest description". Working with the Bulaq corpus, Lane occasionally crosschecked against the Calcutta I and Breslau corpuses. His translation, however, was incomplete. In 1923 a translation by Edward Powys Mathers based on the French translation by J. C. Mardrus appeared.