In Coral Gables, Fla., where I grew up, most of the streets have exotic names – Alhambra, Granada, Zamora – drawn from Moorish Spain. The parks are adorned with elaborate fountains and porticos and the houses are in a “Moorish” style with roofs of red Cuban tile, terrazzo floors, and facades of pale stucco. These surroundings, abetted by my early readings of a child’s version of “The Arabian Nights” with its intricate pseudo-Persian illustrations, whetted my taste for the East; from an early age I was enthralled by arabesques and curlicues, and the taste has stayed with me.

It was a surprise therefore to learn, when I began the serious study of Classical Arabic as an adult, that that teeming cornucopia of fabulous tales known also as “The Thousand and One Nights” is in fact rather despised by Arabists. The language of the tales, I was told, was quite debased, a farrago of colloquial and high-flown Arabic, a “Middle Arabic” interesting enough for linguists but not to be taken seriously as literature. There was also the seemingly insoluble problem of the text itself; the big 19th-century editions, printed in Calcutta, Cairo, or Breslau, were slovenly and unreliable (as well as bowdlerized). No Arabist could take seriously the work of such translators as the flamboyant Richard Burton, who had produced an entire version, in 16 thick volumes (plus supplements), with copious and often rather daffy notes.

This year marks the 300th anniversary of the introduction of the “Nights” into European ken. In 1704, the French man-of-letters Antoine Galland began publishing his “Mille et une Nuits,” a task that took him 13 more years. Though this date is precise, it is strangely hard to grasp. The “Nights” seem to have been with us always. Despite their exotic trappings, we enter them, even for the first time, as if we had long known their peculiar atmosphere, part dream, part gossip, part legend. They have the curious suspension of fairy tales in which anything can, and does, happen in accord with laws which we may not be able to formulate but recognize intuitively.

In the last decade or so, the “Nights” have become more respectable to scholars, and those presenting papers on “Gender and Identity in the Tale of Qamar al-Zaman” at learned conferences are no longer looked at askance, no matter how labored or trendy their analyses. This change probably has to do with the appearance, in 1984, of the first genuine scholarly edition of the text, prepared by the Iraqi-born historian of Islamic philosophy Muhsin Mahdi, long at Harvard. Professor Mahdi toiled over his edition for many years, alongside his more austere investigations of such philosophers as al-Farabi; it was plainly a labor of love. Working with the earliest reliable source – a 14th-century Syrian manuscript – he prepared a critical text quite different from earlier editions of the “Nights.” For one thing, the Mahdi text contains only 10 or 11, not 1001, stories – not as drastic a reduction as it might seem since these stories ramify and unfurl quite extravagantly in their own right. More importantly, his text captures the rough, unpolished, oral quality of the earliest versions; the text is often unpointed (that is, without the diacritical marks that distinguish letters in Arabic) or fragmentary, and it is pungent with colloquialisms.

Another Iraqi-born scholar, Husain Haddawy, has used the Mahdi text as the basis for his translations, two volumes of which have now appeared (“The Arabian Nights,” Everyman’s Library, $20 each). Mr. Haddawy was brought up listening to his great-uncle regale the family with stories, some of them from the “Nights,” and his keen sensitivity to the oral quality of the work is reflected in his translation. It is brisk, clear, fast-moving, and often racy; only his renditions of the many verses that crop up in the tales are lame (and occasionally quite horrible).

Every translator of the “Nights” has harsh words for his predecessors, and Mr. Haddawy is no exception. Richard Burton’s preface to his own versions took rather cruel delight in mocking the failures of earlier attempts, especially that of the prudish William Edward Lane. Lane was a great scholar and his “Manners and Customs of the Modern Egyptians” as well as his colossal “Arabic-English Lexicon,” in eight volumes (left unfinished at his death), remain indispensable today. Mr. Haddawy, in his turn, happily points out Burton’s excesses, such as the following:

Then she thrust her hand into his breast and, because of the smoothness of his body, it slipped down to his waist and thence to his navel and thence to his yard, whereupon her heart ached and her vitals quivered and lust was sore upon her, for that the desire of women is fiercer than the desire of men, and she was ashamed of her own shamelessness.

This may not qualify for the Literary Review’s “Bad Sex” award, but it comes close; the original, as Mr. Haddawy shows, is much plainer, and more convincing for that.

The advantage of Burton is that with him you get all the “Nights,” the well-attested together with the probably spurious (among which, alas, are “Ali Baba” and “Aladdin,” for which no Arabic originals have ever been found). Moreover, while I admire Mr. Haddawy’s translation, Burton’s is in a somewhat different category, being as much the uncontrollable expression of his over brimming personality as a translation tout court. Burton also aimed for something higher; in his inimitable preface, he remarks that “the translator’s glory is to add something to his native tongue, while avoiding the hideous hag-like nakedness of Torrens and the bald literalism of Lane” and so, “I have carefully Englished the picturesque turns and novel expressions of the original in all their outlandishness … and I have never hesitated to coin a word when wanted, such as ‘she snorted and snarked,’ fully to represent the original. “The truth is, I fear, that if you want to celebrate the tricentenary of the “Nights” properly, you will need both the Haddawy and the Burton translations (the latter is available in a Modern Library paperback: “The Arabian Nights: Tales from a Thousand and One Nights,” 872 pages; $13.95).

In one tale, we learn from Burton that the Caliph Harun al-Rashid suffered from insomnia. His clever vizier summoned Ali the Persian to the caliphal chamber. Ali told the story of how he was robbed in the marketplace by a “rascally Kurd” of his travelling bag. Both were hauled before the Qadi, who asked them to describe the contents of the bag. The Kurd launched into a fantastic inventory; the bag contained not only “a cooking pot and 2 water jars and a ladle” but also a “she-cat and 2 bitches and a wooden trencher” as well as a cow and calves, a she-goat and two sheep, a lioness and its cubs, a bear, a jackal, a sofa, two sitting rooms, a kitchen and “a company of Kurds who will bear witness that the bag is my bag.” Not to be outdone, “Ali listed castles, cranes and chess players, two entire villages, a whore and a hermaphrodite, and a brood mare with her colts. The contest becomes ever more far-fetched until the bag contains both the Tigris and the Euphrates and all the earth from Iran to China.” When the Qadi opens the bag, there is only bread and cheese and olives inside.

“The Thousand and One Nights” is like that disputed bag. The whole world bulges inside, conjured out of the homeliest of ingredients. The bag holds Poe and Borges and Calvino and Joseph Roth and Rimsky-Korsakov and Delacroix, and who knows how many others. Bottomless and inexhaustible, it contains all of us, too.