The text eventually known as 1001 Arabian Nights has a frustrating and fascinating manuscript history. The oldest surviving fragment is from pre-Fatimid (9th century) Egypt and written on nicer paper rather than cheaper papyrus. In other words, by the time we have a textual record of Alf layla w layla (which is not yet "w layla"), it's a text with enough popularity and prestige to make it to a cultural backwater. The first mention of that additional night, furthermore, comes from a 12th century Jewish source, further demonstrating the work's broad reach.

But while 1001 Nights has a standard frame narrative (Scheherazad's wager) and title, it's never really accurate to speak of it as "one text." There are two major textual traditions coming out of the Middle Ages, one from Egypt and one from Syria, which transmit overlapping but different sets of stories. Neither of which, by the way, numbers one thousand and one. And this really, really bothered people.

So down to today, people have added different stories onto manuscript and print editions of 1001 Nights. Some of the later additions are among the most iconic "Arabian Nights" stories, too!

With 1001 Night, then, we have an incredibly fluid textual tradition that, from the standpoint of a modern scholar looking back at manuscripts and print, acquires, modifies, and discards stories over the centuries at whim. When turn of the 20th century scholars developed a deep interest in tracing possible sources for various tales, they found a similarly complicated situation. Stories "draw from" Greek, Persian, Indian, Arabian mythology and folklore; there's overlap with stories circulating in medieval Europe; ancient and contemporary allusions abound.

Scholarship on the "origins" of 1001 Nights stories and more recent studies of its textual history, taken together, illuminate a world of stories circulating around the ancient and medieval Mediterranean and Indian Ocean. We now acknowledge just how much merchant activity was flying around the coastlines, creating a true medieval world. They're not just trading goods, either. Surely the stories we're lucky enough to find versions of in texts like 1001 Nights represent a tiny fraction of a rich oral culture.

Sindbad is one of those "iconic" nights that is actually a fairly late addition to the tradition, first appearing in the 17th century. It's also one of the most literary tales, both in terms of its structure and in terms of its mythological/literary allusions. No one would question why its "original" author and later copyists would embellish a story in circulation for inclusion as one of the tale's episodes.

What's interesting, then, is that just like 1001 Nights is a fluid "text of stories, teller's choice," so is Marco Polo. The astounding number of late medieval manuscripts and versions bequeathe us a lot of variations, including in the valley of the diamonds story. Most notably, some versions have serpents and some don't. In other words, this story is swimming around the interconnected medieval world, repeated orally and textually in all sorts of ways.

From the 12th century, the unparalleled Persian poet Nizami has Alexander the Great extract the diamonds. In his version, the difficulty is actually not the snakes but the steep incline of the ravine. Alexander cleverly kills off the serpents with mirrors (...it's a medieval thing), but then still must deal with extracting the diamonds. He throws down the sheep carcasses for vultures to carry back with the loot.

But even though Nizami is specifically writing a Romance of Alexander here, he probably isn't just dropping in a random tale to spice up the conqueror's life. In the 10th century, an anonymous Arab (probably) author's work of natural history has Aristotle himself tell the reader of witnessing firsthand Alexander's cleverness in action. Earlier versions of the Alexander romance don't include the ravine of the diamonds episode--it seems plausible that this "Pseudo-Aristotle" was name-dropping in order to enhance the impression that Aristotle had written his text. The ruse certainly succeeded; its later popularity and influence were absolutely enhanced by the false attribution to The Philsopher.

The presence of the story in a work of natural history, furthermore, points to the constant interflow of stories and ideas among the scholarly elite, the popular elite, and their broader audiences. The tales are shifting as they circulate--eagles, vultures; snakes, no snakes, narratively pointless snakes--but there's a core idea of the lengths to which one must go to obtain the diamonds, and authors of all sorts of genres found use for such a lesson.

The earliest known written version of the legend, in fact, comes from quite another perspective: a Christian bishop on the island of Cyprus, writing at the end of the 4th century. Epiphanius of Constantia writes of a valley somewhere to the east (the name changes, of course) where men are forced to butcher and skin lambs, throw them into the ravine, and then chase down the birds who pick up the bejeweled carcasses to gather up the droppings. (There are no basilisks in his version; the difficulty seems to be chasing the birds over the mountains afterwards).

What use has a bishop for this story? Epiphanius, he tells us, has been asked to write a treatise on the twelve stones in Aaron's robe for another bishop. Being patristic-medieval, the goal is of course to explain the allegorical significance of the twelve gemstones as representing the twelve tribes of Israel. But along the way slash in order to accomplish that task unbelievably thoroughly, the hyper-educated Epiphanius is going to give a run-down of what the stones are. There is a stunning amount of received knowledge displayed in the text.

So Epiphanius gives us the first known written version of the story, but it's clear he's not fabricating it himself. Can we catch hints of an earlier legend?

Like everything else discussed here (surprise, surprise), Epiphanius' work has a fragmentary manuscript history. Among all the enormous number of translations (the most complete version is a 10th century Georgian translation from Armenian, which IIRC was at least translated from Greek although who knows how close to the original), we find one really interesting development in the ravine of the diamonds episode: it's not always diamonds.

So this is really interesting, because it shifts the focus of the story away from "imagining a place where diamonds were known to be found." What do we make of the story if it's not about obtaining diamonds?

Would you accept...cinnamon sticks?

In Book 3 of Herodotus' Histories, he relates the following:

Cinnamon they collect in a yet more marvellous manner than this: for where it grows and what land produces it they are not able to tell...They say that large birds carry those dried sticks which we have learnt from the Phenicians to call cinnamon, carry them, I say, to nests which are made of clay and stuck on to precipitous sides of mountains, which man can find no means of scaling.

With regard to this then the Arabians practise the following contrivance:-- they divide up the limbs of the oxen and asses that die and of their other beasts of burden, into pieces as large as convenient, and convey them to these places, and when they have laid them down not far from the nests, they withdraw to a distance from them: and the birds fly down and carry the limbs of the beasts of burden off to their nests; and these are not able to bear them, but break down and fall to the earth; and the men come up to them and collect the cinnamon.

In all the versions of the story, down through the ages, the location of the valley is always somewhere mysterious; it's always treacherous for humans to access; the payoff is always worth it. "Thus cinnamon is collected," Herodotus finishes, "and comes from this nation to the other countries of the world."

The story need not convey the "accuracy" of a snake-carpeted ravine where diamonds lie glittering. It tells the truth of a coveted good made all the more valuable by its inaccessibility to ordinary men by ordinary means.