Review: The Magic Wife

Some pitfalls of long-range comparative mythology

Think of this post as a study group on Crecganford's recent mythological analysis. I encourage you to watch the 39-minute video and comment. I’ve mentioned Crecganford a couple of times on the blog. He has a PhD in comparative mythology, with a specialty in Indo-European mythology, and now runs a popular YouTube channel that focuses on long-range mythology with many videos on the oldest myths. He also maintains an immensely useful database of world mythology.

The Magic Wife motif is a common story in various cultures where a man marries a magical woman who can transform into an animal. These wives are usually benevolent and do not intend to cause harm. Variations of this motif include wives who transform into swans, ducks, geese, doves, foxes, and other animals. These stories often feature themes of transformation and magical abilities and sometimes involve the husband breaking a taboo, leading to the wife leaving or transforming back.

As you can see below, the motif casts a wide net and is present in hundreds of cultures, though less often in Subsaharan Africa, Siberia, and particularly Australia. This is one area where Crecganford shines because, as someone upkeeping the database, he can explain the dearth in Siberia as a general lack of data from the region (as well as being sparsely populated, historically).

Not much work is done to argue that this theme forms a phylogeny stemming from a root myth. His professional peers are long-range comparativists who are ready to assume transmission, and he has built an audience of lumpers, so that is his purview. So, that is an initial caveat. Why should we assume this was originally a single myth that spread? Maybe stories about magic wives have been invented several (or many) times independently.

(Note that the argument for phylogeny is much stronger with the bullroarer as it leaves archeological data, with the earliest specimens dated 20 kya. It’s also considerably harder to misidentify a bullroarer than a Magic Wife story or to independently invent a bullroarer mystery cult than it is to tell a story about a wife turning into a crane. )

Out of Africa?

After explaining the Out of Africa migration, he says, “Now this allows us to say that myths that seem to be older than 70,000 years must have almost certainly come from Africa.” There are no myths that seem to be older than 70,000 years! Rather, there are myths that seem to have a global phylogeny, which some assume cannot have spread into Australia and Africa in the last 20,000 years. But there is really nothing to suggest myths have survived from 70,000 years ago, and certainly not specific myths (e.g., fighting the dragon, the world egg, the magic wife). This is my frustration with his videos on the Seven Sisters and creation myths, both of which he argues are more than 100,000 years old.

Luckily, in this instance, Crecganford interprets the sparse representation in Africa and Australia to mean it must have entered in the Holocene, so we don’t have to deal with another model of myths magically surviving from before behavioral modernity. Still, as you may have noticed, by the 70 kya date for the Out of Africa migration, he has a bias for the earliest possible dates. For example, if a myth is represented in both the New and Old World, he assumes it is at least 25,000 years old. However, there is no evidence of art or sophisticated stone tools until ~13kya in the Americas. So there could have been some previous group that moved over the Bering Strait 25kya, or even 40kya (there are about a dozen controversial sites that old), that was later replaced. Ultimately, Crecganford puts forth a model of the Magic Wife emerging 35±5 kya in Asia, and probably Southeast Asia. I don’t think you need anything more than 15 kya to get full coverage, and I tend to have a bias for recency because myths decay.

A neolithic connection between Mexico and Southeast Asia?

Building a phylogeny of myths assumes that myths tend to “mutate” slowly. Similar myths are siblings or cousins, and, as a sanity check, they should usually be geographically near one another. However, myths with worldwide distribution often show a connection between Mexico and East Asia. This has come up in several of his videos; in this one, he says:

“The most plausible explanation for this folklore mythological similarity between East Asia and Mesoamerica is the transfer of motifs through Beringia during an early stage of America's original settlement. But while this hypothesis easily explains the wife dog motif, and hints at the flood myth already being known... it is more challenging as an explanation for the motifs G9 and G9a, because they assume a knowledge of agriculture which came in the last 10,000 years, hinting at a deeper historical connection.”

Most would simply say the similarity is a coincidence, but it seems he’s leaving the door open to a relation between Mexican high civilization and Asia—Olmecs of East Asian extraction, Tenochtitlan pyramids derivative of those at Xi’an. This has been a popular line of speculation for over a century and it is more fruitful to say directly. For example, Joseph Campbell wrote:

“Many motifs of the Mayan “historic horizon” suggest specifically contemporary India, Java, and Cambodia; e.g., the trefoil arch, tiger throne, lotus staff and lotus throne, conch shell associated with plants, cross and sacred tree (often with a monster mask in the center and bird in the upper branches), serpent columns and balustrades, seated lions and tigers, copper bells….And are we still to suppose that America remained inviolate?

  If it did, then psychology has a far greater task ahead, comparing the feats of duplication, than archaeology or ethnology would face if it were merely a six-thousand-mile voyage across the Pacific, c. 1500 b.c., that had to be explained.” ~The Masks of God, Volume 1: Primitive Mythology, Joseph Campbell, 1959

He clearly lays out the mystery, and either diffusion or psychology will have to surprise us (and he bets it will be diffusion). With 60 more years of archeological data and now the view provided by genetics, the case for old-world diffusion laying the foundation of Mayan high culture doesn’t look great. It is valuable to show that long-range comparative mythology implies there is diffusion for which we have no evidence. This helps calibrate the best use of different methods and how much we should trust them. Crecganford’s work would be stronger if he treated his model as making predictions and showed when it does and doesn’t pan out.

Why did the myth diffuse and persist?

Let’s grant the root is 35 kya. Is there any cultural development that could explain the Magic Wife motif? Well, shamanism was being developed in Europe, along with a bunch of Venus figurines. Maybe that had something to do with the magic wives¹?

Happy to chat in the comments. What do you think of this type of research?

1. For the record, the Magic Wife motif isn’t that related to EToC, other than examples of magic wives that Crecganford rejects from the grouping. Medea, for example, doesn’t have an animal form. But she does have a potion called the “drug of Prometheus,” which classicist David Hillman says contained: “In addition to viper venom, psychotropic mushrooms, and meadow saffron, Medea’s ios contained the famous ‘purple,’ or πορφυρα, a textile dye obtained from marine mollusks (murex).” She is a clear Eve figure.