Max Muller was one of the greatest religious scholars of the 19th century. Born in Germany, he became fascinated with Eastern religion, and moved to England to be closer to the center of Indian scholarship in Europe. There he mastered English and Sanskrit alike to come out with the first English translation of the Rig Veda, the holiest book of Hinduism.

One of Muller's most controversial projects was his attempt to interpret all pagan mythologies as linked to one another, deriving from a common ur-mythology and ultimately from the celestial cycle. His tools were exhaustive knowledge of the myths of all European cultures combined with a belief in the interpretive power of linguistics.

What the significance of Orpheus' descent into the underworld to reclaim his wife's soul? The sun sets beneath the Earth each evening, and returns with renewed brightness. Why does Apollo love Daphne? Daphne is cognate with Sanskrit Dahana, the maiden of the dawn. The death of Hercules? It occurs after he's completed twelve labors (cf. twelve signs of zodiac) when he's travelling west (like the sun), he is killed by Deianeira (compare Sanskrit dasya-nari, a demon of darkness) and his body is cremated (fire = the sun).  His followers extended the method to Jesus - who was clearly based on a lunar deity, since he spent three days dead and then returned to life, just as the new moon goes dark for three days and then reappears.

Muller's work was massively influential during his time, and many 19th century mythographers tried to critique his paradigm and poke holes in it. Some accused him of trying to destroy the mystery of religion, and others accused him of shoddy scholarship.

R.F. Littledale, an Anglican apologist, took a completely different route. He claimed that there was, in fact, no such person as Professor Max Muller, holder of the Taylorian Chair in Modern European Languages. All these stories about "Max Muller" were nothing but a thinly disguised solar myth.

Littledale begins his argument by noting Muller's heritage. He was supposedly born in Germany, only to travel to England when he came of age. This looks suspiciously like the classic Journey of the Sun, which is born in the east but travels to the west. Muller's origin in Germany is a clear reference to Germanus Apollo, one of the old appelations of the Greek sun god.

His Christian name must be related to Latin "maximus" or Sanskrit "maha", meaning great, a suitable description of the King of Gods, and his surname is cognate with Mjolnir, the mighty hammer of the sky god Thor. His claim to fame is bringing the ancient wisdom of the East to the people of the West - that is, illuminating them with eastern light.

Muller teaches at Oxford for the same reason that Genesis describes the sky as "the waters above" and the Egyptians gave Ra a solar barge: ancient people interpreted the sky as a river, and the sun as crossing that river upon his chariot (perhaps an ox-drawn chariot, fording the river?). His chair at Oxford is the throne of the sky, his status as Taylorian Professor because "he cuts away with his glittering shears the ragged edges of cloud; he allows the...cuttings from his workshop, to descend in fertilizing showers upon the earth."

I could go on; instead I recommend you read the original essay. The take-home lesson is that any technique powerful enough to prove that Hercules is a solar myth is also powerful enough to prove that anyone is a solar myth. Muller lacked the strength of a rationalist: the ability to be more confused by fiction than by reality. This makes the Hercules theory useless, but that is not immediately apparent on a first or even a second reading of Muller's work. When reading Muller's work, the primary impression one gets is "Wow, this man has gathered a lot of supporting evidence."

This is a problem encountered in many fields of scholarship, especially "comparative" anything. In comparative linguistics, for example, it's usually possible to make a case that two languages are related good enough to convince a layman, no matter which two languages or how distant they may be. In comparative religion, we get cases like this blog's recent discussion over the possible derivation of Esther and Mordechai defeating Haman from Ishtar and Marduk defeating Humbaba. The less said about comparative literature, the better, although I can't help but quote humor writer Dave Barry:

Suppose you are studying Moby-Dick. Anybody with any common sense would say that Moby-Dick is a big white whale, since the characters in the book refer to it as a big white whale roughly eleven thousand times. So in your paper, you say Moby-Dick is actually the Republic of Ireland. Your professor, who is sick to death of reading papers and never liked Moby-Dick anyway, will think you are enormously creative. If you can regularly come up with lunatic interpretations of simple stories, you should major in English.

The worst (but most fun to read!) are in pseudoscience, where plausible sounding comparisons can prove almost anything. Did you know the Mayans believed in a lost homeland called Atzlan, the Indonesians believed in a lost island called Atala, and the Greeks believed in a lost continent called Atlantis? Likewise, did you know that Nostradamus predicted a great battle involving Germany and "Hister", which sounds almost like "Hitler"?

Yet it would be a mistake to reject all such comparisons. In fact, I have thus far been enormously unfair to Professor Muller, whose work established several correspondences still viewed as valid today. Virtually all modern mythologists accept that the Hindu Varuna is the Greek Uranus, and that the Greek sky god Zeus equals the Hindu sky god Dyaus Pita and the Roman Jupiter (compare to Latin deus pater, meaning God the Father). Likewise, comparative linguists are quite certain that all modern European languages and Sanskrit derive from a common Indo-European root, and in my opinion even the Nostratic project - an ambitious attempt to link Semitic, Indo-European, Uralic. and a bunch of other languages - is at least worth consideration.

We need a test to distinguish between true and false correspondences. But the standard method, making and testing predictions, is useless here. A good mythologist already knows the stories of Varuna and Uranus. The chances of discovering a new fact that either confirms or overturns the Varuna-Uranus correspondence is not even worth considering.

Mark Rosenfelder has an excellent article on chance resemblances between languages which offers a semi-formal model for spotting dubious comparisons. But such precision may not be possible when comparing two deities.

I have what might be a general strategy for approaching this sort of problem, which I will present tomorrow. But how would you go about it?

20 comments, sorted by
magical algorithm
Highlighting new comments since Today at 4:43 AM
Select new highlight date
Moderation Guidelinesexpand_more

It's worth noting that even if a correlation is untestable in itself, it can still lead to other predictions which ARE testable.

If the Greeks, Indonesians, and Mayans all believed in a lost island, you might start looking for a lost island somewhere. If you believe Nostrodamus predicted World War II, you might expect some of his other predictions to come true in the future. If Varuna and Uranus are the same god, that must imply something about the Greeks and Hindus. The correlation doesn't just exist in a vacuum by itself. At some point it will lead to something testable in the real world.

A correlation is a starting point for knowledge, not knowledge in itself. If it leads to a testable prediction, it's helped you increase your knowledge. If it doesn't lead to any sort of prediction, you don't have any more knowledge, you just have a story.

I published two academic articles showing the relationship between game theory and Greek Mythology. A somewhat objective test of the validity of this relationship is whether Greek mythology can be used to help students achieve a better (as measured by tests) understanding of game theory.

Perhaps a somewhat useful measure of the relationship between two languages is how much learning one helps you master the other.

If it's not a copyright issue/too much trouble, I'd like to see those articles. I've messaged you my email address.

Perhaps a somewhat useful measure of the relationship between two languages is how much learning one helps you master the other.

Sometimes languages end up influencing each other a lot even though they don't share a particularly recent common ancestor.

You can prove ANYTHING if you can ignore disconfirming evidence.

It reminds me of the point in decision making - almost all of our choices are over-determined; that is, there are always good reasons to do almost anything we may want to; to make a good decision we also have to balance the reasons not to choose particular things.

And as to your title question: I am SUPERIOR to a solar deity - at least I actually exist.

Your criticism of comparative linguistics is unfair, because the goal of comparative linguistics is not to determine whether two languages are related, but to measure a degree of relatedness. Comparative linguists try to find a set of transformations (from a list of transformation templates which have been observed to occur) which go from one language's phonology to the other's, where the size of that set is inversely proportional to relatedness, and a percentage of words in common. From these measures, linguists can make falsifiable predictions about how long societies have been isolated from each other, when works were written, and so on.

Okay. I don't know much about linguistics, so I accept your correction.

Would you call the people who are trying to determine whether Nostratic and Amerind are real families, or whether Basque is related to Caucasian languages, and those sorts of things "comparative linguists"? If not, what would you call them?

Once I was trying to invent a character name based on indo-european roots, I kept finding words like cruel, Kaiser, and Caesar coming from IE roots meaning things like 'kill', 'judge' choose', and 'law' I don't have my notes and I don't know the exact roots.

I am not a linguist, so maybe the transforms of the roots I was using wouldn't be likely developments in natural usage, but the meanings fir pretty well. I remember thinking 'This must be what pseudoscience is like, I could be a crackpot with this!'

Isn't this the same problem as overfitting? I think there are techniques against overfitting, even a contest, the Performance Prediction Challenge

My grasp of statistics is weak, but I am missing the connection to overfitting. I always thought overfitting to mean something like interpreting a series of coin flips producing HHTHT to mean the coin was fixed to produce HHTHT. Two characteristics of overfitting, then, are that you can only overfit the same theory to one piece of data, and that getting more data will ruin the overfitting.

But here, you can apply the "sun god" concept to practically any myth you want (if you're a good enough arguer). And I imagine that if someone discovered new information about Hercules, Max Muller would be able to think up a reason why it supported the sun god theory.

We should be careful with Littledale's reductio. Muller's methods obviously have no validity outside of the domain of mythology - his name certainly has little mythological import. However, Muller's methods may be perfectly sound within some domain, and that may be the domain he was working within.

I'm not taking sides on Muller's work, just expressing a meta-level concern.

Develop general algorithms for assessing similarity, apply them to a dataset where the actual historical relationships are known, and then see if they generalize to the rest of the verified datasets. If so, adjusting for multiple hypothesis testing, use those algorithms on the uncertain cases. The key thing is to evaluate the algorithms based on their performance if applied universally, not just in one cherry-picked case.

Moby-Dick is actually the Republic of Ireland

Too bad it was published 65 years before the Proclamation of the Republic. ;-)

all modern European languages ... derive from a common Indo-European root

Except Hungarian and a few others.

Too bad it was published 65 years before the Proclamation of the Republic. ;-)

The author is dead, you see.

Word of warning: I have had a couple glasses of Firefly vodka mixed with lemonade. The everloving devil's brew, I tell you.

Yvain is most likely smarter than me and has the additional bonus of caring intensely about subjects I dabble in. However, he always delivers on the entertainment. That eardrop thing was ludicrously fun.

it seems to me that "a general strategy for approaching this sort of problem" has the same pitfalls as always trusting conventional wisdom over contrarian wisdom.

I'm glad you find my articles entertaining, but I don't see how you can say a general strategy has certain pitfalls without knowing what the strategy is. Or do you think just trying to approach this problem from a rationalist perspective is a mistake?

"In comparative linguistics, for example, it's usually possible to make a case that two languages are related good enough to convince a layman, no matter which two languages or how distant they may be."

The real question is why we think that experts in the field actually know more than laymen. If a shoddy argument has convinced the authorities, why don't we simply conclude that our argument standards are wrong?

The expert should at least know about few examples of incorrect theories based on superficial similarities between languages. (S)he should be also informed about the number of coincidences arising from pure chance when two totally unrelated languages are compared. And (s)he knows miscellaneous facts which provide additional checks. E.g. you wouldn't believe a claim that English is ten thousand years old if you know something about the rate of language evolution or the history of germanic and indo-european languages.

To suppose that the experts don't know more than laymen seems weird. It may be true for some pseudo-scientific disciplines where is nothing to be actually known, but that's not the case of linguistics. I would yet admit that the experts are probably not less biased than laymen since the knowledge of biases is not usually taught.