Throughout Asia, they speak languages that are so strikingly similar it's uncanny.

So it seems obvious to connect the dots and see

if they belong to one and the same language family.

Yet who could've guessed that filling out that family tree would lead to what may just

be the most bitter dispute in the history of historical linguistics, where Altaic would

get labeled a linguistic religious cult?

So, uhm, ever heard language nerds fight?

Well you will today.

Let me describe a language to you, and you guess what it is.

It's spoken on the continent of Asia.

Its normal word order is subject, object, verb.

Its words have a stem and lots of endings built off to the right of that stem.

It has vowels that harmonize, postpositions instead of prepositions, no grammatical gender,

no consonant clusters at the start of words, and pronouns for "me" and "you" that start

with b- and s- or c-.

Did you say Turkish?

You're right, of course!

Thinking back to this video though, you might've guessed Mongolian, which is also correct.

Or maybe you're one of those next-level language enthusiasts and I reminded you more of a language

that proudly adorns the Forbidden City but is today rarely heard as a first language:

Manchu.

Three languages from three different families.

Turkish is Turkic.

Mongolian is Mongolic.

Manchu is Tungusic.

There are other languages in each of these families.

Yet they share such similar traits that

combined with their shared history of the horse, the bow, the Steppe, the Silk Road,

centuries ago, scholars started to wonder if they could link them into one large "superfamily".

I'll spare us some of dates and people involved.

Much is written about the origins of this idea.

Our story starts with what they hypothesized: a family called Altaic, named after the

Altai Mountains.

These Altai mountains where four countries meet: Kazakhstan, Russia, China, Mongolia,

the epicenter of a new, or really old, Turkic-Tungusic-Mongolic superfamily.

Even more languages were about to be added beyond the inland steppes thanks to a simple

observation made in Japan in 1781.

Writer Tō Teikan (藤貞幹), while spilling historical opinions onto the page, at one

point has this aha moment for as far as I know the first time in history: could Japanese

be related to Korean?

The idea seemed to fade with the author's ink, but a century later it was picked up

off the shelf and mounted enough support in Japan that by the mid 1900s, one linguist

was practically alone standing against the idea.

Now in my notes I keep abbreviating this duo as KJ or JK, but it's no joke

and it's not simple.

Sidestories for another day, but, speaking broadly, "Korean" and "Japanese" are actually

clusters of Koreanic and Japonic dialects and languages.

These clusters do have something in common.

They're traditionally considered isolates, lone language families unrelated to anyone

else in the world.

Unless... well, let's go back to my guessing game.

Altaic defenders saw K and J ticking many of the same boxes.

And with that, they looped them into their hypothesis.

And now, every Japonic, every Koreanic, Tungusic, Mongolic and Turkic language was part of Altaic.

By the second half of the 20th century this big concept of Altaic was gaining steam and

winning converts.

As it did, Moscow emerged as the epicenter of this work.

Moscow's Altaicists were faithful but not unquestioning believers.

They spotted a possible issue, one we've also met here before.

Remember that shared history?

Languages living side by side for a long time can end up sharing structure and grammar even

if they don't share a common ancestor, and if they don't share a common ancestor, they're

not part of the same family tree.

Altaicists knew that their languages had this kind of contact throughout history, so here

was their strategy: shun grammar, seize on vocabulary.

They turned their focus increasingly to comparing words and sounds.

In their eyes, they were undertaking exactly the kind of comparative thinking that formed

the bedrock of historical linguistics and had already helped establish language families

like Indo-European: compare words across various languages, weed out borrowed vocabulary to

find possible inherited cognates, establish regular sound correspondences

between the languages.

Easy, right?

That done, they could then reconstruct what they saw as the original shared ancestral

Altaic form of a word.

Everywhere they searched, they found clearly related words and "embarrassing parallels"

leading them to real breakthroughs in their large Altaic project.

The wider historical linguistics community, though, they were... uh, less impressed.

Some brushed Altaic aside.

Others became openly critical.

Let the spats begin.

Too much of Altaic looked to them just like borrowed words and grammar.

By borrowed, I mean they didn't see Altaic words as coming down the ages from a proto-parent.

Instead, they were pretty sure these words had transfered sideways from one unrelated

language to another.

The influence of widespread families like Mongolic and Turkic in particular was strong

across the steppes.

Similar words and similar features actually converged in the present and diverged in history

rather than converging on a common past.

And instead of predictable sound correspondences, historical linguists saw haphazard phonemes

and one-off explanations.

They were increasingly skeptical.

As criticisms mounted, Altaicists kept responding.

They demanded that if experts didn't want to be labeled hypocritical they should apply

the same level of scrutiny to their own families like Indo-European.

Moscow's collection of cognates, words with apparent family resemblances, that keeps

growing and growing.

And by the early 2000s three of the best-known Altaic proponents feel like it's time to publish

their own reconstructions in this, the massive Etymological Dictionary of Altaic Languages.

It might've seemed like a definining moment for Altaic.

Instead, it was fuel for even greater controversy.

You can already sense it brewing in the book's very first chapter, an odd start for a dictionary.

It tries to get out ahead of the competition with responses to critics.

The words herein aren't borrowed between different families, I swear they're really inherited

in one single family.

The critics were unamused and were ready with even more criticism.

Things were about to turn ugly.

In an unexpected twist, one of Altaic's proponents, whose work was cited in this dictionary's

bibliography, turned on the theory.

Vovin charged that this dictionary merely showcased how the whole Altaic project had

been a failure.

Ooh, the betrayal!

To him, something was fishy.

On page 1076 we meet the reconstructed Proto-Altaic pā́li, a kind of fish.

Somehow supposedly this root found its way to Korea and Japan even though Middle Korean

has no ending like this and the Japanese word "clearly" just means flat fish.

To Vovin, this sounded less like skillful etymology and more like summoning creatures

to life through make-believe word history.

And so he lashed out.

The dictionary still does do enough to account for borrowed words, especially between Turkic

and Mongolic.

You're trying to make those lateral moves look like parent-child relatedness.

It doesn't stop with borrowings; you're boldly stretching sounds and meanings until utterly

unrelated words looked related.

Some of your etymologies draw on only two families or even one.

When there were perfectly good internal explanations for the origin of a word, your dictionary

wishes them away as "folk etymologies" and replaces them with concocted Proto-Altaic

ancestors.

Worst of all, the sound correspondences seem random, not predictable like truly

related languages.

But he saved his most scathing judgments for the Moscow Altaicists themselves.

They are in the end little more than the kind of "Proto-Worlders" we met back in my Tower

of Babel animation, long-rangers doing word-list linguistics fishing for resemblances where

a trained mind sees differences and complex histories behind words and texts.

In his eyes, they had ignorantly resorted to a "prescientific" method and were indoctrinating

followers into "a set of beliefs" that had basically become a "religion".

This dictionary amounted to a holy book for their "true believers".

Gauntlet thrown.

Whew.

You know those times when a response requires more words than the

criticism you're responding to?

Well, Moscow felt like this was one of those times.

Two of the dictionary's authors declared Altaic undead, calling for an end to the controversy

over the controversy.

They sum up their sentiments with this Kazakh proverb, they dub their onetime ally turned

opponent AV and give us the sense they'll be taking the high road.

Of course, they could stoop to AV's level and say AV just "switched from one church

to another" and brand his article a "propaganda piece".

But no no, mostly they'lll stand by their dictionary and defend the etymologies therein.

Your hypocritically rigorous expectations would rule out even well-established families

like Indo-European.

Long-rangers are routinely misunderstood anyway, and even if we were forced to throw away

those fifty entries with flat fish folk etymologies, the foundations of Altaic would

still stand strong.

Stronger than ever, I say!

AV was unphased and would keep joining the charge to topple more and more of the "startling"

parallels attributed to Altaic.

The shared words, the pronouns, the grammatical features could all be explained

common ancestry.

AV dared Altaicists to come up with real proposals that would pass even basic historical linguistics

tests.

Six words, just give me six words with the same meaning and completely predictable sound

correspondences between these languages...

c-come on, I dare you!

AV may be unique in his bluster, but today he is in good company.

Time has not been kind to Altaic.

For "traditional" linguists, the current was already flowing downstream.

The family is now commonly labeled "discredited" or even "pseudolinguistic", a has-been.

There are people who still do Altaic.

In the hunch of conventional linguists, their advocacy often extends into other

"fringe" megafamilies.

On the other hand, even among non-Altaicists, there are calls for moving peacefully beyond

devotion or denunciation to a common ground of understanding the languages together.

But outside of dedicated Altaicists, who continue to feel misunderstood, linguists see in these

resemblances a handful of unrelated families with intertwined history and geography and

filled with ancient linguistic tales we're barely starting to recover.

Stick around and subscribe for language.