Hacker News new | ask | show | jobs
by falaki 462 days ago
This article is conflating language and ancestry. The seed of the confusion is in Reich’s research but the WSJ journalist blows it up to preposterous levels. Take India as an example. Most of the population is speaking some variant of an Indo-European (Indo-Iranian to be more precise) language but only a minority is genetically traced to Indo-European steppe people [1]

[1] https://www.science.org/content/article/where-did-india-s-pe...

5 comments

You also see this in places like Egypt. Nearly everyone speaks Arabic, but only a minority of their DNA is from the Arabian peninsula.
Which is not a difficult phenomenon to understand.

The most common ancestry in the US is German, not English, but English is still the dominant language. Language isn't DNA.

This is incorrect. The most common ancestry in the US is in fact English.

https://en.m.wikipedia.org/wiki/English_Americans

That’s an artifact of the 2020 census.

2012: 50.7 million Americans identified as German. 2022: 41 million https://en.wikipedia.org/wiki/German_Americans

I think that numbers relying on self-identification from census can also be far from the actual genetic makeup of the American population. But after some searching I couldn't find any comprehensive study that tried to trace today's Americans to different European ethnicities. And I'm not sure if this is possible anyways, given that Europeans were mixed in many ways.
German? Germany didn't even exist until 1871.
Germans are a distinct ethnolinguistic group that existed prior to the unified German nation state.

Many (most?) countries exist because pre-existing ethnolinguistic groups got their own country. For example, Bengalis have existed for long enough that you can easily identify them as a tightly clustered, distinct group in genetic profiles,[1] but never had their own country until 1971.

[1] https://www.gnxp.com/WordPress/2018/07/09/the-main-interesti...

This has often meant displacing huge numbers of people in order to create contiguous nation states. Some 15 million in the partition of India and a similar number in Eastern and Central Europe.
But it also means freedom and self determination for those who would otherwise be minorities within an agglomerated state.
The German Empire of 1871 is just one of many. Germans have lived in those lands for quite some time. Already Julius Caesar was conducting campaigns in Magna Germania.
that term "gemanic tribes" comes to mind for sure https://en.wikipedia.org/wiki/Germanic_peoples
The place existed, and people existed in the place, and the communication about that place is taking place today, and the way for one person to communicate a reference to that place to another person today, is to use the label Germany.
Before 1871, everyone knew what you meant when you said, "Germany." It just wasn't under a single government.
The “Holy Roman Empire of the German Nation” (its official name) would like a word.
In Florence I've seen an old map from around 1400 that had Germania in it.
Is East Francia (post-Verdun) better term?

   Language isn't DNA
but it's highly correlated. Most people in the US speak Germanic languages, ie english.
Given that English is itself germanic to an extent, yes. But it's also clearly got a lot of latin influence.

All of northwestern europe, of course, had quite a bit of DNA mixing over the centuries, so to what extent some DNA is particularly "German" or "English" largely depends on the time period.

What do you mean to an extent? It's definitely not Romantic.
While English is a Germanic language a huge number of words come from Old French since that was the language of the ruling elite following the Norman conquest of 1066 and continued to be used in administration for a couple of centuries.

Amusingly the Franks after whom French is named were also Germanic but they adopted the Vulgar Latin derived Old French then spoken in Northern France and which the Norse who invaded Northern France adopted before invading England.

https://www.reddit.com/r/asklinguistics/comments/324w60/how_...

English is a Germanic language...
Old english was a Germanic language. Modern English borrowed a little too much to really qualify.
Modern English didn't pop up out of nowhere, or relied on borrowing everything. Old English was a Germanic language, and while modern English borrowed quite a bit from other languages, it didn't start from scratch. It is still based upon Old English, which was Germanic. Even if you mean extremely modern, the base language is Old English, although the language evolves constantly. You would have a harder time comparing what is currently being spoken to Old English, but at the same time, you can't disconnect the two just because comparing the two now sounds entirely different. Language should evolve, since it's meant to communicate, not on it's own merit (as much as those who study language would like it to be). It's not crazy to think that in the future that a language could evolve even further to convey more meaning in a smaller amount of speech.
Modern English is still a Western Germanic language. If you don't speak English, it sounds remarkably like Dutch. The cadence, the intonation, the sounds - they are all distinctly Germanic.

Languages constantly borrow words, but there's a deeper foundation to the language (grammar, phonology, basic vocabulary) that remains.

yes, it has changed, but there's still far enough old german influence for it to be called something else.
This shouldn’t be downvoted. Except for colonizer languages, most languages in the world are coextensive with an ethnic group or closely related ethnic groups. Virtually everyone who speaks Bangla, Japanese, Korean, or Thai is ethnically Bengali, Japanese, Korean, or Thai.
That doesn’t mean that there wasn’t a lot of (often forced) assimilation or worse involved for that to happen.

e.g. it’s not like Japan didn’t have a “native population” that spoke a different language(s). The colonization just happened a few thousand years earlier than in the Americas.

> Except for colonizer languages

Which are spoken by billions of people around the world. That's a huge exception.

This is both pedantic and probably worth the correction. English is a Germanic language, originated in an exceedingly small continental territory nestled among other Germanics. Virtually no one would be able to discern the Germanic people who originated the English language from other Germanic people. If you are referring to the English people from the UK, then of course they are more mixed. But the English language was brought to the UK by the aforementioned continental tribe(s).
Yet ancestry.com can easily tell British with Anglo-Saxon and Brittonic ancestry apart from French with Frankish ancestry.
Nobody can. There's far too much overlap regionally between Britain and North-Western France.

And nobody in Britain has just Brittonic ancestry, or Frankish ancestry in France. For the most part the populations in Europe have been stable since a time that predates the expansions of the Celtic and Germanic linguistic groups.

Studies have found genetic evidence for population movements. You could make an argument for the Welsh having overwhelmingly Brittonic ancestry.

https://peopleofthebritishisles.web.ox.ac.uk/population-gene...

Can they? They clearly want you to think so, but my own personal results are pretty mixed on how accurate that is.
And yet the genetic differences are so insignificant so as to make them pointless to mention in the context of (paraphrasing) "it's super strange that German immigrants speak the Germanic English language that originates from the cultural region of Northern Germany".
I don't see what correction you're making.

The person you responded to explains that the most common ancestry in the US is German, but English is the dominant language.

You seem to be making the point that the most common ancestry in England is from England, but the Germanic language of English is dominant, rather than the Celtic one it replaced.

It's the same phenomenon, not a correction. That languages spread even when genes don't.

I'll be clearer and repost theirs.

>The most common ancestry in the US is German, not English, but English is still the dominant language. Language isn't DNA.

Their specific point is that "language isn't DNA". To support that argument they note that Continental German immigrants to America now speak English.

My specific point is that English originates not only within the language family of German speaking peoples, but that it originated with the DNA pools that comprise Northern Germany.

Therefore, I don't see how "the Germans are speaking English" makes the point that language isn't DNA.

I'm not saying that it is, at all. Its very obviously not. But the example being argued didn't make that point.

When they immigrated to the US, they spoke German. Now, almost all their descendants speak English.

Your objection is that English is a Germanic language. That's interesting, but irrelevant here. German and English are two different languages that are mutually unintelligible, and the fact that immigrants went from speaking one to speaking the other within a few generations illustrates my point that language does not have to (and often doesn't) follow genetics.

It's not only German immigrants to the US who now speak English. Pretty much all immigrants to the US speak English after 1-2 generations, regardless of what language they originally spoke.

> Therefore, I don't see how "the Germans are speaking English" makes the point that language isn't DNA.

Because German isn't English. Speaking one doesn't let you understand the other in conversation, not even a little bit. They're not like e.g. Spanish and Portuguese.

The fact that English is Germanic is a historical fact about where it came from, what it evolved from a millennium ago. But only a small percentage of Germanic influence remains in modern-day English vocabulary. It doesn't have anything to do with Germans in the US learning English.

>> This article is conflating language and ancestry

From the article:

"DNA detectives, including at Reich’s lab, analyzed DNA samples from the remains of around 450 prehistoric individuals taken from 100 sites in Europe, as well as data from 1,000 previously known ancient samples"

Ancestry, not language.

"Reich’s award-winning lab at Harvard has one of the largest ancient DNA databases in the world and uses proprietary gene-analysis software co-developed by Nicholas Patterson, a British mathematician who once worked as a codebreaker for U.K. intelligence services."

Ancestry, not language.

"DNA evidence shows that the proto-Yamnaya population migrated from the Volga region to Anatolia"

Ancestry, not language.

"In many places, indigenous male DNA disappears upon the arrival of the Yamnaya, while indigenous female DNA is traceable in the following generations"

Ancestry, not language.

"Within years of their arrival, some 99% of the indigenous people disappeared, according to Reich’s analysis of DNA samples from the time"

Ancestry, not language.

I rate your claim that "This article is conflating language and ancestry" as false, and I award you no points.

This article's confusion is where it states "half the human beings alive today are descended from the Yamnaya." He thinks because half of the world population speaks an Indo-European language, and because the original speakers of the Proto-Indo-European languages were the Yamnaya culture (as Reich's research suggests), then half of the world population are descendants of the Yamnaya culture.

Is the logical error clear now?

Archive version of the above science.org article "Where did India’s people come from? Massive genetic study reveals surprises Analysis confirms Iranian influx, but also finds genes from Neanderthals and a mysterious human ancestor": https://archive.is/Wd4tP
That article says nothing about the percentage genetic component of the Indo European step people in the Indian population. It does mention a high genetic similarity to Iranians.
And interestingly Iranians are mostly not the descendants of the so-called Indo-Iranian steppe nomads (genetically). But they speak various Iranian languages.
This research specifically incorporated DNA analysis. As is made clear if you actually read the article. I fail to see where the conflation happens.