You don't say!

[Jamie D]

15,000-year-old 'fossil' words reveal ancestral Ice Age language

Would Ice Age man understand us? It may depend on the words we choose. Digging through languages in Eurasia for "fossil" words that have escaped erosion over time, researchers say they have identified an ancestral language that existed as far as 15,000 years ago.

This ancient language, described in Proceedings of the National Academy of Sciences, may have given rise to several different language groups — including Indo-European, which boasts roughly 3 billion speakers and contains such far-flung languages as Spanish and Hindi....

Luckily, some words — like numbers, pronouns and special adverbs that see frequent use — seem to have much longer half-lives of every 10,000 to 20,000 years.

They discovered a number of words — "this," "I," "give," "mother," "hand," "black," "ashes," "old," "man," "fire" — that cropped up in similar form across at least four of the seven language families studied across Eurasia. They traced them back to 15,000 years — right around the time the glaciers would have been melting, allowing humans greater ability to spread out over the globe and for languages to start to diverge.

So if you ever have the unusual opportunity to say this to someone from the Ice Age — "Black ashes? Who is this old man? Mother, I hear fire!" — there's a fair chance they'd get the gist of things.

[Anna++]

Oh!  I read about this the other day!  Then I got distracted learning that we're currently in an ice age...

[cisdad]

Some language fun.  Not sure how far I trust 15,000 year reconstructions of languages.  But, then again, I have a hard time with languages still being spoken.

'ice age' is fun.  By one definition, we came out of one about 15,000 years ago (and, in reading the language paper, that's probably not a coincidental date).  By another, we went in to one about 2.5 million years ago and are still in it (i.e., to have significant ice in the northern hemisphere -- Greenland).  And by another, we've been in an ice age for the last 25 million years or so (significant ice anywhere on the planet -- Antarctica glaciated about then).

[ZoeM]

This just makes me want them to try the same process with other languages - Aramaic, Sanskrit, French - and see how close we can get to the Mother Tongue.


I wonder what the ancient word for Computer was...

[cisdad]

Fezzika: A researcher's name you want to follow up is Cavalli-Sforza.  He's been doing a lot of work on language/genome diversity, connections, and migratory paths.  Plus at least one general reader book on the topic.

The genomics support largely 'out of africa' -- you can trace mitochondrial DNA (mtDNA, passed along only by mothers) back to Africa.  The near east is a center of divergence for those with ancestries from Eurasia.  Similar, but interestingly different in its details, is the Y chromosome divergence structure (passed only through the fathers).  Again, source for everybody is Africa, but for Eurasians, there's a center in the near east.

The (or at least A) connection between Asia and native North Americans has long been pretty confident.  Not as much Japan as Siberia (and, before that, Mongolia), but it's there.  Science or Nature within the last year or two had an article connecting North American languages with a Siberian language (some media attention to this as well), the first time it showed in language rather than just genes.  Seems very much a matter of a large group of people migrating in concert, though perhaps doing so in a series of coastal villages from east Asia all the way down to, say, Seattle.  Also, per linguistic and genetic analysis, seems to have been at least 3 waves of people 12-18,000 years ago.  Improving evidence that they were preceeded by others, but those others may have left no language or gene evidence -- just physical evidence.

Whee!  I like reading the science journals for this kind of thing.  (Not just what I work on, which is ... not languages or genes.)

[brainiac]

Historical linguistics is pretty darn cool. Languages evolve and change over time in predictable ways, which is how they were able to figure this stuff out when combined with archaeological and genetic information.

We can see it happening in English today, too. The subjunctive case is no longer functionally useful, so its use is beginning to fade away in modern speech. It is likely to eventually die out despite people's attempts to preserve it. Another example is the usage of past tense forms of verbs that have both an irregular and regular form (regular means it's +"ed" at the end): previously, the usage of both "dreamed" and "dreamt" was relatively common, but dreamed has increased in popularity while dreamt has lagged behind:



This is an ngram that "graphs the occurrence of dreamed and dreamt in English-language works published between 1800 and 2000", from this page.

(And I do work on languages, but this isn't really my area. )