I’d be fascinated to hear the Dope’s linguists’ takes on this one. Links:
Well I’m convinced.
I am not a linguist. But this is a controversial issue in the fields of linguistics.
Everyone agrees that languages developed from past languages and you can with enough evidence reconstruct these lost proto-languages. The controversy is over how much evidence you need to get meaningful results and how far back the process of reconstruction can be implemented.
The mainstream consensus seems to be that you can go back a few thousand years where you have examples of surviving descendant languages and written documents to work with. But if you go back much past this point, you’re just making things up.
The minority opinion is that you can take the proto-languages that were reconstructed by widely accepted means and apply those same means to the proto-languages themselves to reconstruct the proto-proto-languages the proto-languages evolved from.
IANAL, but the Western European languages I’m familiar with all have different words for those listed in the quoted part. Are those people claiming that the English versions come straight and unmodified from proto-Yuropian, or that they have somehow divined what was the original version and tracked it to different modern versions? What’s next, claiming that the King James Bible was dictated by God the Father and every other version in any language is wrong?
To a very very very very(*) crude approximation, one can imagine that the replacement of vocabulary items is similar to the disintegration of radium atoms. If the half-life of radium is 1600 years, that does NOT mean that all the radium will be gone in 3200 years! Instead 3/4 of it will be gone in 3200 years. After 9000 years 2% of the radium will still be intact.
A comment like “The traditional view is that words can’t survive for more than 8000 to 9000 years” shows ignorance of the statistical principle just described (via the example of radium half-life). If the “half-life” of a basic vocabulary item is 6000 years(*), then one would indeed expect 12.5% of that vocabulary to persist after 18,000 years.
Just as radium has a different half-life than uranium, so “basic vocabulary” items have a different half-life than less-basic, etc.
-
- I’ve written “very very very very crude” rather than “very crude” because some linguists view their science like plane geometry – comprising only theorems which are completely true or completely false. Instead there is very very very much fuzziness in any estimate of “vocabulary half-life” – so much fuzziness I won’t enumerate the very very many sources of fuzziness in this post. As just one example, isolated Icelandic changes very slowly, while Armenian, pressured by neighbors, mutated more quickly. Nevertheless very very crude estimates of the “half-life” of basic vocabulary can be made; 6000 years might not be a bad, though very very very crude, ballpark estimate.
Note that when comparing two languages with a common ancestor 6000 years in the past, only 25% cognates would be expected (very very crudely) since both languages have mutated.
Ruhlen writes on this topic. Agwa (water) and Dik (finger, one, point, hand, five, ten) are two of the ancient words found all over the world. Note that meaning mutates as well; the relationship between the different derived meanings of Dik should be clear. Several cognates of this ancient word are present in English: digit, indicate, decimal.
The article is confused on that issue. The linguists aren’t saying that the current words are the exact same words that were used 15,000 years ago. What they’re saying is the current words are the modern versions of 15,000 year old words.
And that is news how?
Half life or no, I’ll wager there’s a hard limit of between 50,000 and 200,000 years.
I read the French translation of his book The Origin of Language: Tracing the Evolution of the Mother Tongue a few months after it came out.
I can’t say that I was totally convinced by his theory. The fact that he claims that it is backed up by genetic studies is definitely intriguing but on the whole, while I found his approach fresh and interesting, I’m not sure that I accept the whole “disregard finding consistent rules for sound changes, if the words sort of look similar and if the meanings are kind of close, that’s good enough”.
And I was not too fond of his criticism of his colleagues. While I liked his idea to let the reader try to work out the relationships between different languages based on lists of words (that was fun), I found his comments really condescending: he writes several times that if you solve the exercises given, then you, a mere layman, are more capable than most professionnal linguists :rolleyes:.
The claim is that by studying the modern words, it’s possible to figure out what the original word was 15,000 years ago. Obviously, this is only possible if there’s a direct line between that ancient word and the modern one. If at some point, the original word was dropped and replaced by a new word, then the ancient word is now forever lost.
That’s another controversy. There’s an ongoing debate over when human beings started using language. One group says it was as recent as 50,000 years ago. Others say it was much further back.
It’s an important question because it would shed light on the role language had in the development of human beings. Anatomically modern humans arose around 200,000 years ago - a person of that era would be physically identical to a person of 2013. If we started speaking 200,000 years ago then that would be evidence that language was the key development in defining humanity. But if we didn’t start speaking until as recently as 50,000 years ago it would show that humans had developed before there was language and language was just another invention we came up with after we became human.
This study is all over the place, but they seem really reluctant to show the evidence. The sound files are cute, but since we’re dealing with literally hundreds of languages in these language families, how do I know they haven’t cherry-picked the data? Why are they using modern speakers for Indo-European when the proto-language is pretty securely reconstructed: do they disagree with the entire field of historical linguistics, or do they not understand it?
Do they have a model to account for whether this small fraction of vocabulary is inherited from a common ancestor, borrowed, or similar by coincidence? (For instance, do the “ultraconservative” words contain relatively common or rare phonemes?)
Pagel’s paper has 2 entries in the bad linguistics subreddit. I’m no linguistics whiz but it looks like a pretty pathetic paper to me.
Of course it is. It’s fine to hypthesize, but there’s no way at all to tell what language people were using before the first written words.
I read once an article he wrote for laymen. I found it pretty convincing. Later, I read an article written by a linguist disagreeing with his theory (and it’s my understanding that most linguists strongly disagree) and found it even more convincing.
Conclusion : laymen like me are easily fooled and aren’t able to make up their mind on such specialized issues. So, I’m going with the consensus and assuming that the wide majority of linguists are right and that he’s wrong (of course, that implies that I wouldn’t have believed in plate techtonics until the 60s, either, for instance)
IANAL and agree Ruhlen may be wrong on some matters, e.g. his view on Indo-European origin. However his irritation against his detractors may be understandable. The debate between the “lumpers” and “splitters” of linguistics becomes very shrill. Even I, a layman, can see that much of the debate passes beyond reason.
So for the last 15 thousand year I could give anyone on earth a “yo mamma’s a whore” with
“Hear thou, I to give thou old Mother ye Man Worm, to spit.”
? 
I am not surprised that a few words might have passed largely unchanged for 15,000 years. IANA Linguist so this is IMHO but, if you look at words like Father, Pater and Pati they are very similar, not really odd that they would share a common root.
I would think that there would need to be a reason to change a word completely or adopt a new one, otherwise wouldn’t they tend to just drift a little? By drift I mean the process by which you get regional dialects that slowly change to different languages, like the romance languages deriving from Latin.
As I said I have little knowledge on the subject but it does not seem crazy to me, the layman.
Capt
Look. If the title of the article had been “Linguists cannot identify any words that are unchanged over 15,000 years”, the Sydney Herald would not have published it. Sensationalist journalism has it tentacles even in the Science pages of the Media-Industrial Complex.
To be sure, there are linguists of the above bent – but the general public is not going to hear what they have to say about it.
Sure. But how could it be proven? Maybe Pater becomes Father, but that happened in a relatively short time, how could you carry that back 15,000 years and assume any word for ‘father’ would be recognizable today?