PROTO-WORLD.

As a would-be historical linguist, I’ve always had an interest in the topic of the ultimate origin of language, and I’m very pleased to see that Piotr Gąsiorowski of Language Evolution (“How and why language varies and changes”) has started a series on it. He announces it here, and the first post is up today. The title is “Too Many to Communicate,” and that’s his basic point:

Whether the total number of humans was closer to 30,000 or to 300,000 is open to debate, but in any case they were far too many of them to constitute one speech community, especially if the Out-of-Africa migrants were already a separate sub-population somewhere in the Near East, the Arabian Peninsula, and possibly elsewhere in Eurasia and/or Australia (depending on the exact date of the bottleneck). It’s hard to imagine that the same language was spoken in Paleolithic Sub-Saharan Africa and South Asia, no matter how strongly the latter was affected by a demographic crash. No single language, then; at any rate not in anatomically modern humans. We have always been multilingual.

I’m very much looking forward to reading further posts on this fascinating subject.

Comments

  1. can I post a bleg?
    Can anyone identify this language?
    “Okusinziira ku nteekateeka, olutuuka, nga yeesoggerawo akafubo n’abakkulembeze mu kiwayi kya DP ky’akulembera oluvannyuma abanjulire ettu lya NAC.”
    Thanks if you can.

  2. It appears to be Luganda.

  3. Thank you! That was fast.
    I owe you for this one.

  4. J.W. Brewer says

    This all too speculative to be scientific, of course, but this math only tends to establish that it is improbable that a single language was spoken by all then-living proto-humans by the point in time when the entire species was actively using language. But if language was a genuine innovation (and this obv gets very chicken-and-egg in terms of how the necessary anatomical/neurological capacity might have developed before being used for this purpose . . .) it could have initially popped up in one small (and monolingual) band before diffusing outwards. Even if one assumes polygenesis, it seems wildly improbable for it to have been precisely simultaneous polygenesis? There is also of course the boring possibility that some sort of proto-not-quite-language phenomenon evolved into yes-this-counts-as-language-proper so slowly and continuously that if we had video/audio documentation of the relevant millenium we wouldn’t be able to identify the exact transition point, any more than we can point to a breakpoint where the children of Middle English speakers somehow grew up speaking Early Modern English.

  5. JWB:
    The question is not “Was humanity monolingual at some time?” but “Is there evidence that humanity was monolingual at some time?” The answer to the latter question is no, and since there is plenty of evidence of multilinguality, the Occam’s Razor assumption is that it traces back to the beginning.
    Diachronic monogenesis is a question entirely separate, however, from synchronic monolinguality. The languages spoken today might have a common ancestor without that common ancestor necessarily being the first language ever spoken; other languages might have coexisted with it that have no descendants today. Similarly, several thousand years from now when the most recent common ancestor (MRCA) is someone alive in 2013, it won’t be me, but it might be another Hattic (anyone with at least two genetic children).
    I think it’s very likely true that we couldn’t pin down a transition point from non-language to language precisely even if we had the ability to examine the entire past, for the reasons you give. Similarly, there is no point at which non-conscious parents had conscious children. All these things are much more a matter of degree that they appear to be: the recency of the MCRA (only 2000-5000 years ago) makes humanity look much more separated from the rest of the primates than was formerly the case.

  6. “this math only tends to establish that it is improbable that …”: people tend to misunderstand the role of mathematics.
    All that mathematical modelling reveals is the consequences of your assumptions. Change the assumptions and a model yields different results (except in Climate Science of course).

  7. I don’t know about Proto-World, but Upper Paleolithic expansion circa 40,000 BP was very rapid and likely accomplished by population from geographically small area – Levant. They could not have numbered more than 1000 people and probably spoke a single language.
    So here is your “Proto-World” or as close as it gets – ancestor of all European, Asian, North African and American languages except languages of equatorial belt – Australia, Papua New Guinea (and original languages of India and Southeast Asia – but they didn’t survive, so they don’t count)

  8. marie-lucie says

    They could not have numbered more than 1000 people
    Is 1000 people or animals a viable “breeding population”? especially if it is dispersing over a wide area?

  9. —Is 1000 people or animals a viable “breeding population”? especially if it is dispersing over a wide area?
    I think this concept only refers to isolated population which doesn’t interbreed with other populations.
    But Eurasia was not empty at the time of Upper Paleolithic expansion – Neanderthals, Denisovans, Flores hobbits and even anatomically modern humans in equatorial areas. Genetic evidence suggests that modern humans have interbreed with all of them.

  10. A promiscuous lot, aren’t we.

  11. @John Cowan: It’s meaningless to apply Occam’s Razor to the predictions of theories, rather than to the theories themselves. Absent some explanation of how/when/etc. language arose, there’s no way to judge whether we’re making more assumptions in guessing that there have always been many languages, or more assumptions in guessing that there haven’t.

  12. “The question is not “Was humanity monolingual at some time?” but “Is there evidence that humanity was monolingual at some time?”
    Case in point – I happened to look at Ethnologue today and they have obliterated Penutian into its constituent groups and language isolates. That’s not because the connections and simialrities aren’t so obvious Ray Charles could see them but because there is no way any more to determine the nature or the origins of those connections. It’s just been too many centuries of living side by side and the evidence is too muddled to be used as evidence.
    If some kind of wave came out of Asia and only French, English, Welsh and Irish were left, with only some apparently related substratum traits in the language around the Rhine delta, there would be no principled way to prove they were genetically related and formed a valid language family.

  13. Jim, I think there would be. What there wouldn’t be is enough evidence to show how they were related. As I’ve noted before, it isn’t possible to construct a relationship tree from the modern Germanic languages alone, though it’s obvious they are related; borrowings have crisscrossed the North Sea so much to completely muddle the evidence. The older forms of the languages, do provide sufficient evidence.

  14. marie-lucie says

    Jim: they have obliterated Penutian into its constituent groups and language isolates.
    Ethnologue has obviously followed the example of the Smithsonian Institution’s Languages volume (no. 17 of the planned 20-volume series “Handbook of North American Indians” – each volume a hefty tome). That volume was published in 1997 but most of it was written several years before. In the same year, Lyle Campbell (a notorious “splitter”) published his book American Indian Languages: The Historical Linguistics of Native America, which may have been an additional source for Ethnologue to use. But since then there have been new developments (not all of them publicized).
    That’s not because the connections and simialrities aren’t so obvious Ray Charles could see them but because there is no way any more to determine the nature or the origins of those connections. It’s just been too many centuries of living side by side and the evidence is too muddled to be used as evidence.
    Your negative-rich initial sentence here is a bit puzzling. Are you saying the connections and similarities are obvious? And is the “muddle” your conclusion, based on your own work, or are you summarizing “mainstream” opinion as stated in the above volumes?

  15. @marie-lucie
    You probably remember old languagehat thread MUSKOGEAN AND LAMB’S-QUARTERS, where Bill Poser abolished Hokan family out of existence in rather strong terms.
    This thread is now being referenced in academic publication – article by Mikhail Zhivlov (Institute of Linguistics, Russian Academy of Science / RSUH) called “The Hokan family and lexicostatistics”
    See page 3

  16. Thank you all for your kind interest in my blog! I just wanted to comment on the “Out-of-Africa” population estimates. It seems to be the mainstream opinion now that after some failed attempts the first really successful migration left Africa ca. 60 kya, and that despite its eventual success it was affected by a demographic crisis. The study cited in my post estimates the effective population during the bottleneck at 1,200, but in the Paleolithic realities it should probably be multiplied by a factor of ~10 to estimate the census population. I still suspect that even the effective population is underestimated, since some the most “basal” non-African lineages (like those of the Indigenous Australians or Andamanese islanders) were not included in the small — and hardly representative — sample of Eurasian genomes (one Chinese male, one Korean male, three Europeans). The African sample was still worse, if that’s a consolation (just two Yoruba males). Coalescent theory is a pretty exact science, but of course the quality the data matters a lot.

  17. This thread is now being referenced in academic publication
    I have not blogged in vain! But I wonder what Bill Poser would have to say about that article.

  18. marie-lucie says

    SFR, I had forgotten the thread on Muskogean, etc, and when I opened it and saw the date 2005 I thought it must have been from before I discovered languagehat. Reading through, I was surprised to see a comment with my signature, part of a later discussion involving yourself, Etienne, and others, less than a year ago!

  19. Hat reopened all those old threads for his blog anniversary, so there are many 2012 comments on much older articles.

  20. Well, not that many, but it was worth doing just for the Muskogean thread. (It was JC who noodged me into reopening them, so we have him to thank for it.)

  21. @m-l,
    “But since then there have been new developments (not all of them publicized).” Would you expand on this a bit? I recall an issue of IJAL in 1997 with several articles dedicated to the reassessment of Penutian. How much advancement or reassessment would you say has been since then?

  22. marie-lucie says

    Y, one of those 1997 articles is mine, still valid in general, and I have presented other papers since and continue to work on Penutian comparison, focusing on different aspects.
    The Wikipedia article on Penutian languages appears to have been partially rewritten more than once, and the latest version is not bad. The article used to present several different versions of the classification, without discussing their origin or their merits, resulting in great confusion for people not familiar with any of the languages. I don’t know who came up with the current “core” and “periphery” subclassification, and I don’t think it will be confirmed in the long run. Among other things in the article, I was surprised to see a partial reconstruction of consonants for some parts of the “core” done by Lyle Campbell, who used to be one of the most adamant disbelievers (and, as far as I know, has never done serious work on any of the relevant languages). Except for the most obvious correspondences (eg all m’s), I would take those supposed reconstructions with a large helping of salt.

  23. marie-lucie says

    p.s. A co-authored paper has been quoted in Mithun 1999 (The languages of native North America) and on Wikipedia.

Speak Your Mind

*