17 April 2011

Origin of Human Language?

Filed under: Culture, Language, Psychology — ktismatics @ 1:46 pm

I’ve been thinking about the rationale behind the recently-released study by Quentin Atkinson tracing the original human language to southwest Aftrica. Atkinson predicated his work on the “serial founder effect” in biological ecology. When a small group splits off from a larger population of a species, that group carries with it only a fraction of the genetic diversity present in the entire species. When this fragmentation occurs frequently, relatively homogeneous divergent subpopulations of the same species arise. When the fragmentation occurs serially, then the diversity of each newly-budded subgroup is lower than that of its predecessor.

It’s known that the human species evolved in Africa and spread from there, and that the serial founder effect reduced the genetic diversity within human subpopulations branching off during this spread. Atkinson contends that the serial founder effect works not just with human genetics but with human language as well. Rather than looking at vocabulary or grammatical elements, Atkinson studied phonemes, which are the smallest units of sound used in building meaningful utterances (e.g., in English the “k” sound is a phoneme). It turns out that the languages of southwest Africa use the largest number of phonemes (including various clicking noises). Atkinson presents evidence demonstrating that the longer it took for any given subpopulation of humans to migrate from Southwest Africa, the smaller the number of phonemes there are in that subpopulation’s language.

Let’s assume that Atkinson’s findings stand up to subsequent empirical scrutiny. The question is why the serial founder effect would work for languages. I can see why a small group of migrants might between them use only a fraction of the vocabulary of their native tongue. Even so, wouldn’t they have occasion to use most if not all of the distinct sounds of their language? English has something like 45 phonemes. I suspect that even a 3-year-old knows enough words to use all of those phonemes. Why would even a very small subset of language-users drop start dropping their phonemic diversity? And how in the world did the Hawaiians, way out on the farthest frontier of the serial-founder dispersion from Africa, manage to lose all but 13 phonemes? Someone else reviewing Atkinson’s work cited Dave Barry’s hypothesis:

The Hawaiian language is quite unusual because when the original Polynesians came in their canoes, most of their consonants were washed overboard in a storm, and they arrived here with almost nothing but vowels. All the streets have names like Kal’ia’iou’amaa’aaa’eiou, and many street signs spontaneously generate new syllables during the night.

I came across a study suggesting two possibilities accounting for this phonemic erosion in subpopulations. Maybe the leaders of these splinter groups are so powerful that, if they get sloppy with their phoneme use, no one around them is likely to risk correcting them. Instead, the followers adopt the regressed linguistic practices of the leaders. Alternatively, maybe splinter groups are already so cohesive that they understand one another without having to enunciate their meanings and intentions clearly. As a consequence the language within the entire subgroup deteriorates. Both of these ideas presume that less complex social structures demand less complex language use, a proposition that is supported by another line of research showing that more widely-spoken languages tend to use more phonemes than languages spoken by smaller numbers of people.

If that’s the case, then why do the southwest African languages use twice as many phonemes as English, when English is clearly a more widely-spoken language? Maybe it’s because English is still a relatively new language. Maybe it took tens of thousands of years for the African languages to achieve their level of phonemic complexity. The phonemic deterioration resulting from the serial founder effect must be rapid, and the recovery slow.

Is this argument persuasive? Not to me it isn’t, though I might be missing something. Is the empirical evidence persuasive? I suppose I should actually read the original article in Science, but I’m sure this one study won’t stand without challenge in the field.



  1. Interesting those remarks about the Hawaiian language being even more ‘vowelized’ than the Polynesians’ language who came there. Heyerdahl was refuted long ago, of course, about Chile, etc., but the Tahitian language is itself very vowelish. I’ve got a couple of ancient LP’s I found at an oldie store about 2001-2 that have the Tahitian written out so you can see it.

    It’s one of those weird ironies that Hawaii is so much more overdeveloped for tourism than is even Tahiti (the second most desired destination in the Pacific, it still has only 3% of the tourists Hawaii has), but that there are only the rarest corners of any of the islands of Hawaii where the language is still spoken. Tahitian is spoken all over French Polynesia, including by many French; and all the Tahitians speak French too, of course. But you also see many more Polynesians of pure or mixed blood, from what I’ve read (not having been to Hawaii.) They are literally everywhere, and alarmist things I read as much as 40 years ago are very misleading. Even on the main island, if you go to ‘le district’, which is about 75 miles to the south of the capital, Papeete, down to the smaller half of the hourglass (the larger is Tahiti-nui, the smaller Tahiti-iti), you find it almost all Polynesians. And even with French still usually involved with administration, the Polynesians still own some 85% of the land. Tell me how much the ‘pure Hawaiians’ own. I bet not even .01%. This is that kind of thing that always fascinates me for some reason–like the way at least a tiny few fragments of ancient Greek music exist, but not a single one of later Roman music has been found.

    I realize this isn’t the main topic of the post, which I’ve barely skimmed, but I thought it was interesting that Hawaiian could be even more VOWEL.


    Comment by Illegal Dances of New York City — 17 April 2011 @ 2:29 pm

    • I believe I recall counting SIX consonants even being missing from the Tahitian on that LP.


      Comment by Illegal Dances of New York City — 17 April 2011 @ 2:32 pm

  2. One implication of your observation is that ownership of the land protects ownership of the language. It’s like species that go extinct because of loss of habitat. Certainly that happened with the American Indian languages. I can’t remember the numbers exactly, but of the something like 6000 remaining languages spoken on earth, half are expected to go extinct by the end of the century.


    Comment by ktismatics — 17 April 2011 @ 2:40 pm

  3. “half are expected to go extinct by the end of the century.”


    Another irony is that the Marquesas, the least populous islands of French Polynesia and the furthest away from continental land mass of any major (is)land areas in the world also have a language of their own, which the Tahitians don’t speak, and often don’t understand. But they understand Tahitian, and most probably speak French too (but it’s possible some of these don’t even. But the strangeness comes in because European colonization in the 18th and 19th century wiped out about 95% of the Marquesan population, through the introduction of new diseases and grog shops. However, once France ‘stole back’ the Society Islands (Tahiti, etc.) and the Marquesas Islands and the Tuamotu Islands, the Marquesas bottomed out and began to gradually regain population. Which is not to praise the French over the English in all cases with their less-obtrusive colonization: the nuclear testing in the Tuamotus was grotesque and only ended about 1997, or sometime during Chirac’s tenure; and the most outrageous thing that ever occurred here in this category was once when DeGaulle was visiting back in the 60s, and the winds were blowing the wrong way a second day. He said ‘I’m a busy man’, and they went ahead with the blast. And it blew right straight onto the main island of Tahiti, including the capital. So sometimes there are disasters that are recovered from, albeit gradually, sometimes there just seem to one after the other, and you could be right that in non-Polynesian ownership of most of Hawaii, they did lose their language. When I was in Tahiti the second time, they had even successfully elected a Tahitian president for the first time in history, but he was thrown out after 4 months, and replaced by a French one.


    Comment by illegal dances of new york city — 17 April 2011 @ 3:24 pm

    • So that it’s hard to know what ‘ended up for the best’. I believe it was in the 1850’s that France managed to pull this part of Polynesia from the English, and they’ve always rather loved it (who wouldn’t?) New Zealand is the Maori, and a related kind of Polynesian, but I always tend to think of it as ‘very English’, more than I remember that they have a huge Polynesian population. The weirdest example of all is Pitcairn, very central in the Bounty Mutiny, with the descendants of Fletcher Christian and Captain Bligh–and an absolute nightmare of incest and extreme ill health and ignorance. Only about 60 inhabitants, and in the early 90s, and English writer managed to infiltrate the place (they wouldn’t allow tourists there) and wrote a book about it. Not good style of writing, but amazing information. The Polynesians were run out of Pitcairn long ago, although there are still sculptures, carvings, paintings, etc., they were administered by England for a long time, maybe still are, or by Aus. or NZ. Not one Pitcairner has any of the teeth (s)he was born with. Whereas the Marquesans still have their native culture mostly intact, a few of the most paradisical islands don’t even have televisions and no boat or air service except once a month; and all the medicines they’d need from Tahiti, and the benefits of French organization when they need it. I met several Marquesans, and they often come to Tahiti to do business.


      Comment by illegal dances of new york city — 17 April 2011 @ 3:34 pm

  4. Here’s John Hawks, a paleoanthropologist, expressing skepticism about using the serial founder effect to track back to the original language:

    “Why should the origin of languages have had the largest inventory of phonemes? If small populations typically lose phonemic variation, why would sparse hunter-gatherer populations of Africa have built up the largest store of sounds just as they were getting started talking? Atkinson suggests that African populations have had more time to recover diversity after a bottleneck at the origin of language. That seems an inauspicious suggestion, considering that the genetic model of a founding bottleneck in Africa has taken some serious body blows this year.”

    In this post Hawks remarks offhandedly that the serial founder effect has also come under increasing criticism within evolutionary biology, which in his view makes Atkinson’s extension of the theory to cultural evolution that much more problematic. To be clear, I harbor no objections to the possibility of southwest Africa being the place where humans first started using language, or even that spoken language would have been invented only once in the history of the species. To the contrary: language would probably have been crucial for humans to achieve widespread geographical distribution over the course of only a few thousand years. Of course I’m no expert in this field; I’m just skeptical about the theory and methodology deployed in this particular study.


    Comment by ktismatics — 18 April 2011 @ 12:13 pm

  5. Here’s a guy who proposes that the number of phonemes in a language is related to the prevalence of adults who must learn that language. Little kids acquire their native language almost effortlessly; adults, in contrast, find it much harder to learn a new language. One way to simplify the adult learning task is to simplify the language, and one way to achieve simplification is to reduce the number of phonemes. In what circumstances would there be large numbers of adults trying to learn a language? I’d think it would be in situations where one culture that already speaks a language finds it advantageous to learn the language of a second culture, and/or where the second culture finds it advantageous to teach its language to adults from the first culture. Examples might include cross-cultural trade relations or conquest or intermarriage.

    The important point is that phoneme reduction could imply the coexistence of two already-existing languages, rather than the serial founder effect’s model of an exploratory subgroup’s streamlining of a single language. It’s an alternative hypothesis that Atkinson apparently did not consider in evaluating his data. This alternative hypothesis strikes me as more plausible than the serial founder effect. That doesn’t mean it’s true, of course.


    Comment by ktismatics — 19 April 2011 @ 1:40 pm

  6. My hypothesis is that when two tribes merge ( crossing over ) some sort of standardization happens.

    1. The phoneme sets are unified first. 
    2. Among those with a low phonetic distance, one of them might be eliminated ( r vs l in Japanese for example ). 
    3. A phoneme might also be eliminated according to the "adults learning" hypothesis you mentioned above. 

    This is still nostalgic in the sense that an initial fond only ever decreases and only loss is experienced. Fixing initial conditions ( an origin ) may have this effect. In contrast one can add mutations to the big picture. The mentioned unification/reduction can also act as a dialectic force: instead of choosing one or the other a third one bridging them will be selected.


    Comment by Kay — 19 April 2011 @ 10:14 pm

  7. These too are plausible alternative hypotheses, and they could conceivably be subjected to empirical evaluation. Another amateur blogger observes that there are significant geographical clusterings of phoneme distribution that can’t be attributed to historical time since leaving Africa. He claims, for example, that tropical languages tend to use more different vowel sounds, which he suggests might be attributable to humans imitating the wider diversity of bird calls audible in tropical climates. That seems far-fetched, but it might have some effect. Someone else proposed that cultures trying to establish their difference from (and superiority to) neighboring cultures would start incorporating phonemes into their languages that differ from neighboring languages.

    In Atkinson’s regression model, the raw number of phonemes accounts for 20-30% of the variance in languages’ geographic distance from southwest Africa. That is a strong result, so there’s probably something to it, even if the serial founder effect proves not to be the best explanation. Still, 70-80% of the variance remains unexplained. In all likelihood future studies will explore multivariate predictor models building on and refining Atkinson’s work.

    Particularly startling would be empirical evidence supporting the origin of language somewhere other than Africa, or even multiple origins. Written language was invented maybe three discrete times in human history, but I think virtually everyone assumes that language use is integral to humanity’s distinctiveness as a species. I assume it too, even though it’s clear that language is a cultural artifact and not a biological inheritance — hence the multiplicity of languages that have proliferated in the world, languages that are utterly incomprehensible to most humans.


    Comment by ktismatics — 20 April 2011 @ 9:26 am

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

Create a free website or blog at WordPress.com.

%d bloggers like this: