African ur-language reconsidered

By Razib Khan | April 16, 2011 3:22 pm

Mark Liberman at Language Log has looked through the Science paper Phonemic Diversity Supports a Serial Founder Effect Model of Language Expansion from Africa. Overall he seems to think it is an interesting paper, but he has some pointed criticisms. Here’s the utility of the post: Liberman uses analogies to domains (e.g., genomics) which are comprehensible to me. My main issue with linguistic evolution is that I’m so ignorant that I barely understand the features being discussed. I may know their dictionary.com definition, but I have pretty much no deep comprehension with which to test the inferences against. By analogy, imagine trying to evaluate a morphological cladistic model with no understanding of anatomy. Here’s the part which may be of particular interest to readers of this weblog:

However, this combination of coarse binning into ranges, for functionally-defined subsets of elements with radically different numbers of members, seems to me to be much more problematic for Atkinson’s purposes. It’s as if a human genomic survey made geographically localized counts of the number of alleles involved in color vision and in blood physiology, divided each set of counts into a few bins (“a little variation”, “a medium amount of variation”, “a lot of variation”), standardized the binned counts for each functional class separately, and averaged the results, thus giving as much weight to each color-vision variant as to several orders of magnitude more blood-physiology variants. This might be OK, but choosing to give this kind of boost to features that happen to be enriched in one region or another will obviously push the results around by a considerable amount

Even if you can’t evaluate the technique in its guts, it is easy to spot some possible issues in the way the data you input into the method is coded or categorized. I hope in the near future this will be less and less of an issue, but it’s a problem which I can understand pretty easily without being very aware of the linguistic details. Also, Liberman’s last paragraph is funny. Though in defense of this paper I think we need to evaluate its plausibility in terms of the overall conditional probability; we often have strong prior models of the origin and expansion of modern humanity, and so we give a particular specific significance to this result. That can of course lead us to greater error than would otherwise be the case if our priors aren’t quite as robust as we’d thought.

Looking forwarded to A Replicated Typo’s take. Also see Dienekes’ opinion.

CATEGORIZED UNDER: Culture, Linguistics
MORE ABOUT: Language, Linguistics
  • http://washparkprophet.blogspot.com ohwilleke

    Language Log’s identification of the source data and description of the methodology was very helpful. Analysis of a couple of additional maps from the same author that Atkinson relied upon, in particular, suggest a middle ground.

    In the cases of several classes of consonants and at least one class of vowels, a serial founder effect theory can plausibly explain the absence of these classes of phonemes from a particular area.

    The loss of frictive consonants is a major factor in the low number of consonants found in Australia, Papuan languages, and consonant poor languages of the Americas. The high consonant count African languages are often notable for including labial-velar consonants (found elsewhere only in a small number of Papuan languages). The high vowel count in African languages are often notable for including a distinction between nasalized and non-nasalized vowels (otherwise found mostly in the Americas and sporadically elsewhere, e.g. French but no other IE languages). The high consonant count in the Caucasian languages is largely attributable to the pharyngeal consonants used there, in some Afro-Asiatic languages, and in few other places where they could have been developed independently. Click consonants are also found only in one small, regional subset of Africans and presumably never found their way out of Africa. The geographic pattern of these phonemic shifts with regard to specific classes of phonemes are consistent with a serial founder effect, and are clear enough, and have few enough exceptions, to fit the theory in a robust way that isn’t very model specific. These five cases alone probably drives most of the overall trend and may illuminate interesting pieces of pre-history and deep linguistic relationships.

    On the other hand, the serial founder effect theory doesn’t do a good job of explaining the number of other vowels in a language, the presence or absence of tone systems, or the presence of the “th” sound in a language. It also doesn’t do a good job of explaining the circumstances under which consonants are added to languages (which seems particularly prone to happen in languages which are vowel poor and lack a tone system).

    Thus, the theory may be half-right, but really would make more sense narrowed into a collection of theories about specific classes of phonemes (which seem to come and go in sets rather than individually), rather than being a global grand theory.

  • http://www.kinshipstudies.org German Dziebel

    @oh-willeke

    Andrew, as usual, you show encyclopedic ignorance.

    Examples: “the high consonant count African languages are often notable for including labial-velar consonants (found elsewhere only in a small number of Papuan languages).”

    In reality, labiovelars are found in Amerindian languages (e.g., Tlingit), they are reconstructed for proto-Indo-European and survive as q in Latin and Germanic.

    “Click consonants are also found only in one small, regional subset of Africans and presumably never found their way out of Africa. ”

    Clicks are African-specific and hence emerged in Africa after Africa had been colonized. Whistling languages, on the other hand, are much more interesting from the point of view of global distributions.

    “The loss of frictive consonants is a major factor in the low number of consonants found in Australia, Papuan languages, and consonant poor languages of the Americas.”

    These consonants are called “fricative.” And again they are found in the Americas (e.g., Plateau languages).

    “The geographic pattern of these phonemic shifts with regard to specific classes of phonemes are consistent with a serial founder effect…”

    How can it be consistent with the serial founder effect if, as Ives Goddard put it in a comment on Language Log, “53 ‘large’ consonant inventories (with 34 and up) [are found in the following areas – G.D.] Asia 16 (Caucasus 7, others 9), Africa 13 (click languages 7, others 6), North America 12 (Northwest and northern California 9, others 3), Europe, South America, and Oceania 4 each. So, large consonant inventories (on these data) are not more apt to be found in Africa, but they do have a distribution that is non-random and unexplained. And of course the numbers could be increased if WALS had correctly classified Abkhaz (say 58) and Slavey (37) and had included, for example, Ubykh (80), Tanaina (36), and conservative Peel River Eastern Gwich’in (53).”?

  • http://washparkprophet.blogspot.com ohwilleke

    For “labiovelar” distribution I rely on the maps of rare and omitted common consonsants from the same WALS data set used in Akinson’s paper, prepared by the same author as the charts upon which he relies. If this data is incomplete, than so is mine. But, the fact that Indo-European might have once had labiovelars which it lost supports the theory that this particular class of consonants experienced a serial founder effect, and the fact that there is a tiny vestigal exception in a couple branches of the linguistic tree doesn’t argue against that theory either, any more than isolated borrowed words argue against a common genetic origin for two languages.

    “How can it be consistent with the serial founder effect if, as Ives Goddard put it in a comment on Language Log, “53 ‘large’ consonant inventories (with 34 and up) [are found in the following areas . . . ”

    Because there is more going on than just serial founder effects. Some consonant classes show strong serial founder effects, some don’t. In the case of the Caucasus, click languages and fricative consonants, the case that founder populations elsewhere lacked certain classes of consonants that those populations held onto seems pretty plausible.

    One means by which one can get more consonants, is at language borders (including substrate-superstrate interactions), where areal effects cause more than one language family’s worth of consonsants to be used in the same area. Another factor in increasing consonant sets is vowel paucity and lack of tone that creates a necessity to make sounds in some other way to differentiate words; but mostly this is done in the context of existing phoneme schemes. A language is more likely to fill out the roster of potential consonants in classes that already exist than it is to spontaneous add a whole class of phyrangeal ones, for example.

    The serial founder effect theory of Atkinson, is like using only one reasonably important term to describe something that is really the sum of half a dozen factors. The problem is not that there aren’t significant founder effects, it is that there are other processes that are equally important that are omitted and look like noise when not expressly considered.

    “Clicks are African-specific and hence emerged in Africa after Africa had been colonized.”

    This makes no sense. Why in the world would Africa specific clicks that emerged there be post-colonial?

    “These consonants are called “fricative.” And again they are found in the Americas (e.g., Plateau languages).”

    I didn’t say otherwise. The branch of coastal route migrants that went to Papua New Guinea and Australia lost them, but they continued into the Americas only to be lost in a group of languages far from the point of entry that are consonant poor – not all languages of the Americas. But, in both cases, almost every language that departs from a main stream across a particular geographic line loses the same set of consonsants.

    As for imperfect spelling, etc., that will improve greatly in the day when I can afford to hire a copy editor for my blogging, until then I disclaim all responsibility for typos and blame it on the gremlins in my keyboard.

  • http://www.kinshipstudies.org German Dziebel

    You’ve just mentioned “founder effects” 6 times. The simplest way to talk about it is that phonemes come and go. Rare phonemes (clicks, pharyngeals, th, etc.) are likely relatively recent local convergent and diffused developments possibly associated with population growth. The loss of more typical phonemes (such as fricatives) can be correlated with isolation and population size decline.

    “The branch of coastal route migrants that went to Papua New Guinea and Australia lost them, but they continued into the Americas only to be lost in a group of languages far from the point of entry that are consonant poor – not all languages of the Americas. But, in both cases, almost every language that departs from a main stream across a particular geographic line loses the same set of consonsants.”

    You just retrofit data into your cosmological vision of people exiting Africa and entering America, Australia, etc. There’s indeed a paucity of fricatives in Australian languages but America and Papua New Guinea (Arapesh, Asmat, Dani, etc.) have plenty of cases of fricative series. As for Australia, it’s isolation and in-situ drift in small populations has led to the decay of this consonant class. There’s no “founder effect” and “distance from Africa” involved.

    “Why in the world would Africa specific clicks that emerged there be post-colonial?”

    I meant the colonization of Africa by modern humans. The fact that clicks aren’t found outside of Africa suggests that they evolved in Africa AFTER it had been colonized by modern humans. The out of Africa theory is based on a misunderstanding of the data.

    “For “labiovelar” distribution I rely on the maps of rare and omitted common consonsants from the same WALS data set used in Akinson’s paper, prepared by the same author as the charts upon which he relies. If this data is incomplete, than so is mine. But, the fact that Indo-European might have once had labiovelars which it lost supports the theory that this particular class of consonants experienced a serial founder effect…”

    It’s mocassin radio, Andrew. Mathieson simplified the picture for his purposes, Atkinson simplified it further for his purposes, you simplified their simplifications even further. In Proto-Indo-European there were two interesting classes of phonemes: labiovelars and palato-velars. Labiovelars survived and evolved in the centum languages, while palato-velars survived and evolved in satem languages. Labiovelars got lost in satem languages only, while palato-velars got lost in centum languages only. I don’t understand why one would need to introduce the term “founder effect” to obfuscate a well-understood, albeit complicated process of consonant evolution in IE languages.

    “As for imperfect spelling…”

    It’s not your spelling, Andrew. You’re just trying to hit too many birds with one stone…

NEW ON DISCOVER
OPEN
CITIZEN SCIENCE
ADVERTISEMENT

Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at http://www.razib.com

ADVERTISEMENT

See More

ADVERTISEMENT

RSS Razib’s Pinboard

Edifying books

Collapse bottom bar
+

Login to your Account

X
E-mail address:
Password:
Remember me
Forgot your password?
No problem. Click here to have it e-mailed to you.

Not Registered Yet?

Register now for FREE. Registration only takes a few minutes to complete. Register now »