April 30, 2012

Tracing the origin of language in Africa (again)

A recent paper on the origin of language in Africa was subjected to heavy criticism not long after its publication. Now, the African origin of language hypothesis is championed by a new pair of authors who also base their claim on phonemic diversity of African languages.

I am sure that there will be a new round of criticism by linguists shortly, but for my part, I will simply note a couple of things:

First, the paper assumes rather than prove a tree-like divergence of human languages. It is not at all clear to me that language does have a common origin in humans. For one thing, research on Neandertal anatomy has suggested that they had the ability to vocalize, and Neandertals diverged from other humans long before the emergence of H. sapiens or even H. helmei in the African Middle Stone Age.

It could very well be that the ability to talk may have been invented twice or even more times, evolving perhaps from a simpler communication system available to species such as H. heidelbergensis. If that is the case, then a tree model of the accumulation of phonemic diversity need not only involve Out of Africa migration, but also contact with other human groups with languages of their own, possessing separate phonemic inventories.

A second point that needs to be made is that I do not really understand the inference of law-like accumulation of phonemic diversity on which the paper is based. It could well be argued, for example, that migration creates separate language communities that undergo different processes of phonemic evolution, some of them retaining ancestral phonemes, some of them developing new ones. Indo-European is a case in point, where languages such as Greek and German have added fricatives, and dispensed with aspirated stops while Hindi has not. The totality of the phonemic inventory in Indo-European languages today may be greater than it was in PIE.

But, there is also a process of simplification within languages leading to a loss of phonemes (e.g., the simplification of the Greek vowel system and the loss of initial aspiration). It is not clear to me whether simplification or diversification wins out in the end: perhaps our ancestors had a much richer phonemic diversity in the past which was reduced as people started becoming part of larger social units than the ancestral hunting tribe. In such units people might get rid of their ancestral way of speaking and accumulate into a new and simpler way, in the same manner that e.g., the phonemic inventory of the entire world is reduced after a couple generations into the sounds of American English by people who become part of the American social unit.

Finally, the paper assumes Out-of-Africa 60-70ky ago. I have been a vocal critic of that idea for quite some time now, and a bird tells me (or should I say tweets me?) that a new synthesis will soon be published that takes into account some recent and exciting archaeological discoveries, that point to a pre-100ky event and apparently exclude a post-70ky one. But, more on that when it sees the light of day and I read it for myself.

 PLoS ONE 7(4): e35289. doi:10.1371/journal.pone.0035289

Dating the Origin of Language Using Phonemic Diversity

Charles Perreault, Sarah Mathew


Language is a key adaptation of our species, yet we do not know when it evolved. Here, we use data on language phonemic diversity to estimate a minimum date for the origin of language. We take advantage of the fact that phonemic diversity evolves slowly and use it as a clock to calculate how long the oldest African languages would have to have been around in order to accumulate the number of phonemes they possess today. We use a natural experiment, the colonization of Southeast Asia and Andaman Islands, to estimate the rate at which phonemic diversity increases through time. Using this rate, we estimate that present-day languages date back to the Middle Stone Age in Africa. Our analysis is consistent with the archaeological evidence suggesting that complex human behavior evolved during the Middle Stone Age in Africa, and does not support the view that language is a recent adaptation that has sparked the dispersal of humans out of Africa. While some of our assumptions require testing and our results rely at present on a single case-study, our analysis constitutes the first estimate of when language evolved that is directly based on linguistic data.



Rob said...

On the general topic of out of Africa, several excellent papers have suggested that the current molecular evidence suggesting an apparent ROA could represent merely a recent gene flow out of many older ones, or a "process' of modern adaptations rather than a single speciation event, and the mtDNA could represent a recent selective sweep, etc. However, what such potentially valid theories don;t discuss is why, almost every time, such flow nevertheless stems out of Africa, each and (almost) every time

eurologist said...

Africa is large and has diverse climates and as such allows humans as well as other living beings to exist in relatively large numbers regardless of changes.

On the flip side, I don't think one can make the case that humans, let alone modern humans, solely evolved in Africa. Theory and records indicate that every 50,000 or 100,000 years or so, climate was favorable for ancient humans to live in and disperse from what currently are deserts - parts of the Sahara, Arabia, and generally the Near and Middle East. So, genetic exchange between Europe, the West Asia, and Africa likely was the norm during many of such periods. For example, the very existence of a geographically widely dispersed Heidelbergensis category wold be impossible, otherwise.

Andrés said...

Simplification is illogical. If languages tend to lose complexity (be it phonemes, declension, whatever), how did they become complex in the first place?

German Dziebel said...

Agree with all your points, Dienekes. More from me here: http://anthropogenesis.kinshipstudies.org/2012/04/phonemic-diversity-and-out-of-africa-again-the-myth-is-gaining-a-momentum/

Slumbery said...

Blogger Baldric said...

Simplification is illogical. If languages tend to lose complexity (be it phonemes, declension, whatever), how did they become complex in the first place?

The always increasing complexity is the same funny joke as the (forever) sustainable development of economy. :)

Languages evidently reached the complexity necessary to work and being useful a very long time ago.
On what base you say that there is any preferred direction of complexity change after this point? I do not know such a natural law, also can't see how could simplification be illogical.

Also I do know that language simplification happens, since I am a native speaker of a language that lost most of its tenses it the last several centuries.(Hungarian. We have past and present/future now, but educated people recognize other tenses in texts barely 200 years old and even more in medieval texts. Also this went completely against IE (latin and german) language influence.)

Andrew Oh-Willeke said...

I'm very convinced that all human languages are derived from African languages, since the notion of language acquisition from archaics or a new language constructed from scratch in the Middle Stone Age are very implausible. I'm also strongly incline to think that even if there were a few original human languages from different bands in early speaking hominin groups, that the groups that gave rise to modern humans were at least part of a sprachbund and strongly influenced by each other at the phonetic level. But, those aren't conclusions that flow from the linguistic evidence itself.

Bottom line, I think that there has simply been too much language churn over time and in the natural process by which languages evolve for contemporary, or event historically attested linguistic evidence to be discernable at even a recent Out of Africa time depth. The deepest linguistic connections discerned to date are in the 6,000-14,000 years old and tone as well as a few other phonemic features seem to be areal linguistic features (i.e. shared by languages in a region regardless of their parent languages) rather than parts of linguistic family tress to some extent at least. And, the options are fewer than they seem because phonemic features interact with each other rather than being independent.

Also, even if Out of Africa is really 100,000 years ago, it is quite plausible that a language that swept out of Africa 60,000-70,000 could have supplanted whatever language was there before with modest population genetic impact (Hungary, Alexandrian Greek, Hittite, Arabic, Bantu in Pygmies, Hebrew in Israel, and the Romance languages provide just a few examples).

Out of Africa is an oldest possible most recent common Eurasian linguistic ancestor, but a more recent date is possible and not even all the implausible.

Jim said...

"Simplification is illogical. If languages tend to lose complexity (be it phonemes, declension, whatever), how did they become complex in the first place?"

What is illogical is to assume the either simplification of increasing complexity is unidirectional. in fact it is cyclical.

A complex consonantal system can simplify with compensatory increased comnplexity of the vowel system. Or the langaage may evolve tones (often from devoicing of voiced consonants, or from lost laryngeal or pharyngeal consonants.

Or the language may lose consonant complexity and compensate with compounding or some other way of making longer mophemes. If these collapse through vowel loss, newly adjoining consonants may combine into new consonants that add to the consonant inventory. Or adjoing consonants may influence and change the articulation of a consonant.

A language can go through a cycle of having no tones, developing tones, simplifying the tone system, compounding words to compensate for or forestall the loss of distinctive units, which then simplify into still complex but shorter units, which then lose some consonants, making tones necessary to retain distinctions in the system. This cycle has been going in most if not all the Chinese langauges for at least 2,000 years. Shanghainese has gone the farthest, with only two tones left and some fairly long (2-3 morphemes) strings of formerly independent morphems that now function like single morphemes.

German Dziebel said...


"What is illogical is to assume the either simplification of increasing complexity is unidirectional. in fact it is cyclical."

Agree. No need to contrast cyclical and linear, though. It's both cyclical and unidirectional: the cycles of simplification and complexification result in the loss of archaic phonetic properties and the spread of novel phonetic properties, but language doesn't step into the same river twice. E.g., the studies that use click languages as examples of phonemic complexity overlook the fact that other languages (e.g., Piraha in South America) have "whistled" sounds. Whistled sounds look just as odd, childish and auxiliary as clicks. It's possible that whistled sounds are archaic phonemes that got lost in most languages. Some languages lost whistled sounds and didn't gain clicks, others lost whistled sounds and gained clicks. If this hypothesis is true, then whistling to no-whistling or whistling to clicking would be a unidirectional change.

terryt said...

"tone as well as a few other phonemic features seem to be areal linguistic features (i.e. shared by languages in a region regardless of their parent languages)"

That could actually be a result of 'language acquisition from archaics'. We can assume that any language archaics spoke was regionally distinct by the time 'moden' humans arrived there. I think that any influence of archaic languages on modern languages would lead to much the same as has happened with the spread of modern languages around the world. Pronunciation and expressions from 'original' regional languages have survived to influence regional accents within English, French, Spanish etc.

Jim said...

terryt, that is called a substrate effect. Anyone who has tried to learn a second langauge, actually anyone who had a grandparnet who had some other native language, has experienced it.

Tone does tend to spread, but it can arise on its own. The Baltic is one area where languages are developing tone systems, across genetic lines, uninfluenced by any external model.