ASPM & Microcephalin & tonal languages?

Share on FacebookShare on Google+Email this to someoneTweet about this on Twitter

Note: The authors have a website which summarizes their research (via Language Log).

Speaking in tones? Blame it on your genes:

People who carry particular variants of two genes involved in brain development tend to speak nontonal languages such as English, while those with a different genetic profile are more likely to speak tonal languages such as Chinese.

In tonal languages, which are most common in South East Asia and sub-Saharan Africa, subtle differences in pitch can change the meaning of vowels, consonants and syllables. Nontonal languages, which prevail in Europe, the Middle East and North Africa, use pitch only as a way of conveying emphasis or emotion.

He cautioned, however, that the research had so far found only an association that appears to be more than chance, and that more work was needed to confirm a causal effect.

I first started hearing stuff this sort of research (i.e., the correlations between particular alleles and language forms) in 2006, so I’m not too surprised. We’ll see how this pans out, look for it in PNAS.

But, which alleles on which genes? From Scientific American:

The new research, published this week in Proceedings of the National Academy of Sciences USA ties this difference to two genes, ASPM and Microcephalin. The exact functions of both genes are still open to debate, but they are known to affect brain size during embryonic development. “They presumably have something to do with brain structure, because there are deleterious mutations of the genes that lead to microcephaly” (a condition in which a person’s brain is much smaller than the average size for his or her age), says senior study author, Robert Ladd, a professor of linguistics at the University of Edinburgh in Scotland.

Ladd and Dediu compared 24 linguistic features-such as subject-verb word order, passive tense, and rounded vowels-with 981 versions of the two genes found in the 49 populations studied. Most of the language contrasts could be explained by geographic or historical differences. But tone seemed to be inextricably tied to the variations of ASPM and Microcephalin observed by the authors. The mutations were absent in populations that speak tonal languages, but abundant in nontonal speakers.

First, note that the authors here imply that derived (more recent) forms of the genes in question under strong selection are connected to the spread of non-tonal languages around 6,000 years ago. There’s a lot of question begging here, are tonal languages ancestral? Is this association causal? But look, there’s one thing that jumps out at me from a cursory examination of the various times suggested here: it sees possible that the rise to prominence of non-tonal languages (assuming their derived character) dovetails with the posited expansions of Indo-European and Afro-Asiatic languages around the core of the World-Island (Eurasia + Africa).

Here is the wiki info on the geography of tonality:

In Europe, only Norwegian, Swedish, Scottish Gaelic, the Limburgish language (according to some a dialect of Dutch), Lithuanian, Serbian, Croatian, and some dialects of Slovenian possess tonality, and only Lithuanian regularly marks it in text other than in dictionaries. The tones of Lithuanian are believed to be especially authentic, as they agree for the most part with the tones of Vedic Sanskrit, its ancient cousin. Another Indo-European tonal language, spoken in the Indian subcontinent, is Punjabi.

Most languages of sub-Saharan Africa (notably excepting Swahili in the East, and Wolof and Fulani in the West) are tonal. Hausa is tonal, although it is a distant relative of the Semitic languages, which are not.

There are numerous tonal languages in East Asia, including all the Chinese “dialects”, Thai, Vietnamese and Burmese (but not Mongolian, Cambodian, Malay, standard Japanese or standard Korean). In Tibet, which is riven by harsh geography, the Central and Eastern dialects of Tibetan (including that of the capital Lhasa), are tonal, while the dialects of the West are not. Much speculation has been generated over the reasons for this partial tonogenesis.

Some of the native languages of North and South America possess tonality, especially the Na-Dene languages of Alaska and the American Southwest (including Navajo), and the Oto-Manguean languages of Mexico. Among the Mayan languages, which are mostly atonal, Yucatec, with the largest number of speakers, has developed tones.

Finally, from the authors’ website:

The next step is to do experiments in which we look for evidence of the nature of the predisposition or bias. The work of Patrick Wong and his colleagues provides one possible lead here: they have shown that some monolingual adults find it much harder than others to learn an artificial language vocabulary that makes use of tone or pitch distinctions, and that the differences between these groups show up in subtle differences of brain structure as well. If we could show that these differences also reflect differences in genetic make-up, it would go some way to showing that the correlation we have found is based on a real causal link.

Related: This is Bruce Lahn’s brain on ASPM and MCPH1. ASPM, Microcephalin, and intelligence. Developmental cell biology of ASPM. Brain size & genes.



  1. I’m a little skeptical on Scottish Gaelic being tonal, as although I don’t speak it, it is derived from and closely related to Irish Gaelic, so much so that I can readily read and understand most of it.  
    Any other sources on this?

  2. Fascinating! Where do “click” languages fit in?

  3. Just did a Google news search for “language aspm” and the picture for the second grouping was Razib and his cat. :)

  4. The geography of tonal languages seems to make the genetic explanation implausible. On the one hand, genetically very similiar peoples (Swedes and Germans, Japanese and Chinese) differ. On the other hand, large numbers of people speak have different mother tongues than their ancestors did even 1000 years ago. 
    Tone in SE Asia is suspected to be an “areal feature”, acquired by originally non-tonal languages because of proximity to tonal languages (especially Chinese.) 
    Any statistical study would be distorted by the 500-lb Chinese gorilla. Something like a fifth of all people are Chinese, genetically related and speaking a tonal language.

  5. john, two points 
    1) there might be a disjunction between total genome content (overall ancestry) and this locus (obviously there is to some extent insofar as some populations exhibit polymorphism between ancestral and derived, so on this locus they’d be ‘more closely related’ to similar people in other populations). LCT, the locus which has the mutant for adult lactose digestion might be an example of this, north indians might have a north eurasian allele despite the fact that they’re genetically far more similar to south indians than they are to north eurasians. in other words, if the derived alleles started spreading the last 10,000 years they are going to be an overlay upon the overall genome (just like the derived LCT might be). in much of the word the overall genome structure might have been pretty much established 10,000 years ago, but with the rise of agriculture and what not portions of the genome might have been subject to periodic selection sweeps which tie together populations in various ways due to history and ecology. 
    2) the association might be strong for some populations than others. i am particularly interested in correlations between the alleles and tone discernment on the individual level in this case.

  6. What about populations that speak non-tonal languages with a “lilt”? Irish English or Mexican Spanish for example. 
    And note that some now tonal languages were at one time non-tonal. Last I heard this included Cantonese, and not so long ago.

  7. What about populations that speak non-tonal languages with a “lilt”? Irish English or Mexican Spanish for example. 
    guys, let’s be careful about trying to account for all the variation just by this association. i’m assuming that the correlation is proportional to  
    a) the extent of tonality 
    b) the length of time of tonality 
    and, i’m also assuming even these parameters controlled you can still seem variation.

  8. There’s a point of detail: a “tonal language” is one within which words (or morphemes) are distinguished by tone alone. Thus “at” with a rising tone would be a different word than “at” with a falling tone. 
    There are lots of tone puns in Chinese. For example, “yizi” with one tone means “chair”, and with a different tone means “sister-in-law”. A guy I knew actually did introduce his sister-in-law to someone as his chair, and got a laugh. 
    English has sentence tone. For a bare example, “Yes?” is a rising tone, and “Yes!” is a falling tone. In English questions are often marked with tone (though as you know, some people have the annoying habit of pronouncing declarative sentences as questions). 
    Anyway, languages where there’s a lilt aren’t tonal languages in terms of linguistics. The wiki either is new information or is wrong: when I was studying only Swedish and Scandinavian languages were said to have tone. However, I’m not an expert.

  9. p.s., my comment above presupposes some causality. the authors are careful not to assume this. we know that strong selection coefficients seem to have drive ASPM derived alleles to higher frequencies, but we don’t know exactly what. it is something aside from tones, but which is correlated moderately or strongly with non-tonality, which is driving the selection for ASPM & microcephalin then it stands to reason that the mapping will be imperfect as it is only a side effect.

  10. Here is !Kung tone pun: chi’ (high tone) is water, while chi_ (low tone) is to copulate.  
    A !gu !a is a well that never runs dry, that is it is a place where one can always chi’. It is also the term for a woman of easy virtue, somewhere one can always go to chi_. 

  11. From Wikipedia: Some Indo-European languages are usually characterised as tonal, such as Lithuanian, Slovenian, Serbo-Croatian, Limburgish, Swedish and Norwegian; more correctly, however, they are pitch accent languages, as only the tone on the stressed syllable can have any effect on the meaning. 
    Pitch accent.

  12. Maybe this concern is entirely unfounded if I were to see the methods section of their paper, but I hope we don’t have a chopsticks gene on our hands here.

  13. I recall back in the 80s there was an argument on the BBC about sociobiology between Robert Trivers and Richard Lewontin. Lewontin used the example of language to show the unimportance of genetic differences between human populations, asserting it as a self-evident truth that children of all different races were equally capable of learning all languages. Trivers replied that he wasn’t sure about that, and he didn’t think any research had been done!

  14. Rosko, 
    Yeah, I think this could easily be a case of a “Chopsticks gene”… 
    But it occurred to me that in Irish Gaelic there is a pattern of long versus short vowels, and they are always harmonized within words, so that a lilting effect is produced. This, along with other mnemonic devices was used for millenia by Seanachaí (storytellers) to remember stories, often many hours long. Could it be that tonal languages are used in a similar way? 
    If so, then maybe the advent of writing obviated the need for tone, and those people – such as Middle Easterners and populations descended at least partially from them, such as Europeans – could utilize brain areas given over to verbal/tonal memory in new ways – leading to differences in ASPM and Microcephalin distributions??

  15. Is there any relation between speakers of tonal languages and the prevalence of perfect pitch? I’ve read perfect pitch is far more common in China than it is in European populations.

  16. Is there any relation between speakers of tonal languages and the prevalence of perfect pitch? I’ve read perfect pitch is far more common in China than it is in European populations. 
    i’ve heard the same.

  17. the authors here imply that derived (more recent) forms of the genes in question under strong selection are connected to the spread of non-tonal languages around 6,000 years ago 
    Historically, this runs counter to the linguistic consensus: In Chinese, tones are derived – proto-Chinese was atonal. From Wikipedia: An interesting question is how tones arise in a language, i.e. tonogenesis. In the Chinese languages they arose as a reinterpretation of initial and final consonants. Middle Chinese, for example, had three tones (rising, “departing”, and level), which are said to have arisen from Old Chinese final consonants (/?/, /s/, or neither of these). I think this is a case of looking until you find data that matches your hypothesis…

  18. Evidently, haloscan doesn’t like the form of the link above, since it’s correct in the database. It’s this: 

  19. Here is a paper on speakers of tonal language and ‘perfect pitch’: 
    I have my doubts on this ‘perfect pitch’ at least in the sense that it is generally used. 
    I majored in linguistics (but have limited knowledge of and experience with tonal language. I’m also a musician, have studied piano in a conservatory, and have a very well-developed musical hearing, but no ‘perfect pitch’; that is, I need a reference note to be sure of which note I’m hearing, but I’m able to transcribe melody lines and harmonies accurately by ear once I have such a reference. 
    Based on my experience (which includes accompanying people of many different backgrounds singing at weddings etc, and working regularly with a singer who is a speaker of a tonal language) the notion of almost all speakers of tonal languages having ‘perfect pitch’ in the usual, musical sense strikes me as highly unlikely – I tested a Vietnamese-born and Vietnamese-speaking colleague who plays the guitar, and he couldn’t tune it correctly. The lead singer I usually work with is an African, a native speaker of Lingala (a tone language) and a gifted musician who plays the piano and the guitar; but he needs reference notes, too (also when he has to do an a capella intro, for example). 
    On the other hand, I’ve noticed that amateur singers with no notion of pitch or musical scales whatsoever often hit the ‘right’ first note if they sing a song they are familiar with without accompaniment; the same individuals are utterly unable to hit a note in an unfamiliar piece, or harmony part. Possibly their hitting the right note has to do with their ‘feeling’ and reproducing a certain amount of vocal cord tension. The same phenomenon could be at work with the Vietnamese and Mandarin speakers.  
    Also, the explanation for the acquisition (or retention) of perfect pitch does not convince me: 
    “our findings lead us to conjecture that the potential for acquiring absolute pitch may be universal, and that it can be realised by the association of pitches with meaningful words very early in life. “ 
    Some linguistic points: 
    - Vietnamese and Mandarin are ‘contour tone languages’. It is the shape of the pitch contour, not its absolute pitch, that determine the identity of a speech sound. 
    - even in ‘level-tone languages’ (e.g. many African languages) the sounds are not associated with exact pitches; for instance, they are subject to downdrift, so the exact pitches will vary depending on where in the utterance they occur (and on the downdrift characteristics of the speaker. The degree of downdrift also varies with utterance length. 
    - different individuals have very different vocal ranges; the same word pronounced by two speakers will usually be pronounced at different pitches as well, but remain recognizable. The vocal range of individuals also changes over their lifetime. So when a child acquires vocabulary in a tonal language from different indivuals, e.g. mother and father (whose vocal ranges are likely to differ by about an octave on average) it doesn’t “associate pitches with meaningful words very early in life”, at least not absolute ones.

  20. David Boxenhorn I wouldn’t be completely sure about that Wiki piece. The Chinese tone systems have varied historically and vary between neighboring dialects and subdialects, but I don’t think that a non-tonal early phase has been established. Unfortunately, I’m not up to date.

  21. David Boxenhorn I wouldn’t be completely sure about that Wiki piece. 
    That wasn’t the source of my information. I pointed to it because it is on line.

  22. John, a non-tonal early phase of Sinitic has been established for half a century. Karlgren noticed it. Haudricourt gave us a time of 2,000 years ago.

  23. Oh yeah, we even have the case of a tonal Chinese language (Wu Chinese) evolving into a non-tonal sub-dialect in the city of Shanghai in the last 50 years. As I’ve said before, this new ASPM variant could even be shown to make people eat cheese, so let’s call it more accurately the cheese gene instead of the chopstix gene.