The Unz Review - Mobile
A Collection of Interesting, Important, and Controversial Perspectives Largely Excluded from the American Mainstream Media
 Gene Expression BlogTeasers
Language (Culture) and Genes Evolve Differently
Email This Page to Someone

 Remember My Information



=>
Search Text Case Sensitive  Exact Words  Include Comments

9780226520438 Evolutionary process can be modeled in both genes and culture. The former is defined by vertical transmission, while the latter can be vertical and/or horizontal. Unlike heritable biological traits, cultural phenotypes have no discernible units of inheritance in a straightforward fashion which can be easily mapped. But some of the formal models common in evolutionary genetics are also utilized in social evolution and behavioral ecology.

One of the easiest aspects of culture to gain a comprehension of is language. Unlike other cultural phenomena, such as religion, language is clear and distinct. Many believe that in some way it is a deep biological competency, and in fact would put it outside of the domain culture altogether because of its unique role at the center of the propagation of cultural “memes.” A new paper in PNAS explores the correlations of language and genes and geography, A comparison of worldwide phonemic and genetic variation in human populations (open access!). In short, the authors find that the differences in transmission of genes and language result in differences in their patterns of distribution. The correlation between genes and geography scales over the whole world. The more distant a population is from a focal group of interest, the more genetically different it is. In contrast the signal of linguistic affinity (or lack thereof) exhibits spatial limits, beyond which the linear relation decays. Beyond 10,000 kilometers more distant languages are no more dissimilar.

bravo There are a few issues to unpack here. First, they used a database of phonemes. I have no idea how one would categorize differences using syntactic features, but it strikes me that someone without more familiarity with this field might argue that looking at variation in phonemes is a bit like looking for the key under the lamp. Interestingly the authors found that phoneme similarities transcend language family. In other words, nearness breeds familiarity through horizontal transmission even if the linguistic groups are dissimilar rather than being part of a dialect continuum.

Second, they suggest that one aspect of phonemes and how they differ from genes is that isolated populations exhibit more richness and diversity, rather than less. This illustrates that there are differences between genetic and cultural process. Not only is there a great deal of horizontal transmission, but cultural processes are subject to a greater “mutation” rate, and selection can be much more efficacious. The latter is why group level selection is more mathematically plausible for culture than genes; competing demes can be much more distinct in culture than genes because minimal gene flow can equilibrate biological differences, while biased transmission of culture can result in insulation of different groups from homogenization (e.g., inheriting your cultural traits from your father, rather than your mother, who may have been kidnapped from an enemy tribe).

Finally, in line with the high mutation rate of language the authors reject earlier findings that it follows the same serial founder model detected in a 2005 paper from some of the same authors. I have to jump in here to suggest that we need be careful about assuming that this paper is a robust result upon which we should build up our model. See Towards a new history and geography of human genes informed by ancient DNA for a slight revision. In any case, the results from the language patterns suggest that Europe is the source of human language, using the same framework as genes where there is a decay of diversity from the ancestral homeland. The authors point out that this is a artifact of the fact that phoneme richness is very low in Oceania and South America, and Europe is equally distant from both regions. In other words language is too protean to gain a signal of the “Out of Africa” movement. I do agree with this. It strikes me that those who attempt to reconstruct language as it was 50,000 years ago are grasping for straws. For example, I do not think that we can presume that clicks are ancestral just because the Khoisan have clicks in their language.

The relationship of patterns of genetic variation and cultural variation are essential to elucidate. That is because I believe that we can’t understand patterns of genetic variation without a clear grasp of the common cultural processes by which human genes propagated over time and space. Language is probably the cultural trait that’s lowest down on the tree, so hopefully researchers will keeping picking at it until the big questions get resolved.

 
• Category: Science • Tags: Genes, Language 
    []
  1. Language is not a proxy for culture. Cultures, among humans, involve a range of economic systems, as well as various social, religious, and political ways of organizing community life. Thus a culture is not just codes of conduct, systems of explanation, stories, or symbolic representations. Language permits coded information transfer. It also permits synthesis and communication of cumulative knowledge, and this serves as a mechanism of quorum sensing. All social animals appear to have various local adaptive behaviours that are learned by the young. Human cultures have much expanded replicative capacity, besides imitative behaviour and systems of social control (which can convey a lot of information) human language bumps the replicative potential to a whole new level. I suspect that language originated in some form – along with gestures, signs, song and dance, as well as ritualized behaviours, quite early – perhaps coinciding with the transition we distinguish as the beginning of the genus Homo.

    Read More
    • Replies: @JayMan

    Language is not a proxy for culture. Cultures, among humans, involve a range of economic systems, as well as various social, religious, and political ways of organizing community life. Thus a culture is not just codes of conduct, systems of explanation, stories, or symbolic representations.
     
    On that point, we agree. But language is easiest to "quantify," so that's why they studied it. But you're right, for that reason and others mentioned in the post, they have a long way to go.
    ReplyAgree/Disagree/Etc.
    AgreeDisagreeLOLTroll
    These buttons register your public Agreement, Disagreement, Troll, or LOL with the selected comment. They are ONLY available to recent, frequent commenters who have saved their Name+Email using the 'Remember My Information' checkbox, and may also ONLY be used once per hour.
    Sharing Comment via Twitter
    /gnxp/language-culture-and-genes-evolve-differently/#comment-860039
    More... This Commenter This Thread Hide Thread Display All Comments
  2. Bill says:

    What about the data that show that the number of phonemes in a language decay with serial founder events? The Khoisan have clicks because they have had the largest effective population size over the last 100,000 years. Native Americans and Pacific Islanders have fewer phonemes because with each bottleneck they lose diversity.

    Read More
    • Replies: @pseudoerasmus
    the new study does not support Atkinson :

    A regression-based analysis of phoneme inventory size (15) concluded that a global sample of 504 languages fit a serial founder effect model of expansion out of Africa (but see refs. 16– 19). Using a similar approach, we found that phoneme inventory size decreased with geographic distance from northern Europe (Fig. 3); we do not conclude that this supports an origin for lan- guage in Europe for several reasons. Although a population’s genetic diversity reflects the number of its founders, the re- lationship between the number of founders of a population and its language’s phonemes is more complex (18, 21, 25, 27, 43–46). Furthermore, only a subset of the model’s predictions apply to languages (16), and the mutation rate of phonemes may be high enough that signatures of ancient divergence are erased faster in phonemes than in genes (39, 57). In contrast to previous studies (15, 43), speaker population sizes did not explain a significant proportion of variation in phoneme inventory size
    , @pseudoerasmus
    I could never believe in Atkinson because it's just so obvious that languages can adopt a whole new set of phonemes. At some point Indo-Aryan acquired retroflex consonants from Dravidian, and there's tonogenesis in China and SE Asia (little more shrouded in mystery).
  3. Robert Ford says: • Website

    So, should I stop making fun of linguists who fetishize linguistic diversity? I remember smirking during the credits of “The Linguists” as the audience clapped so enthusiastically…maybe I should have joined them.

    Read More
  4. @Bill
    What about the data that show that the number of phonemes in a language decay with serial founder events? The Khoisan have clicks because they have had the largest effective population size over the last 100,000 years. Native Americans and Pacific Islanders have fewer phonemes because with each bottleneck they lose diversity.

    the new study does not support Atkinson :

    A regression-based analysis of phoneme inventory size (15) concluded that a global sample of 504 languages fit a serial founder effect model of expansion out of Africa (but see refs. 16– 19). Using a similar approach, we found that phoneme inventory size decreased with geographic distance from northern Europe (Fig. 3); we do not conclude that this supports an origin for lan- guage in Europe for several reasons. Although a population’s genetic diversity reflects the number of its founders, the re- lationship between the number of founders of a population and its language’s phonemes is more complex (18, 21, 25, 27, 43–46). Furthermore, only a subset of the model’s predictions apply to languages (16), and the mutation rate of phonemes may be high enough that signatures of ancient divergence are erased faster in phonemes than in genes (39, 57). In contrast to previous studies (15, 43), speaker population sizes did not explain a significant proportion of variation in phoneme inventory size

    Read More
  5. @Bill
    What about the data that show that the number of phonemes in a language decay with serial founder events? The Khoisan have clicks because they have had the largest effective population size over the last 100,000 years. Native Americans and Pacific Islanders have fewer phonemes because with each bottleneck they lose diversity.

    I could never believe in Atkinson because it’s just so obvious that languages can adopt a whole new set of phonemes. At some point Indo-Aryan acquired retroflex consonants from Dravidian, and there’s tonogenesis in China and SE Asia (little more shrouded in mystery).

    Read More
  6. No surprise that it does not support Atkinson, as the entire Gray & Atkinson approach to historical linguistics is utter nonsense. For a full exposition, see the forthcoming book The Indo-European Controversy: Facts and Fallacies in Historical Linguistics (Cambridge, March 2015). Most of the issues raised in this post are addressed in this book as well.

    http://www.amazon.com/Indo-European-Controversy-Fallacies-Historical-Linguistics/dp/1107054532/ref=sr_1_1?ie=UTF8&qid=1422469409&sr=8-1&keywords=the+indoeuropean+controversy

    Read More
  7. The use of phonemic inventories strikes me as entirely welcome. Phonemes (unlike vague semantic categories represented by English words written in upper case letters, e.g. WATER) are well-defined entities which can be objectively compared between languages and the set of all possible phonemes is sufficiently small to permit the use of all available data. If anything I would like to see a study on these lines which further analysed phonemes into their phonological features such as voicing or pharyngeal articulation, since sound change generally operates on features rather than entire phonemes.

    Read More
    • Replies: @Asya Pereltsvaig
    There's a lot of problems with using phonemic inventories in this way. What Razib mentions in the post, that phonemes are subject to more horizontal transmission than, say, grammatical structures is part of the problem of course. Other issues are discussed here:

    http://languagesoftheworld.info/bad-linguistics/phonemic-diversity-prove-africa-theory.html

    and in my co-authored response to Atkinson's paper in Science 335(6069): 657
  8. JayMan says: • Website
    @Helga Vierich
    Language is not a proxy for culture. Cultures, among humans, involve a range of economic systems, as well as various social, religious, and political ways of organizing community life. Thus a culture is not just codes of conduct, systems of explanation, stories, or symbolic representations. Language permits coded information transfer. It also permits synthesis and communication of cumulative knowledge, and this serves as a mechanism of quorum sensing. All social animals appear to have various local adaptive behaviours that are learned by the young. Human cultures have much expanded replicative capacity, besides imitative behaviour and systems of social control (which can convey a lot of information) human language bumps the replicative potential to a whole new level. I suspect that language originated in some form - along with gestures, signs, song and dance, as well as ritualized behaviours, quite early - perhaps coinciding with the transition we distinguish as the beginning of the genus Homo.

    Language is not a proxy for culture. Cultures, among humans, involve a range of economic systems, as well as various social, religious, and political ways of organizing community life. Thus a culture is not just codes of conduct, systems of explanation, stories, or symbolic representations.

    On that point, we agree. But language is easiest to “quantify,” so that’s why they studied it. But you’re right, for that reason and others mentioned in the post, they have a long way to go.

    Read More
  9. Jefferson says:

    I just got my 23AndMe results and found out that I am 99.5% European. To be more specific my ancestral breakdown is 80.4% Italian, 10.7% Non Specific Southern European, 6% Ashkenazi, 1.5% Balkan, 0.9% Non Specific Northern European, 0.4% Middle Eastern, and 0.1% unassigned.

    Read More
  10. ohwilleke says: • Website

    Coincidentally, I had just posted a lengthy analysis of a smaller subset of phonemes (glottal consonants, clicks and tone) and historical linguistics a few days ago before learning that this study existed, based upon the source data at WALS, and some additional spot research on extinct or earlier forms of modern languages, on languages not covered in that survey, and to confirm or question WALS data.

    http://dispatchesfromturtleisland.blogspot.com/2015/01/do-glottal-consonants-tell-important.html

    I reach only a couple of global conclusions: “Glottal consonants, like click phonemes, appear to be easier for a language to lose than to gain in the absence of strong areal or substrate influences.”, and “The geographic distribution of languages with tone systems is similar, although not identical, to the geographic distribution of languages with glottal consonants.” Click languages are also highly likely to have glottal consonants.

    Also, it isn’t just the case that culture or language evolve differently from genes. Different features of languages evolve differently, in part for functional reasons. For example, languages with more phonemes tend to be spoken more slowly than languages with fewer phonemes, that the semantic transfer can take place at similar rates (Japanese is an apparently outlier exception where the small number of phonemes isn’t compensated for completely by a faster speech rate).

    SVO language order seems to be more mutable than phoneme inventory. Phoneme shift tends to happen as a package deal, operating on an entire set or major subset of phonemes, rather than on individual phonemes in isolation. The roots and grammar for core common words are more resistant to change than those for less commonly used words (grammatically irregular words are usually commonly used words). Certain subject matter words are less resistant to change (e.g. toponyms and indigeneous flora and fauna names), while others are easily borrowed (e.g. trade goods and words related to newly acquired technologies). Major grammatical categories appear to be more stable than individual moderate to low frequency words. Independent innovation does happen, although borrowing is common too. Superstrate cultures have dramatically more linguistic influence than genetic influence in most cases, and fathers in superstrate cultures tend to have more linguistic influence on their children than mothers. Language shift tends to produce grammatical simplification, while languages tend to get more elaborate with stability over time and increasingly complex societies (points that are sometimes at odds with each other and sometimes reinforce each other).

    While I reach few global conclusions from my analysis, I reach half a dozen conclusions about the linguistic prehistory of Africa (although some new questions are also raised in the process related to the perennially vexing question of the relationship of the half dozen major subfamilies of the Afro-Asiatic macrolinguistic family). In particular, phonemes seem to be particularly useful suggesting when to hypothesize substrate influences in outlier languages in a language family.

    In the New World, phoneme data sheds light on the nature of the Na-Dene and Yenesian language connection (Na-Dene languages probably acquired its glottal consonants from a Native American substrate language), and also provides suggestive evidence regarding the languages of the founding population of the Americas (that probably had a full range of glottal consonants). It also provides insight into the prehistory of a Siberian language spoken near the Bering Strait (the only living language of several dozen in Siberia with glottal consonants) suggesting areal influence on it from Alaska.

    In Asia, the analysis tends to corroborate links between groups of Southeast Asian and South Chinese languages already suspected to be related on other lexical, grammatical and population genetic grounds, and it also tends to favor the deep level language family relatedness of Korean and the Japonic languages.

    It also provides another data point suggesting the strong relatedness of the Caucasian languages, either due to common origins, or strong areal influences at some point, and tends to attest to their antiquity.

    Finally, it adds new perspective to the question of whether the Proto-Indo-European language contained several laryngeal consonants that are absent in all attested Indo-European languages (the current majority view) except the Anatolian languages which had a substrate population that spoke languages with those consonants, although it doesn’t resolve the question definitively.

    @1 I would normally consider language to be part of the content of culture, in addition to often being a litmus test that is usually, although not always accurate, for a host of highly correlated cultural content. Coherent cultures and communities are the environment in which languages come into being and live.

    @2-7 With regard to Atkinson, one of the big flaws in his historical linguistic studies is a failure to adequately consider the evidence that language change peaks during the period of language formation and intentional differentiation from its parent language, and the powerful impact of substrate and areal influences without which languages evolve much more slowly.

    Read More
    • Replies: @jtgw
    Your argument that Anatolian borrowed the laryngeals from Caucasian doesn't address the correspondence between the Anatolian laryngeals and various vowels in other Indo-European languages. Did Hittite just happen to borrow the sounds precisely where they would be predicted to exist by the comparative method?
  11. jtgw says:
    @ohwilleke
    Coincidentally, I had just posted a lengthy analysis of a smaller subset of phonemes (glottal consonants, clicks and tone) and historical linguistics a few days ago before learning that this study existed, based upon the source data at WALS, and some additional spot research on extinct or earlier forms of modern languages, on languages not covered in that survey, and to confirm or question WALS data.

    http://dispatchesfromturtleisland.blogspot.com/2015/01/do-glottal-consonants-tell-important.html

    I reach only a couple of global conclusions: "Glottal consonants, like click phonemes, appear to be easier for a language to lose than to gain in the absence of strong areal or substrate influences.", and "The geographic distribution of languages with tone systems is similar, although not identical, to the geographic distribution of languages with glottal consonants." Click languages are also highly likely to have glottal consonants.

    Also, it isn't just the case that culture or language evolve differently from genes. Different features of languages evolve differently, in part for functional reasons. For example, languages with more phonemes tend to be spoken more slowly than languages with fewer phonemes, that the semantic transfer can take place at similar rates (Japanese is an apparently outlier exception where the small number of phonemes isn't compensated for completely by a faster speech rate).

    SVO language order seems to be more mutable than phoneme inventory. Phoneme shift tends to happen as a package deal, operating on an entire set or major subset of phonemes, rather than on individual phonemes in isolation. The roots and grammar for core common words are more resistant to change than those for less commonly used words (grammatically irregular words are usually commonly used words). Certain subject matter words are less resistant to change (e.g. toponyms and indigeneous flora and fauna names), while others are easily borrowed (e.g. trade goods and words related to newly acquired technologies). Major grammatical categories appear to be more stable than individual moderate to low frequency words. Independent innovation does happen, although borrowing is common too. Superstrate cultures have dramatically more linguistic influence than genetic influence in most cases, and fathers in superstrate cultures tend to have more linguistic influence on their children than mothers. Language shift tends to produce grammatical simplification, while languages tend to get more elaborate with stability over time and increasingly complex societies (points that are sometimes at odds with each other and sometimes reinforce each other).

    While I reach few global conclusions from my analysis, I reach half a dozen conclusions about the linguistic prehistory of Africa (although some new questions are also raised in the process related to the perennially vexing question of the relationship of the half dozen major subfamilies of the Afro-Asiatic macrolinguistic family). In particular, phonemes seem to be particularly useful suggesting when to hypothesize substrate influences in outlier languages in a language family.

    In the New World, phoneme data sheds light on the nature of the Na-Dene and Yenesian language connection (Na-Dene languages probably acquired its glottal consonants from a Native American substrate language), and also provides suggestive evidence regarding the languages of the founding population of the Americas (that probably had a full range of glottal consonants). It also provides insight into the prehistory of a Siberian language spoken near the Bering Strait (the only living language of several dozen in Siberia with glottal consonants) suggesting areal influence on it from Alaska.

    In Asia, the analysis tends to corroborate links between groups of Southeast Asian and South Chinese languages already suspected to be related on other lexical, grammatical and population genetic grounds, and it also tends to favor the deep level language family relatedness of Korean and the Japonic languages.

    It also provides another data point suggesting the strong relatedness of the Caucasian languages, either due to common origins, or strong areal influences at some point, and tends to attest to their antiquity.

    Finally, it adds new perspective to the question of whether the Proto-Indo-European language contained several laryngeal consonants that are absent in all attested Indo-European languages (the current majority view) except the Anatolian languages which had a substrate population that spoke languages with those consonants, although it doesn't resolve the question definitively.

    @1 I would normally consider language to be part of the content of culture, in addition to often being a litmus test that is usually, although not always accurate, for a host of highly correlated cultural content. Coherent cultures and communities are the environment in which languages come into being and live.

    @2-7 With regard to Atkinson, one of the big flaws in his historical linguistic studies is a failure to adequately consider the evidence that language change peaks during the period of language formation and intentional differentiation from its parent language, and the powerful impact of substrate and areal influences without which languages evolve much more slowly.

    Your argument that Anatolian borrowed the laryngeals from Caucasian doesn’t address the correspondence between the Anatolian laryngeals and various vowels in other Indo-European languages. Did Hittite just happen to borrow the sounds precisely where they would be predicted to exist by the comparative method?

    Read More
  12. @Jefferson
    I just got my 23AndMe results and found out that I am 99.5% European. To be more specific my ancestral breakdown is 80.4% Italian, 10.7% Non Specific Southern European, 6% Ashkenazi, 1.5% Balkan, 0.9% Non Specific Northern European, 0.4% Middle Eastern, and 0.1% unassigned.

    please stay on topic :)

    Read More
  13. As far as tracing language back o the Ur-language goes:

    1. Our first language records go back only 4500 years and the earliest records as non-phonetic. The affiliation of Sumerian is a mystery, whereas Egyptian seems clearly to be Semitic, but very different than any contemporary Semitic language. Language itself originated 100,000 years ago by a conservative guess. So we have a sketchy record of 5% of that time.

    2. If you look at the record of written languages, a high proportion of them extant in 1000 BC are extinct by now, though some have distant relatives. If you look at the extant languages, some belong to recognizable families with knowable histories but there still are many isolates and unclassifiable languages. An incredible amount of the record has been destroyed.

    3. There’s something called “areal effects” (the Sprachbund). This means that historically / genetically unrelated languages in an area will pass common traits back and forth. (Romanian, Bulgarian, Hungarian, Greek, and Turkish are a striking example).

    4. Many languages are pidginized and then creolized — stripped down to the simplified basics and then regularized into a working language. When this happens, evidence is lost.

    5. I have speculated that you might have complexification processes when a creolized language becomes the mother tongue of a defined group which continues in time. This is less accepted than the first three points but it seems quite reasonable and linguists I have talked to about it haven’t been negative.

    6. 4 and 5 together amount to alternating mixing and letting settle, with the mixing destroying information (pidgin-creole phase) and the complexification creating new forms and structures.

    7. After two or three cycles of 4 and 5, with the effects o 3 added, how much of the original information would be left? My guess is, not much. So my guess is that about 5-10,000 BC is the absolute maximum of what we could hope to know, probably less. Nowhere near the Ur-language.

    Read More
    • Replies: @syonredux

    The affiliation of Sumerian is a mystery, whereas Egyptian seems clearly to be Semitic, but very different than any contemporary Semitic language.
     
    That's not quite accurate.Egyptian is a branch of the Afro-Asiatic family of languages, which includes the Semitic languages, Omotic, Berber, Chadic, etc.
  14. @Philip Neal
    The use of phonemic inventories strikes me as entirely welcome. Phonemes (unlike vague semantic categories represented by English words written in upper case letters, e.g. WATER) are well-defined entities which can be objectively compared between languages and the set of all possible phonemes is sufficiently small to permit the use of all available data. If anything I would like to see a study on these lines which further analysed phonemes into their phonological features such as voicing or pharyngeal articulation, since sound change generally operates on features rather than entire phonemes.

    There’s a lot of problems with using phonemic inventories in this way. What Razib mentions in the post, that phonemes are subject to more horizontal transmission than, say, grammatical structures is part of the problem of course. Other issues are discussed here:

    http://languagesoftheworld.info/bad-linguistics/phonemic-diversity-prove-africa-theory.html

    and in my co-authored response to Atkinson’s paper in Science 335(6069): 657

    Read More
    • Replies: @pseudoerasmus
    I follow your blog but I missed that one. A great post !
    , @Philip Neal
    The points you make there are well taken but I think this whole approach is still awaiting its Fisher. I have just found where you went after Geocurrents - excellent site.
  15. eric says:

    Chomsky’s brief ‘Language and the Problems of Knowledge’ describes his basic theory of native language rather clearly. I remember as a child diagramming sentences, and looking back think they thought that, per Chomsky, they could find some basic structure. I think they gave up on that, as diagramming sentences didn’t really help learn syntax. Indeed, the whole Chomsky project just seems to present a puzzle, how something so complex is wired and how it got that way, rather than something we can build upon, say by using such insights in computer languages, which are being created all the time now (ruby, perl, python).

    Read More
  16. syonredux says:
    @John Emerson
    As far as tracing language back o the Ur-language goes:

    1. Our first language records go back only 4500 years and the earliest records as non-phonetic. The affiliation of Sumerian is a mystery, whereas Egyptian seems clearly to be Semitic, but very different than any contemporary Semitic language. Language itself originated 100,000 years ago by a conservative guess. So we have a sketchy record of 5% of that time.

    2. If you look at the record of written languages, a high proportion of them extant in 1000 BC are extinct by now, though some have distant relatives. If you look at the extant languages, some belong to recognizable families with knowable histories but there still are many isolates and unclassifiable languages. An incredible amount of the record has been destroyed.

    3. There's something called "areal effects" (the Sprachbund). This means that historically / genetically unrelated languages in an area will pass common traits back and forth. (Romanian, Bulgarian, Hungarian, Greek, and Turkish are a striking example).

    4. Many languages are pidginized and then creolized -- stripped down to the simplified basics and then regularized into a working language. When this happens, evidence is lost.

    5. I have speculated that you might have complexification processes when a creolized language becomes the mother tongue of a defined group which continues in time. This is less accepted than the first three points but it seems quite reasonable and linguists I have talked to about it haven't been negative.

    6. 4 and 5 together amount to alternating mixing and letting settle, with the mixing destroying information (pidgin-creole phase) and the complexification creating new forms and structures.

    7. After two or three cycles of 4 and 5, with the effects o 3 added, how much of the original information would be left? My guess is, not much. So my guess is that about 5-10,000 BC is the absolute maximum of what we could hope to know, probably less. Nowhere near the Ur-language.

    The affiliation of Sumerian is a mystery, whereas Egyptian seems clearly to be Semitic, but very different than any contemporary Semitic language.

    That’s not quite accurate.Egyptian is a branch of the Afro-Asiatic family of languages, which includes the Semitic languages, Omotic, Berber, Chadic, etc.

    Read More
  17. Tim says:

    Speaking of click languages, I know that most English speakers use clicks as “non-verbal” language. Especially when interacting with children and animals, and also when trying to be subtle about someone misbehaving. I know that my children picked up on this at a very early age, and talked to their stuffed animals with clicks.

    Is this common in other languages?

    Read More
    • Replies: @ohwilleke
    Making clicks in that context is not equivalent to using clicks as a linguistic phoneme. In the same vein, many languages use tone in a manner that has meaning (for example, in English, a raised tone at the end of a sentence can indicate that you are asking a question), but this is not equivalent to using tone as a phoneme substitute as Chinese and many SE Asian languages do.
  18. ohwilleke says: • Website
    @Tim
    Speaking of click languages, I know that most English speakers use clicks as "non-verbal" language. Especially when interacting with children and animals, and also when trying to be subtle about someone misbehaving. I know that my children picked up on this at a very early age, and talked to their stuffed animals with clicks.

    Is this common in other languages?

    Making clicks in that context is not equivalent to using clicks as a linguistic phoneme. In the same vein, many languages use tone in a manner that has meaning (for example, in English, a raised tone at the end of a sentence can indicate that you are asking a question), but this is not equivalent to using tone as a phoneme substitute as Chinese and many SE Asian languages do.

    Read More
  19. @Asya Pereltsvaig
    There's a lot of problems with using phonemic inventories in this way. What Razib mentions in the post, that phonemes are subject to more horizontal transmission than, say, grammatical structures is part of the problem of course. Other issues are discussed here:

    http://languagesoftheworld.info/bad-linguistics/phonemic-diversity-prove-africa-theory.html

    and in my co-authored response to Atkinson's paper in Science 335(6069): 657

    I follow your blog but I missed that one. A great post !

    Read More
  20. @Asya Pereltsvaig
    There's a lot of problems with using phonemic inventories in this way. What Razib mentions in the post, that phonemes are subject to more horizontal transmission than, say, grammatical structures is part of the problem of course. Other issues are discussed here:

    http://languagesoftheworld.info/bad-linguistics/phonemic-diversity-prove-africa-theory.html

    and in my co-authored response to Atkinson's paper in Science 335(6069): 657

    The points you make there are well taken but I think this whole approach is still awaiting its Fisher. I have just found where you went after Geocurrents – excellent site.

    Read More
    • Replies: @Asya Pereltsvaig
    Thank you for the kind words about my blog, Philip. It was my blog before GeoCurrents, so now I went back to it. I hope you find more interesting stuff there.

    But what do you mean by "this whole approach is still awaiting its Fisher"?
  21. @Philip Neal
    The points you make there are well taken but I think this whole approach is still awaiting its Fisher. I have just found where you went after Geocurrents - excellent site.

    Thank you for the kind words about my blog, Philip. It was my blog before GeoCurrents, so now I went back to it. I hope you find more interesting stuff there.

    But what do you mean by “this whole approach is still awaiting its Fisher”?

    Read More
    • Replies: @Philip Neal
    I was referring to RA Fisher - his advances in statistics and his role in putting natural selection on a mathematical basis. We know a great deal about what sound changes have actually taken place and little about why. When books about historical linguistics are as mathematical as the work of William Labov it might be time to think of language change in terms of evolution.
  22. @Asya Pereltsvaig
    Thank you for the kind words about my blog, Philip. It was my blog before GeoCurrents, so now I went back to it. I hope you find more interesting stuff there.

    But what do you mean by "this whole approach is still awaiting its Fisher"?

    I was referring to RA Fisher – his advances in statistics and his role in putting natural selection on a mathematical basis. We know a great deal about what sound changes have actually taken place and little about why. When books about historical linguistics are as mathematical as the work of William Labov it might be time to think of language change in terms of evolution.

    Read More
  23. I see what you meant now, thanks for the clarification. I don’t agree that we know little about why sound changes happened the way they did. Nor that language change is entirely parallel to biological evolution; the following concerns words, but phonemes aren’t good parallels to genes either, as I discussed elsewhere:

    http://languagesoftheworld.info/bad-linguistics/words-just-like-genes.html

    Read More

Comments are closed.

Subscribe to This Comment Thread via RSS Subscribe to All Razib Khan Comments via RSS
PastClassics
Confederate Flag Day, State Capitol, Raleigh, N.C. -- March 3, 2007
The major media overlooked Communist spies and Madoff’s fraud. What are they missing today?
Are elite university admissions based on meritocracy and diversity as claimed?
The “war hero” candidate buried information about POWs left behind in Vietnam.
The evidence is clear — but often ignored