What’s the Oldest Language?

A very silly and pointless article:

The globe hums with thousands of languages. But when did humans first lay out a structured system to communicate, one that was distinct to a particular area? […] Language comes in different forms—including speech, gestures and writing—which don’t all leave conclusive evidence behind. And experts use different approaches to determine a language’s age.

Tracing the oldest language is “a deceptively complicated task,” says Danny Hieber, a linguist who studies endangered languages. One way to identify a language’s origins is to find the point at which a single tongue with different dialects became two entirely distinct languages, such that people speaking those dialects could no longer understand each other. “For example, how far back in history would you need to go for English speakers to understand German speakers?” he says. That point in time would mark the origins of English and German as distinct languages, branching off from a common proto-Germanic language.

Alternatively, if we assume that most languages can be traced back to an original, universal human language, all languages are equally old. “You know that your parents spoke a language, and their parents spoke a language, and so forth. So intuitively, you’d imagine that all languages were born from a single origin,” Hieber says.

But it’s impossible to prove the existence of a proto-human language—the hypothetical direct ancestor of every language in the world. Accordingly, some linguists argue that the designation of the “oldest language” should belong to one with a well-established written record.

So we get a roundup of Sumerian, Akkadian and Egyptian (“the oldest languages with a clear written record”), Hebrew and Arabic (“both belong to the Afroasiatic language family, whose roots trace back to 18,000 to 8,000 B.C.E. […] contemporary linguists widely accept Afroasiatic as the oldest language family”), Chinese (“The language likely emerged from Proto-Sino-Tibetan, which is also an ancestor to Burmese and the Tibetan languages, around 4,500 years ago, although the exact date is disputed”), Sanskrit (“‘In my view, Sanskrit is the oldest continuous language tradition, meaning it’s still producing literature and people speak it, although it’s not a first language in the modern era,’ Patel says”), and Tamil (“Tamil [speakers] have been especially [enthusiastic] in trying to separate the language as uniquely ancient”). The conclusion:

Disagreements about the age of Sanskrit and Tamil illustrate the broader issues in pinpointing the world’s oldest language. “To answer this question, we’ve seen people create new histories, which are as much political as they are scientific,” Patel says. “There are bragging rights associated with being the oldest and still evolving language.”

The curious reader will learn pretty much nothing from all this, which seems pure clickbait. As Craig, who sent me the link, says: “Disappointing from Scientific American.”

Comments

  1. Belgian, duh.

  2. David Eddyshaw says

    KONGO!

    Actually, given the deeply stupid premises of TFA, Greek has a better claim than either Sanskrit or Chinese.

    https://en.wikipedia.org/wiki/Mycenaean_Greek
    https://en.wikipedia.org/wiki/Vedic_Sanskrit
    https://en.wikipedia.org/wiki/Oracle_bone

    And Hausa is obviously the oldest AA language, and therefore WINS!!!

  3. David Eddyshaw says

    Of course, that nice paper on the evolution of complex grammars showed that “Atlantic-Congo” is itself one of the two primary branches of Proto-World, the other being Everything Else.

    https://languagehat.com/the-evolution-of-complex-grammars/#comment-4546920

    So KONGO it is, after all. AA is a mere mayfly.

    Kusaal is therefore the oldest language, not Hausa. (I knew it all along.)

  4. David Eddyshaw says

    I am unwilling to click through to the actual article on hygiene grounds. Does it miss the trick of labelling Khoisan as the oldest, Because Clicks? You would have expected such an author to throw that in somewhere …

  5. @DE So you’re accusing the clickbait of being Click-bait?

  6. David Eddyshaw says

    Indeed.

  7. No, I provided all the candidates.

  8. Sanskrit Is The Oldest, because moon landing.

  9. David Eddyshaw says

    Good point.

  10. David Eddyshaw says

    I was expecting someone to suggest Welsh: however, this common belief is based on a confusion.

    Welsh is the oldest language in Europe, having been brought originally by Neanderthal Man.*

    This is why Welsh is VSO.

    * And, in all probability, Neanderthal Woman, although it is possible that she was Basque.

  11. This is why Welsh is VSO.” – Neanderthal man of action.

    I dunno, the excerpts here seem totally innocent. Yes, linguists don’t call langauges old (occasionaly someone may call something “young”).

    But a review of possible meanings of “old” as applied to langauge sounds like a good idea…

  12. Is this the thread to ask about this piece of research that’s busting my news feeds? This surely identifies who it was speaking the world’s oldest language.

    Results showed that human ancestors went through a severe population bottleneck with about 1280 breeding individuals between around 930,000 and 813,000 years ago. The bottleneck lasted for about 117,000 years and brought human ancestors close to extinction.
    [needs subscription; Guardian report here or syndicated almost everywhere]

    This appears _not_ to be saying that at some time within a 117,000 year timespan, human population dropped to 1280. But rather it remained at that level for like a third of a million generations(?)/longer than recorded history of languages. Ample time to go from agglutinative to inflected to tonal to clicks to analytic and back again.

    Is there any population of any (mammal) species that’s dropped that low for that long and survived?

    “The gap in the African and Eurasian fossil records can be explained by this bottleneck in the Early Stone Age as chronologically. It coincides with this proposed time period of significant loss of fossil evidence,” says senior author Giorgio Manzi, an anthropologist at Sapienza University of Rome.

    Is this merely absence of evidence? Or does it legitimately stand as evidence of absence?

  13. for like a third of a million generations(?)
    In 117,000 years? That would mean a generation lasting less than 4 months… or did I misunderstand your calculation?

  14. Trond Engen says

    117 000 years is about 4000-5000 generations.

    Hu et al: Genomic inference of a severe human bottleneck during the Early to Middle Pleistocene transition, Science, 2023

    Editor’s summary
    Today, there are more than 8 billion human beings on the planet. We dominate Earth’s landscapes, and our activities are driving large numbers of other species to extinction. Had a researcher looked at the world sometime between 800,000 and 900,000 years ago, however, the picture would have been quite different. Hu et al. used a newly developed coalescent model to predict past human population sizes from more than 3000 present-day human genomes (see the Perspective by Ashton and Stringer). The model detected a reduction in the population size of our ancestors from about 100,000 to about 1000 individuals, which persisted for about 100,000 years. The decline appears to have coincided with both major climate change and subsequent speciation events. —Sacha Vignieri

    Abstract
    Population size history is essential for studying human evolution. However, ancient population size history during the Pleistocene is notoriously difficult to unravel. In this study, we developed a fast infinitesimal time coalescent process (FitCoal) to circumvent this difficulty and calculated the composite likelihood for present-day human genomic sequences of 3154 individuals. Results showed that human ancestors went through a severe population bottleneck with about 1280 breeding individuals between around 930,000 and 813,000 years ago. The bottleneck lasted for about 117,000 years and brought human ancestors close to extinction. This bottleneck is congruent with a substantial chronological gap in the available African and Eurasian fossil record. Our results provide new insights into our ancestry and suggest a coincident speciation event.

    Dmitry will actually know something about this, but I was surprised that they’re able to estimate the duration of the bottleneck episode. I would have thought that a prolonged bottleneck would show up as a smaller effective population size at a point a little before the population expanded. All the more power to genomic inference if they can.

    The paper is not open access, but the Materials and Methods supplement is. I’ll try to read it.

  15. Giacomo Ponzetto says

    The published version is not open access, but there is a 2021 preprint on BioRxiv: Genomic inference of a human super bottleneck in Mid-Pleistocene transition.

    SUMMARY
    The demographic history is a foundation of human evolutionary studies. However, the ancient demographic history during the Mid-Pleistocene is poorly investigated while it is essential for understanding the early origin of humankind. Here we present the fast infinitesimal time coalescent (FitCoal) process, which allows the analytical calculation of the composite likelihood of a site frequency spectrum and provides the precise inference of demographic history. We apply it to analyze 3,154 present-day human genomic sequences. We find that African populations have passed through a population super bottleneck, a small effective size of approximately 1,280 breeding individuals between 930 and 813 thousand years ago. Further analyses confirm the existence of the super bottleneck on non-African populations although it cannot be directly inferred. This observation, together with simulation results, indicates that confounding factors, such as population structure and selection, are unlikely to affect the inference of the super bottleneck. The time interval of the super bottleneck coincides with a gap in the human fossil record in Africa and possibly marks the origin of Homo heidelbergensis. Our results provide new insights into human evolution during the Mid-Pleistocene.

  16. … or did I misunderstand your calculation?

    Errk. It’s me that didn’t understand my calculation, so thanks. And thanks @Trond for a better number. That’s still way more generations than I can get my head round. Is it even the same species after that much generationing?

  17. Trond Engen says

    @Giacomo: Thanks!

    That summary seems to answer one of the questions I had: How do you see an earlier bottleneck through more recent bottlenecks? The answer is that you don’t, but the results of the old bottleneck will still be evident in all descendant populations. Here they used that effect to treat non-African (i.e. later bottleneck) populations as test cases.

  18. Trond Engen says

    @AntC: The authors speculate that what came out of the bottleneck was effectively a different species, Homo heidelbergensis: A prolonged bottleneck episode is a major selection event. Most of the selection will just be genetic drift, random selection, inbreeding effects, whatever you choose to call it, but it’s also a chance for fast species-wide spread of useful traits (genetic or, I guess, cultural). Every time that happened, an initially small subset of the population would (by definition) have had a strong advantage in the competition for mates and/or resources. A selected subset producing a large proportion of the productive descendants would have reduced the effective population size.

    One such useful trait might have been sophisticated language.

  19. One such useful trait might have been sophisticated language.

    You mean we’ve found the Merge gene?

    The hypothesis supposes the emergence and fixation of a single mutant (capable of the syntactic operation Merge) during a narrow historical window as a result of frequency-independent selection under a huge fitness advantage in a population of an effective size no larger than ~15 000 individuals.
    [They go on to conclude:]
    Our results cast doubt on any suggestion that evolutionary reasoning provides an independent rationale for a single-mutant theory of language.

    De Boer et al need to re-run their numbers for “1280 breeding individuals”.

  20. Trond Engen says

    No, I mean some useful but essentially random assembly of genes and cultural traits. The smaller the population, the more would such an advantage mean in the descendant population. Even if the advantage was entirely cultural, as the origin of human language may or may not have been, its random collocation with genes would mean that those genes had an advantage. The larger the population, the more time for a piggybacking set of genes to be watered out.

  21. David Eddyshaw says

    [Berwick and Chomsky] argue that evolutionary considerations represent an independent motivation for the single-mutant theory, because the narrow historical time-window in which the human language faculty is said to have emerged rules out the emergence and fixation of more than one language-relevant genetic anomaly.

    The form of this argument appears to be

    A. Human language ability reduces to just one central cognitive ability, which is either fully present or completely absent

    B. Human language arose de nihilo rapidly because it must have done, because it has only one unique feature, and that feature can’t have evolved slowly because was either present or absent.
    [There is no actual evidence that human language emerged rapidly in essentially its current state, but I’ll let that slide.]

    Therefore

    Human language ability reduces to just one central cognitive ability.which is either fully present or completely absent.

    Such piercing analysis leaves me awestruck. Speechless, even. (The syllogism appears to be logically valid. And yet …)

    Incidentally, while it is futile to attempt to engage intellectually with Chomskyanism, Homo heidelbergensis is surely much too early to fit ANC’s fantasy scenario. His “evidence”, I think, depends on circular reasoning about the Upper Palaeolithic.

  22. Trond Engen says

    I mean, the introduction of agriculture goes hand in hand with a major change in the genetic composition of the population. Still, nobody suggests that the genetic change is what led to agriculture. It’s just that agriculture arose in certain random subsets of the total human population, and those subsets got what turned out to be an evolutionary advantage.

    But also, as agriculture spread and time went by, in what was now a global human population, the random founder genome would be more and more watered out by admixture with pre-agriculture populations.

  23. I mean some useful but essentially random assembly of genes and cultural traits.

    The form of this argument appears to be …

    Chaps! Call off the heavy artillery! I haven’t suddenly started drinking the ANC Kool-aid.

    (I didn’t think it would be that easy to wind yous up. On the Internet nobody can hear you snark.)

  24. “ a population super bottleneck, a small effective size of approximately 1,280 breeding individuals between 930 and 813 thousand years ago”

    This phrase could be read as indicating the uncertainty in the timing of the bottleneck, without specifying its duration.

  25. David Eddyshaw says

    It certainly goes with the startling lack of genetic diversity.

    No wonder Martians say that we all look alike. Practically Clone People. It probably accounts for the way we all speak the same language, too. Nightmare stuff, really. You can’t really blame them for hiding from us while they gather strength. It is only too natural to fear the Attack of the Merge People.

    Resistance is futile! You will be Merged!

  26. @AntC: Not wound (up or otherwise), and appreciating the antichomskian snark. I just realized that I needed to explain myself better.

  27. Me: those subsets got what turned out to be an evolutionary advantage

    I meant “a competitive advantage”. If I didn’t, I should have,

  28. david: This phrase could be read as indicating the uncertainty in the timing of the bottleneck, without specifying its duration.

    Very much so. It’s one of the things one hopes to be wiser about after getting around to read the actual paper.

  29. This phrase could be read as indicating the uncertainty in the timing of the bottleneck, without specifying its duration.

    The former, surely. A population of 1280 people wouldn’t remain that size for 53,000 thousand years. It would either expand or vanish.

    These numbers, moreover, seem unreasonably precise, which makes me doubtful of the analysis.

  30. just chiming in for this coinage: “fast infinitesimal time coalescent process (FitCoal)”.

    i’m sure* it has a perfectly reasonable technical meaning, but it reads as pure 1980s sf vat-grown technobabble, probably about collapsing state vectors using laser-assisted temporal condensation or resolving multiple-timeline contradictions through superstring engineering. i love it! (and the determinedly opaque abbreviation only makes it better)

    .
    * for a value of “sure” asymptotically approaching zero.

  31. jack morava says

    well done rozele!

  32. This phrase could be read as indicating the uncertainty in the timing of the bottleneck, without specifying its duration.” – The duration is specified elsewhere… (the comment right above Giacomo’s)

    Very much so.” – Am I missing something?

    ___
    I’m seriously surprised that duration can be determined. Hopefully, Dmitry will confirm that it can….

  33. Is this merely absence of evidence? Or does it legitimately stand as evidence of absence?

    Sometimes the first stands in for the second, as in Jimmy Carter and the American Nazi Party.

  34. Results showed that human ancestors went through a severe population bottleneck with about 1280 breeding individuals between around 930,000 and 813,000 years ago. The bottleneck lasted for about 117,000 years and brought human ancestors close to extinction.

    I didn’t notice that explanation (and I can’t subtract, evidently). But I find this conclusion very improbable. A population remained stable at a small number like that for over 100,000 years? For a few generations, sure, but for that length of time? As a matter of population dynamics it seems preposterously unlikely. In the modern world, when an animal population falls to that sort of size, they are either on their way to imminent extinction or can be sustained only with substantial help and safeguards.

  35. It is still effective population size.

    Obviously, when you always marry within your family and compete with other similar groups (eventually replacing them) the effective size must be no larger than the size of your family.

    I suspect there must be many other scenarios that will make your measure (whatever it is) identical to what it would be for a total of 1000 people on Earth….

  36. the effective size must be no larger than the size of your family

    You (or rather your descendants only a few generations on) will end up breeding with cousins. I was under the impression inbreeding is a Bad Thing. (We could point to the royal families in Europe: Albert was first cousin to Victoria — a not unusual proximity amongst royals at the time. I cast not aspersions on those two, but the current shower seem several pennies short of the full shilling.) At least somebody got some things right: haemophilia seems to be eliminated.

    “Animals avoid incest only rarely.” says wp; “Common fruit fly females prefer to mate with their own brothers over unrelated males.”

    Then does inbreeding have deleterious affects only on mammals? On hominids? Is it all a myth?

    So I’m awaiting the Hattery’s full brains trust to weigh in on whether the claims in that paper make any sort of sense. If the whole breeding population dwindles to 1280 — even if only briefly — has that gotten below some sort of threshold? Even if the population later recovers, I guess everybody’s passing around the same narrow selection of genes?? Or does random variation wipe out that effect pretty quickly?

    We should also remember homo sapiens interbred with Neanderthals fairly promptly when they encountered them — around 15% of our genes from that is the number I hear.

  37. Trond Engen says

    @David L: I took the rounder numbers in the editorial summary as a mild critique of the too precise numbers in the paper.

    @AntC: Inbreeding is bad because deleterious genes may accumulate and spread quickly through the breeding community, and also because small communities will keep losing genetic diversity by chance. But if a useful innovation shows up, it too will have a better chance of spreading through the community,

    I’ve been thinking that the more a species is relying on offspring quality — i.e. few births, long childhood and long life — the better its strategies to avoid close-kin mating. The more it relies on quantity — i.e. quick regeneration in large numbers — the less that means.

    I’ve read the 2021 preview. I gave up on the maths, but I’ll try again. But one reaction is that the bottleneck may be real, but we shouldn’t necessarily trust the estimate of timing or population size. The conclusion in the paper that the bottleneck can’t be seen in non-African populations, but instead shows up as a reduced effective population after the bottleneck compared to African populations, makes me think that e.g. multiple bottleneck events in the same population might be merged into one and would yield a smaller effective population estimate than if the actual events could be singled out.

    It’s not known if Neanderthals and Denisovans branched off before or after the great bottleneck. Current estimates, using other genetic methods, are too wide to tell, and this work is primarily based on African populations with little Neanderthal or Denisovan admixture.

    @drasvi: The effects of positive selection are said to be ruled out by excluding quickly changing genes from the analysis. Good, I think, but I don’t know how that would work for old selection for now species-wide genes. And even if that can be ruled out as well, a population replacement based on cultural or technological advantage would have had a whole genetic profile piggybacking along, but it may not have been genetic selection at all.

  38. David Marjanović says

    Is there any population of any (mammal) species that’s dropped that low for that long and survived?

    I don’t think the fossil record is ever good enough for that.

    Is it even the same species after that much generationing?

    Yes under some definitions of “species”, maybe under others, no under some…

    [Berwick and Chomsky] argue that evolutionary considerations represent an independent motivation for the single-mutant theory, because the narrow historical time-window in which the human language faculty is said to have emerged rules out the emergence and fixation of more than one language-relevant genetic anomaly.

    …did they believe mutations have to wait for each other?

    Then does inbreeding have deleterious affects only on mammals? On hominids? Is it all a myth? […] has that gotten below some sort of threshold?

    Inbreeding doesn’t cause any mutations. What’s going on, mainly, is that every one of us has one copy of a few mutations that are lethal if they’re present in both copies of the genes in question. If parents have the same such mutations, for example because they inherited them from a common ancestor, the likelihood that their offspring will inherit one such mutation from both sides and die from it is higher.

    So, there are no thresholds, only probabilities.

    around 15% of our genes from that is the number I hear.

    News to me. It’s around 2%, never higher than 3% since the Last Glacial Maximum.

    * for a value of “sure” asymptotically approaching zero.

    The coalescent is perfectly cromulent.

  39. If parents have the same such mutations, for example because they inherited them from a common ancestor, the likelihood that their offspring will inherit one such mutation from both sides…

    This is something that always perplexed me: I would expect frequency of such alleles to drop in such population. I don’t know how much and whether we would just obtain a population with a larger proportion of healhy children this way, it it will be smaller proportion or just the same, but definitely there will be some selective pressure.
    I’m just too lazy to read more about this, so what perplexed me is the popular understanding of the process (which is: inbreeding leads to what you just described).

    (Anyway, there are endogamous groups on Earth)

  40. David Marjanović says

    I would expect frequency of such alleles to drop in such population.

    Yes, once enough individuals have died.

    And then the next lethal mutation that appears will have the same effect…

  41. @DM, which is relatively important if the expectation of the number of children whose health is damaged by it is comparable to the number of children whose health is damaged by other mutations which will happen over the average period between two of them.

    Practically there are communities where marrying uncles and first cousins is rather the rule than an exception (e.g. some Arabs), so this is quite interesting.

  42. Trond Engen says

    I saw a study not long ago that found very little close-cousin mating in analysed ancient DNA. It seems that cousin marriages is a fairly recent invention. That rises questions of how and when it arose, and what made it the dominant strategy in some societies.

  43. Trond Engen says

    Ringbauer et al: Parental relatedness through time revealed by runs of homozygosity in ancient DNA, Nature 2021

    Abstract
    Parental relatedness of present-day humans varies substantially across the globe, but little is known about the past. Here we analyze ancient DNA, leveraging that parental relatedness leaves genomic traces in the form of runs of homozygosity. We present an approach to identify such runs in low-coverage ancient DNA data aided by haplotype information from a modern phased reference panel. Simulation and experiments show that this method robustly detects runs of homozygosity longer than 4 centimorgan for ancient individuals with at least 0.3 × coverage. Analyzing genomic data from 1,785 ancient humans who lived in the last 45,000 years, we detect low rates of first cousin or closer unions across most ancient populations. Moreover, we find a marked decay in background parental relatedness co-occurring with or shortly after the advent of sedentary agriculture. We observe this signal, likely linked to increasing local population sizes, across several geographic transects worldwide.

    I suppose there could be sampling bias. There’s still not much DNA from ancient Egypt and Persia, where there’s written evidence that even sibling unions may have been socially acceptable, also outside the ritual marriages of the royal families, or from South India, where the “Dravidian” cross-cousin marriage system got its name.

  44. @drasvi, David Marjanović: There are basically two reasons that deleterious mutations persist in a population. One was mentioned above, that new mutations arise, and eventually an equilibrium may be established, at which point the rate of new appearances of a mutation equals the rate at which they are lost from the gene pool, due to the death of individuals who are homozygous for the mutation.

    In spite of the comparatively inbred character of nineteenth-century European royalty, their most famous genetic abnormality, the hemophilia mentioned above, did not have anything to do with inbreeding. The gene arose as an example of a new mutation, probably a germ line mutation in Victoria’s father. Sperm quality decreases with a father’s age, since the gametes are produced from cells that continue to divide throughout his lifetime, accumulating mutations along the way.* (After two generations of legitimate Hanoverian heirs were wiped out with the loss of Princess Charlotte and her child in childbirth, George III’s remaining unwed sons, even though they were middle-aged, rushed to marry and have children, to keep the dynasty going.) Moreover, inbreeding did not make the hemophilia problem worse, either. Since the condition is sex linked, meaning it corresponds to a mutation in a gene on the X chromosome, it manifests itself among males who have inherited just one defective gene from their mothers.**

    When we talk about a mutation remaining present in a population because mutant alleles keep being generated by replication errors, we should understand, of course, that the new mutations are not typically identical with the old ones. At the level of the DNA, they are typically completely different, but they have in common that they make the protein the gene codes for nonfunctional (completely or partially). However, there is also another, more subtle reason why a mutant alleles may persist in a population. That happens if a mutation makes a gene produce a protein that is nonfunctional for its normal purpose, but which is useful in a different way! (Actually, it may be enough if having the mutant version just slows down the normal function of the protein, as in the example described below.) If progeny that are homozygous for a mutant form die, but heterozygotes are actually more fit than the average population, there can again be an equilibrium level of the mutation present in the gene pool. A notable example is Tay-Sachs disease, which is found in a number of European-descended populations, especially Ashkenazi Jews. Children with two copies of the defective gene suffer slow neural degeneration and death, due to an inability to manage the composition of phospholipids in their nerve cells. It’s a terrible disease. However, heterozygotes actually turn out to be more resistant to tuberculosis than people without the allele. In this case, the mutant version of the protein involved probably does not actually have any alternative enzymatic function. Instead, the small, subclinical changes in neuronal membrane composition that arise from having only half the usual number of active copies of the enzyme probably enhance TB resistance. The precise mechanism is unclear, but Tay-Sachs is actually one of several genetic diseases that involve changes to phospholipid management and which were historically found in areas where TB was endemic.

    * In contrast, for a mother, all the eggs complete meiosis I before she is even born, then wait decades to undergo meisos II and be released during ovulation. This preserves genome quality, which makes sense. Men can practice either r or K strategies for reproduction; many high-status men have traditionally engaged in both. However, women are tied to a K strategy by the long gestation and nursing periods.

    ** I remember thinking that, had Paul Alexis, the victim in Have His Carcase, understood the inheritance pattern of hemophilia, he might not have been murdered.

  45. @Trond, for Arabs one obvious correlate is inheritance (though cousin marriages can also happen in exogamous societies when two clans systematically intermarry, so it is not just Arabs).

    Particularly, among Arabs parallel cousin marriage is widespread (differently from exogamous societies). Muslim women inherit a share of property, so one can see how such marriage can be conveineint for at least land owners. But neither it is limited to Muslim socieites in the ME (besides it existed in pre-Islamic times and attested in the Bible – maybe Brett/rozele/Y can add something here) – nor it is common to all Muslim societies.

    So the explanation does not seem to work well, even though synchronously “property” is indeed one of concerns.

    @Brett, equilibrium: but drift. There is no reason why an allele will have an equilibruim frequency in absence of extra pressures, it is supposed to just walk randomly, and in a small population it may reach the point where it is the only allele or the point where no one has it in a reasonable time. The alleles in question techically can’t reach the former point, and hardly can reach even 50% (everyone has it, but only one copy)….

  46. David Eddyshaw says

    Cousin marriage is (or was) common among the Akan, on account of their being matrilineal: the purpose indeed being to keep the paternal-side property together,

  47. David Eddyshaw says

    I had some Tamil friends who were cousins and married to each other; common in their culture, they told me. (They were Christians, but of Buddhist background,)

  48. By the latest estimates, Easter Island had a stable population of ca. 3,000 for several centuries.

  49. David Eddyshaw says

    Charles Darwin was the grandson of first cousins, and married his own – he later carried out research into inbreeding in plants, motivated by how prone his his family was to disease and infertility.

    https://therake.com/stories/code/pedigree-chums-a-history-of-aristocratic-incest/

  50. Darwin married his cousin too:)

    P.S. Didn’t see DE’s comment, but won’t delete it.

    P.P.S. WP about Darwins:
    The writings of Scottish deputy commissioner for lunacy Arthur Mitchell claiming that cousin marriage had injurious effects on offspring were largely contradicted by researchers such as Alan Huth and George Darwin.[87][88] In fact, Mitchell’s own data did not support his hypotheses and he later speculated that the dangers of consanguinity might be partly overcome by proper living. Later studies by George Darwin found results that resemble those estimated today. His father, Charles Darwin – who married his first cousin – had initially speculated that cousin marriage might pose serious risks, but perhaps in response to his son’s work, these thoughts were omitted from a later version of the book they published. When a question about cousin marriage was eventually considered in 1871 for the census, according to George Darwin, it was rejected on the grounds that the idle curiosity of philosophers was not to be satisfied.[89]
    ….
    Higher total fertility rates are reported for cousin marriages than average, a phenomenon noted as far back as George Darwin during the late 19th century.

  51. Trond Engen says

    Yes, yes, inheritance of land. but as Ringbauer et al make clear — surprisingly, to me — it was rare in the stone age, and it did not become more common with agriculture anywhere in their sample. We need a better theory or at least the confounding factors that make cousin marriage a reasonable strategy in an agricultural society

    Matrilineal inheritance is mentioned, but that doesn’t quite explain it. You may just find your son a partner with land elsewhere, just as well as your daughter. My just-so story is that it’s about maintaining existing patriarchal alliances. With matrilineal inheritance of land, your son will have to make new alliances if he moves far away, and so will your son-in-law who’s supposed to uphold the inherited property of your daughter, so you may eliminate a lot of uncertainty by just exchanging sons with the next patriarch over. Do that twice and its cousin marriage. Do it thrice and it’s a stable system for keeping the clan together. And at some point you may decide that it’s much easier to move the daughters.

    But that doesn’t explain it either. The just-so story needs something that makes the existing alliance and the undivided property more important than it used to be. Maybe it’s agricultural methods that allow the use of the same land over generations (and thereby property as non-degradable value), maybe it’s scarcity of new land and opportunities elsewhere (increasing the value of property on hand), maybe it’s endemic warfare (increasing the value of stable alliances). Maybe they’re connected and it’s all of the above.

  52. it was rejected on the grounds that the idle curiosity of philosophers was not to be satisfied.

    Quite right too. Let them satisfy their own idle curiosity!

  53. By the latest estimates, Easter Island had a stable population of ca. 3,000 for several centuries.

    Several centuries is a whole lot less than 117,000 years. And Easter Island is an isolated place with limited resources and no predators that would affect the human population.

    I suppose it’s possible that 1280 people lived on an island in the middle of a lake and occasionally some bright spark would say, hey, I wonder what’s going on in that jungle on the other side of the water, maybe we should go check it out. And that person was quickly put to death, but after the 3,647th person said that, the elders said, huh, why not?

    Seems unlikely, though.

  54. David Eddyshaw says

    You may just find your son a partner with land elsewhere, just as well as your daughter

    Not if (like any good patriarch) you want your son to inherit your (or your father’s) property.

    The Akan are matrilineal, but they are not matriarchal. (Unfortunately, actual old-time matriarchal societies seem to belong to feminist wishful thinking.)

    Of putative prehistoric societies I cannot speak (and neither can anyone else, I reckon.)

  55. Trond Engen says

    Oh, I can speak, but not with any credibility.

  56. Trond Engen says

    David L. I suppose it’s possible that 1280 people lived on an island in the middle of a lake

    It may have been a metaphorical island in a metaphorical lake. The writeup in Forskning.no (and now also Dmitry on Facebook) relates this to the merger of what is still two chromosomes in our closest relatives. Forskning.no links to:

    Poszewiecka et al: Revised time estimation of the ancestral human chromosome 2 fusion, BMC Genomics 2022.

    Abstract

    Background
    The reduction of the chromosome number from 48 in the Great Apes to 46 in modern humans is thought to result from the end-to-end fusion of two ancestral non-human primate chromosomes forming the human chromosome 2 (HSA2). Genomic signatures of this event are the presence of inverted telomeric repeats at the HSA2 fusion site and a block of degenerate satellite sequences that mark the remnants of the ancestral centromere. It has been estimated that this fusion arose up to 4.5 million years ago (Mya).

    Results
    We have developed an enhanced algorithm for the detection and efficient counting of the locally over-represented weak-to-strong (AT to GC) substitutions. By analyzing the enrichment of these substitutions around the fusion site of HSA2 we estimated its formation time at 0.9 Mya with a 95% confidence interval of 0.4-1.5 Mya. Additionally, based on the statistics derived from our algorithm, we have reconstructed the evolutionary distances among the Great Apes (Hominoidea).

    Conclusions
    Our results shed light on the HSA2 fusion formation and provide a novel computational alternative for the estimation of the speciation chronology.

    This merger could have isolated a sub-population reproductively. People crossing the metaphorical lake in any direction may just not have had reproductive offspring. The 46 chromosome population lingered on the brink of extinction for a long time, until something happened and it started spreading.

    OTOH, a small population is exactly what you need for such a strange and probably detrimental mutation (at least while in minority) to take over, so maybe the isolation came first. Or maybe (see me above) there were several events.

    David E.: Not if (like any good patriarch) you want your son to inherit your (or your father’s) property.

    That’s just a cultural convention that might have been otherwise, but you have more trust in the loyalty of sons than sons-in-law. Also, of course, a daughter may die in childbirth, and then the son-in-law takes a new wife. Either way, the patriarchy decides that it will be better off exchanging daughters.

    The Akan are matrilineal, but they are not matriarchal. (Unfortunately, actual old-time matriarchal societies seem to belong to feminist wishful thinking.)

    Matriarchal societies seem to be a romantic fantasy, but matrilineal and matrilocal societies do exist. I want to say (but see my previous comment) that in agricultural societies where women do the cultivation, matrilocality makes so much sense that it should be expected and patrilocality need an explanation.

  57. David Eddyshaw says

    The Kusaasi are patrilineal and patrilocal (and you have to marry outside your own clan.)

    However, in Kusaasi culture and many neighbouring cultures, there is a specially close relationship between a man and his sister’s children, who refer to each other by related (in some languages, identical) terms unconnected with other family-relationship words; this is all the more striking as many core family relationship words, including all those for siblings, cousins, grandparents, grandchildren and parents-in-law, take no direct account of the sex of the person referred to. Even ‘son’ and ‘daughter’ are by default both usually just expressed as biig ‘child.’

    Whether all this reflects some previously matrilineal system I do not know. AFAIK there’s no other evidence for it.

  58. Matriarchal societies seem to be a romantic fantasy” – I think it is very difficult to define what is X-archy.

    Also imagine for a moment that the traditional European account of male and female forms of power (I mean, there are always people who say that it is women who actually contorol everything) is accurate, that most women are simply not interested in specifically male institutions of power, but that those institution don’t represent the Real Power. HOW do you determine whether this is correct or not?

  59. A database of hypohteses:

    https://hraf.yale.edu/ehc/variables?q=cousin+marriage

    Unfortunately you can’t just click “cousin marriage” and see a list of all hypotheses about it, “cross-cousin marriage” has a page of its own, and so does “cross-cousin marriage preference” etc. (e.g. “Flinn, Mark V. 1981 Cross-cousin marriage preference will be related to source of altruism (456). Supported “), and lists of hypotheses are different.
    ___
    “Marriage for love” contains a tautological hypothesis: “The degree to which a woman’s marriage is arranged by her family will be negatively related to the use of marriage for love (17).”

    Arabs appear here

  60. TE’s just so story makes intuitive sense to me, and my bid for the something more – or, maybe better, the materiality of the “all of the above” – is that the kinds of societies that conventionally get counted as “agricultural” (grain-tax-based states, basically*) are based on tying cultivators to the land in one or another way (so relationships/alliances have to be local to make a difference), result in massive reductions in the variety and quantity of resources available to cultivators (so the strength and durability of relationships/alliances matters more), and have structural imperatives towards organized violence** (again increasing the importance of strong local relationships).

    separately: i’ve got no speculation about the reasons***, but close-cousin-marriage was unremarkable if not preferred in much of the traditional yiddish jewish world. i’ve got at least one that i know of in my family tree, i think a generation before emigration (so late 19thC). my understanding is that that (in combination with the prevalence of TB, presumably) is where the concentration of Tay-Sachs in those communities comes from.

    another element in all the cousin-marriage stuff is, it might be worth mentioning, the ways that different kinship systems work. through some varieties of ethnographic lens, “first cousin marriage” can look preferred or avoided without being anything of the kind. all it takes is a context where “parents’ siblings’ children” isn’t a category of relationship, but where all or some parents’ siblings’ children fall into categories structured in other ways that have implications for the un/desirability of a match. and that’s easy.

    .
    * as opposed to non-state societies whose nutritional webs include various forms of cultivation (as usual, i’m leaning on scott’s Against the Grain)

    ** between grain-tax states competing for land and populace; against non-state polities to ‘recruit’ cultivators or alter trade relations; against cultivators seeking to leave.

    *** outside of the “yikhes” caste system of the economic/religious elite, where, as with the european royal lineages and any hereditary elite, in-marriage is always about consolidation of power.

  61. The coalescent is perfectly cromulent.

    well, damn. i suppose so.
    if anyone wants to tell me that the “Fit” part is also sensible, i’ll be tied to the mast with wax in my ears.

  62. Matrilineal inheritance is mentioned, but that doesn’t quite explain it. You may just find your son a partner with land elsewhere, just as well as your daughter.

    Fragmentation. When both boys and girls inherit land.

    (but again the actual distribution doesn’t match the pridiction… And who siad that sexual behaviour of humans and associated customs must be determined by economy? )

  63. @rozele another element in all the cousin-marriage stuff is, it might be worth mentioning, the ways that different kinship systems work.

    [Rudi] Giuliani married Regina Peruggi, his second cousin, whom he had known since childhood, … The Giuliani-Peruggi marriage legally ended in two ways: a civil divorce was issued by the end of 1982,[482] while a Roman Catholic church annulment of the marriage was granted at the end of 1983,[481] reportedly because Giuliani had discovered that he and Peruggi were second cousins.

    ‘Please explain’ how you could know someone since childhood but not know you were related. Or is this the sort of murkiness as to facts [**] I should expect from a Catholic-Italian immigrant family? Didn’t anybody from the sprawling family point it out at the time? (I’d have to say exactly the same murkiness goes on in Catholic-Irish families: a niece born out of wedlock brought up as a sister. Heck there’s one in my (extended) (ex-) family.)

    [**] Or is this specifically Rudi-oriented murkiness as to facts? — which he seems to make something of a speciality.

  64. that’s just rudy lying (probably both about knowing her since childhood and not knowing they were cousins). he’s remarkable in that you don’t have to see his lips moving to know that he’s doing it.

  65. The whole institution of Roman Catholic annulment (at least in America) is based on willful blindness (on the part of the institutional church) and disingenuousness (on the part of the principals).

  66. Trond: the abstract you cite suggests (to me, anyway, not having read the paper) that the 48 to 46 switch happened at some time within the given time range, i.e. that it was a relatively discrete event. That’s different from suggesting that there was an extended period when the early population remained tiny, which I continue to find implausible.

  67. David Marjanović says

    we estimated its formation time at 0.9 Mya with a 95% confidence interval of 0.4-1.5 Mya.

    Whoa! Awesome.

    This merger could have isolated a sub-population reproductively.

    That takes a while. Wild boars across Eurasia routinely have different numbers of chromosomes in the same population with no known effect on fertility.

  68. PlasticPaddy says

    @dm
    Would it not depend on details of what could unzip and reconnect wrong? Or is 24+23 always 46 in DNA arithmetic? Or 47, something like XYY?

  69. David Marjanović says

    Stretches of chromosome that are similar enough team up during meiosis. It doesn’t matter if one side is a continuous chromosome and the other is interrupted. What matters is how similar the sides are.

    And then it’s still a probabilistic matter, not either-or. Occasionally a mule is fertile after all.

  70. Trond Engen says

    Me: This merger could have isolated a sub-population reproductively.

    David M.: Wild boars across Eurasia routinely have different numbers of chromosomes in the same population with no known effect on fertility.

    In the case of humans, there’s no such variation. The merged chromosome is alone.

    But thinking about it, it’s not straightforward and more interesting. If there was only one merger event, and the population carrying it was isolated quickly, there should be very little genetic variation on the merged chromosome, and I haven’t seen anyone claim that. How is the variation on chromosome 2 compared to other chromosomes? Would it be possible to estimate the number of separate mergers that gave us the modern variation?

  71. David Marjanović says

    In the case of humans, there’s no such variation. The merged chromosome is alone.

    Yes, the merged chromosome has reached fixation in the population. An extreme population bottleneck would of course account for this nicely.

    If there was only one merger event, and the population carrying it was isolated quickly, there should be very little genetic variation on the merged chromosome

    Why? Mutations don’t care if they happen on a merged chromosome.

    Would it be possible to estimate the number of separate mergers that gave us the modern variation?

    There are only 4 ways for 2 chromosomes to merge, so it’s not likely the most parsimonious hypothesis – a single event – could be disproved.

  72. @David Marjanović: You might expect the merged chromosome to have a bit less genetic diversity, because it is known to have been inherited from a just a single member of the population. Crossing over with homologous regions of unmerged chromosomes inherited from other ancestors would be somewhat suppressed—although it’s hard to know how much. Meanwhile, the other chromosomes could still carry much of the genetic diversity of the population at the time of the merger event.

    On the other hand, we know that plenty of genes (or, for example, the whole Y chromosome) are descended from single individuals more recently than 0.4 Mya. So there might actually be no observable signature in the diversity of loci on chromosome 2 after all.

  73. David Eddyshaw says

    did they believe mutations have to wait for each other?

    You wait a billion years for a language-enabling mutation, and then three turn up at once!

  74. Of course, we all know the oldest language is actually Dutch.

  75. What’s the world’s oldest language? Whatever these carpenters were using to organise their teams, “476 ± 23 kyr ago”.

    The paper’s intro includes a survey of previously-discovered wooden artefacts:

    This construction has no known parallels in the African or Eurasian Palaeolithic. The earliest known wood artefact is a fragment of polished plank from the Acheulean site of Gesher Benot Ya’aqov, Israel, more than 780 ka …

    Video from the U of Liverpool Professor. (Rather cutesy, but nice visuals.)

  76. So the oldest sentence ever spoken is “Hand me that hammer.”

  77. David Eddyshaw says

    St Ludwig was ahead of the game here, as so often:

    Let us imagine a language for which the description given by Augustine is right. The language is meant to serve for communication between a builder A and an assistant B. A is building with building-stones: there are blocks, pillars, slabs and beams. B has to pass the stones, and that in the order in which A needs them. For this purpose they use a language consisting of the words “block”, “pillar’’, “slab”, “beam”. A calls them out; B brings the stone which he has learnt to bring at such-and-such a call. – Conceive this as a complete primitive language.

    (Philosophical Investigations, §2, tr. Elizabeth Anscombe.)

    This is evidently a stone-age language.

  78. PlasticPaddy says

    This is wrong. The first sentence is “Watch what you are doing, you &*!@ ! You could have *&%!£ killed me.

  79. Or “You’re doing it wrong. Here, let me show you.”

  80. J.W. Brewer says

    But how long did it take to evolve more complex syntactic structures like “Don’t crush that dwarf; hand me the pliers”?

  81. I’d just like to note how badly these quotes were taken out of context. I spent the entire interviewing trying to get the journalist to understand they were asking a bad question, and encouraging them to frame the article differently.

  82. Thanks very much for dropping by and providing that necessary context, which doesn’t surprise me a bit. Journalists…

Speak Your Mind

*