Bloody Shovel 3

We will drown and nobody shall save us


I dig linguistics, and I dig HBD, so how do you join them both? I've had this idea for some time now, but I hadn't written about it lest some guy stole it and wrote a book before I did. 

It seems I'll have to give up on that, as science is fast catching up with my awesome blog (see, I just pulled a Half Sigma here). Razib Khan quotes a recent study in Northern Australia that documents how some Abo kids came up with a new language just for the kicks. It fast became cool, and now the young kids of the tribe have a different language from their elders.  As it looks it's a fully new language, with some grammar changes too, not just a bunch of jargon to fool their parents so they can avoid being eaten. A smart guy in Razib's Twitter also linked to an experimental study where they put people to compete in a game, and prompted to develop their own secret speech, which they did.

When you ask a layman they'll tell you that languages are to communicate. But that's patently false, if we wanted to optimize communication we'd all speak the same language. And languages wouldn't change over time. What the common theory is lacking is just a simple modifier. Languages are to communicate within the group. In fact this little modifier explains most of the mysteries of human psychology. Happiness correlates with income within the group. People are naturally cooperative within the group. Take the modifier away and you get the Cathedral.

Many academic theories about language posit that language evolve to aid better coordination, say for hunters. You go left, I go right, I throw the first spear, etc. Chimps seem to be able to coordinate without speaking, but it sounds reasonable that talking does help coordinate better. But if the idea is to be able to coordinate hunters, then why are men worse at language than women? Women do 70% of the talking, and it's mostly inane gossip. It has extremely little information density. Woman conversation is most of the time a status confirmation task, all they do is say get a group, say something and listen carefully to the tone of voice of all the participants, to check what everyone thinks of each other. If the mother hen suddenly is rude to you, well you know you're in trouble. You better find ways to raise your status or undermine hers. I taught a girlfriend that all her speech was an unconscious status confirmation task, and that she should stop caring as she will always be high status in my eyes. She never nagged me again.

For all I know language did start as a way for men to coordinate hunts better, but over time it's obvious that it evolution found other uses for it. Language itself is a big, a huge shibboleth, a simple way of knowing which tribe people belong to. Babies stop telling apart sounds not used in their native language by 10 months, before they even start talking themselves. And the ability to properly learn new phonemes dies permanently after age 10. With years of effort you can learn to communicate in a foreign language, but your accent will always give you away. And that's being lucky, most people just don't have the capability. And of those who do, a big majority are women, whose tribal membership is always tenuous. After all they never knew when they would be exchanged to a different tribe, or kidnapped and taken away.

A big puzzle of linguistics has always been the relationship between languages. Why are some language families so big, and others so small. One big language family extends from Ireland to Bengal. Yet dozens of different languages of 4 unrelated language families linger in close proximity in Southwest China. Not to mention the Papuans, with hundreds of languages of a dozen families. And those in the know say that most family groupings are very suspect.

Why don't the Papuans get their shit together and talk the same language? Because they don't want to. For thousands of years they have had no need of talking with the neighboring tribe. The neighbors were there to raid, kill, and occasional cannibalistic feast. Austronesian languages are famously extensive, from Hawaii to Madagascar. Yet the Philippines or Borneo are a patchwork of small tribal languages which are not intelligible by the nearby villages. It surely has something to do with the fact that every year, the able bodied males of a given village would raid the neighboring tribe, cut their heads off and bring them home as a trophy.

Farming changed the normal dynamics of tribal speech, with cooperation forced top down to vast masses of people engaged in farming and trading. First you had tens of thousands of people speaking the same language. Then millions. But massive, empire-wide koinés are tied to their empires, and always die and fragment. Ancient Greek died, Latin fragmented, as did Tang Chinese. The Middle Ages brought regional dialects, mostly sharing local market areas, the Enlightenment chose one dialect and artificially transformed it in the national tongue.

The ideology behind national tongues, nationalism, is dead, but national tongues are still around. Of course they are far too useful, and they are too strongly linked to the nation states who created them. But in the same way that the nation state is slowly losing relevance, so national languages are fading too. 50 years ago you would never have listened a regional accent on national TV, today the BBC makes a point of casting Scottish scientists for their documentaries. Italian dialects are making a comeback.  Even in Japan a big part of movies are voiced in regional dialects, some quite obscure. All while every country on earth is putting ever more resources into English education.

If languages were to communicate, we would have an English speaking world in no time. Instead what we will have is a global English speaking elite, lording over masses speaking bad English to their masters, and revived regional dialects to themselves. Given Google Translate and PRISM, it wouldn't surprise me if vernacular writing dies out, with most speech being done in untranslatable dialect, and writing done in English. A massive Hong Kong style diglossia. It might be the only feasible resistance against what's coming.


Leave a Reply
  • Balkanization of the language happened in the old days because travel was limited. Language is dynamic. Given the same starting point, and if the groups are sufficiently isolated, different accents, dialects and eventually different but related languages will evolve. U.S. underwent a consolidation of accents since the advent of TV. Dan Rather did more to standardize accent then just about all the government edicts.

    While there are backlashes as you pointed out, I am not sure as a whole for China and U.S. they are going the way of further balkanization as you said. Certain groups who wanted to did so. Ebonics is an example, but by large, different groups like the Mexicans coming to the states and adopt the standard language as their own.

      • While language evolves over longer periods of time, in the short run it is resistant to change. What is more, the direction and speed of the change is need based. Surely inside China, you see the moderation of the local dialects in favor of the standard mandarin. In Hong Kong, the will of the people do want a separation as they feel that they are not a part of the main land for a long time. I think it could go both ways but the big trend I see is integration.

        • Integration only happens under massive, strenuous state coercion. When China goes postmodern like everyone else, dialects will spring back again.

    • "Balkanization of the language happened in the old days because travel was limited."

      Pre-immigration, working class accents in London changed (slightly) every mile or so. It was a group identity thing.

  • Despite the opening sentence, you haven't joined linguistics with HBD in any way with this discussion. As far as I know, nobody has yet discovered a case where kids of some particular ancestry would be unable to acquire a given language as perfect native speakers. So linguistic identity merely proves that you grew up in tribe, not shared ancestry.

    • I'm not claiming that. HBD is not only about group differences.

      And I still think a Papuan kid might have trouble acquiring proper English.

  • Of course, given the unintelligibility of the concept of meaning (in the pomo-sense I'm afraid), words can only stand in for a chain of references, ending with an X to which both speaker and hearer must have some form of non-verbal access (whether it be through memory, personal or collective, sensory perception or something else). Given that, the phenomenon at the end of the chain is likely to be something tribal, or at least local.

    Linguistics is not that enigmatic for those who never bought the psychotic (judaic) chomskyan idea of universal grammar. Why does some intangible concept always have to be dragged into it? Why this hysterical craving for the invisible? Neither Wittgenstein, nor Derrida nor, for that matter, Dan Everett can come as a real shock to someone who has a sober, hands-on view of language. In essence, it's clever monkeys making sound (and later ink blots).

  • I dig linguistics, and I dig HBD, so how do you join them both?

    The best explanation of language I've seen invokes "HBD". It is in "The Mating Mind" by Geoffrey Miller, which is one of the best books I've read and superior to the oft-compared "The Red Queen". Along with other exceptional human traits, Miller explains language as a sexually selected adaptation. Like the peacock's tail, male humans' verbal ability evolved at least in part as a survival handicap to impress the opposite sex. An item of evidence in favour of this hypothesis is that all natural languages have a vocabulary far in excess of utilitarian necessity. "Basic English", which has just 850 words, is only slightly less efficient than the full 60,000 word English vocabulary.

    Another interesting hypothesis is due to Robin Dunbar; he suggests that language evolved because it's an efficient form of grooming in humanity's relatively large social groups. Both these explanations probably have merit. Geoffrey Miller concludes:

    "For males, verbal self-advertisement appears to be a fairly constant function of speech, while for females, it may be an occasional function, more limited to one-on-one conversations with desired mates. A complete theory of the evolution of language will probably have to combine sexual selection and social selection, integrating the gossip-as-courtship theory with the gossip-as-grooming theory."

    why are men worse at language than women?

    Here is Miller's discussion, in terms of his sexual selection hypothesis:

    "Most tests of human verbal abilities are tests of language comprehension, not tests of language production. Given a strict male-display, female-choice mating system, we should expect female superiority in language comprehension and male superiority in language production.

    For example, females should recognize more words, but males should use a larger proportion of their vocabulary in courtship, biasing their speech towards rarer, more exotic words. In this simple picture, more women might understand what "azure" really means (so they can accurately judge male word use), but more men might actually speak the word "azure" in conversation (even if they think it means vermilion). Standard vocabulary tests measure only comprehension of word meaning, not the ability to produce impressive synonyms during courtship. Reading comprehension questions are more common than creative writing tests. Women are faster readers and buy more books, but most books are written by men."

    The relationship between language, tribal interactions and politics is a separate matter that takes us outside HBD. However, I criticised "tribalism" as a frame of analysis here. In short, what I've learned about tribes doesn't support the popular notion that various factional phenomena are "tribal". This word lacks explanatory power.

    I could add to that essay an insight from Miller, which in that in the Paleolithic era, instead of tribes of nuclear families, an alternative depiction of human social organisation is that women were self-sufficient gatherers, and men hunted animals mainly to impress women. Leks of young men would attempt to monopolise groups of women, and male instincts related to this type of sexually homogeneous faction, which one would not call a tribe, may contribute in an important (but incomplete) way to phenomena that are called "tribal".

    • I have to say that I find sexual selection theories quite unconvincing in general. And very circular. Males do something because women like it. But why would they like it? They didn't like it for millions of years. Nobody thought it important.

      Women on average speak more and have better vocabularies. I am very skeptical that most mating by primitive people is done through verbal seduction, but if you have some data I'll be glad. The best seducers I know don't have better vocabularies than I do, nor than the women they seduce.

      I mean, something is surely wrong when you conclude that men hunt animals to impress women. Surely men hunt animals to eat meat which is the most nutritious food on earth. That women are impressed is just a byproduct of their wanting to eat it.

      As for "tribal", shibboleths are tribal. The ability to pronounce the word depends on the tribe you belong to. If that's not a useful concept you can tell the Ephraimites.

    • Verbal ability is really, really handy for manipulating the people around you, getting them to do stuff for you, tricking them into accepting moderately lopsided deals, convincing them that you have high status and convincing them that they don't want to kill you when they do figure out that they've gotten a bum deal.

      It still works rather well today. Humans are cooperative animals, but some people profit from that cooperation a lot more than others. All else being equal, those people would have had higher reproductive fitness, up until birth control changed everything.

      Women have better verbal skills because they are / were more dependent on manipulating the people around them.

      • That (if you can't count on your muscles to elbow your way through the hardships of life, you'll reply on deception more), but also women have a more emotional mind and, at least till short ago, verbal language was the vehicle of expression for emotions.

        Conversely, men developed an higher general intelligence because their superior physical strength put them in front of more problems requiring to be solved.

  • When you ask a layman they’ll tell you that languages are to communicate. But that’s patently false, if we wanted to optimize communication we’d all speak the same language.

    Language evolved among small populations, in non-cosmopolitan environments where the coefficients of relatedness are high enough for altruism to arise and for communication to take place.

    The evolution of communication -- as opposed to manipulation -- is very much akin to the evolution of altruism. It is subject to the same exploits in a cosmopolitan environment. So manipulation via signaling taking on the appearance of communication is far more common in the present environment (of greater cosmopolitanism) than in the environment in which communication evolved. We're in an evolutionary arms race to defect faster than the other guy while maintaining keeping active his obsolete instincts to think communication is still viable.

    The end result of civilization is a bifurcation: 1) communication terminates as a genetic capacity and/or 2) communication takes a form similar to that in eusocial species (e.g. pheromone manipulation) even as civilization takes that form.

  • As Dan Sperber points out, the function of language is to manipulate other peope's internal states.

    Clearly, over time, some of us have grown to be relatively immune to such manipulation. There will still be many who can be manipulated, of course.

    If those people do not know your mother tongue, they will not be very effective at manipulating you.

    • It's quite natural for someone from a cosmopolitan background like Dan Sperber to view that language is about manipulation. It's certainly the case that in more cosmopolitan environments, language will increasingly take on the character of manipulation, rather than communication. The inherent genetic costs and benefits of diverse environments inhibit communication, just as they inhibit altruism more generally, and favor manipulation and exploitation.

      But cosmopolitan environments are evolutionarily novel. Civilization and cities only 10,000 years old. The modern, extremely cosmopolitan environments have only arisen over the past century with modern transport and communication technologies. Language arose among much smaller, more homogeneous populations where the cost and benefits would have been much less skewed towards favoring manipulation over communication.

  • > But if the idea is to be able to coordinate hunters, then why are men worse at language than women? Women do 70% of the talking, and it’s mostly inane gossip. It has extremely little information density. Woman conversation is most of the time a status confirmation task, all they do is say get a group, say something and listen carefully to the tone of voice of all the participants, to check what everyone thinks of each other.

    Women talking more is not women being better at language. You probably don't really think that since you immediately follow it up saying women talks are bad for expressing information (or in some case, incapable of expressing information) and serve different purposes.

  • I'd wager that part of the reason the Chinese languages diverged so much in spite of the long literary tradition is because Chinese characters aren't tied to phonetics at all.

    • I wouldn't say "at all". There was this research by Zhao Yuanren on villages where 破绽 was read as po-zhan and others as po-ding.

      • Fair enough, I exaggerated there.

        I asked my mother why this discrepancy could be, since I'm illiterate in Chinese. Apparently it's a misreading, as ding4 (定) is a similar character to zhan4 (绽), except without the radical. Perhaps that's the source of confusion?

        • Exactly. Which means that the phonetic information in the character won over the oral transmission. And that's for a very basic word which means a split seam.

          Given the huge amount of barbarians and their languages that the Chinese had to assimilate, it's a wonder that Chinese dialects are as similar as they are. Cantonese character readings are almost perfectly consistent with Mandarin once you remember a set of rules. The languages are way closer than the genotype.

          • Fascinating.

            Though Chinese characters are largely not phonological, it seems to me that the strong tradition of rime dictionaries (such as Qieyun) helped stabilise varieties of spoken Chinese and resisted huge changes. It's also noteworthy that, to my knowledge, there have historically been no massive external migrations into areas already established as Chinese.

            • 90% of characters are 形声, and even though some are misleading like the po-zhan example above, most are a good guide to how a certain word should be pronounced. I guess in each village there was always some literate elder who people could ask when necessary.

              Southbound emigration was almost invariably triggered by barbarian invasions in the northern plains. Invaders were never enough to replace the locals, but it's recorded on history that they were there for centuries, and they surely left their mark in the local dialects. People look different the further north you go; the purest Henan people can be quite dark in the villages.

  • I speak decent German, and during a recent trip to Germany, I got into a conversation with a couple Swiss girls who were also on vacation. Although we conducted the entire conversation in Hochdeutsch (Standard German), I was surprised to hear them tell me that Hochdeutsch was not their second language, but their third. The Swiss German dialect was of course their first language and what came the most natural to them, but they both admitted that English came far more easily and naturally to them than Standard German. Although they needed Hochdeutsch to communicate with Germans and Austrians, as well as to understand broadcast media and written German (as Swiss German is not generally used for writing, outside of chatspeak), the importance of English obviously superseded that to the point that it became easier to use than the national language of their northern neighbors, in spite the fact that Hochdeutsch is linguistically far closer related to the Swiss German dialect than English is.

    • Interesting anecdote.

      I wonder how good was their English really is. It's one thing for them to think of English as their second language, a different thing that their English is actually better than their Hochdeutsch. Even if they're watching CNN every waking hour, their exposure to Hochdeustch surely is 50x bigger than to English.

  • The ideology behind national tongues, nationalism, is dead, but national tongues are still around.

    Is that ideology as gone as you say? It may be dead in dead nations, nations whose majority ethnicity culture is expiring. Why extend this to the whole world.

    "It might be the only feasible resistance against what’s coming."

    ? How is your crystal ball?

  • 6 pingbacks