4 ms·

Article mentions Altaic theory without mentioning that Altaic theory is utter horseshit. Still an interesting article though.

stanrivers 6y ago
Thank you for this; I wish we could be friends
- stanrivers 6y ago
  Ok - this chain is interesting; didn’t have this background; I deserve the downvote. Thanks for context... will go read some more
cmrdporcupine 6y ago
To be fair, if I recall it's only in the last 15-20 years that the Altaic theory seems to have gone from 'speculative' to 'not at all likely.' Those of us who studied this stuff 20 years ago could easily spout off this 'might-be fact' if we didn't bother to verify first. I guess at this point similarities between the two languages are assumed to be a result of regional influence?
- eindiran 6y ago
  Many mainstream linguists have been against the hypothesis since the 1950s, but there are still plenty of linguists who contend that the Turkic, Tungusic and Mongolian families form a family; a smaller number have arguments in support of the stronger claim that the Koreanic, Ainu and Japanese–Ryukyuan languages are connected as well. See eg: [0] Vovin set out to demonstrate that the latter, stronger claim was true and ended up writing this: [1] Arguably, the biggest blow struck to the hypothesis in the last 15-20 years came when Vovin, previously a supporter of the hypothesis, wrote "The end of the Altaic controversy"[2]. Among the more notable arguments in it is that the non-traditional methods[3] used to argue that Altaic is a family are rejected by traditional comparative linguistics because they overweight lexical comparisons. This was a rather prescient point, as some of the methods that have developed since then that have been used to argue in favor of the Altaic hypothesis suffer from this problem quite badly; this is perhaps especially true of the Bayesian phylogenetic inference, sometimes called Bayesian phylolinguistics when applied to historical linguistics. With this technique, used in eg [0], it is pretty hard to argue that this method wouldn't overweight cognates and loan words gained through contact unless you specifically control for that, which will cause you to end up erroneously marking sprachbunds as true families. See [4] for another usage of the technique, this time to support the Dravidian family, which is already very very well support by traditional comparative methods. All of this ultimately points to what I think is the real answer to the issue; outside of some arguments about archaeology that I don't have enough background to evaluate [5], most of the shared bits (shared pronouns, lexical stuff) between the languages here exist because of contact. The languages here, if they belong grouped at all, should be grouped as a sprachbund, not as a family. The possibility that there is a true family with "micro-Altaic" is small, but I think it is still a much greater possibility than the "strong" hypothesis (w/ Japanese, Korean, etc). This paper [6] has a very cool approach to evaluating the weaker form of the hypothesis that I think most HN readers would find interesting. So to answer your question, yes, I believe your assessment is correct, and mainstream historical linguistics does as well. But I wouldn't go so far as to say there are any nails in the Altaic hypothesis' coffin just yet. [0] https://academic.oup.com/jole/article/3/2/145/5067185 https://academic.oup.com/jole/article/3/2/145/5067185 [1] https://www.academia.edu/4208284/WHY_JAPONIC_IS_NOT_DEMONSTRABLY_RELATED_TO_ALTAIC_OR_KOREAN_pre_publication_version_ https://www.academia.edu/4208284/WHY_JAPONIC_IS_NOT_DEMONSTR... [2] https://www.academia.edu/6345901/The_end_of_the_Altaic_controversy https://www.academia.edu/6345901/The_end_of_the_Altaic_contr... [3] "non-traditional" = anything outside the comparative method (https://en.wikipedia.org/wiki/Comparative_linguistics https://en.wikipedia.org/wiki/Comparative_linguistics). See eg: https://www.reddit.com/r/linguistics/comments/6tg6cr/why_is_the_uralaltaic_hypothesis_largely_rejected/ https://www.reddit.com/r/linguistics/comments/6tg6cr/why_is_... [4] https://pure.mpg.de/pubman/item/item_2564924_3/component/file_2564936/shh973.pdf https://pure.mpg.de/pubman/item/item_2564924_3/component/fil... [5] https://www.cambridge.org/core/journals/evolutionary-human-sciences/article/tracing-population-movements-in-ancient-east-asia-through-the-linguistics-and-archaeology-of-textile-production/440213710888C5D9488901B27E047D03 https://www.cambridge.org/core/journals/evolutionary-human-s... [6] https://www.ling.upenn.edu/~ceolin/Diachronica.pdf https://www.ling.upenn.edu/~ceolin/Diachronica.pdf
garmaine 6y ago
The Altaic theory isn't necessary to postulate a connection between Japanese and Korean. It's known that genetically the Japanese population was seeded by Korean settlers which largely displaced the native Ainu populations. Japanese is likely a descendent of the language spoken by these colonists from the Korean peninsula. The two languages are very divergent, and the usual mechanisms for classifying languages are difficult to apply because the dominant Chinese influence came to each separately after they diverged.
- Mediterraneo10 6y ago
  While a relationship between Japan and certain lost languages of the Korean peninsula is agreed upon by nearly all scholars, that still doesn’t necessarily mean that Japanese and Korean are related. One recent school of thought, for example, is that pre-Proto-Japanese entered the Korean peninsula from mainland China (Shandong province) while Korean came down from the north, so they would not be genetically related in spite of being neighbours.
  - garmaine 6y ago
    I was referring to human genetics.
- danans 6y ago
  > It's known that genetically the Japanese population was seeded by Korean settlers which largely displaced the native Ainu populations It's even more complicated than that. In addition to the Ainu there were the Jomon people, a non-agricultural but nonetheless settled society who are believed to be the first people in the world to use pottery - which is surprising for a non-agricultural people. They might in turn have been an offshoot of the first migration of modern humans out of Africa along the south coast of Asia, whose haplotypes [1] occur today at some of the highest frequences in Japan and as far north as Mongolia and as far south as Australia. Some sub-branches spread west all the way to the edge of Europe. 1. https://en.m.wikipedia.org/wiki/Haplogroup_M_(mtDNA) https://en.m.wikipedia.org/wiki/Haplogroup_M_(mtDNA)
  - 6y ago
naniwaduni 6y ago
For an amateur's blog post in 2009 to call Altaic theory utter horseshit would been rather hubristic to begin with, and moreover, not even the point. The Altaic hypotheses were higher-profile proposals that happened to include common origin for Japanese and Korean. Even dismissing the former, the latter is not exactly out of consideration. And, as the post points out, the whole line of argument is a curiosity that has little concrete bearing on the fact that the two are very similar from a learner's perspective.

stanrivers 6y ago

Thank you for this; I wish we could be friends

stanrivers 6y ago
Ok - this chain is interesting; didn’t have this background; I deserve the downvote. Thanks for context... will go read some more

cmrdporcupine 6y ago

To be fair, if I recall it's only in the last 15-20 years that the Altaic theory seems to have gone from 'speculative' to 'not at all likely.' Those of us who studied this stuff 20 years ago could easily spout off this 'might-be fact' if we didn't bother to verify first. I guess at this point similarities between the two languages are assumed to be a result of regional influence?

eindiran 6y ago
Many mainstream linguists have been against the hypothesis since the 1950s, but there are still plenty of linguists who contend that the Turkic, Tungusic and Mongolian families form a family; a smaller number have arguments in support of the stronger claim that the Koreanic, Ainu and Japanese–Ryukyuan languages are connected as well. See eg: [0] Vovin set out to demonstrate that the latter, stronger claim was true and ended up writing this: [1] Arguably, the biggest blow struck to the hypothesis in the last 15-20 years came when Vovin, previously a supporter of the hypothesis, wrote "The end of the Altaic controversy"[2]. Among the more notable arguments in it is that the non-traditional methods[3] used to argue that Altaic is a family are rejected by traditional comparative linguistics because they overweight lexical comparisons. This was a rather prescient point, as some of the methods that have developed since then that have been used to argue in favor of the Altaic hypothesis suffer from this problem quite badly; this is perhaps especially true of the Bayesian phylogenetic inference, sometimes called Bayesian phylolinguistics when applied to historical linguistics. With this technique, used in eg [0], it is pretty hard to argue that this method wouldn't overweight cognates and loan words gained through contact unless you specifically control for that, which will cause you to end up erroneously marking sprachbunds as true families. See [4] for another usage of the technique, this time to support the Dravidian family, which is already very very well support by traditional comparative methods. All of this ultimately points to what I think is the real answer to the issue; outside of some arguments about archaeology that I don't have enough background to evaluate [5], most of the shared bits (shared pronouns, lexical stuff) between the languages here exist because of contact. The languages here, if they belong grouped at all, should be grouped as a sprachbund, not as a family. The possibility that there is a true family with "micro-Altaic" is small, but I think it is still a much greater possibility than the "strong" hypothesis (w/ Japanese, Korean, etc). This paper [6] has a very cool approach to evaluating the weaker form of the hypothesis that I think most HN readers would find interesting. So to answer your question, yes, I believe your assessment is correct, and mainstream historical linguistics does as well. But I wouldn't go so far as to say there are any nails in the Altaic hypothesis' coffin just yet. [0] https://academic.oup.com/jole/article/3/2/145/5067185 https://academic.oup.com/jole/article/3/2/145/5067185 [1] https://www.academia.edu/4208284/WHY_JAPONIC_IS_NOT_DEMONSTRABLY_RELATED_TO_ALTAIC_OR_KOREAN_pre_publication_version_ https://www.academia.edu/4208284/WHY_JAPONIC_IS_NOT_DEMONSTR... [2] https://www.academia.edu/6345901/The_end_of_the_Altaic_controversy https://www.academia.edu/6345901/The_end_of_the_Altaic_contr... [3] "non-traditional" = anything outside the comparative method (https://en.wikipedia.org/wiki/Comparative_linguistics https://en.wikipedia.org/wiki/Comparative_linguistics). See eg: https://www.reddit.com/r/linguistics/comments/6tg6cr/why_is_the_uralaltaic_hypothesis_largely_rejected/ https://www.reddit.com/r/linguistics/comments/6tg6cr/why_is_... [4] https://pure.mpg.de/pubman/item/item_2564924_3/component/file_2564936/shh973.pdf https://pure.mpg.de/pubman/item/item_2564924_3/component/fil... [5] https://www.cambridge.org/core/journals/evolutionary-human-sciences/article/tracing-population-movements-in-ancient-east-asia-through-the-linguistics-and-archaeology-of-textile-production/440213710888C5D9488901B27E047D03 https://www.cambridge.org/core/journals/evolutionary-human-s... [6] https://www.ling.upenn.edu/~ceolin/Diachronica.pdf https://www.ling.upenn.edu/~ceolin/Diachronica.pdf

garmaine 6y ago

The Altaic theory isn't necessary to postulate a connection between Japanese and Korean. It's known that genetically the Japanese population was seeded by Korean settlers which largely displaced the native Ainu populations. Japanese is likely a descendent of the language spoken by these colonists from the Korean peninsula. The two languages are very divergent, and the usual mechanisms for classifying languages are difficult to apply because the dominant Chinese influence came to each separately after they diverged.

Mediterraneo10 6y ago
While a relationship between Japan and certain lost languages of the Korean peninsula is agreed upon by nearly all scholars, that still doesn’t necessarily mean that Japanese and Korean are related. One recent school of thought, for example, is that pre-Proto-Japanese entered the Korean peninsula from mainland China (Shandong province) while Korean came down from the north, so they would not be genetically related in spite of being neighbours.
- garmaine 6y ago
  I was referring to human genetics.
danans 6y ago
> It's known that genetically the Japanese population was seeded by Korean settlers which largely displaced the native Ainu populations It's even more complicated than that. In addition to the Ainu there were the Jomon people, a non-agricultural but nonetheless settled society who are believed to be the first people in the world to use pottery - which is surprising for a non-agricultural people. They might in turn have been an offshoot of the first migration of modern humans out of Africa along the south coast of Asia, whose haplotypes [1] occur today at some of the highest frequences in Japan and as far north as Mongolia and as far south as Australia. Some sub-branches spread west all the way to the edge of Europe. 1. https://en.m.wikipedia.org/wiki/Haplogroup_M_(mtDNA) https://en.m.wikipedia.org/wiki/Haplogroup_M_(mtDNA)
- 6y ago

naniwaduni 6y ago

For an amateur's blog post in 2009 to call Altaic theory utter horseshit would been rather hubristic to begin with, and moreover, not even the point. The Altaic hypotheses were higher-profile proposals that happened to include common origin for Japanese and Korean. Even dismissing the former, the latter is not exactly out of consideration. And, as the post points out, the whole line of argument is a curiosity that has little concrete bearing on the fact that the two are very similar from a learner's perspective.