Talk:Altaic languages

Article requires rewrite based on latest DNA revelations

Multiple studies now published including data from the Reich lab at Blavatnik @ Harvard, part of the human atlas project funded by the John Templeton foundation, have conclusively demonstrated that the once so-called Altaic theory is no longer a theory, but fact. The language developed approximately 9000 years ago in what is today China, before splitting into three sub branches. This article is therefore completely wrong. Historiaantiqua (talk) 07:04, 28 November 2023 (UTC)

No it doesn't need a complete rewrite because DNA is NOT language. The Altaic hypothesis is not proven by DNA evidence, nor is any other language family proven by DNA evidence. This is just a way for people to ignore actual linguistic evidence that doesn't prove what they want it to prove. Historical linguistics has its own scientific methodology that is independent of biology, just as DNA evidence is independent of language (there is no "language" gene since language is behavioral, not physical). When DNA evidence matches historical linguistic evidence, then talking about DNA evidence makes sense, but the historical linguistic evidence ALWAYS comes first when talking about language families. DNA evidence never proves a linguistic hypothesis without the solid linguistic evidence coming first. The majority of historical linguists still find the Altaic hypothesis unconvincing based on the linguistic evidence. Thus DNA is a trivial issue and never conclusive. It is clear from your comments and your user page that you are not a linguist, but a historian, so you should leave linguistics to the linguists when they tell you that your assertions are not based on the science of linguistics. The DNA evidence shows that the modern humans of the area derive from older humans in the area, but that proves absolutely nothing about what languages they were speaking as they interbred and reproduced. --TaivoLinguist (Taivo) (talk) 09:37, 28 November 2023 (UTC)

Leaving aside the validity of Historiaantiqua's original argument, Wikipedia is free to edit by anyone. You don't need to be an expert or "leave linguistics to the linguists," we simply have to cite authoritative sources, eg linguists. There were some large edits last year that removed content due to "undue weight" without much discussion, I'm proposing that they either be restored with help tags and/or be discussed here first. AnandaBliss (talk) 19:46, 22 April 2024 (UTC)

I'm confused about the "expert needed"-tag. What makes you believe that Taivo lacks expertise in order to remove undue details cited from primary sources? If it's about consensus, I support this removal. –Austronesier (talk) 20:51, 22 April 2024 (UTC)

I'm going to re-remove the table, if that's okay. You appear to have missed that a lot of the discussion on that wasn't here, but was part of a much bigger discussion on the linguistics wikiproject about these tables being included on articles where the familial relationship is broadly doubted. That means we're eating up a lot of space either presenting in-universe evidence which isn't actually showing anything real underlying it and which may confuse the readers, or we're demonstrating a sprachbund using the tools we (as in Wikipedia) typically present familial relationships. Either way, it's a big WP:UNDUE issue here as it is in other macrofamily pages. Warrenᚋᚐᚊᚔ 00:35, 23 April 2024 (UTC)

The unfortunate reality is that ethno-politics play into this matter. I agree with your proposed course of action. Emotional investment in topics of this sort is a sure indication of one's own bias. Historiaantiqua (talk) 17:18, 30 April 2024 (UTC)

DNA is not language, but language is an objective reality, as is the DNA, while religious and ethnic categories are mostly non-objective. To suggest that Turkmenic languages are a unique and autochthonous development unrelated to any macro-linguistic family based on the period in which they appear and the locale, and its significant connection to the present day is to propose a fantastical displacement of material reality before never seen in any other language development. Historiaantiqua (talk) 17:21, 30 April 2024 (UTC)

Nobody who considers the Altaic hypothesis as unproven argues that the Turkic language family (I assume that's what you mean when taking about "Turkmenic languages") is "a unique and autochthonous development". That's the polygenesis fallacy that is often evoked by macro-comparatists. When we say that a language or language family is not known or cannot not proven to be related to any other language or language family in the world, we mean exactly that. Languages inevitable change with time, and indeed can change so drastically by modifying/eroding the original shape of words and affixes or simply replace them with entirely new ones, to the point that any genealogical connections to other languages get unrecognizable. And even if some points of resemblance remain that actually do go back to a shared ancestral language, they might be too singular to be distinguishable from the noise of concindental resemblance. In the latter case, historical linguists consider the evidence as inconclusive and hence any claims of language relationship as unproven.

In the case of the Altaic languages, we additional have secondary resamblances as a result of prolonged contact between these individual language families. Robbeets and her collaborators argue that they can unravel the 'real' shared features that go back to common inheritance behind the veil of areal convergence. Some fellow linguists are convinced, some aren't. Many of the arguments that have been brought forward against Robbeets's predecessors can also be applied to her material.

As for the DNA evidence, it is compatible with either scenario: shared inheritance or areal convergence. Since languages and genes are transmitted in completely different ways, a shared genetic ancestry between let's say Mongolic and Turkic speakers might either be indicative of parallel genetic and linguistic inheritance, or a result of language shift and language convergence between groups speaking unrelated languages as a result of interaction and interbreeding with each other. –Austronesier (talk) 19:16, 30 April 2024 (UTC)

Proto-Altaic Forms

There’s been a minor edit war brewing with @AnandaBliss over the inclusion of in-universe lists of possible etymologies for Altaic (and other macrofamily proposals). There’s was a discussion on the linguistics wikiproject and to a lesser extent one at WP:FTN about this and the general sense was that since these lists aren’t actually representing real data, as opposed to the controversial interpretations of individual academics, that including these big lists becomes an issue with WP:FALSEBALANCE, WP:PARITY, and WP:PROFRINGE.

While Altaic’s position is better than, say, Nostratic within academia, this article is still not doing a stellar job at making it clear that Altaic as a language family is not widely accepted. These big lists of etymologies run a risk of simply misinforming a reader when compared to, say, Proto-Indo-European, where the etymologies are broadly considered real reconstructions by the linguistics community even if they’re revised over time as more information comes out. What this article doesn’t need more of as far as I can tell is “if Altaic is a real family, here is why.” Warrenᚋᚐᚊᚔ 19:19, 30 April 2024 (UTC)

How can people like yourself continue to say it is not widely accepted while there are more supporters to critics listed in the article and the theory is taught as fact in schools and universities of every Altaic country? The article needs editing, for sure, but you seem to be removing this content just because you don't like it or the theory. The purpose of this page is to explain the theory and then at the end, the criticism. 2A02:C7E:3011:FC00:68AB:3511:3ACB:5016 (talk) 20:08, 30 April 2024 (UTC)

How can people like yourself continue to say it is not widely accepted while there are more supporters to critics listed in the article

Because, intentionally or not, the article has been heavily edited by people who either are misinformed about the academic status of Altaic or who want to make a case to others. That's not what wikipedia is about. There's also a historical issue of Altaic having more broad support in the past than it does now, so there's a disconnect in the quality of sources between the past and present, and figuring that out can be quite tricky for nonexperts.

the theory is taught as fact in schools and universities of every Altaic country?

It can be taught and still wrong. See: Intelligent Design.

The article needs editing, for sure, but you seem to be removing this content just because you don't like it or the theory.

No, the field of historical linguistics doesn't like the theory due to the lack of evidence. There is absolutely evidence that Altaic is a result of a sprachbund, but not actually a language family itself.

The purpose of this page is to explain the theory and then at the end, the criticism.

You may want to read WP:FALSEBALANCE, I don't think you'd find many supporters for this proposal on this page or any other macrofamily page. Warrenᚋᚐᚊᚔ 20:15, 30 April 2024 (UTC)

This article, without a doubt, has had far more anti-Altaicist edits. If there's more critics you can think of, please list them. They would have listed as many as they possibly could.

And it can be taught and correct, which is more likely considering the many academics who agree with the theory. There's not a single foreign person that I know of who disagrees with the theory that speaks 2 or more Altaic languages fluently. Not a single one. The only people who do are generally Japanese and Koreans who dislike eachother. People who speak Turkish and learn Japanese, or people who speak Korean and learn Turkish are stunned by the similarity in grammar and syntax to the point many of the most used particles are exactly the same. Such as the destination and locations particles. And many others are very phonetically similar, too. Sure, languages in proximity borrow many words, but do they borrow grammar, suffixes and features like vowel harmony? No. Never been proven.

Yes, there's evidence. People are trying to show some of that evidence here and you're trying to hide it.

You can go ahead and check but before removing sourced content you should reach a reasonable consensus. So far, it's just you. 2A02:C7E:3011:FC00:68AB:3511:3ACB:5016 (talk) 20:24, 30 April 2024 (UTC)

This article, without a doubt, has had far more anti-Altaicist edits.

And the article on medicine spends little time on the humours. This is to be expected when academic scholarship moves on from an idea. "Anti-Altaicists", or at least those generally negative to the idea without an affirmative opinion, represent the mainstream academic consensus.

And it can be taught and correct, which is more likely considering the many academics who agree with the theory.

Then the article should be updated following WP:RS on this. So should the community of historical linguists, in this particular case.

There's not a single foreign person that I know of who disagrees with the theory that speaks 2 or more Altaic languages fluently. Not a single one.

I don't know a single one who does, but Wikipedia's sourcing requirement is beyond either of our anecdote.

Yes, there's evidence. People are trying to show some of that evidence here and you're trying to hide it.

You need to take a big step back and read WP:AGF. Warrenᚋᚐᚊᚔ 20:33, 30 April 2024 (UTC)

How can I assume good faith when you're removing completely reliable sourced content and keep citing a conversation that you keep refusing to give a link to? 2A02:C7E:3011:FC00:68AB:3511:3ACB:5016 (talk) 20:44, 30 April 2024 (UTC)

Sorry to double-comment, but per what I said below, could this article then be re-oriented to mostly focusing on, as you say, "...absolutely evidence that Altaic is a result of a sprachbund, but not actually a language family itself," with a section denoting that a very small minority view it instead as a family, again with not much weight? AnandaBliss (talk) 15:02, 1 May 2024 (UTC)

You are continuing to WP:Edit war and we're now at WP:3RR. To address your edit comment; there has been a discussion. There's even references to it above in a different talk page thread, but do not continue to revert and engage in an edit war while we're talking about this here. Note I am not making these edits in a vacuum, I've been eliciting the help of the Linguistics Wikiproject to clean up some of these macrofamily articles with a huge focus excplicitly being these lists of proto forms. By all means, get involved in the wikiproject or make your case here that Altaic specifically should be an outlier in that discussion (I'm open to different perspectives!) but think we'd need a third party from the linguistics wikiproject to comment if anyone else is reading along (@TaivoLinguist?) Warrenᚋᚐᚊᚔ 20:23, 30 April 2024 (UTC)

Nope, not "we're at", just you are. You've reverted it 3 times. You've mentioned a page 4 times now that you've still not linked. 2A02:C7E:3011:FC00:68AB:3511:3ACB:5016 (talk) 20:25, 30 April 2024 (UTC)

I'm not sure "edit war" is a remotely appropriate characterization, as I'm pretty sure I haven't reverted anything twice, let alone three times. Having said that, I must admit I'm a bit of an inclusivist, so in my mind laying out both arguments will reflect scholarly consensus, because the reasoning and support behind both reveals the views of the field writ large. I figured adding the table back would be an easy way to show what the discussion's about. AnandaBliss (talk) 01:03, 1 May 2024 (UTC)

The truth isn't somewhere in the middle on this one, though. An accurate reflection of the scholarship will result in readers concluding that the consensus is Altaic isn't real as a language family. The table has been part of a long-running issue on macrofamily articles where people present long-ranged word pairs in huge table forms which take up a lot of space in the article for what is essentially historical linguistics fanfiction. See @TaivoLinguist's reply above.

re: Edit warring, I had assumed you were the IP above but logged out, and since that appears to be wrong I'm sorry for that. I was basing that off the edits you both made being back to back and identical, but I'll try to be more careful about that. Warrenᚋᚐᚊᚔ 01:52, 1 May 2024 (UTC)

I'd been unaware that there was such unified consensus in that field, and that there was a history of speculative table-style edits. The article fluctuates between controversial, minority opinion and disproven so much that it was hard to make heads or tails of what the "discussion" really was.

In that case, would you present it as a small, minority opinion that doesn't hold much weight (but still "sincere," if that makes any sense), or as a fringe position (i.e. deceptive, universally rejected by scholars)? AnandaBliss (talk) 14:55, 1 May 2024 (UTC)

Both, depending on the scholar in question. There are serious altaicists who are well respected linguists and simply disagree with the majority read of the data, but there are also a lot of people who work from the first principle that Altaic is a thing and just plow ahead on that front, which very much can get into fringe territory.

The fact that this article is unclear about the academic standing of the theory is a reflection on both how hard it is to communicate a lot of the subtle nuance here as well as the fact that a substantial number of people are operating on the base assumption from their schooling of the veracity of Altaic. Warrenᚋᚐᚊᚔ 18:39, 1 May 2024 (UTC)

Thanks for pinging me. Reconstructed "Proto-Altaic" forms have no more place in Wikipedia than reconstructed Proto-Klingon-Romulan forms. Both are fictional. Based on the contemporary majority view within the historical linguistics community, "Altaic" is a Sprachbund and not a genetic family. If the anon IP wants the discussion, then it is incumbent upon the anon IP to look up the discussion in the archives. Every generation of linguists seems to bring another crop of budding historical linguists who wish to make a name for themselves by proving the unprovable, namely the existence of an Altaic family. Sadly for them, the linguistic evidence still does not exist and they are never able to expand the supposed proto-vocabulary beyond the forms that were tentatively "reconstructed" decades ago. Thus, they go to non-linguistic methodology like DNA and proto-agriculture to try to prove their case. DNA evidence and agricultural evidence are utterly irrelevant to linguistic reconstruction since there is no gene for language and no cultigen that is language-specific. Altaic is still unproven and is no closer to proof than it was 50 years ago. --TaivoLinguist (Taivo) (talk) 22:54, 30 April 2024 (UTC)

It seems that user just got a two year ban for repeated edit warring. That said, and I'll try to bring this up at the linguistics wikiproject again but we really need a standardized approach to proposed macrofamily articles. There's this huge shotgun approach to handling them and you get situations where people point to the handling Altaic to boost the credibility of proto-Turkic-American Sign Language-Nivkh or whatever the flavour of the day is. Warrenᚋᚐᚊᚔ 23:33, 30 April 2024 (UTC)

...we really need a standardized approach to proposed macrofamily articles No, we don't. Each proposal has its own merits and non-merits. As I have said before in WT:LING, proposals of relationships beyond the currently established macrofamilies are not just an amorphous mass of bullshit. In our presentation of these hypotheses and their acceptance we go by secondary sources (per WP:RS/AC). This also includes evaluation of notability for standalone articles and the present question about how much of the proposed evidence we want to include here (and similar articles). If proposals are generally well-favored in secondary sources, it will be less of a WP:DUE-issue when we extract material from primary research articles than in the case of controversial proposals (including mostly rejected proposals). For the latter, we should ideally go by secondary sources when presenting the proposed evidence, although personally, I believe that a short paragraph with material from primary sources that is explicitly tagged as proposed evidence is also fine, especially with unquestionably notable proposals such as Altaic (Transeurasian).

Hence, I don't entirely reject the addition of the contested material from the Oxford Handbook. But:

1. The endless bloated list (following the table) with reconstructions of agriculture-related terms in proto-languages of the indivual language families taken from ch. 43 ("The homelands of the individual Transeurasian proto-languages") does not represent evidence for the Altaic/Transeurasian proposal and is thus off-topic. For uninitiated readers, it's visual prominence alone suggests "hey, the evidential base for Altaic/Transeurasian is huge" whereas it's not.
2. The gigantic table taken from ch. 44 does not just contains "some lexical reconstructions of agricultural terms argued by Robbeets (2020)", but in fact copies verbatim the first two columns all tables with macro-comparative reconstructions in that chapter. So it is in fact a blatant piece WP:COPYVIO. In place of it, I propose a short paragraph that goes "Robbeets (2020) proposes the following PTEA reconstructions of agricultural terms: *foo 'bar' (list proto-forms for the individual families here), ...."

@Dumuzid: you have done the last revert as an "uninvolved" editor. Since you have already commented in the 3RR-post and also because you are one of the few regulars of FTN that cares about non-STEM topics, I'd love to hear your take on it. –Austronesier (talk) 11:34, 1 May 2024 (UTC)

”As I have said before in WT:LING, proposals of relationships beyond the currently established macrofamilies are not just an amorphous mass of bullshit”

Austronesier, I deeply respect this reply but we’ve had a few interactions start out slightly spicy with you thinking we disagree more than we do. This seems to be one of them, and if that’s me communicating poorly with you I apologize and will work on that.

I’ll repeat again that I’m broadly a lumper, I just take serious issue with others who are more convinced by the lines of evidence which exist at present editorializing the acceptance of that evidence. Heck, I think there’s a chance Altaic is real, just fundamentally pre-evidential as far as a convincing application linguistic methods are concerned.

”No, we don't. Each proposal has its own merits and non-merits.”

I can bring this up back there again, but my concerns stem from the fact that there appears to be what, four or five of us monitoring this entire very niche sub-discipline with any regularity (at least insofar as talk page activity? The only reason we’re lucking out here with this discussion not lasting weeks without an RfC is it’s Altaic and not Amerind or something else similar. Perhaps there’s a middle ground between winging it each time and a strict rubric? Warrenᚋᚐᚊᚔ 12:04, 1 May 2024 (UTC)

Austronesier, I fear you think too much of my opinions! By way of quick disclaimer, I do have a background in linguistics, but that was a long long time ago and I have kept a toe in only intermittently. At this point I am best considered a dilettante. That out of the way, to the narrow point of the proposed addition--I agree that the recent work deserves mention, but I also agree that the reverted versions were far too much. Perhaps it is my dilettante-ness (dilettantitude?), but my lodestar approach is to remember that this is a generalist encyclopedia, and so I default to thinking of an intelligent adult with no background in the subject at hand. Therefore I think a couple of properly couched and footnoted sentences would be appropriate. Obviously I agree that the whole field of "Transeurasian" is far from a prevailing theory, but it doesn't quite fit "fringe" for me and feels more like a significant minority opinion. Again, though, I will certainly defer to my more learned colleagues here.

As to the wider question of a standardized approach, I am absolutely sympathetic to both advanced positions here. Warren is quite right that this is, even among the rarefied nerddom of Wikipedia, an area with a tiny interest group. So some sort of standard approach would seem desirable. But at the same time, I am sympathetic to Austronesier insofar as the obvious ideal is to evaluate each theory on its own merits. I don't think anything like a strictly mandatory approach would work, either philosophically or practically. But I think if perhaps an essay or other document could be put together it might be of some real help. I, for instance, would feel like I was on much more solid ground if I had a bit more guidance in how to approach and think about these subjects. TaivoLinguist is absolutely correct that these theories seem to fly fast and thick, with many of the old ones never really going away. However, as something of an Indo-European guy, I am reminded that William Jones had something of a kooky theory that ended up pretty helpful.

I don't know if much can be gleaned from the foregoing, but I would like to thank everyone involved here for their time and expertise which allow me to keep that toe in the water. Cheers, all. Dumuzid (talk) 14:03, 1 May 2024 (UTC)