Talk:Uralic languages/Archive 5

Stem variants in comparative vocab table
We list some stem variants for Finnic nominals, where this is an important feature. They are listed for verbs as well, though here most classes of verbs are not underspecified in their basic forms the way e.g. Finnish syli can be the nominative of both syle- "fathom" or syli- "bosom".

I recently added stem variants also for the Samic words for "to pee", to make clear that they are verbs and not nouns. However for consistency's sake, should we perhaps also: I'd be happy listing just the actual stems for verbs, but from past experience, this seems to attract editors who then inconsistently correct one or two entries to their citation forms. -- Trɔpʏliʊm • blah 15:37, 5 July 2014 (UTC)
 * 1) indicate stem forms for the other Samic verbs as well?
 * 2) list the citation forms for Mordvinic/Mari/Permic as well? E.g. for "to go", Mari miješ (mija-), Komi mynyny (myn-).

Classification history
I've rehauled the overview of the traditional classification system. I believe it should be apparent from this that there is no one traditional classification; rather there is a "maximal binary" arrangement, of which rather few people seem to have ever been 100% convinced.

Note that all cited sources are tertiary accounts, not works directly addressing the nature and evidence of the subgroups — with the possible exceptions of Janhunen and T. I. Itkonen, who passingly mention a small number of lexical innovations.

— This is moreover in fact a temporary fork from a draft article I am working on: Draft:Classification of Uralic languages. Contributions from other editors are welcome! -- Trɔpʏliʊm • blah 16:17, 9 September 2014 (UTC)

Map
The map is quite distorted. I initially didn't recognize any countries. --2.245.188.228 (talk) 22:41, 4 June 2015 (UTC)
 * It isn't "distorted". It takes Earth's curvature into account and is viewed from a particular point above. --JorisvS (talk) 08:55, 5 June 2015 (UTC)
 * Agree with above. See for example the article Russia for a very similar map right on top there, just zoomed out a little. --hydrox (talk) 14:43, 5 June 2015 (UTC)

Selected cognates carpathian.
Note: in hungarian 'fire' is 'tűz' (or 'tu"z' if the special character is unreadable). Why is it left empty (-) in the table? 82.131.155.95 (talk) 23:07, 22 November 2015 (UTC)
 * Because it's not cognate with the other forms for 'fire'. --Taivo (talk) 23:17, 22 November 2015 (UTC)
 * So you say it is a branch dravidian import word?
 * Hand (from wrist down) is 'kéz' (ke'z) in hungarian and 'arm' (the whole limb) is 'kar'. Both could be listed in the table, since both seem to fit?
 * I can't get why 'fathom' has a line in the table? The Urals isn't exactly a place to dive and most of the finno-ugrians only ever saw much water in a well. A less maritime concept could possibly replace it for a more useful table? Anyhow it's hungarian cell currently looks weird. The 'öl' (o:l) is an old unit of measurement app. 1.9 meters long, but it also means 'kills'. Yet 'ölel' (o:lel) means embraces. 82.131.155.95 (talk) 23:27, 22 November 2015 (UTC)
 * You need to learn the meaning of the technical linguistic term cognate. It's not just "similar", but related by regular sound correspondences and descended from a common term in the proto-language.  "Kéz" is cognate, "kar" is not.  And the terms for "fathom" are all cognate, no matter what you think about the appropriateness of the meaning.  Dravidian has nothing whatsoever to do with this table, so I don't know why you keep bringing it  up.  The Uralic languages are not related to the Dravidian languages.  --Taivo (talk) 23:40, 22 November 2015 (UTC)
 * The intended sense of fathom here is not a unit of depth in particular, but the unit formed by the opened and stretched-out arms. Do we have a better term we could gloss this with?
 * Hungarian öl 'kill' is, of course, unrelated entirely to the 'fathom' group. (It's instead related to Mansi äl-, Khanty wel- with the same meaning.)-- Trɔpʏliʊm • blah 00:40, 23 November 2015 (UTC)
 * Actually the "fathom" set is based on Björn Collinder, An Introduction to the Uralic Languages (1965, University of California Press) with the glosses "bosom, outstretched arms, fathom" and includes Hungarian öl, which is not the verb 'kill', but the noun 'lap' or the noun 'six feet, cord (of wood)' (either one fits within the gloss and they are undoubtedly polysemous at a very deep time depth, although they usually have separate entries in Hungarian dictionaries). --Taivo (talk) 01:13, 23 November 2015 (UTC)
 * Yes, sorry if I was unclear: of the three basic meanings of öl in Hungarian, the first two are related and belong in the cognate set under discussion, but 'kill' does not. (My Hungarian-Finnish dictionary in fact only lists two entries anyway and consideres 'fathom', 'lap' polysemous.) -- Trɔpʏliʊm • blah 00:06, 24 November 2015 (UTC)

And so castles made of sand, slip into the sea . . . (apparently NOT)
" All your hard won edits will be eventually eroded and washed away like a sand castle. It's a bit like Sisyphus eternally attempting to roll a rock to the top of the hill only for it to roll back down. --Nug (talk)
 * - did not occur! — Preceding unsigned comment added by 50.111.46.194 (talk) 22:01, 25 October 2016 (UTC)

New candidate of the homeland
In the History section, Homeland subsection, someone (ABCEditor) recently added: Recent ancient DNA analysis revealed that Uralic haplogroup N1 (Y-DNA) was originated from northeastern China, Liao river region, which is a new candidate of the homeland. This is referenced to a Eupedia page (Eupedia) that doesn't actually say that. It just says that the Proto-Uralic peoples were N1C1. The Eupedia page is about haplogroup N1c, and discusses the origins of N in general. Zyxwv99 (talk) 23:19, 16 November 2017 (UTC)


 * If a reliable source can be found for this, the content can remain, but I'm skeptical, if Eupedia is the only source. Tagged the material as better source needed. Mathglot (talk) 02:35, 17 November 2017 (UTC)
 * Haplogroup N1 in general or even N1c1 is not "the Uralic haplogroup"; for starters, it is much too old (see Haplogroup_N-M231). This piece of information, while probably sourceable, is either irrelevant for this article or a covert attempt at WP:SYNTH. -- Trɔpʏliʊm • blah 20:57, 17 November 2017 (UTC)
 * Agree with Tropylium. I came here to comment the exact same thing. Also, genetics don't determine which language people speak. Genetics is only useful in understanding linguistics when both (i.e. haplogroup and language) can reasonably be attached to a specific archaeological culture. The three disciplines must be investigated in context and, as Tropylium points out, N1 isn't "the Uralic haplogroup" anyway so this information is irrelevant to this article.--William Thweatt TalkContribs 22:55, 17 November 2017 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on Uralic languages. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20111002150529/http://helimski.com/2.140.PDF to http://helimski.com/2.140.PDF

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 13:51, 2 December 2017 (UTC)

older statement still needs a reference
" In the late 15th century, European scholars[who?] noted the resemblance of the names Hungaria and Yugria, the names of settlements east of the Ural. They assumed a connection but did not seek linguistic evidence.[citation needed] " - The tags are going on three years old or more.50.111.62.5 (talk) 06:34, 6 December 2017 (UTC)


 * Sometimes the easiest thing to do, is to just to go look it up yourself, and add it. That took me all of about 30 seconds to find.  Cheers, Mathglot (talk) 11:36, 6 December 2017 (UTC)
 * Historian, not a linguist - I don't edit non-SME articles.50.111.62.5 (talk) 17:38, 6 December 2017 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on Uralic languages. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20091218022542/http://finland.fi/Public/default.aspx?contentid=160054&nodeid=41807&culture=en-US to http://finland.fi/Public/default.aspx?contentid=160054&nodeid=41807&culture=en-US

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 04:44, 13 January 2018 (UTC)

IPA
added some additional IPA for the example vocabulary table. It already has some of these, so this could be a good idea to expand on (though I must disagree with adding IPA for reconstructions — it is not known whether e.g. *ć was [tsʲ] or [tɕ] or [c] or what). I think we should consider for a moment what information to include, though. -- Trɔpʏliʊm • blah 15:20, 3 October 2018 (UTC)
 * Obviously the languages written in the Latin alphabet should show the orthography, but what about the ones written in Cyrillic?
 * Since Uralic Phonetic Alphabet transcription is still common in Uralic studies, at least for languages that are unwritten or only written in Cyrillic, this could still be given in addition to IPA. If we were being sticklers about only following reliable sources, these would be in many cases the only transcriptions you could reasonably find anywhere.
 * Should IPA be given as phonemic or phonetic? Some languages have significant allophony, e.g. the Erzya for 'hand' is phonemically /kedʲ/ but phonetically [kʲedʲ]. But this would be hard to get right in all cases, and would probably be distracting details rather than anything especially relevant for the point.
 * Glosses could be added for divergent cases, like 'nest' > 'egg' in Udmurt or '5' > '10' in Samoyedic.

Answering
 * Linguistics is a science, and it should use the appropriate tools for scientific descriptions, in this case the widely accepted IPA.
 * Orthographies are not scientific, be it in the Latin, Cyrillic or Uralic Phonetic Alphabet, in which case a universal system (again, the IPA) is always preferred. Language-specific orthographic conventions can be included in addition to the IPA, but never instead of.
 * What does one gain from using a symbol like *ć? At least [tsʲ] or [tɕ] or [c] mean something specific that can be understood and tested by linguists. Based on the general acceptance that Uralic languages show palatalized consonants, a reconstruction using [tsʲ] is the most likely.
 * Using sub-sub-field specific notation (Uralic within Historical Linguistics within Linguistics) makes that field less accessible to other linguists and more prone to error.
 * This table is inconsistent; why use <ŋ> but not <æ>? Why use  instead of, ,  etc.? This is a mixture of two systems, again confusing. The IPA symbols, once learned, serve for every spoken language.
 * The paragraph that comes right after the table explains the cumbersome symbols using the IPA! Why make the job of interpreting the table less straightforward than it could be?
 * The IPA can be used for both phonetic and phonemic transcriptions; reconstructions are neither but may be considered closer to phonemic, since one would never know what a segment sounded like.


 * Watch the peacock terms — sure IPA "is scientific", but so is UPA. Their main differences are due to them being developed for different uses, for which different tools are appropriate. Nonspecific notation for reconstructed phonemes has precisely the benefit that it doesn't require nailing down a phonetic value, something that as you note yourself is not necessarily available. Sometimes, e.g. for *δ, there is no consensus even about a rough phonological value: some propose /ð/, some /d/, some /ɬ/, some /rʲ/, some /ɽ/, etc. Only the sound correspondences are established. Yet we need to be able to talk about the proto-phoneme without repeating ten different caveats at every turn. The usual convention is to use some degree of nonphonetic transcription. You can see this with most older proto-languages: *ḱ *ǵ *ǵʰ *h₁ *h₂ *h₃ for PIE; *ʒ *c *c̣ *ʒ₁ *c₁ *c̣₁ for Proto-Kartvelian; *ḫ *ḥ *ṣ *ṣ́ *ṯ *ḏ *t̬ for Proto-Semitic; etc.
 * Anyway I agree that all this makes Uralic studies and maybe historical linguistics in general somewhat less accessible, but even then a key point is that these systems are what reliable sources use. IPA is surely preferrable if available, but we cannot just pull some arbitrary IPA from our asses and declare it automatically better: it would have to be based on a source of some kind. (I myself happen to disagree that [tsʲ] would be "most likely" — only the Mordvinic languages directly point to this; most languages actually point to a more backed, postalveolar or palatal value — but of course, WP:OR.)
 * I also agree that giving IPA would be a good idea for the modern languages where phonetic studies do exist. I have reverted your edit on this per your insertion of arbitrary IPA into the reconstructions in the same edit, not due to being categorically opposed. However, before re-adding them, I'd appreciate a comment on the issue of allophony I noted above.
 * Some detailed problems in particular are:
 * the transcription of Komi/Udmurt orthographic ӧ. UPA usually uses e̮, which covers all IPA in the [ɜ ~ ə ~ ɘ ~ ɤ ~ ʌ] range. I've already earlier opted for ə, but this could be misleading. Some Udmurt dialects even have a "real" /ə/ corresponding to standard /ɨ/, in contrast to /e̮/).
 * the Komi/Udmurt palatals: some sources claim that ć ś ź are palatalized alveolar, some others that they are palatalized postalveolars . AIUI the source of this confusion is that "narrow" UPA does not make this distinction. However, "close" UPA always uses č́ š́ ž́, indicating the latter. But almost none of this is yet documented in our base articles!
 * reduced vowels ă ˇŏ etc. in Khanty?
 * almost everything about the Tundra Nenets vowel system: should we give vowel lenght always, or just for í ú, or also for ā? What do we do with the ə versus ° distinction?
 * -- Trɔpʏliʊm • blah 12:23, 5 October 2018 (UTC)

Table showing Relative numbers of speakers of Uralic languages
This table says that 2.5% of Uralic speakers are speakers of Moksha, even though only about 2,000 people speak Moksha. At the same time it says that only 2% are Mari speakers, a language which has over 500,000 speakers. What's going on? Dylanvt (talk) 23:46, 28 February 2019 (UTC)


 * That’s a good point. The number of Erzya speakers is 36,726. This gives a total of 40k for both Mordvin languages which is ridiculously low when compared with the number of ethnic Mordvins (ca. 800k). The Erzya and Moksha figures are from the Ethnologue, which cites data from the Russian 2010 census. I have checked the 2010 census data at the official Russian government website, which indeed has 36,726 speakers for  Erzya (эрзя-мордовский) and 2,025 for Moksha (мокша-мордовский), but also 392,941 people who responded as “Mordvin“  (мордовский) speakers without specifying which variant they speak. The latter were thus completely ignored in the Ethnologue figures, which were copied into the WP articles without further scrutiny or cross-checking. Austronesier (talk) 13:26, 1 March 2019 (UTC)


 * The "2000" figure is nonsense, as noted above, although due to the poor census methodology it would be somewhat challenging to get more accurate ones. I think we used earlier a rough assumption of Mordvin-NOS dividing as ⅓ Moksha, ⅔ Erzya, but this would really need an in-article discussion with sources (so far neither Moksha language or Erzya language does this). -- Trɔpʏliʊm • blah 15:24, 4 March 2019 (UTC)

Internal classification
What is NPOV when it comes to presenting the internal classification of a language family in the infobox? Is it a long-standing classification that is reflected in most textbooks and overviews, even though many of its components are increasingly contested? Or is it a list of the unconstested low-level subgroups, thus playing safe at the smallest common denominator, even though some larger groupings are still defended by contemporary scholars? And if it is something inbetween, by which criteria can we decide which larger groupings are more well-founded than others?

I'm asking this because the long-standing version of the infobox (contested by, restored by ) has Ugric as a primary branch, but not Finno-Ugric, Finno-Saamic, Volga-Finnic or other "traditional" branches. I have added Salminen as a source for the radical low-level approach, and could also further add a pre-print article by Aikio (to appear in an OUP overview volume). Ugric has been contested by various scholars, just like Finno-Saamic or Finno-Ugric in toto, so can we still justify to have it in the infobox? –Austronesier (talk) 14:58, 3 December 2019 (UTC)
 * Hi, you see it right. And we would enter into the (proto)-Indo-European field or Indo-Iranian, a domino-effect may arise vica-versa. So we should be careful and think twice.(KIENGIR (talk) 15:06, 3 December 2019 (UTC))


 * Not actually long-standing. The long-standing consensus version of the infobox was the agnostic nine-branch version, which was at some point silently changed to include the disputed "Ugric" branch. See Salminen (2002), and Aikio (2019) on why Salminen's conclusion still stands. --Florian Blaschke (talk) 18:47, 3 December 2019 (UTC)


 * Yes, Ugric was removed years ago.
 * There are parallels with IE, but they're in Italo-Celtic, Satem and Indo-Hittite, not in Indo-Iranian. Even Balto-Slavic (or Baltic itself), while there are doubts, those doubts have been around for a century and the general consensus is still that it's valid. It's not like someone just realized it was unsupported. Uralic (and Austroasiatic) are different -- there in recent years consensus has shifted away from binary branching. So in both cases we give an agnostic comb classification. In a few years, once the evidence is reevaluated, it may turn out that some of the mergers are salvageable, and that might include two or more of the Ugric languages (though if it's Hungarian and just one of the Ob-Ugric languages, I don't know what we'd call it). Meanwhile, best to restrict the debates to the text.
 * A possible compromise, if there is particular debate about Ugric, would be to add "(Ugric)" or "(Ugric?)" after Hungarian, Khanty and Mansi. I personally don't know if Ugric is more defensible than the other traditional branches. Is it perhaps a remnant of the Ugric-Turkic War? — kwami (talk) 20:05, 3 December 2019 (UTC)
 * Not long-standing then, which makes it all the better. I did some random clicks in the history trail, and apparently was unfortunate enough to find versions with Ugric in the infobox. Anyway, my main purpose of calling you here was to get a consensus so we can be patrolling the same thing and not against each other.
 * Note: the "Classification" section in the main article starts with the agnostic (Aikio's term: "shallow") classification, but only has a dead link to "Salminen (2009)", so we need a better ref here. I think Salminen (2007) will do the job: it is in English, and already published.
 * Personally, I agree with leaving all disputed nodes out of the infobox, including Ugric, which has no privileged position among other proposed groupings within Uralic. –Austronesier (talk) 21:33, 3 December 2019 (UTC)
 * Exactly. Indo-Iranian and Balto-Slavic are consensus now (see Fortson's textbook), and reconstructions exist whose distinctness from PIE reconstructions is blatantly obvious. Not so in the case of Ugric, let alone higher-level groupings like Finno-Permic or Finno-Ugric, whose distinctness from Uralic has never been demonstrated (Aikio's main point). There are no certain, generally accepted common innovations that could be used to define these groupings, and Finno-Permic or Finno-Ugric reconstructions do not differ relevantly from Proto-Uralic reconstructions. Proto-Ugric reconstructions? I haven't even seen any. The evidence is flimsy, and there is simply no consensus for these groupings (except, maybe, as areal groups), so they aren't suited for the infobox. Moreover, nine branches are hardly an overwhelming number, so that's not an argument against the comb, either. --Florian Blaschke (talk) 23:08, 6 December 2019 (UTC)

I do not have an immediate opinion on infoboxes in particular, but this reminds me that I used to have a draft for an article Classification of the Uralic languages around. It got deleted as stale sometime ago, but I could already revive it into reasonable article shape if there is interest for it. At least it might redirect some of the back-and-forth editing on this article into something more productive. — Also, at least our article on Ugric really ought to cover Honti (1997) in some detail, he's been the most accomplished defender of Ugric over the late 20th century, and I think it might be true to say that Ugric remains consensus within Hungary even if maybe not worldwide.

The "Ugric-Turkic War" is largely unrelated actually, early Hungarian Finno-Ugricists taking part in this used the term ugor idiosyncratically as referring to the entire Uralic (or at the time, Finno-Ugric) family, not specifically to a Hungarian + Ob-Ugric subgroup. In particular Josef Budenz actually supported a "North Ugric" group which he saw as comprising Sami, Permic, Ob-Ugric and Hungarian (and what we today think of as the clearest Ugric innovations he often saw as nondiagnostic common retentions). The narrow Ugric hypothesis is even older, e.g. Rasmus Rask and M. A. Castrén used to support it on relatively flimsy non-linguistic grounds. -- Trɔpʏliʊm • blah 10:07, 4 December 2019 (UTC)


 * Please do consider reviving your draft. An article like this could be very useful to limit the size of this one. Just give a short summary of the debate here and leave the nitty-gritty details to the specialised article.
 * Interestingly, Parpola mentions here on p. 142 that Janhunen supports Häkkinen's radical new proposal (West, Central and East Uralic), but Aikio evidently remains unconvinced. --Florian Blaschke (talk) 23:29, 6 December 2019 (UTC)

Counterarguments to Marcantonio
Given the large number of critical reviews, we can probably state that Marcantonio's book has been a notable attack against the Uralic theory. The criticism itself also establishes that it is, in fact, a fringe position that should not be given WP:UNDUE weight. In a relatively long article, we can probably spend some time on discussing it however. I additionally believe some of the objections will be helpful also for readers trying to assess other fringe proposals. Having re-read of all six critiques of Marcantonio, I have now identified and summarized the four most common objections against the book's main thesis.

There are many more repeating critiques that could be mentioned, e.g. but unless other editors are in favor of including more of these in the article itself, I will leave them here for now.
 * Marcantonio incorrectly claims that the Ugric hypothesis would be based only on the similarity of the names Hungary and Jugra, while completely ignoring the arguments collected by Honti (1979, 1999), indeed despite including these works in her bibliography (Kulonen, Laakso)
 * M. incorrectly claims that historical linguistics would be incapable of distinguishing inherited vocabulary from loanwords (Bakro-Nagy, Georg, Kallio)
 * M. mistreats the limited word corpus of Janhunen (1981) [n.b. selected specifically for illustrating the relationship of Samoyedic and the western Uralic languages; --Tropylium] as the entire etymological evidence base behind the Uralic family (Bakro-Nagy, Georg)
 * M. professes a conceptually mistaken belief that some contemporary languages would be "young" while others would be "old" (Bakro-Nagy, Laakso)
 * M. refers to the fact that the Hungarian people were called "Turks" in mid-1st millennium Arab sources, and incorrectly treats this as evidence for the affiliation of the Hungarian language (Bakro-Nagy, Laakso)

-- Trɔpʏliʊm • blah 19:05, 5 December 2019 (UTC)


 * Thanks. I guess it's not possible to avoid Marcantonio completely here, but she's basically a complete outsider who proved ignorant about historical linguistics, not only but especially on Uralic historical linguistics, so her criticism does need to get short shrift. What is currently in the article is just about still acceptable, but could probably use cutting down further.
 * Vajda is not a Uralicist, right? I'm not sure why his surprisingly positive opinion of Marcantonio's book (basically, she's asking valid questions somehow) is relevant here. --Florian Blaschke (talk) 23:18, 6 December 2019 (UTC)
 * Most of the reviews actually agree that usually a book like hers should be dismissed as much too amateurish, and that they are dedicating a longer review to it only due to it having been published by an established scientific publisher in a generally reliable comparative linguistics series, i.e. there is no clear a priori way for a non-expert to screen the book off as a probably non-reliable source. It is IMO telling how already a non-Uralicist like Vajda does not manage to realize any of the fundamental problems. -- Trɔpʏliʊm • blah 12:28, 7 December 2019 (UTC)


 * Agreed, but given that more than a few lay readers who have come across her name will probably wonder about her and expect coverage or at least a mention, and we have plenty of RS to contextualise her criticism, I'm okay with her being in the article. --Florian Blaschke (talk) 12:18, 8 December 2019 (UTC)
 * Oh wait, you were actually trying to argue that we definitely need to cover her criticism, right? Sorry, I completely misunderstood you. Okay, fair point. --Florian Blaschke (talk) 12:20, 8 December 2019 (UTC)
 * Oh yeah, that was your original point after all. D'oh! Just expand the section if you like. --Florian Blaschke (talk) 12:22, 8 December 2019 (UTC)

Khanty-Mansi.
This article fails ro recognise ANY proximity of Khanty and Mansi branches within Uralic? --Yomal Sidoroff-Biarmskii (talk) 22:40, 19 February 2020 (UTC)
 * Look "Classifications" section, there is written about proximity of different branches. --Minnekon (talk) 09:44, 20 February 2020 (UTC)

Assessment data removed
This article is not appropriate for WP Estonia, but there was the following assessment: --Estopedist1 (talk) 05:52, 12 April 2021 (UTC)

Uralic ≠ Uralian
In the article we can read that Uralic languages are sometimes called Uralian, which could cause misunderstanding.

The Uralic language family is not the same family as the Uralian language family.

I'm not sure if it's 100% correct so I decided to write it here so more qualified people could do something with that. Teh kaczuch (talk) 10:24, 23 November 2023 (UTC)


 * They are the same. The handbook by Sinor (1988) tells so explicitly. (Preface, p. x) Jähmefyysikko (talk) 13:29, 9 January 2024 (UTC)