Talk:Proto-Balto-Slavic language: Difference between revisions

Text and/or other creative content from this version of Balto-Slavic languages was copied or moved into Proto-Balto-Slavic language with this edit. The former page's history now serves to provide attribution for that content in the latter page, and it must not be deleted as long as the latter page exists.

Languages C‑class High‑importance

This article is within the scope of WikiProject Languages, a collaborative effort to improve the coverage of languages on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.LanguagesWikipedia:WikiProject LanguagesTemplate:WikiProject Languageslanguage articles C This article has been rated as C-class on Wikipedia's content assessment scale. High This article has been rated as High-importance on the project's importance scale.

Proto-Balto-Slavic is controversial bordering on pseudoscience. At one time it was a legitimate topic, like the Phlogiston theory in chemistry. The latest thinking is that Proto-Slavic is very recent (Iron Age), apparently a hybrid of an unknown North Iranian language (something like Scythian) and an unknown west-central Baltic language (something in between Lithuanian and Old Prussian). Zyxwv99 (talk) 19:21, 3 December 2011 (UTC)[reply]

That slavic is influenced by some iranian language is nothing new to the field, but how does that make the grouping together of baltic and slavic languages to a common ancestor pseudoscience? Actually, you say yourself that slavic share a comon ancestor with baltic and that is exactly what makes a language subgroup. The controversy around this idea has rather been if they share a common ancestor at all or if they just have influenced one another so much due to geographical proximity so that they appear to be more closely relate. But even this former controversy is not a topic among specialists anymore. The current view is that the baltic and slavic languages share a common ancestor and that both, but especially slavic, has been influenced by iranian languages in the past. Amilah (talk) 11:19, 7 December 2011 (UTC)[reply]

I completely agree with Amilah. Your objection demonstrates confusion, not a solid understanding of the problem. Proto-Slavic, as in the most recent common ancestor of the attested Slavic languages, was (according to the present scholarly consensus) spoken at some point in the early to mid first millennium AD (about 600 AD according to Holzer) – that's the Iron Age or even the Migration Period.

The Proto-Slavic language was not a hybrid (although isolated scholars with confused understandings about language contact and glottogenesis vs. ethnogenesis have apparently proposed that, but they would also consider Middle English a pidgin language), but it absorbed a number of Iranian loanwords (due to some extent of Slavic-Iranian bilingualism) and was apparently adopted as a lingua franca by a motley crew of ethnic groups in Eastern Europe in the Migration Period, among which were also some Iranian-speaking groups. (The dialects of Iranian that Slavic was in contact with seem to have been of the Scythian group, which includes Sarmatian, Alanic and modern Ossetic, but the loanwords do not exhibit any specific characteristics which can be used as diagnostic criteria for any particular form of Iranian.)

While traditional Proto-Slavic reconstructions look almost indistinguishable from Old Church Slavonic and Old East Slavic, there is a lot of evidence (which has actually been known for a long time) that a "Great Vowel Shift" as well as the liquid metathesis and pleophony took place in Common Slavic (a large coherent dialect continuum which was only weakly differentiated) as late as about 800 AD, and some scholars, including Holzer, prefer to reconstruct a much more archaic Proto-Slavic which happens to look extremely similar to the attested Baltic languages. (Unfortunately, our article on Proto-Slavic fails to even mention this development, but Proto-Slavic borrowings uses it, for reasons that become obvious once one examines the loanwords.)

While the issue of Balto-Slavic has seen a lot of controversy in the 20th century, the findings of accentological research especially have lately convinced a seeming majority of scholars that the Balto-Slavic hypothesis is hard to ignore (and there is even important additional evidence), and Fortson's influential textbook teaches Balto-Slavic as the state of the art. That enough is a reason to consider Balto-Slavic as part of the established consensus in the field now, and dissenters as representing a distinct minority.

Proto-Slavic does not descend from any known form of Baltic, and even if it did, that would only mean a stronger version of the Balto-Slavic hypothesis, far from a refutation: it would make Balto-Slavic effectively identical with Baltic, and Proto-Slavic a Baltic language.

The only remaining issue is subgrouping. Slavic, Latvian-Lithuanian and Old Prussian are clear branches, but their precise classification remains unclear. Proto-Slavic, (Pre-)Proto-Latvian-Lithuanian and a direct ancestor of Old Prussian seem to have been coordinate (extreme?) points in an extensive Balto-Slavic dialect continuum which mostly disappeared in the course of the Middle Ages, due to the spread of some language groups, especially the Slavic expansion. (Due to the close similarity, any speaker of a Baltic/Balto-Slavic dialect, and perhaps also of many pre-Slavic ancient Balkanic dialects, must have found it easy to acquire Early Slavic, which may well have considerably aided its spread as a vehicular.) Regrettably, very limited knowledge is available about the disappeared dialects, while names and locations of several groups have been reconstructed, and there are estimates of their extinction. One of the disappeared dialects is (Old) Curonian, which remains as a substrate in western dialects of Latvian and Lithuanian. How exactly the extant and well-attested Balto-Slavic languages were related to the extinct ones remains unknown and even unknowable pending the turning up of more evidence for them. --Florian Blaschke (talk) 15:32, 24 April 2012 (UTC)[reply]

The following rant was added in the article by Drosofilos. It is obviously not encyclopedic material, but I am copying it here on the talk page for posterity. --Trɔpʏliʊm • blah 00:37, 12 January 2015 (UTC)[reply]

“Balto-Slavic” is a rather doubtful concept and its current political context of Russian expansionism it’s impossible to ignore. The use of Balto-Slavic or Baltic & Slavic is probably more about contemporary political stance than about any science. The authors on Wikipedia side with those who say Balto-Slavic and the Baltic Languages entry in Encyclopedia Britannica, for example, doesn’t even mention the concept.

Many Lithuanians who attended school attentively in the Soviet times would be surprised to find out that their language is Balto-Slavic. Not surprisingly, one of the authors in the references of the Wikipedia entry Balto-Slavic languages, Thomas Olander, named his PhD Det baltoslaviske problem – Accentologien. Problem implies a solution. The same Wikipedia has an entry Final Solution. The choice of the title is astonishing.

The articles on Wikipedia (Baltic languages, Balto-Slavic languages, Proto-Balto-Slavic language and Indo-European languages) assert the existence of Balto-Slavic and do not provide any scientific evidence for this, if you take a closer look. The biggest argument I encountered was that the majority of scholars uphold that view.

The section Historical dispute of the entry Balto-Slavic languages in Wikipedia lists seven sources. Anyone who has ever come into contact with mathematical statistics would know very well that one doesn’t make any scientific conclusions from seven observations. The entry Indo-European languages in Wikipedia mentions four sources, one of them from the XIX century.

Nevertheless, the arguments themselves are not provided. The reader has to have trust: “Beekes (1995:22), for example, states expressly that "[t]he Baltic and Slavic languages were originally one language and so form one group".[8] Gray and Atkinson's (2003) application of language-tree divergence analysis supports a genetic relationship” (from Balto-Slavic languages in Wikipedia).

The accuracy and scientific validity of these articles is questionable. The entries on the “Balto-Slavic problem” may be more interesting as raw data, as a possible indication of political bias. Maybe to play a scientist even wasn’t very difficult, because the discourse about Indo-European linguistics takes us into the midst of the XIX century.

“In the 16th century, European visitors to the Indian Subcontinent began to suggest similarities between Indo-Aryan, Iranian and European languages”, says Wikipedia (Indo-European languages). I have just returned from Sapmi and kept listening to Sami radio at home afterwards. Remarkably, I found it strikingly similar and often understandable to a Lithuanian for a language that is not even Indo-European.

Let’s return to the basics: “A media text is made by a particular media institution and this will also affect the way that it is constructed and the meaning it communicates” (from Revision Express AS and A2 Media Studies, Mr Ken Hall & Philip Holmes). Sometimes it’s great to know who bakes your bread.

Since this continues to be reinserted, here is some actual discussion on the points here.

The most obvious point is that this is an opinion piece and it unambiguously has no place in the main article. Wikipedia categorically should not have phrases like "let's return to the basics". Self-references to things like "the authors on Wikipedia" are also out of place. Yet, all of this is right at home here on the talk page, so:

1. "Current political context of Russian expansionalism": I fail to see how this is even remotely relevant. The hypothesis is much older than the current political environment. Also, why single out Russian? The Balto-Slavic hypothesis equally well proposes a distant affinity of Lithuanian with e.g. Polish or Bulgarian. There is zero implication of a special relationship between Baltic and Russian specifically. (I get an impression that this piece has been written by someone with a political axe to grind.)
2. "The same Wikipedia has an entry Final Solution." Pardon? Certainly we have an article Final Solution (as we should), but it's about a topic completely unrelated to the current article, and nowhere does our article even mention the word 'solution'. Hence, allow me to offer a marginally more convincing argument:
   "The article claims that there was a Proto-Balto-Slavic word for 'salt'. Do you know who else ate salt? The Nazis, that's who."
3. "The articles on Wikipedia (…) do not provide any scientific evidence for this." Have you even read the article you are editing? There is an abundance of shared features listed and sourced, including e.g. the ten-point list beginning with "RUKI law" and ending with "*wl, *wr > *l, *r word-initially".
   — I do think that more discussion of the evidential value of all features involved would be good. As noted, most of these are simply widespread areal features across the late Indo-European dialect continuum, and of the ten bullet points, only Winter's law has a clear chance of being a defining synapomorphy.
   (Arguments such as "Winter's law preceded the voiced stop merger, therefore the latter was a specific B-S innovation" are hogwash, since the IMO superior third option left unmentioned — an analysis of the voiced stop merger as an areal phenomenon affecting various IE dialects such as Baltic, Slavic, Albanian, Iranian, etc. Alas, most historical linguistics dealing with subgrouping has been written from an oversimplified cladistic point of view, where all similarities must be either genetic or unrelated.)
4. "[O]ne doesn’t make any scientific conclusions from seven observations." Aside from the transparent failure to understand that one source does not equal one observation, the flaw in this critique is that Wikipedia is a work in progress. If you think there are important references against the idea missing, feel free to bring them up. If you think you would like to have more citations for specific claims, feel free to request some.
5. "Remarkably, I found [Sami] strikingly similar and often understandable to a Lithuanian for a language that is not even Indo-European." Feel free to work out a couple of similarities between Lithuanian and Sami that come anywhere near the similarities between Baltic and Slavic; maybe you can then release an article in a linguistic periodical where you demonstrate that the similarities between Lithuanian and Russian are no greater than the similarities between Lithuanian and Sami, which we can then cite here as a critique. I am, however, skeptical.
6. "A media text is made by a particular media institution and this will also affect the way that it is constructed and the meaning it communicates". Of course, which is exactly why we preferrably do not use media texts as sources, but instead peer-reviewed scientific literature. --Trɔpʏliʊm • blah 20:42, 13 January 2015 (UTC)[reply]

User Tropylium does not appear to understand the nature of sound change and the evidence used for grouping languages. A group of shared innovative sound changes, in the same order and not collectively found in nearby languages, is strong evidence of a clade grouping them.

Actually, I do know this. No need to get personal here.

The mere fact that change X is also found in language X and change Y is also found in language Y is not counterevidence unless you find another language sharing all of them (or at the very least, almost all). In fact, I doubt that all the other changes are actually found in a "wide variety of languages"; e.g. the change of ṛ -> ir is not. Hence this is OR. On top of this, not all Balto-Slavic scholars even accept Winter's law, but all agree that the groups form a clade. Benwing (talk) 05:44, 22 February 2013 (UTC)[reply]

The value I found in the statement is exactly what you are saying as well: if items 5 thru 12 can be dated as posterior to Winter's Law, then that demonstrates that any changes in there shared with other languages (perhaps not all of them, yes, but many certainly are) regardless cannot constitute evidence for any grouping wider than or alternativ to Balto-Slavic. Without pointing out this about the chronology, a reader might be lead to wonder how e.g. RUKI + delabialization + satemization + possibly o → a could not be used instead to define a "Proto-Satem" group etc.

For that matter, some of the orderings listed are somewhat suspicious and would benefit from elaboration. o → a has nothing to do with RUKI, so how DOES one show that the former precedes the latter? --Trɔpʏliʊm • blah 15:36, 22 February 2013 (UTC)[reply]

What evidence is there for reconstructing a laryngeal H in Proto-Balto-Slavic? None of the descendants have it, so why is it necessary to reconstruct it? CodeCat (talk) 23:52, 14 June 2013 (UTC)[reply]

It is known the the "acute accent" in Balto-Slavic was more like an acute register distinction, which could occur on all syllables, not just the stressed one. This is still the case in conservative dialects of Latvian, esp. those that preserve the three-accent system on long vowels (acute, circumflex, broken), which can potentially occur in all syllables; the broken tone was originally an allophone of acute register on unaccented, esp. pretonic, syllables, cf. Latvian gal̂vâs "on the heads" < *galˀvāˀséˀ or similar with acute register apparently on all three syllables and accent on the last one, cf. Lithuanian galvosè. This suggests that some reflex of a laryngeal was still present in Proto-Balto-Slavic, although unclear whether as an actual consonant or as a suprasegmental phoneme (e.g. glottalization of a long vowel or diphthong). Kortlandt at least, however, reconstructs an H still present and actually has different outcomes depending on position, including a different outcome for CHV vs. CV, unlike other non-Indo-Iranian IE langs. (However even in his theory it's possible that it could be written so that these H's were actually present in a pre-BS stage, and had developed into a supersegmental features by the BS stage proper. Have to look at his proposed sound changes to see when the Latvian vs. Lithuanian vs. Slavic split happened relative to the changes sensitive to the exact position of H.) Benwing (talk) 08:12, 15 November 2013 (UTC)[reply]

So Latvian has a register distinction on non-accented syllables, and this distinction is inherited from PBS? I didn't realise that... CodeCat (talk) 14:52, 15 November 2013 (UTC)[reply]

Someone else showed me this paper, which was published in a well-known Baltic linguistic journal. It suggests a regular Balto-Slavic sound change can be posited, in which word-final o > u before word-final -s and -m. Since this is a new publication (only a year old), I don't know how widely accepted it is, but I do think it's worth mentioning in this article at least. CodeCat (talk) 01:40, 17 April 2014 (UTC)[reply]

That was me. Given that Wikipedia does not only report on generally accepted textbook knowledge, but also on peer-reviewed hypotheses proposed by specialists, I see no fundamental reason to exclude this hypothesis. What is crucial is that the reader can tell the difference between consensus knowledge and proposals. --Florian Blaschke (talk) 16:17, 17 April 2014 (UTC)[reply]

By the way, there is a whole sequence of sound changes proposed in that paper: an Osthoff-like shortening (following Hirt's law), a specific apocope of */i/, a loss of */d/ under certain conditions, and finally the raising. (Oh, and the first "word-final" in your second sentence should be stricken.) --Florian Blaschke (talk) 16:27, 17 April 2014 (UTC)[reply]

Admittedly, our policy WP:UNDUE is a possible reason against that I neglected to keep in mind. Recentism and crystal-balling could also be criticised. However, even if there may be a lack of reactions and citations so far, I would like to point out that a crucial hallmark of fringe views (in the broadest sense!) is absent here: there is no serious contradiction with established views, quite the opposite; a raising like this (but with different conditions) has been proposed before, by Kortlandt, and the paper adds to the already mainstream support of Balto-Slavic. So the paper is entirely within the mainstream. Ultimately it comes down to editorial judgment, I think. --Florian Blaschke (talk) 13:05, 18 April 2014 (UTC)[reply]

I do think it's the most plausible argument so far, and I like it because it explains many phenomena in Balto-Slavic in one go. I was always puzzled by how Slovene somehow has -mo in the 1st person plural of verbs, while OCS has -mŭ. And there are other similar things that are explained very neatly. So I think we should include it in the article for its explanatory power if nothing else. Of course Wikipedia can't decide which theory is correct, but we can still decide which theories have merit based on existing research. And like you said, this explanation fits into it, it's not a fringe theory, just a more recent extension of existing knowledge. CodeCat (talk) 13:53, 18 April 2014 (UTC)[reply]

My point exactly: we're not knowledge-compiling robots, we can apply our specialist knowledge – within reasonable limits – to judge sources for their merit. Another problem which I recall mentioned in this context was the ending -o found in (medieval and later) South Slavic o-stem proper names. Perhaps this approach can even help explain the mysterious Novgorod/Pskov Slavic o-stem nom. sg. ending -e, although this would only be regular as the umlaut/iotated variant of -o. --Florian Blaschke (talk) 14:07, 18 April 2014 (UTC)[reply]

Does this -e ending trigger palatalisation at all? CodeCat (talk) 14:26, 18 April 2014 (UTC)[reply]

Apparently not, although I don't know what the ending of soft o-stems is. Interestingly, the accusative is as in other Slavic languages, so the case forms have not merged. This appears to be the most problematic feature of the Old Novgorod dialect overall. --Florian Blaschke (talk) 15:47, 18 April 2014 (UTC)[reply]

Ah, here's an appealing attempt to explain the ending, which is compatible with Hill's proposals. (Several pages are not visible online, but the conclusion reveals the general idea.) --Florian Blaschke (talk) 15:57, 18 April 2014 (UTC)[reply]

Some Slavic languages have almost completely eliminated the vowel alternation that resulted from fronting. Slovene for example only kept the o ~ e alternation, and only some instances of it at that, while it has eliminated all other vowel alternations generally in favour of the soft endings. It also has no more consonant alternations in nominal declensions (verbs still have it). So the -e might simply be a result of levelling in the same way. It's actually more likely that levelling occurred earlier in Old Novgorod than in any other Slavic languages, because palatalisation became unproductive much earlier in that language than in the others. Once it had become unproductive, the alternations would have become subject to levelling. CodeCat (talk) 16:18, 18 April 2014 (UTC)[reply]

Good point. This factor could have been part of the motivation for the change, but the involvement of the vocative looks plausible to me since the -e nominative is attested mainly in proper names in the early period, as pointed out in the paper I've linked to. --Florian Blaschke (talk) 16:15, 19 April 2014 (UTC)[reply]

The vocative was inherited -e rather than -o fronted to -e from a preceding soft consonant, though. So for the vocative you'd expect the first palatalisation, which occurred earlier in Slavic and is present in Old Novgorod without exceptions as well (I think?). On the other hand, several of the items listed at Old Novgorod dialect show that it had a general tendency to replace hard endings with soft endings, much like Slovene. So replacing hard nominative -o with soft -e just fits into that same tendency. CodeCat (talk) 16:25, 19 April 2014 (UTC)[reply]

Ah yes, now I get your drift! So perhaps the -o ending was generally preserved longer in proper names in Old Slavic, and survived longest at the geographical periphery. Neat! --Florian Blaschke (talk) 13:44, 20 April 2014 (UTC)[reply]

From the paper about final-syllable -o-, we know that at least for some time, there must have been a distinction between stressed nominatives in -us and unstressed in -as in early Slavic. Those would have become -ъ and -o, and would have had soft-stem variants -ь and -e. -o/e survived only in names in most Slavic languages, and was completely ousted by -ъ/ь in all other words. But in Old Novgorod, it was not ousted, and instead -o survived until it was replaced by the soft-stem equivalent -e, along with other endings which were also replaced in that way. The survival of -o/e was not limited to names in Old Novgorod, it appears on all masculine o-stems. So the question that remains to be answered is why -ъ/ь replaced -o/e in the rest of Slavic. I think the answer is similar to the conditions that led to soft endings replacing hard ones in some Slavic languages. Here, the conditioning was accent, and Late Common Slavic obviously had a lot of changes affecting accent, in particular Ivšić's law, which caused accent to retract away from final yers. At that point, the accent conditioning would no longer have been obvious, setting the stage for analogical replacement of the ending. The replacement spread across most of Slavic territory, but the Old Novgorod dialect escaped it because of its isolation. CodeCat (talk) 17:19, 20 April 2014 (UTC)[reply]

I've rearranged and rewritten large parts of the article. As it stood originally, the article seemed more like an unconnected collection of sections rather than a single flowing whole. There were even points where there was contradiction with things said earlier. I've attempted to clean this up a bit, and I hope it's an improvement. More can probably be done still. CodeCat (talk) 21:39, 18 April 2014 (UTC)[reply]

* PIE *pl̥h₁nós > Proto-Balto-Slavic *pilˀnas (Slavic *pьlnъ, Lithuanian pìlnas), > Proto-Germanic *fullaz (< earlier *fulnaz, with regular *n̥ > *un; English full), but > Latin plēnus (*n̥ > en normally).

Wait. What's going on in Germanic in this word is clearly not *n̥ > *un, but *l̥ > *ul (just as regular). But what happened in Latin to produce the long vowel? While I am at it, the article currently says Holzer found 12 sound changes from PIE to Proto-Balto-Slavic – and then it lists only 10. What are the other two? David Marjanović (talk) 23:48, 23 October 2014 (UTC)[reply]

Germanic has been corrected, but Latin is still wrong, it should say that syllabic l > ol or ul normally. As for plēnus, that isn't the regular outcome; it should be *plānus, compare Proto-Celtic *ɸlānos. What happened is evidently that the expected vocalism was changed in analogy to plērus and the verb plēre, perhaps already in Proto-Italic. --Florian Blaschke (talk) 05:28, 30 October 2014 (UTC)[reply]

The article states that there are 12 common innovations but only lists 10? 1 appears to be final -m to -n as stated shortly after, that should be fixed I think

Also it states that all the sound laws except winter's appear in other indo-european branches and i'm wondering what other indo-european branches did changes 8 through 10 happen?

Change 8 also occurred in Italic and Celtic I believe. 9 is unique to Balto-Slavic, but the insertion of specifically a closed vowel is shared with Germanic (which inserted -u- in all cases). 10 is quite a common change and probably occurred in several branches although I'm not sure which. I do know that Celtic and Germanic don't have it, although many Germanic languages (notably Old Icelandic, High German, more recently English) did undergo the change in historical times. CodeCat (talk) 16:21, 19 November 2014 (UTC)[reply]

See Talk:Winter's law.