Proto-(Balto)-Slavic/PIE

 

Godmy

Sīmia Illūstris

  • Censor

Location:
Bohemia
A thread with random questions for Hemo Rusticus

So I've been going over some Proto-Balto-Slavic and Proto-Slavic declensions, comparing them and trying to imagine how one ending evolved to another, usually I can form some "phonetic" map in my head, how can the vowels move and restructure themselves, etc.

But here's something which just seemed mysterious to me, so these two tables of the same word ("snow"):

https://en.wiktionary.org/wiki/Reconstruction:Proto-Balto-Slavic/snaigas

https://en.wiktionary.org/wiki/Reconstruction:proto-Slavic/sněgъ

How in the hell did the instrumental singular gain the "m"?? :D It seems to me like it was merged with some new suffix or something... Or was there some sort of nasalization etc.?

Also, I was surprised to learn that in Czech the masculine declension of the type hrad, hradu (castle, of castle) - I don't know if I can call it "u" declension (maybe it's not the right term, we call it "vzor hrad") having its two possible locatives: hradě and hradu - that the "u" is actually the dual ending and "ě" is the singular one, while I think that "u" has a preference often (at least with certain prepositions with this very word at least). With our word for "snow" the "ě" is impossible and only "sněhu" is the valid locative... Then we have a variant inanimate declension for the pán, pána (dominus, domini) which is les, lesa (silva, silvae) and there are two equally used locatives "lese" and "lesu"... I guess again the "e" is the original singular and "u" the dual... it's really awesome how the dual endings got smuggled to the singular if I'm understanding that correctly :p


Anybody else feel free to join for Slavic comparative/historical discussion! (as long as it's not in the spirit of false slavic friends and swear words...)
 
 

Godmy

Sīmia Illūstris

  • Censor

Location:
Bohemia
The other thing that captures my attention in this table https://en.wiktionary.org/wiki/Reconstruction:Proto-Balto-Slavic/snaigas were the identical endings in genitive plural and the original accusative singular. That means that if the modern Czech didn't have an issue with animate/inanimate and had the original accusative for this word, it should sound then the same today like our modern genitive plural, then it should be sněhů ... which is so strange and awesome in the same time :p

Although I guess each ending has often several reconstructions so... but it's at least some preliminary conclusion I made based on these limited data ; P
 

Glabrigausapes

Philistine

  • Civis Illustris

Location:
Milwaukee
How in the hell did the instrumental singular gained the "m"??
That's a good question. The common thought seems to be that of the three reconstructed PIE instrumental endings, * -eH1, -bhi, and -mi, the first was probably the 'original' and the latter two were highly unproductive innovations, and so your
It seems to me like it was merged with some new suffix or something...
is essentially what others (like Schmalstieg) think.

Consider the second ending -bhi, which in Greek is non-paradigmatic (being reserved for specific words in Homer, like ὄχεσφι 'with chariot(s)' or ἶφι 'with strength', cf. L vis), but is paradigmatic in Sanscrit as an instrumental plural (and so the pluralizing morpheme '-s' is added to it, as in dēvībhis 'with goddesses'). Such was probably the case with Slavic, beginning to expand upon the type -mi (which like you said was probably just a suffix to start with) after separating from PBS, while Baltic kept the 'original' ending, as seen in your paradigm and Lithuanian 'sniegu' = R снегом.

As to why modern Rus and presumably other dialects have 'hard' -m in instrumental sg., where OCS most often has the inherited soft, I have to guess analogy with dative plural. But isn't it true that Polish, for example, has a front vowel there? Like šniegem or something?

it's really awesome how the dual endings got smuggled to the singular if I'm understanding that correctly
Также думаю. Although I'm inclined to think your examples, like леса, лесу, are borrowings from u-stem singular (but I know nothing about Czech), Slavic is ever so clever in keeping old duals for regular plurals where it makes sense, like глаза for '(two) eyes' (so also older очи, cf. G ὄσσε < PIE *H3ekwiH1) and врата '(two) gates'.
 
 

Godmy

Sīmia Illūstris

  • Censor

Location:
Bohemia
Thanks for the response! I know in person too few (if any) PIE scholars focused on Slavics (usually they focus on Proto-Italic/Greek...).

That's a good question. The common thought seems to be that of the three reconstructed PIE instrumental endings, * -eH1, -bhi, and -mi, the first was probably the 'original' and the latter two were highly unproductive innovations, and so your

is essentially what others (like Schmalstieg) think.
Ah, cool then!

Consider the second ending -bhi, which in Greek is non-paradigmatic (being reserved for specific words in Homer, like ὄχεσφι 'with chariot(s)' or ἶφι 'with strength', cf. L vis), but is paradigmatic in Sanscrit as an instrumental plural (and so the pluralizing morpheme '-s' is added to it, as in dēvībhis 'with goddesses'). Such was probably the case with Slavic, beginning to expand upon the type -mi (which like you said was probably just a suffix to start with) after separating from PBS, while Baltic kept the 'original' ending, as seen in your paradigm and Lithuanian 'sniegu' = R снегом.
So... the latter two PIE reconstructed highly unproductive instrumental (singular) endings had to be carried (to Greek partially and Sanskrit) over the PBS all the way to PS, even though the reconstruction I referred to pretends the -mi variant doesn't exist anymore. But then, you yourself say (which I hadn't checked before), that Lithuanian instrumental sg. does come from this original ending, so it's rather that the table I saw simply didn't show the variant either in the PBS itself or the variant was absent in this very word, perhaps being in other words/declensions and PS simply "shifted" it... right?

Therefore one can't simply draw connections between the PBS and PS tables (at least if the tables are simplified as on the links I sent) not taking the PIE tables into consideration at all, since the "-mi" was clearly a "hidden" suffix there : P (kind of)

As to why modern Rus and presumably other dialects have 'hard' -m in instrumental sg., where OCS most often has the inherited soft, I have to guess analogy with dative plural.
Hmmm... how do we phonetically differentiate here between soft and hard -m, is it somethig with voicing? (Like, I know an unvoiced "m" or "n" appear(s) often in some PIE endings...)

Anyway, I can't serve much when it comes to this or Polish, but I guess it would be another interesting question (once I evolve "enough" for it :D).

Также думаю. Although I'm inclined to think your examples, like леса, лесу, are borrowings from u-stem singular (but I know nothing about Czech), Slavic is ever so clever in keeping old duals for regular plurals where it makes sense, like глаза for '(two) eyes' (so also older очи, cf. G ὄσσε < PIE *H3ekwiH1) and врата '(two) gates'.
Right, that makes a little bit more sense: the "dual" not being carried into singular just like that on the semantic grounds, without the noun marking some substance or consisting of two parts (dveře/vrata) or without the dual simply becoming a plural like in some of the nouns you mention (oči vs. oka; uši vs. ucha; ruce vs. ruky; -ama (inst.) vs. -ami (nohama vs. nohami) ).

Hmmm... ok, I should have thought that would be too simplistic solution what I suggested.

Anway, thanks again for the reponse. Feel free to spam this thread with Slavic stuff of this kind anytime another thread gets derailed too much again! This could be the universal "(Historical) Slavic thread" .
 
 

Godmy

Sīmia Illūstris

  • Censor

Location:
Bohemia
...it should sound then the same today like our modern genitive plural...
So, after checking the Latvian and Lithuanian tables on Wiktionary for this word, it's interesting that Latvian confirms this, but Lithuanian still makes distinction between acc.sg. and gen.pl (in this word). So the PBS reconstruction of this word, again, hides something which would be probably only explained by PIE tables + related pieces of information. What would you make of these two cases in Lithuanian?
 

Glabrigausapes

Philistine

  • Civis Illustris

Location:
Milwaukee
I know in person too few (if any) PIE scholars focused on Slavics
I remember there was only one member of the Slavic department interested in this stuff, but he's gone now.
I saw simply didn't show the variant either in the PBS itself or the variant was absent in this very word, perhaps being in other words/declensions and PS simply "shifted" it... right?
Yeah, that sounds about right.
between soft and hard -m, is it somethig with voicing?
Palatalized (soft) vs. unpalatalized (hard). That is, OCS дроугомь vs. R другом(ъ), < PBS *draugomi
Of course, in Proto-Slavic these sounds which became the твердый and мягкий знак were actual short vowels, like L optumus/optimus.
Feel free to spam this thread with Slavic stuff of this kind anytime another thread gets derailed too much again! This could be the universal "(Historical) Slavic thread" .
Конечно спамлю. (I shall spam):D
 
 

Godmy

Sīmia Illūstris

  • Censor

Location:
Bohemia
I remember there was only one member of the Slavic department interested in this stuff, but he's gone now.
Wow, so that's even worse than I thought. I suppose people deal with Slavic languages proportionately with their interest in Sanskrit (be it purely philologically or because of PIE), since Sanskrit seems to show often a great overlap (with Slavics) which is not exactly seen in the same way with Latin or even Greek. (or at least that's my impression)
 

Glabrigausapes

Philistine

  • Civis Illustris

Location:
Milwaukee
but Lithuanian still makes distinction between acc.sg. and gen.pl (in this word)
Well, yes, but the Lith word you're using, which maybe you got from the Latvian Wiktionary page, is better replaced with this.

But still, the question you raise is one that Schmalstieg does some considerable 'head-scratching' over in his historical Old Church Slavic grammar. It seems that, at least in the o-stems (like the word we're using), although the accusative sg. is supposed to be < PIE *-om, and the gen. pl. < *-ōm, they both end up undergoing nasalization (and in Lithuanian, even though the ogonek is used, the vowels aren't nasalized, but rather 'compensatorily lengthened') and reduction to short *-u in Proto-Slavic, often written with the yer symbol, and then to -ø in OCS.

A further problem is that we're not even positive that the gen. pl. ending in PIE had a long vowel, because some of our most useful attested forms (Latin included) are ambiguous. It is therefore possible that PBS had, to begin with, the same ending in acc. sg. and gen. pl., that is short *-om. This is why some historical linguists prefer to give the gen. pl. proto-ending as -(short/long)o (I can't even find the symbol for it: o with 'breve' and 'macron').
 

Glabrigausapes

Philistine

  • Civis Illustris

Location:
Milwaukee
The Lithuanian paradigm, which has acc. sg. -ą and gen. pl. -ų, supports the gen. pl. proto-ending having a long vowel, but then we would also expect something different in Slavic.
I suppose people deal with Slavic languages proportionately with their interest in Sanskrit (be it purely philologically or because of PIE), since Sanskrit seems to show often a great overlap (with Slavics) which is not exactly seen in the same way with Latin or even Greek. (or at least that's my impression)
I think you hit the nail on the head. I suppose I am myself one of them. :cool:
 

Glabrigausapes

Philistine

  • Civis Illustris

Location:
Milwaukee
It's a shame too, because the Slavic languages are every bit as rich and diverse as, say, the much-more-philologically-popular Germanic family.
 
 

Godmy

Sīmia Illūstris

  • Censor

Location:
Bohemia
The cases issue: wow! I didn't know I inadvertently came upon something which actually still turns out to be an open question in PIE/PBS/PS linguistics. I also guess the opinions that try to remove a hypothetical PBS language as a midstep (and the two sub-families relationship) altogether don't help clearing it too much either. But thanks for the explanation of the problem. I will also pay more attention now to the short/long o with both breve and macron!)

Then, on the other hand, Latvian (at least from that one table I saw & linked) seems to follow the suit of identity of the two endings at least in some proto-stage - I guess that also isn't too helpful.

Thanks for recommendations for the possible resources (Schamlstieg) too!

I noticed the PGMC also observes the gen.pl ac.sg. distinction...

Well, yes, but the Lith word you're using, which maybe you got from the Latvian Wiktionary page, is better replaced with this.
Ah, I know almost nothing about Lithuanian, yeah, I just followed the link - is there some problem with Lithuanian not using the former word much or similar?


I think you hit the nail on the head. I suppose I am myself one of them. :cool:
I guess I had a number of friends similar enough in who(m) I observed a heightened interest in, let's say Russian, after studying Sanskrit for some time (maybe it's also that Russian provides lots of resources/dictionaries for it too?), so I made a guess :)

It's a shame too, because the Slavic languages are every bit as rich and diverse as, say, the much-more-philologically-popular Germanic family.
Indeed! So Slavic languages are still the ugly duckling, but I would say a pretty large duckling : P ...
 
 

Godmy

Sīmia Illūstris

  • Censor

Location:
Bohemia

Glabrigausapes

Philistine

  • Civis Illustris

Location:
Milwaukee
That's another good one.

The riddle is solved by Schmalstieg as a difference of toneme in the proto-language. Proto-Slavic inherited different tonemes from PIE, and some (most? including Czech?) Slavic langs continue these tonemic differences. The terms used that I've seen are 'ascending/descending' or 'rising/falling' (represented, for example, by Greek Ζεύς/Ζεῦ, that is 'Zeus' in the nominative (ascending) and vocative (descending) respectively).

The nom. pl. in Slavic inherited the ascending accent of PIE supported by Greek evidence (to provide which we need a word accented finally, e.g. σταθμοί 'farmsteads'), so also did Baltic (although in Lithuanian grammar the 'circumflex' is used for this purpose, e.g. vilkaĩ 'wolves'). This ascending diphthong in *-oi became PBS *-aí, and monophthongized in Slavic: > S *-ī > OCS -и (incidentally, this monophthongization is later, and causes the 'second palatalization of velars' in Slavic). On the other hand, the locative was supposed to have a descending accent, which also monophthongized but with a different result (almost naturally, the vowel midway between low front 'a' and high front 'i', that is the 'ē' seen in attested forms). The wiki link should give some examples.
 
 

Godmy

Sīmia Illūstris

  • Censor

Location:
Bohemia
Thanks! So, in PBS the difference not shown is the accents/tones (albeit with same vowels) which fact exhibits itself in the daughter languages in vowel difference...

I wanted to check also the modern Lithuanian declension on this, then I remembered those words of my friend who said there should be some locative case borrowing in Lithuanian somewhere from Finnish (possibly with the original locative having been lost?).
 

Glabrigausapes

Philistine

  • Civis Illustris

Location:
Milwaukee
who said there should be some locative case borrowing in Lithuanian somewhere from Finnish (possibly with the original locative having been lost?).
I've heard that also, but I can't give definite examples 'cause my Lithuanian is pretty poor. Evidently, some dialects influenced by Finnish/Estonian even have a few extra cases (like illative and allative).

But it is true that the Lith locative vilke does not reflect PIE *wḷkwoi (> PBS *vḷkai); it should rather be *vilkai (where diphthong ai is unaccented).
 

Glabrigausapes

Philistine

  • Civis Illustris

Location:
Milwaukee
which fact exhibits itself in the daughter languages in vowel difference...
Does (standard) Czech have tones/tonemes?
 
 

Godmy

Sīmia Illūstris

  • Censor

Location:
Bohemia
In the standard/general Czech there should be no pitch accent... but in the Ostrava dialect (a big city near the Polish borders) they do some strange stuff: shortening most of the long vowels, shifting the accent position from the beginning rather to the end and I also think that the quality of the accent is slightly pitch-like :p But they are still a minority accent-wise.
 

Glabrigausapes

Philistine

  • Civis Illustris

Location:
Milwaukee
I've been thinking what language within Indo-European would be the most challenging for an English monoglot to learn & speak convincingly (i.e. like a native), and I'm inclined to think a Slavic dialect of Slovene or Serbo-Croat. Not only would the English learner struggle with the complex ('older') morphology, grammar & syntax, but quantitative vowels (absent in standard North American) and tonemics (absent in much of IE as a whole).
 
Top