Reconstruction talk:Proto-Slavic/tьlknǫti

Yer

@Useigor All descendants reflect ъ, none reflect ь. Why the move? —Rua (mew) 17:32, 27 April 2019 (UTC)

• @Rua: The Baltic data (Lithuanian tìlkti) demands pBSl *-i-. I suppose it's just a convention to reconstruct proto-Slavic *-ь- in such cases. Note that the present stem of *telkti is also reconstructed with *-ь- (by Vasmer, et al) even though the available evidence consistently points towards -ъ-. Bezimenen (talk) 12:00, 29 April 2019 (UTC)
  • Yes, Baltic i is cognate with Slavic ь. But by moving the page, you've also moved the problem. Now, the Proto-Slavic form agrees with Baltic but not with its own descendants, which seems worse to me than Slavic agreeing with its descendants but not with Baltic. It may be possible that the descendants do agree with the Slavic after all, I'll have to see. —Rua (mew) 12:04, 29 April 2019 (UTC)
  • Ok, I had a bit of a look around. If *vьlkъ is any indication, the sequence *ьl becomes ъl in East Slavic, so at least that part is regular. That leaves OCS as the only irregular descendant. —Rua (mew) 12:07, 29 April 2019 (UTC)
    • @Rua: In East Slavic, the distinction between front and back vowels was generally lost before -l- (cf молоко (moloko) < *melko). And in West South Slavic, -ь- and -ъ- merged in all environments. In Serbo-Croation, in particular, *-ьl-/-ъl- > -u-, *-ьr-/-ъr- > -ṛ- and *-ъ-/-ь- (in strong positions) > -a-. As you said, only the Church Slavonic data is somewhat indicative of the actual state of affairs. Even there, though, the situation is somewhat chaotic. If we compare texts from different places and eras, we'll see that there is no consistency on the writing of emphatic -ь- and -ъ- (in non-palatal environments). Apparently, the distinction between the two had begun eroding already in 8-9th century. Check these examples - Examples of тлькнѫти in Church Slavonic texts [unfortunately, I cannot tell what the abbreviations М, З, A, CK, С, etc stand for]. In some places -ъ- is used, in others -ь- is used. There is no consensus. Bezimenen (talk) 13:13, 29 April 2019 (UTC)
      • Ok, thank you for that. We can keep the page where it is then. Regarding OCS, perhaps we should standardise on using the historically accurate yer just for consistency. —Rua (mew) 13:25, 29 April 2019 (UTC)
    • @Rua, Bezimenen: perhaps sometimes Polish distinguishes 'ьl' from 'ъl' (as 'il'):

    (words from Category:Proto-Slavic_nouns)

    0 ъL || kiełb chełm, [zełwa/zołwa] kiełbasa, mełnia || ... ||

    1 ъL || pułk [dług], mołwa, słońce, tłumacz || bark garb kark targ karm oskard garnek, gardło, barć barszcz garść karczma, marchew ||

    1 ьL || wilk, wilga wełna, pilść || wierch czerw sierp, ziarno, wierzba, wiersza, ścierń śmierć pierść pierś pierć szerszeń, serce, świder ||

    0 ьL || ... || czart, sarna, twardy czwarty ||

    ? ьL || ... || tarn/cierń parst/pierścień ||

    Also some verbs: vilgnąć milczeć-milknąć, dłubać bełtać pełzać

    —Игорь Тълкачь (talk) 20:52, 29 April 2019 (UTC)

    • According to some sources, cases with 'il' occur with labials in front which is not the case for this word.—Игорь Тълкачь (talk) 19:06, 30 April 2019 (UTC)

• @Rua: I agree. On assumption that ъ is secondary: 1) usually root1 and root2 are both front/back vowels (*gъnati-*ženǫ is exception if 'gъn' is not secondary), for this reason тлѣщи-тлъкѫ (tlěšti-tlŭkǫ) looks unusual, 2) some authors reconstruct *tьlkǫ though i don't see what descendants support this vowel. I don't mind if the page and related pages will be renamed to *tъlk-.—Игорь Тълкачь (talk) 18:50, 29 April 2019 (UTC)
  • @Useigor: I think we are overthinking the whole business with *-ьRC- vs -ъRC-. In most cases, the yer in those environments is emphatic and reflects a PIE syllabic sonorant (exceptions exist of course, e.g. *sъlnьce). I don't think the ablaut pattern had much to do with what yer arises. More important was the palatalization environment when it was inherited, e.g. in *gъnati-*ženǫ < PIE *gʷʰénti or *pъlkъ < Germ. *fulkaz. PIE *tl̥k- is palatalization neutral, though, so both -ъ- and -ь- are possible. In such cases, the general practice is to reconstruct what suits the external comparison.
  • Moreover, note that there are some Church Slavonic texts which use -ь- in тльк-. It's not like all the evidence points towards -ъ-. My guess is that Medieval scribes tended to write -ъ- to emphasize that #t- was non-palatalized, not so much that the word was pronounced with ъ. Words like тьма (tĭma) were also occasionally written with -ъ-, even though we have plenty of evidence that it had *-ь-. Bezimenen (talk) 23:44, 29 April 2019 (UTC)
    • Your analysis of these as palatalisation is incorrect. You need to look at Balto-Slavic before you can look at Slavic. There is no such thing as palatalization in Balto-Slavic nor in PIE, so "palatalization neutral" is meaningless from a historical point of view. In Balto-Slavic, syllabic sonorants received an epenthetic vowel before them, which was usually i but sometimes u for unknown reasons. These vowels were directly inherited in Slavic as yers. They were very much real vowels and still are in the Baltic languages, which have no such thing as a "yer" anyway; they are real short vowels as much as any other. Palatalisation only came into play in Slavic, when the first palatalisation triggered changes to velar consonants when a front vowel followed. Palatalisation was thus also triggered where Balto-Slavic had earlier added an i before syllabic sonorants, but not in the rare cases where it added u.
    • In the context of South Slavic, palatalisation is also an anachronism. What we know as general palatalisation from languages such as Polish or Russian did not exist in Proto-Slavic, and never existed in South Slavic either. Its development was limited to West and East Slavic, and then it was later lost again in Czech and Slovak. Consequently, what we find in OCS has nothing at all to do with palatalisation. —Rua (mew) 11:38, 30 April 2019 (UTC)