Template talk:sla-noun
Latest comment: 11 years ago by CodeCat
Shouldn't this be {{sla-pro-noun}}? —Μετάknowledgediscuss/deeds 02:26, 30 December 2012 (UTC)
- Technically yes, but there will never be a conflict because "sla" is not a language code but a family code. And this is shorter. :) —CodeCat 02:31, 30 December 2012 (UTC)
- Well, as long as it gets you to make more Proto-Slavic entries, I guess I'm fine with it. (I bet Category:Yiddish terms derived from Proto-Slavic and its contents need some cleaning up, BTW.) —Μετάknowledgediscuss/deeds 02:35, 30 December 2012 (UTC)
- Why would it need some cleaning up? —CodeCat 02:40, 30 December 2012 (UTC)
- In my experience, nobody agrees on how to notate Slavic reconstructions. Whether that's just orthography or real reconstructional differences is beyond me, because I've never studied it. —Μετάknowledgediscuss/deeds 02:50, 30 December 2012 (UTC)
- From my own experience they are easier than Germanic reconstructions because there is much less time between Proto-Slavic and the attestation of Old Church Slavonic. In a sense, OCS still is Proto-Slavic, albeit with significant Bulgarian dialectal features. It is very easy to reverse the OCS changes and find the original form, if it differs at all from the OCS form. And when there are differences usually there are clear patterns that are easy to recognise. Such as East Slavic -oro-, South Slavic/Czech -ra-, Polish -ro- which implies Proto-Slavic -or-. The differences in notation are usually just a matter of denoting the sounds since most linguists seem to agree on which sounds there were, just not on the exact details or progression of sound changes. For example I've chosen to represent the sound that results from earlier *kt and *tj as *ť, but others may just write *kt and *tj (even though these two sounds are indistinguishable when reconstructed), while one book I found writes *tj for both. —CodeCat 03:14, 30 December 2012 (UTC)
- Hmm. I think that Proto-Polynesian is still way easier than that, but then again in PPN we run into the problem of inadequate and/or misleading wordlists, because every daughter lect is an LDL. I noticed that OCS and Proto-Slavic seemed similar, but in a situation like *-or-, is OCS alone sufficient to reconstruct it? How much else do you even need? —Μετάknowledgediscuss/deeds 04:00, 30 December 2012 (UTC)
- A proto-language by definition is the common ancestor of all of its descendants. So if the descendants differ, you need as many of them as you can get to make a reconstruction. In the case of -or- that certainly applies; with only South Slavic attestations it's not possible to reconstruct it, because you can't know whether the word had -ra- originally or -ra- from earlier -or-. Sometimes certain languages make reconstruction very easy because they preserve distinct outcomes of sounds that have merged elsewhere. In Russian for example the PS sound *ť merged into *č so with Russian alone you can't tell the difference. However, if you have Serbo-Croatian and/or Macedonian you can tell because they have ć/ḱ for *ť and č for *č. This is also possible with OCS/Bulgarian but you have to be careful because the outcome of that sound is št in that language, which is also a possible outcome of *šč. Nevertheless, once you have a few attestations in different languages and compare them, you start to see patterns which give away the original phoneme. This is also possible with vowels but somewhat trickier, in particular with nasal vowels. Polish and Sorbian are the only modern languages that preserve them (early Russian and OCS preserve them too), so if you see a nasal vowel in either of those there's almost certainly one in the Proto-Slavic word. But if those languages are not available you can look at the patterns in different languages. If you see u in East Slavic and Serbo-Croatian, but ǎ in Bulgarian, then *ǫ is the only possible reconstruction. Similarly, ja in East Slavic but e in South Slavic tends to go back to *ę. —CodeCat 13:24, 30 December 2012 (UTC)
- It sounds like it's pretty regular... are you able to, say, predict a Ukrainian reflex by seeing a Bulgarian reflex? Why would the languages react like that to a nasal vowel? Seems rather odd, because the languages I've seen getting rid of nasals just denasalize them and don't mess with the vowel itself beyond regular shifts. Also, how exactly did you get interested in Proto-Slavic in the first place? —Μετάknowledgediscuss/deeds 17:25, 30 December 2012 (UTC)
- Nasality can often lead to changes in vowel quality, though. In North and West Germanic, word-final nasal *ǭ was lowered to *a while non-nasal *ō was raised to *u, and in Old English nasal *ã was rounded to *õ. In French, nasal ẽ was lowered to ã and nasal ĩ lowered to ẽ. It's not possible in general to predict a Ukrainian reflex from Bulgarian, because of phoneme mergers like metathesis. A Bulgarian word with -ra- in it could have -ra- too in Ukrainian, or it could have -oro- instead. But it's possible to make a small list of possible Ukrainian cognate forms, and then use other Slavic languages to eliminate the impossible ones. For example a Polish word with -ro- where Bulgarian has -ra- generally implies Ukrainian will have -oro- and not -ra-. And if you have a Russian word with -oro- Ukrainian always will have it too. —CodeCat 18:22, 30 December 2012 (UTC)
- There is more information about the dialectal differences in Slavic at w:Proto-Slavic. —CodeCat 18:23, 30 December 2012 (UTC)
- It sounds like it's pretty regular... are you able to, say, predict a Ukrainian reflex by seeing a Bulgarian reflex? Why would the languages react like that to a nasal vowel? Seems rather odd, because the languages I've seen getting rid of nasals just denasalize them and don't mess with the vowel itself beyond regular shifts. Also, how exactly did you get interested in Proto-Slavic in the first place? —Μετάknowledgediscuss/deeds 17:25, 30 December 2012 (UTC)
- A proto-language by definition is the common ancestor of all of its descendants. So if the descendants differ, you need as many of them as you can get to make a reconstruction. In the case of -or- that certainly applies; with only South Slavic attestations it's not possible to reconstruct it, because you can't know whether the word had -ra- originally or -ra- from earlier -or-. Sometimes certain languages make reconstruction very easy because they preserve distinct outcomes of sounds that have merged elsewhere. In Russian for example the PS sound *ť merged into *č so with Russian alone you can't tell the difference. However, if you have Serbo-Croatian and/or Macedonian you can tell because they have ć/ḱ for *ť and č for *č. This is also possible with OCS/Bulgarian but you have to be careful because the outcome of that sound is št in that language, which is also a possible outcome of *šč. Nevertheless, once you have a few attestations in different languages and compare them, you start to see patterns which give away the original phoneme. This is also possible with vowels but somewhat trickier, in particular with nasal vowels. Polish and Sorbian are the only modern languages that preserve them (early Russian and OCS preserve them too), so if you see a nasal vowel in either of those there's almost certainly one in the Proto-Slavic word. But if those languages are not available you can look at the patterns in different languages. If you see u in East Slavic and Serbo-Croatian, but ǎ in Bulgarian, then *ǫ is the only possible reconstruction. Similarly, ja in East Slavic but e in South Slavic tends to go back to *ę. —CodeCat 13:24, 30 December 2012 (UTC)
- Hmm. I think that Proto-Polynesian is still way easier than that, but then again in PPN we run into the problem of inadequate and/or misleading wordlists, because every daughter lect is an LDL. I noticed that OCS and Proto-Slavic seemed similar, but in a situation like *-or-, is OCS alone sufficient to reconstruct it? How much else do you even need? —Μετάknowledgediscuss/deeds 04:00, 30 December 2012 (UTC)
- From my own experience they are easier than Germanic reconstructions because there is much less time between Proto-Slavic and the attestation of Old Church Slavonic. In a sense, OCS still is Proto-Slavic, albeit with significant Bulgarian dialectal features. It is very easy to reverse the OCS changes and find the original form, if it differs at all from the OCS form. And when there are differences usually there are clear patterns that are easy to recognise. Such as East Slavic -oro-, South Slavic/Czech -ra-, Polish -ro- which implies Proto-Slavic -or-. The differences in notation are usually just a matter of denoting the sounds since most linguists seem to agree on which sounds there were, just not on the exact details or progression of sound changes. For example I've chosen to represent the sound that results from earlier *kt and *tj as *ť, but others may just write *kt and *tj (even though these two sounds are indistinguishable when reconstructed), while one book I found writes *tj for both. —CodeCat 03:14, 30 December 2012 (UTC)
- In my experience, nobody agrees on how to notate Slavic reconstructions. Whether that's just orthography or real reconstructional differences is beyond me, because I've never studied it. —Μετάknowledgediscuss/deeds 02:50, 30 December 2012 (UTC)
- Why would it need some cleaning up? —CodeCat 02:40, 30 December 2012 (UTC)
- Well, as long as it gets you to make more Proto-Slavic entries, I guess I'm fine with it. (I bet Category:Yiddish terms derived from Proto-Slavic and its contents need some cleaning up, BTW.) —Μετάknowledgediscuss/deeds 02:35, 30 December 2012 (UTC)