Wiktionary talk:About Proto-Turkic

Definition from Wiktionary, the free dictionary
Jump to navigation Jump to search

@Borovi4ok, Anylai, 몽골어 물리 (watch this page so that we can communicate Turkologic matters) Crom daba (talk) 10:23, 21 February 2018 (UTC)

Classification[edit]

For the Proto-Turkic entries, we, the Turkic contributors' community, need to agree on a consistent classification and terminology. In particular, I suggest that Tatar and Bashkir (and Siberian Tatar idioms) be grouped under Northern Kypchak. Other options (Ural, Volga-Ural Kypchak etc.) are based on narrower geographic concepts and have inherent inaccuracies. Borovi4ok (talk) 08:52, 21 February 2018 (UTC)

(the above is copied from my talk page, there's probably template to use for these cases but I couldn't find it.)
I'm alright with Northern Kipchak, I originally put in Ural since I couldn't think of a better name.
This is the sort of information we should put on this page.
Another thing that I'm not sure what to do about is Kirghiz-Altai. I've put them under Kipchak as according to Tenišev E. R., editor (2002) Sravnitelʹno-istoričeskaja grammatika tjurkskix jazykov: Regionalʹnyje rekonstrukcii [Comparative Historical Grammar of Turkic Languages: Regional Reconstructions] (in Russian), volume 5, Moscow: Nauka, but I'm hearing conflicting information on this, and even in the very same book at that: On page 733 we have this tree (based on glottochronology) which contradicts earlier text in placing Kirghiz-Altai under Siberian Turkic.
Also note that Kumandin and Chelkan (Kuu) are placed in the same group, and that Chulym is placed within the Yenieseian group. This is consistent with the text, although it contradicts Johanson's (1998) scheme that seems popular on Wikipedia.
And I'm not sure if we should keep Medieval either. It seems very odd to me to classify Old Turkic as Karluk or Siberian since I'd think it predates most of the innovations that define these groups, but Medieval also has faults, after all Old Anatolian Turkish and Cuman are also medieval but we don't doubt that they're Oghuz and Kipchak.
Crom daba (talk) 10:17, 21 February 2018 (UTC)
Crom daba, thanks for starting this.
Before we start with the specifics, I think the following needs to be pointed out.
From what I've seen in Turkology, and for whatever reasons, there is still no definitive and universally accepted classification of Turkic languages. Existing classifications keep being challenged and revisited; this has happened up to date, e.g. in Russian Turkology, and will like continue so in the near future.
In this situation, and in view of our narrow practical purposes here in this project, I think we need to agree on some type of practical classification that would be simple, appropriate, sound, consistent, straightforward, honor the knowledge existing to date, and would not be misleading or give room for possible misconceptions or misinterpretations. Borovi4ok (talk) 10:49, 21 February 2018 (UTC)
Good points, under such conditions I suggest avoiding larger groupings which might be disproven later such as South Siberian, and maybe putting Chulym in no subgrouping smaller than Siberian considering its (apparently) controversial position (in truth it seems to contain both Altai-Kipchak and Yeniseian dialects). Perhaps also removing Oghuz subgroupings (Western, Eastern Southern).
I'd also keep Kirghiz-South Altay-North Altay as a subgrouping within Kipchak as they all seem to at least have the change of *-d- into -y- and the raising of *y- (in addition to some Siberian features). This would be in accordance with SIGTYA (disregarding the glottochronology chapter) and Poppe, but apparently at odds with most other classifications (check out Tekin, A new Classification of the Turkic languages) most of which however don't recognize Siberian as a grouping at all. Crom daba (talk) 14:24, 21 February 2018 (UTC)

Slight reclassification of Azerbaijani[edit]

I suggest adding Old Anatolian Turkish as an ancestor of Azerbaijani. Allahverdi Verdizade (talk) 00:53, 22 February 2018 (UTC)
Requests like this are for the Wiktionary:Beer_parlour. As to whether this is a good idea I have no idea, I generally found little literature about OAT in English and Russian. Crom daba (talk) 14:21, 22 February 2018 (UTC)
I thought this was the new headquarters of the Turkic social club =( Allahverdi Verdizade (talk) 14:57, 22 February 2018 (UTC)
It is, maybe Anylai will have an informed opinion on the relationship between OAT and Azerbaijani when he logs in. Crom daba (talk) 15:26, 22 February 2018 (UTC)
If you ask me, Old Anatolian Turkish (13th to 15th century) is direct ancestor of Turkish. Because I think the difference between what is called Ottoman Turkish and Turkish is just the script and perhaps the abundant number of loans in Ottoman, deliberately selected to make poetry very deep and unintelligible. Now this is lacking in "Turkish" because this need seems to be vanished. But my knowledge on OAT is also very limited. One might argue Azerbaijani is also descendant of this language, and due to its proximity it is hard to say anything. But there might indeed be dialects of this language that is a lot like ancestor of Azerbaijani. Such as loss of /y-/ and /-q-/ --> /-x-/ which is absent in modern Turkish. Most of the data is on nishanyan's website, tracing the Turkish words to their first attestation. --Anylai (talk) 19:06, 22 March 2018 (UTC)
Yes, you are right. Although OAT was not a unitary language in the sense of other languages, but rather a dialect continuum, two dialect zones are usually discerned: Western and Eastern. Eastern is the one that would be a good candidate for ancestor of Azerbaijani. Guzev has written a book on that. It is true that he is only concerned with OAT as an ancestor language of Ottoman Turkish (he calls OAT "Old Ottoman"), it really can be understood as if Azerbaijani is descendant of OAT as well, he makes several remarks that allow us to make such a conclusion. It is somewhat confusing how this could be the case given the present day geography of Azerbaijani, but keep in mind that its area of distribution lied more to West than it is today prior to Battle of Çaldıran.
Why it would be practical to set OAT as an ancestor for Azerbaijani is, among other reasons, because the first "officially" Azerbaijani texts appear in early XVIth century, in poetry of Fuzûlî. However, the proportion of non-native words in his works is tremendously higher than that of any works in OAT-corpus. Drawing from the latter could help to give attestations of many Azerbaijani words not as "cognates to", but as as descendants of those words in OAT. Allahverdi Verdizade (talk) 00:19, 23 March 2018 (UTC)

The Great Proto-Turkic Orthography Reform of 2018[edit]

Anylai and I are considering moving some aspects of our PT notation away from the StarLing model and towards the more traditional. For example using *y instead of *j and *ï instead of *ɨ. It would be good to discuss the details of this and finally make an official recommendation on this project page, preferably with a table summarizing differences between the most prominent sources and our solution. Crom daba (talk) 10:32, 21 February 2018 (UTC)

I welcome this discussion too; in fact this discussion is long due.
I support these two specific propositions. Let's discuss all the details of PT notation here. Borovi4ok (talk) 10:56, 21 February 2018 (UTC)
I don't think I'd change anything else. Perhaps using *j or *ǰ for Clauson's c and *ž for Clauson's j in Common Turkic words that have them.
I'm ambivalent about * but we might as well keep it.
*ń and *đ could maybe be used for Pre-Proto-Turkic mentioned in Etymology sections in those cases where PT *y corresponds to Mongolic *n or *d (see my subpage on this). Crom daba (talk) 15:21, 21 February 2018 (UTC)
There's also ă, I am unsure for what it is reconstructed.
It would be very nice to be able to talk about Pre-Proto Turkic for the mentioned word initial consonants. However there is one more correspondance which baffles me. Turkic word initial "y" sometimes corresponds to Mongolic "s" from the proposed cognates I saw in EDAL, and Starostin has /z-/ in such cases. Could Turkic "y" can also go to an earlier ǰ or ž? I believe I have read a paper discussing some Turkic words of Sinitic origin and proposing "ž-" for such cases.
I also propose alternative reconstructions with "h-" when Khaladj evidence suggests so. --Anylai (talk) 18:49, 22 March 2018 (UTC)
I'm working on a source correspondance table at User:Crom_daba/Proto-Turkic_orthography, I did Clauson, EDAL and Räsänen for now, feel free to edit it everyone. What system does Nishanyan use?
I also support reconstructing *h (as you can see from the page)
Most of these **z- etymons in EDAL seem much less convincing than **đ- etymons, if they are real their branching point must be much deeper back in time (Proto-Altaic?). If you could find that Sinicisms paper, I'd be very grateful. Crom daba (talk) 23:14, 22 March 2018 (UTC)
Oppose. The notation that is closer to the IPA is generally better and less ambiguous. Allahverdi Verdizade (talk) 00:43, 22 February 2018 (UTC)
@Allahverdi Verdizade I understand, here are my reasons for supporting this change:
  • y is less ambiguous than j, someone could understand j- to mean an affricate, especially since there are so many affricate reflexes.
  • ɨ is a rare symbol, most people who aren't into phonetics won't be familiar with it, and I'm not even sure what warrants using it instead of ɯ, I don't remember coming across it elsewhere in Turkological literature (except as an allophone of /i/ in Uzbek in that one grammar).
  • I'd like to establish our reconstructions stylistically as distinct from EDAL's, copying their stuff verbatim was simpler at a time, but I think we can now do better than that.
Crom daba (talk) 23:52, 22 March 2018 (UTC)
  • y can be confused with all sorts of things. Haha, where do I even start? Depending on what tradition you come from you could mistake it for all kinds of semivowels, vowels, and even our phoneme of choice, ɨ/ɯ.
  • ɨ is not that rare. It is used in Slavonic and Semitic linguistics as well (these are only those I can think of). True, people might not now what it means at first. But, first, the same goes for the proposed replacement, and second, once they do know, they will not confuse it with anything else, not with any long version of i or anything of that sort at least. I mean, does anyone in Turkology even use ï? Like, anyone? If we are to make up things, we'd better have some real good reasons for that.
  • Why exactly is it important to have notation "stylistically as distinct from EDAL's"? We do copy their work, after all. Allahverdi Verdizade (talk) 18:50, 1 April 2018 (UTC)
First, I apologize for participating in this orthography change before thoroughly discussing this with you, the discussion was going slow so I just went along with it.
  • Fair enough, although in positions in which it occurs it is unlikely to be mistaken for a vowel, while j being mistaken for an affricate is more probable.
  • At a quick glance I can find examples of Poppe, Doerfer, Erdal, Tekin, Shcherbak, Rachewiltz, Róna-Tas, Roos and Nugteren using ï, although not all of these are in Proto-Turkic as such (many are in transcription of attested Turkic languages, but generalizing them to Proto-Turkic is unproblematic). Other alternatives are y, ı and ы, which have their own problems.
  • We should strive to copy it less. We should compare sources available to us (including EDAL) to arrive at an independant conclusion, EDAL is prone to adjusting their results to better fit their ideas of Altaic reconstruction.
Crom daba (talk) 11:43, 2 April 2018 (UTC)