Talk:թշուառ
Latest comment: 10 years ago by Ivan Štambuk
I was just about to ask whether this is related to Iranian duš- found in թշնամի (tʻšnami)! --Ivan Štambuk (talk) 16:40, 27 July 2013 (UTC)
- Iranian duš-, duž- has left many traces in Armenian. թշուառ (tʻšuaṙ), թշնամի (tʻšnami), թշնամ- (tʻšnam-), դժգոյն (džgoyn), դժխեմ (džxem), դժկամ (džkam), դժկերպ (džkerp), դժնեայ (džneay), դժոխ (džox), դժուար (džuar), դժպարտ (džpart) all contain it. --Vahag (talk) 17:49, 27 July 2013 (UTC)
- I was thinking about the idea of having all these reflexes (both inherited and borrowed) organized in a lexicon in the appendixspace, grouped by the highest common ancestor (PIE, Proto-Semitic, or e.g. Proto-Indo-Iranian or Proto-Slavic if the etymon can be etymologized only up to the PII or PS period), and in a treelike format. So basically providing something similar to the appendices of AHD by Watkins, and the UoT website (though the latter is grouped by semantic domains, and multilingual). Those lexicons would be language specific, i.e. only listing reflexes in the language of choice (e.g. English, Armenian, Serbo-Croatian), via whatever route the word has arrived.
- It would both solve problems of having "deep" etymologies linking to one another (e.g. English kshatriya and satrap, which I added the other day, or e.g. Slavic dъždь "rain" and Balkanism dušman both connected via PIE *dus- that we have here), and having a very large list of cognates via various routes listed in etymologies, like the one you provided above. These semantic associations by roots could enhance learning, plus we could do it 10x better and more extensive than AHD or anyone else. And have such roots linked in etymologies once they achieve sufficiently extensive coverage. --Ivan Štambuk (talk) 18:13, 27 July 2013 (UTC)
- Module:etymtree. Right now it only does descendants but it could easily be made to generate the standard (x from x from x) etymology sections. DTLHS (talk) 18:18, 27 July 2013 (UTC)
- I was thinking a more language-specific approach, by making trees that only pertain to a specific target language. E.g., in case of Template:etymtree/ine-pro/*wódr̥, those would be (for English as a target language):
- So basically we could have it on a single giant page (every inherited and borrowed word to every language), and have module generating on PIE appendix pages only inherited trees and leaves (final nodes), as well as English, and for other languages in such lexicon entries as I described only those subtrees that their target leaves appear in.
- However, given that these pages could possibly hold thousand upon thousand of entries, for hundreds of descendant languages, as well as for hundreds of non-IE languages that IE languages have acted as donors, perhaps it's better for start to create such entries manually so that we can escape performance problems (think of [[water]]) as well as complex layout that could easily get broken due to bad formatting of a single item. --Ivan Štambuk (talk) 21:09, 27 July 2013 (UTC)
- I've had the idea for a long time myself. Whichever route we decide to do it, I hope we can avoid duplication. --Vahag (talk) 20:27, 27 July 2013 (UTC)
- Duplication would only be for inherited words, that would appear both in proto-pages as well as such entries. However, I suspect that borrowings would actually constitute a much larger chunk, and that it's worth it for them alone. You guy really shouldn't be such duplication-phobes - think about etymtree module as we have it now - IMHO it solves the non-problem of having a constant list of 10-20 items listed on two different places, by having them listed on a single place. We have shitload of dup^H^H^Hmultiplication already on various descendants list of intermediate languages, list of cognates and similar. --Ivan Štambuk (talk) 21:09, 27 July 2013 (UTC)