Talk:अप्

From Wiktionary, the free dictionary
Jump to navigation Jump to search

@AryamanA Why Hapš?? -- माधवपंडित (talk) 01:47, 22 December 2017 (UTC)[reply]

@माधवपंडित: I don't know, User:Victar moved it. @Victar How does *Hā́pš work exactly... And surely Proto-Iranian *Hā́fš shouldn't have *f, most of the descendants have -b or -p endings? I'm not trying to sound confrontational or anything, I'm genuinely curious about the p-stem and the -š. —AryamanA (मुझसे बात करेंयोगदान) 05:25, 22 December 2017 (UTC)[reply]
@AryamanA, माधवपंडित: Sanskrit (and by extension PIA, I assume) only retained the plural of *Hā́pš, *Hā́pas. To find the nom.sg. you have to look at Avestan 𐬁𐬟𐬱 (āfš). Many of the descendents in PIr also stem from the plural, and some even were transfered to ī-declensions, i.e. OP, thus the p. --Victar (talk) 05:54, 22 December 2017 (UTC)[reply]
Oh and *š is a regular outcome after *p and *bʰ in PII. I just learned that as well, really. --Victar (talk) 06:04, 22 December 2017 (UTC)[reply]
@Victar: So I was just looking at some Sanskrit words and none of them show the -pṣ- one would expect from a PII -pš-. In a few cases where a PII -ps- has occurred, it's always been -ps- in Sanskrit. So maybe the shift from ps --> pš --> fš was just in Iranian. There's no evidence for it in Indo-Aryan. -- माधवपंडित (talk) 01:42, 27 December 2017 (UTC)[reply]
@माधवपंडित: I'm going off Skjærvø (2007), page 864. --Victar (talk) 05:59, 27 December 2017 (UTC)[reply]
@Victar: I see. Wish he explained more. Still, nice paper. -- माधवपंडित (talk) 06:18, 27 December 2017 (UTC)[reply]
@माधवपंडित: It's a major go-to of mine. That said, he could be wrong. It may also only be word-final. --Victar (talk) 06:36, 27 December 2017 (UTC)[reply]

Cognates[edit]

@Bhagadatta: Re diff, I'd say that's a remnant of the past, when we had few reconstruction pages and it was convenient to lump cognates in the etymology section. But the increasingly common practice (and I'd like that it become a policy) is to remove them when a reconstruction page exists. I do this, Victar does this, @Fay Freak, @Rua too, and probably others. Canonicalization (talk) 09:46, 24 December 2019 (UTC)[reply]

Remnant? I think entries with cognate words in the etymology vastly outnumber those with just the proto-form ancestor. If it becomes a well defined policy, I have no issues complying with it. The thing I am saying is, almost all other entries conventionally list cognates so why single this entry out? I also feel that listing cognates will help users understand the etymology better, it is better for comparative analysis and it is an integral part of etymology. But I am willing to follow the policy - I'm only saying we must make it official, ask people and then uniformly apply it to all concerned areas. -- Bhagadatta (talk) 10:08, 24 December 2019 (UTC)[reply]
  1. When no direct parent entry exists, cognates in the etymology are completely appropriate and expected.
  2. Formatting guidelines that apply to English words do not necessarily apply to foreign words, so rules of uniformly with such entries are inapplicable.
  3. We're an online dictionary, not a print one. If people want more information, they can click through to it.
  4. There is no policy that states that there should be cognates, so there is no policy to counter. This is simply the current way in which we fashion new foreign entries.
--{{victar|talk}} 13:38, 24 December 2019 (UTC)[reply]
Yes, remnant. It is a form of atavism. Writing a print etymology dictionary of an individual language it was natural to present cognates, as no reconstruction entry is presented. And people are used to it and imitate it here, but a dictionary of every language on the internet has grounds against what one is used to from elsewhere, and may and must do things differently. It is not a policy but, more: it’s the format itself the implications of which people have insufficiently understood, stuck to the codified twentieth century. @Bhagadatta, we are doing our best to help users understand the etymology better, and that is particularly by ensuring that there aren’t less perfect copies of analyses going around on descendant pages of which the editors lose the plot; pages being out of sync or needing synchronization is a rather great problem. That the information is an “integral part” misses the point of why we do that and of course does not convince even yourself that it is too cumbrous to do a click forward if you think what’s behind the words. Fay Freak (talk) 14:47, 24 December 2019 (UTC)[reply]
Victar, with all due respect, if a policy does not mandate it, then a particular practice is just the editing style of some editors. There are still hundreds of non-English entries with well functioning parent entries that still list cognates. Asking for consistency is not unreasonable. If you ask more people and get their opinion, I will follow the people's will. -- Bhagadatta (talk) 14:51, 24 December 2019 (UTC)[reply]
@Fay Freak: Consider me convinced. How then do we account for the fact that just THIS entry in just THIS language is being subject to this not-a-policy-but-a-sentiment? Erase cognates from all the entries which have parent entries. Even the entries listed on this page as cognates follow the same pattern. -- Bhagadatta (talk) 15:02, 24 December 2019 (UTC)[reply]
Seriously, it is not the rationale I am contesting anymore. Just apply these changes to all entries and I will follow it. Plus, while I agree with Victar that English is handled differently than non-English languages here, I don't see what makes listing cognates ok for English but not for foreign words. -- Bhagadatta (talk) 15:06, 24 December 2019 (UTC)[reply]
That’s great to hear. All are subject to the same considerations, it’s just not yet done as it should be as we remove cognates and add reconstructions manually (and unpaid). Sometimes etymologies have possible cognates and the like where no reconstruction is possible, so of course even under perfect conditions we still have cognates here and there – sometimes an entry looks like Old Armenian կաղամախի (kałamaxi) –, but they do not need to be shoved into the view all the time when it’s boilerplate; this entry is one of these easy cases. Fay Freak (talk) 15:14, 24 December 2019 (UTC)[reply]
Cool. While I anyway would not engage in an edit war even if Victar removes the cognates again now, I do want to know why he thinks English entries like father (and the entries of its cognates in other languages) are fine with long lists of cognates cluttering the etymology section. -- Bhagadatta (talk) 15:41, 24 December 2019 (UTC)[reply]
@Bhagadatta: It isn’t fine 😄, I don’t think he thinks so, and I have removed the cognate lists of many descendants of the mother-word specifically some months ago, but perhaps we have not even looked into English. Note also that this removal should somewhat mitigate our memory/pageloading-time problems. Fay Freak (talk) 15:51, 24 December 2019 (UTC)[reply]
@Fay Freak: Hmm he said formatting guidelines that apply to English entries do not necessarily apply to foreign language entries. While it is true as a generalized statement, I am curious how it works here, in this particular issue. -- Bhagadatta (talk) 15:58, 24 December 2019 (UTC)[reply]
@Bhagadatta: No, it's not just this language. I'd rather cognates in such cases be removed from English entries as well, but my point is, as a general principle, the guidelines applied to English entries don't always necessarily apply to foreign ones. --{{victar|talk}} 03:25, 25 December 2019 (UTC)[reply]