User talk:Malku H₂n̥rés

Feel free to talk with me !

Why some user-page templates don't work

Hi. You asked why some of your user-page templates don't work. Please see WT:USERPAGE#User_pages. Equinox ◑ 13:09, 19 June 2020 (UTC)

You also asked for something to show your contributions on your user page. I think the best you can do is to add this link: Special:Contributions/Malku_H₂n̥rés. Equinox ◑ 15:56, 20 June 2020 (UTC)

Nostratic

Just so you know, mentions of Nostratic theories are deleted on sight here on en.Wikt. --{{victar|talk}} 09:59, 20 June 2020 (UTC)

Ok sorry I didn't know. But I saw you removed all the notes I put about PAA *ʔariĉ̣- instead of only the last sentence about Nostratic, but isn't it interesting to mention this resemblance with PIE and PT reconstructions and to balance with arguments for and against a relation ? Malku H₂n̥rés (talk) 13:59, 20 June 2020 (UTC)

Furthermore, on earth there's written that it is probably not related with PSem *'ars-. I wonder three things: * why isn't it removed, or just changed to "probably related";

• why is it opposed to what is written on the reconstruction page of *'ars-, i.e. that "It has been hypothesized" that it has IE cognates;
• if PIE and PSem reconstructions are linked, why isn't there any mention of it on PIE *h1er- ?

Eventually, PT *yer and PIE *h1er- have this regular sound correspondence, like with PT *yed and PIE *h1ed (both mean "to eat"). And compare the /-d/ in Arabic 'ard, German Erde and Turkish yerde (for some declension cases), whereas there's no /-d/ in the three reconstructed nouns/roots. I understand you reject Nostratic (I do it too), but there are too much evidence for this word that borrowing must be considered, at least accompanied with a "probably" as I see sometimes in reconstructions entries. Malku H₂n̥rés (talk) 14:26, 20 June 2020 (UTC)

One syllable false cognates are a dime a dozen and these words even fall short of that. They simply don't belong on the entries because they have absolutely nothing to do with one another. --{{victar|talk}} 15:52, 20 June 2020 (UTC)

Reconstructions with empty descendant templates

Creating reconstruction entries consisting mostly or exlusively of long lists of empty descendant templates is a bad idea, please refrain from doing that. Allahverdi Verdizade (talk) 18:10, 20 June 2020 (UTC)

I apologize for have done this. I was completing/creating pages for Proto-Turkic numerals but I suddenly had to stop. Now I will continue by adding reflexes of altï. Malku H₂n̥rés (talk) 18:28, 20 June 2020 (UTC)

I would recommend that you first create the page as a user sandbox page, for example User:Malku H₂n̥rés/Reconstruction:Proto-Turkic/altï. That way, you can work on it at your leisure and when it's ready, you can delete your page and create the public one with that content. It also makes the edit histories less messy. --{{victar|talk}} 18:53, 20 June 2020 (UTC)

Ok Victar thank you for the advice (it's way easier like that!). As a newcomer, there are basic things I ignore. Am I allowed to move my page back or only admin can to this.

Apparently it's hard to move page. Isn't it better to simply copy-paste the content of my sandbox page ? I prefer waiting you for doing it (I don't like to spend time on something and see it was reverted in some seconds because it was wrong). Malku H₂n̥rés (talk) 19:38, 21 June 2020 (UTC)

When it's ready, you can simply copy-paste the content to Reconstruction:Proto-Turkic/altï and replace the content of User:Malku H₂n̥rés/Reconstruction:Proto-Turkic/altï with {{delete}}. I noticed however you still have some blank entries and others in an incorrect script. --{{victar|talk}} 19:20, 22 June 2020 (UTC)

@Victar I've finished the page, so as you told me, I recreated the page and copied-pasted the content which I reorganized to put descendants at the good place and I replaced all my "draft" by {{delete}}.

Now I'd like to make the same thing for the PT numbers without entry (I.e. 7,9,10,20,...,100 and 1 000). How can I create a "draft" page to work on it without bothering the true entry ? Malku H₂n̥rés (talk) 12:33, 23 June 2020 (UTC)

Finally I found how to create user subpages so there's no problem anymore about that, but you just reverted my descendants reorganization on *yer, but the problem is that I copied a Turkic tree 4 days ago at my starts, I blanked it and full it because I didn't know Turkic. After studying Proto-Turkic descendance I decided to reorganize *yer in so as to correct my mistake. Malku H₂n̥rés (talk) 14:57, 23 June 2020 (UTC)

Glad you figured it out. The order in which you put the entries is non-standard, which is why we reorganized them after you made the entry. I created this while back: User:Victar/Turkic. It's not fully agreed upon, but it's somewhere in the ballpark. --{{victar|talk}} 15:27, 23 June 2020 (UTC)

The Proto-Turkic category [[1]] provides a standard tree (at least on Wiktionary), which disagrees partly with your tree and a lot with *yer's tree, for instance Karakhanid is shown like if it descended directly from PT on *yer. Malku H₂n̥rés (talk) 15:58, 23 June 2020 (UTC)

The wikitiony tree is not a good representation of the current academic understanding. Also many of your edits were incongruent with even that tree. {{victar|talk}} 03:51, 24 June 2020 (UTC)

Reconstruction:Proto-Afroasiatic/ʔariĉ̣-

I have deleted this page. The StarLing database and related Russian sources are not reliable for Proto-Afroasiatic, so the only way we can have PAA entries here is if the evidence for them is built up from actual descendants or reconstructions made by experts in the groups being reconstructed. In general, we try to follow the highest standard of scholarship here, rather than maximise long-range comparison. —Μετάknowledge^{discuss/deeds} 19:17, 20 June 2020 (UTC)

Proto-Indo-Aryan entries

Please don't create Proto-Indo-Aryan entries. They're unnecessary unless they have Hurrian (Mittani) borrowings. --{{victar|talk}} 15:49, 30 June 2020 (UTC)

Oops sorry. I wasn't aware. I saw that there was a redlink and that everything was already put so I had decided to create and fill the entry. But both questions I wonder are why was there a redlink if it mustn't be created, and why is the Sanskrit tree duplicated (I think it's because desctree takes only one line but I'm not sure). Last but not least, could you tell me if there are other languages I shouldn't create redlinks ? I'd like to fill some sourced proto-languages redlinks. Malku H₂n̥rés (talk) 19:37, 30 June 2020 (UTC)

In principle, if there is a redlink, then that means an entry is desired there. If no entry is desired, there should be no link. But there is no particular rule against creating entries for Proto-Indo-Aryan, as it's a language like any other on Wiktionary. So feel free to create entries for it. —Rua (mew) 20:13, 30 June 2020 (UTC)

As someone that actually works in PII and PIA, we simply don't do that, much for the same reason we don't create Italic or Hellenic entries unless they have multiple descendants. If this is a hill to die on for you, then start a discussion about it. --{{victar|talk}} 21:00, 30 June 2020 (UTC)

Ok thank you Rua. So I can create Proto-Turkic, -Semtitic, -Afroasiatic, -Bantu, etc entries (with at least one or two good references of course). I think Victar is applying like me a bytes saving principle, but I only apply it on my subpages (I prefer to expand entries). Malku H₂n̥rés (talk) 20:59, 30 June 2020 (UTC)

No Victar haha I don't need a entire discussion to catch the thing. Then I just have to remember that PIA, PHel and PIt are useless if they haven't Hurrian, Faliscan, Umbrian or Mycenaean too. Malku H₂n̥rés (talk) 07:20, 1 July 2020 (UTC)

Duplicating etymologies

They're no need to duplicate etymologies found on the parent's entry onto child ones, especially deep PIE etymologies. --{{victar|talk}} 16:13, 30 June 2020 (UTC)

Actually I told me it was useless, although I often see duplicated entries, for instance English words inherited from PIE, listing everything until PIE. I will not restart this. The only thing I would like is the explication of duplicated etymologies sometimes elsewhere. Malku H₂n̥rés (talk) 19:51, 30 June 2020 (UTC)

Uh just two things: for छुरी you removed the duplicated etymology for Hindi... but not for Sanskrit (you forgot and we revert it or is Sanskrit VIP?), and for PIE tuh2, you removed everything and put all the blame on me, but the first part wasn't of me (and by the way it was about Eurasiatic and for this case it matched, but ok I'll stop even for Swadesh list words and I admit my mistake). Malku H₂n̥rés (talk) 20:12, 30 June 2020 (UTC)

Good you see the reasons of things instead of just imitating. Unlike one might expect, English etymology entries are often the worst. which is because they are the oldest, often from a time when there were few pages for proto-languages or exotic source languages, and the are the most numerous, so cleaning up has gone very slow, and people copied together etymologies from subpar sources and without particular understanding of etymology or foreign languages just to start with some mass while the large vocabulary of English was being created. Tip: You might copy etymologies “one level upwards” to remove duplications but remark at the same time that you have not checked the validity of the claim but have copied it from a certain place (which you should probably note anyway for the sake of traceability and sometimes copyright). Concentrating dubious etymologies away from many entries into the entries where they really should be treated is probably unblamable work. Fay Freak (talk) 20:20, 30 June 2020 (UTC)

I've removed the duplicated etymology on the Sanskrit as well now. @Fay Freak is right in the reasoning -- the more elaborate the etymology on a single entry, the more chances of it becoming out of sync with other entries, resulting is erroneous and conflicting iterations. This is more so the case with reconstructions, where the research is always advancing. You're correct, *túh₂ already did have some spurious garbage on it, but that's no reason to pile on more. --{{victar|talk}} 21:13, 30 June 2020 (UTC)

Thank you Fay Freak. It makes one more thing I'll add to my list of stupid things not to do! When I see such an English entry, should I leave it like that or should I remove asap the long paragraph about etymology (will it be helpful) ?

Well... I didn't catch your tip. What I understood is that it's better that each page displays only the etymology just before but not the rest, so an Eng page links only to an MEng, linking to OEng, to PWGmc etc (concentrating etym where they should be) rather than the english direcly giving a link to PIE (entries like earth, horse, water and so on).

But I don't understand the middle about validity of the claim, traceability and copyright. What claim, what traceability if not linking ultimately to anything, and what copyright if I remove something which, moreover, is on the Wiktionary (what a person added isn't its property, only the fact that it added, it has the action, not the result) ? Malku H₂n̥rés (talk) 06:48, 1 July 2020 (UTC)

Oops I didn't saw your answer Victar (didn't refreshed the page since yerterday). Ok so, small, clean, clear and simple etymologies only, not the list of cognates for each stage. And the only way having big paragraphs about etymology is when it's controversial (e. g. aryas, *h1ekwos, Ασία, Ευροψ). You're right sometimes etymologies links to something a bit different but don't exist, so simpler etym would help them to be synchronized. Eventually for tuh2, you saw everything wasn't from me but everything was bad (even if the comparison with Hittite was interesting). Malku H₂n̥rés (talk) 07:09, 1 July 2020 (UTC)

Proto-Nuristani

Please don't create Proto-Nuristani entries. There isn't enough research done of this family to create credible entries. Also as a general rule, maybe don't create entries for languages you're completely unfamiliar with. That's a sure way to get into hot water. --{{victar|talk}} 12:47, 4 July 2020 (UTC)

Етегән#Etymology

Hi, I'm reverting your edit on subj - it is problematic for two reasons. In case you have any sources to back up - feel free to provide them. Borovi4ok (talk) 18:12, 9 July 2020 (UTC)

Hi @Borovi4ok. On Етегəн, I saw an etymology was provided (*yetigän) but the language of this etymology was undertermined. How could one know the exact etymology but not the language ? Cognates in other Turkic languages are given, and they are enough far geographically one to each other to think they have a Proto-Turkic etymology rather than borrowing through language contact. The first part is easy, just ете from Proto-Turkic *yẹt(t)i (in which I put also *yẹt(t)igën when I created it), and, because Ursa Major has seven star, I guessed гəн has a meaning related to star or sun, and I found on gün that there's justly a Proto-Turkic word matching consonants, the vowel (admitting ü>ë by vowel harmony next to *yẹt(t)i fronted last vowel), and meaning of sun. As in most compounds, giving the morphemes is enough if they are verified. For *yẹt(t)i there are a lot, but *gün unfortunately has no references, but I trust Wiktionarians. Of course as a native Bashkir you know your language far better than me. To conclude, there's no source directly for *yẹt(t)igën, but it fits etymology, which is ultimately from Proto-Turkic, and as a Bashkir speaker you can know what word compounds the second part and thus provide a secure etymology. Malku H₂n̥rés (talk) 07:48, 10 July 2020 (UTC)

Hi Malku H₂n̥rés, 1. I would readily and happily specify the reconstruction language as Proto-Turkic, but, unfortunately, there are no cognates in Chuvash, or any evidence of cognates in other Oghur languages (which are all now extinct, except Chuvash). Without them, there is no ground for a Proto-Turkic reconstruction.

2. As for your suggested ü>ë change in the last syllable, vowel harmony does not work like this in Turkic languages. ü is already front from Turkic phonology standpoint. If we hypothesize such a compound word, then I would expect that ü would have preserved.

3. Lastly, gün has the meaning of 'the Sun\day', not 'star'. Of course, the Sun is a star in our modern-day knowledge, but IMHO this knowledge is quite advanced to expect that such a lexical shift could have been possible at an Ancient Turkic stage. Rather, we should IMHO expect yẹt(t)i+yulduz for "seven stars".

I have a couple of ideas as to this reconstruction. However, I cannot support them with any sources, so I don't provide them here.Borovi4ok (talk) 18:39, 10 July 2020 (UTC)

Inviting @Crom daba, @Allahverdi Verdizade, @Anylai, @몽골어 물리, @Fay Freak into this discussion.

@Malku H₂n̥rés: Generally, some sources would be nice here. You put in a bunch of barely attested language forms but no sources. Where did you get them from? ("elegant dictionary" is not a primary source). And where did you get the derived terms from? *yẹt(t)igën doesn't belong here for sure. Allahverdi Verdizade (talk) 19:33, 10 July 2020 (UTC)

@Malku H₂n̥rés:, the Soviet-time Etymological Dictionary of Turkic Languages has one paragraph on this etymon, however it simply provides the cognates and states that it is derived from yet(t)i '7' plus (some) affix -gen. Apparently, there is no obvious explanation, nor a consensus. It quotes earlier research on it by Räsänen, Bang, Clauson - if you want to dig further, I can give you the references.Borovi4ok (talk) 19:47, 10 July 2020 (UTC)

Also, yet(t)i+gün would simply mean "seven days" and hardly anything beyond that.Borovi4ok (talk) 19:47, 10 July 2020 (UTC)

Hi again @Borovi4ok ! That's a first problem, we can securely reconstruct a Kipchak etymom but it would be original research. But sometimes having descendants in only one branch doesn't mean it didn't exist in a upper-order reconstruction. Because Turkic people were able to see Ursa Major ~2.5 kya ago, there's nothing refraining them from having a word for it. The last thing I would add for the first point is that native cognate can be replaced but they still existed before, for instance Turkic has a calque meaning Great Bear, ultimately from Calque of Latin Ursa major (actually I don't know was had Ottoman Turkish or Old Anatolian Turkish for this word).

Uh, yes you're right <ü> transcribes /y/ which is a front vowel, but it's rounded unlike ë and vowel harmony can occur for that too, let's stay on seven and Turkish has, for ordinal -inci (front unrounded), -üncü (rounded) and -uncu (back), similarly to Azərbaijani but @Allahverdi Verdizade is the expert for that. Thus I'd say *ü > *ë / Template:front unroundedC_ but that's not secure.

For the meaning, I also thought to string because a constellation is more the strings that links stars than the star themselves, and moreover it explain the meaning of an instrument with 7 strings in I-dont-remember-which-language. But I didn't find anything for that (and I surely badly searched). There was nothing else than *gün so I took it.

I am eager to see your ideas for reconstructions. As a newcomer (I still learn new templates and tricks after a month), I ignore if there are "reconstruction meeting" as you summon four users as experienced as you.

The Soviet reference isn't a book I might find or read properly. For the references I gather on a sandbox the easiest cognates on Wiktionary, I copy the reference unless they seem strange, and I use them to seek further. I tried to put as much reliable sources as possible (few Proto-Turkic entries have a lot references). I don't know of it's good or bad to provide descendants for so many languages (few entries have almost all Turkic languages, but there are plenty of redlinks). I searched in all English Wiktionary, but some had no references. Elegant Dictionary provided entries for 70 languages, I used it to fill as many {{desc}} templates as possible and added some supplementary languages without code, and the tree is from Victar (who told Wiktionary Category has a bad tree). For the reconstruction, I created the redlink with t(t) (from Starostin's reconstruction) but explained in "reconstruction" that it's rather tt (from Clauson) than t (actually it's evident but the more information there's the better it is). Lastly, for the compounds, we'll delete *yẹt(t)igën as you asked, for yetmil (70) it's from Wikipedia and a second source but I don't remember which one and by analogy with altmil (60), and for yet(t)imci it's just the ordinal, with the ordinal suffix -imci whence -mes in Chuvash and something like -inci through assimilation in other branches. I apologize if my justifications are naturally long and soporific. See you tomorrow ! Malku H₂n̥rés (talk) 21:58, 10 July 2020 (UTC)

Etymologies are not definitions

Re: diff. — Mnemosientje (t · c) 16:49, 11 July 2020 (UTC)

User:Malku H₂n̥rés/Trees to fill

Why is your user page generating so many term requests? Those requests end up generating cleanup categories where there is in fact nothing to clean up. I would suggest that you insert a hyphen to avoid this. —Μετάknowledge^{discuss/deeds} 06:12, 19 July 2020 (UTC)

@Metaknowledge the purpose of this page is that people creating an entry could simply take already made and verified trees and fill them easily. I saw unfortunately that, in languages entry maintenance, this page was here automatically. If I put a hyphen it would prevent this page from being on those categories, but I'll have to do it manually (don't have access to the replace function or other Regex-replacing things) and users would have to manually remove every hyphen. Is not there any other solution, like some template I could put at the top of the page and it would make that the categories would bypass the whole page (like the nocat parameter on the {{com}} template, but once for all the page) ? (By the way, as an admin, could you change *d to *þ in Proto-Germanic *twifaldaz entry name ?) Malku H₂n̥rés (talk) 07:36, 19 July 2020 (UTC)

To “manually remove every hyphen” is hardly an effort; there is always something to edit. You could also use code-tags. <code> </code> Fay Freak (talk) 12:16, 19 July 2020 (UTC)

Of course you have access to a replace function; literally everyone with access to the internet does. Here you go. As for the Proto-Germanic page, I suggest that you use WT:RFM instead of asking somebody who doesn't edit that language. —Μετάknowledge^{discuss/deeds} 20:03, 19 July 2020 (UTC)

@Metaknowledge yeah of course there are tons of websites that can do a replace with regular expressions, but I saw somewhere on Wiktionary that there's a replace thing, but it's not a template and its access is very limited (however that's not a problem now). I'm disappointed that Fay Freak's suggestion has such a horrible look for trees and that nocat is a parameter in some templates to decategorize but that a {{nocat}} template doesn't exist.

On Special:WhatLinksHere/User_talk:Malku_H₂n̥rés there are the entries where I left a message, but what should I do when it's not about a move/merger/split (for instance the {{reconstructed}} template doesn't links to the Proto-Tocharian language on Wikipedia.Wikipedia for Proto-Tocharian entries) ? Thank you for your answer and your sympathy. Malku H₂n̥rés (talk) 08:42, 20 July 2020 (UTC)

Why should it link to the WP page? {{reconstructed}} doesn't normally do that. —Μετάknowledge^{discuss/deeds} 17:22, 20 July 2020 (UTC)

@Metaknowledge the {{reconstructed}} template provides a little text with links on it, and the first one, just after "This", links to the Wikipedia page or "About" on Wiktionary. On an entry on PIE, P.afroasiatic, PAustronesian etc, PGmc, PSino-Tibetan, etc., the link links to something, but on a Proto-Tocharian entry, there's no link on "This Proto-Tocharian entry contains ...", for instance see Proto-Tocharian *wəkʷsó. Malku H₂n̥rés (talk) 09:23, 21 July 2020 (UTC)

Yes, but it doesn't link to Wikipedia. It's not linking because WT:About Proto-Tocharian hasn't been written yet. —Μετάknowledge^{discuss/deeds} 18:04, 21 July 2020 (UTC)

Let's create it ! I have an intermediate knowledge of Proto-Tocharian, so with the Wikipedia article and Wiktionary transcription for Proto-Tocharian, I could create it tomorrow or later. By the way, do you know someone knowing this reconstructed language who could help expanding significantly this page ? Malku H₂n̥rés (talk) 22:01, 21 July 2020 (UTC)

I advise caution. The writing of these About pages is rather unlike writing a Wikipedia page. If you have domain-specific knowledge, you are welcome to share it by adding to the Wikipedia article, but the About page is aimed at editors and concerns the standard treatment of a language on Wiktionary. You will probably want to wait until you understand our practices better. —Μετάknowledge^{discuss/deeds} 03:51, 22 July 2020 (UTC)

Ok so I won't create it right now. The purpose to create this page was, according to me, to justly set a standard, bc Wiktionary has Proto-Tocharian entries but no standard on them, so anyone can create an entry with the transcription of an author then with the one of another author. Actually there's no particular knowledge to have, the tree is evident, transcription is logic, present on Wikipedia and can be adapted following already existing entries, pronunciation is not needed bc it's reconstructed (even if PGmc and Proto-Austronesian bypass this rule I still don't know why), and the rest is standard like other entries (put {{reconstructed}}, put * bc non-attested, etc). Else I propose to create it in sandbox and to ping you when I'll have finished everything. Malku H₂n̥rés (talk) 08:24, 22 July 2020 (UTC)

Can you help editing?

I started the programs of Manchu Wikinews/Even Wikivoyage, Even Wiktionary/Even Wikibooks, Evenki Wiktionary/Evenki Wikibooks, Oroqen Wiktionary/Oroqen Wiktionary/Oroqen Wikibooks, Negidal Wiktionary/Negidal Wiktionary/Negidal Wikibooks, Udege Wiktionary/Udege Wikibooks, Oroch Wiktionary/Oroch Wiktionary/Oroch Wikibooks, Nanai Wiktionary/Nanai Wikibooks, Ulch Wiktionary/Ulch Wiktionary/Ulch Wikibooks, Orok Wiktionary/Orok Wikibooks on 14th July, 2020. If you know at least one of these Tungusic languages, can you help editing? (talk) 12:17, 24 July 2018 (UTC)

Reconstruction:Proto-West Semitic/milḥ-

Unfortunately, we haven't developed good standards for Proto-Semitic entries yet, but I doubt this entry would meet reasonable standards. The solid attestation is Western Semitic, and Kogan reconstructs it as Proto-Western Semitic only, saying that the Akkadian mil'u ("potassium nitrate") is too late and rare to be anything but a loanword (plus, the sound correspondence does not match regular inheritance). @Fay Freak, Victar, Rhemmiel —Μετάknowledge^{discuss/deeds} 21:52, 31 July 2020 (UTC)

Good afternoon @Metaknowledge. Thanks for deleting *milḥ. I saw that Phoenician, Arabic, Aramaic, Hebrew and Classical Syriac entries for salt were linking to each other and giving the same PS redlink. Because of its presence on all those entries, I've thought it was a good idea to create it. There's no About:Proto-Semitic (actually it links to its Wikipedia article) so it could be created (we have transcription, tree of descendants and declension templates, so putting them there would standardize new entries). West Semitic provides solid cognates for a Proto-West Semitic *milḥ- as I've put. Actually I ignored Akkadian had mil'u for salt rather than ~tabtu, but you've wrote with ' for a plosive whereas <ḥ> transcribes a fricative. A way we could be sure it existed in PS is finding Afroasiatic cognates or simply Eblaite cognate (which is East Semitic as Akk). Can we create a Proto-West Semitic entry ? I was looking for references, but Ehret (1995) doesn't give anything for this root and I don't have the earlier book. Starling is ,as you said one month ago, not reliable, for PS it hesitates between /a/ and /i/ and for PAA there are Chadic cognates but PS isn't cited. So either we find a way to prove it existed in true PS, or a sysop moves the page to PWestS, if it's possible... Malku H₂n̥rés (talk) 14:16, 1 August 2020 (UTC)

This reveals to me that you are not yet qualified to create Proto-Semitic entries. There are a few problems here. Firstly, just because you see a redlink in multiple entries does not mean you should create it. Many such links are not in fact valid entries, and you need to refer to standard reference works in order to determine whether an entry is justified. Ehret's reconstruction is also not reliable, although it can sometimes be a helpful place to look for potential cognates. We don't seem to have any Proto-West Semitic entries, but perhaps we should, following Kogan. —Μετάknowledge^{discuss/deeds} 18:51, 1 August 2020 (UTC)

@Metaknowledge I am sorry to disappoint you. What should be done, knowing we can't move the page to PWS and that we can't rely it for PS? Do what should be done and I will follow your example. Kogan gives p.239 the PWS *milḥ- so the entry name is reliable. According to you, Ehret is not reliable, so which references are considered reliable for PS? Isn't it enough to add a reconstruction note saying only West Semitic is attested and putting the following reference template ?

• Kogan, Leonid (2011) “Proto-Semitic Lexicon”, in Weninger, Stefan, editor, The Semitic Languages. An International Handbook (Handbücher zur Sprach- und Kommunikationswissenschaft – Handbooks of Linguistics and Communication Science; 36), Berlin: De Gruyter, →ISBN, page 239 Malku H₂n̥rés (talk) 09:46, 2 August 2020 (UTC)

I think we probably should move it to PWS, which means the creation of some relevant infrastructure; I'll get to that shortly. Ehret is not an altogether bad scholar, by the way; he's just not a Semiticist and his 1995 reconstruction of Afroasiatic contains a fair number of significant mistakes. I'm not sure any source is perfectly reliable, but Kogan's lexical work (as opposed to his classificatory work) seems pretty solid outside of Aramaic. Of course, we should always check all the descendants in language-specific dictionaries. —Μετάknowledge^{discuss/deeds} 18:29, 2 August 2020 (UTC)

Creating Proto-West Semitic would involve plenty of changes. Even if it slightly differs from PS, it would need to reconstruct PWS in order to name these new entries, to reorganize descendants trees of most PS entries to distinguish East and West Semitic (using a bot) then to move most part of descendants trees entries hitherto and using {{desctree}}, and to create a new declension template, an "About:" page, a PWS category, an update of {{inh}} template, new {{R:sem-wes-pro:}} references templates, a good Wikipedia article, etc., and several serious studies on it to determine clearly changes in every level. But, well, you mean you actually have the power of introducing new languages on Wiktionary and then to do all of this ?! Malku H₂n̥rés (talk) 14:12, 3 August 2020 (UTC)

I'll undo it if needed, but I wanted to move it for use as an example. As for your other requirements, some should be done anyway (the organisation of descendants, creating an about page), some are irrelevant to our purpose or just unneeded (a WP article, making redirected reference templates), and some are important, but you misunderstand them ({{inh}} doesn't need changing, but the language modules do). —Μετάknowledge^{discuss/deeds} 16:32, 3 August 2020 (UTC)

Reconstruction:Proto-Kartvelian/arwa-

Why would Proto-Kartvelian borrow the Proto-Semitic for "four" as its word for "eight"? Why does this entry not have any sources whatsoever? @Vahagn Petrosyan, Victar, Dixtosa —Μετάknowledge^{discuss/deeds} 23:06, 16 August 2020 (UTC)

Dunno, but it had been written so already at the Georgian page რვა (rva) by @Samarra11, so he just moved it whither it would belong, in his favour. Fay Freak (talk) 23:13, 16 August 2020 (UTC)

And that too had no source and seems strange enough to require something to back it up. It does not defend what Malku did. —Μετάknowledge^{discuss/deeds} 23:26, 16 August 2020 (UTC)

I added the sources. The Kartvelian is supposed to be from the dual. PIE *oḱtṓw too may be a dual. --Vahag (talk) 08:06, 17 August 2020 (UTC)

For the refs that's not a problem, I was trying to edit to add Fähnrich p3 & Klimov p38 when Vahagn Petrosyan added 4 refs 5 minutes ago. For the relationship between 4 and 8, see that the PK word for 4 was borrowed from the PIE word for 8 (which is itself the dual of a stem meaning 4 fingers) and conversely the PK word for 8 could likely have been borrowed from the PS word for 4 (in the dual). Other examples exists like natives numerals 4 and 8 in Japanese. Last but not least, the development b > v is regular for PK loanwords, just see the number before (*šwid-) whose /w/ corresponds to Akk /b/. Malku H₂n̥rés (talk) 08:10, 17 August 2020 (UTC)

Error: dev b > w not v Malku H₂n̥rés (talk) 08:12, 17 August 2020 (UTC)

Proto-Cushitic, Proto-Chadic, Proto-Omotic

Nobody has ever made workable, systematic reconstructions in these languages, and nobody here knows them well enough to do it either (myself included). The best solution is probably to make them etymology-only languages, but I haven't done that yet. —Μετάknowledge^{discuss/deeds} 18:18, 26 August 2020 (UTC)

Additionally “Omotic” may in whole or in parts not belong to Afro-Asiatic. Is Omotic Afro-Asiatic? It may be some Cushitic languages mixed with neighbouring non-Afro-Asiatic languages, being Nilo-Saharan or related to Hadza, as mentioned on the better German Wikipedia article on Omotic. This vague affiliation of the “Omotic” languages indicates that they cannot help with Proto-Afro-Asiatic reconstructions and that there will never be “Proto-Omotic” reconstruction. Fay Freak (talk) 21:22, 26 August 2020 (UTC)

I wouldn't take Theil too seriously; his view is a fringe one. Otherwise, I agree. —Μετάknowledge^{discuss/deeds} 21:28, 26 August 2020 (UTC)

The creation of the redlinks categories for these languages was to see if there were interesting proposed etymologies on the Wiktionary but feel free to delete them if they're useless. For, PAA reconstructions, the Omotic family actually never appears, but at least some serious scholar material might exist for the Chadic and Cushitic families. It makes of these families a still almost unstudied part of Afroasiatic field. Malku H₂n̥rés (talk) 08:18, 27 August 2020 (UTC)

Outlandish etymologies

Bruh, if you're going to make outlandish dubious etymologies, you better bring it with tons of sources. --{{victar|talk}} 02:21, 28 August 2020 (UTC)

@Victar: Please don't insult me, I don't remember to have done it twelve days ago and I pressed myself to remove the equivalent stuff on the Proto-Semitic entry. I don't do anymore the imprudence of unsourced edits since around this date, please believe it is one of the last times one will have to revert my edits for absence of references. Malku H₂n̥rés (talk) 09:10, 28 August 2020 (UTC)

Great! Then please clean up after yourself. --{{victar|talk}} 16:56, 28 August 2020 (UTC)

āzͅrəf

I deleted this. You have got to stop creating entries in languages you don't know anything about. Only resources for the language in question should ever be used. That orthography is terrible, and it's fine as a temporary romanisation, but no good for an actual entry. —Μετάknowledge^{discuss/deeds} 06:39, 3 November 2020 (UTC)

Oh boy, I see now that you created a bunch more, with Hausa descendants not following normal Hausa orthography either. I don't have time to clean your mess up right now, but you really need to avoid doing this. —Μετάknowledge^{discuss/deeds} 06:42, 3 November 2020 (UTC)

@Metaknowledge Since Tamasheq had not a single entry on English Wiktionary, I expected someone to correct them few time after: figure out that I voluntarily created less than 10 entries so that it would be short to correct. I wondered why were the Hausa borrowings as redlinks instead of entries. It seems that the website used in {{R:WOLD}} isn't reliable for Hausa. However if you have some material about Tamasheq (with its native writing system), I'd be glad to add it. Malku H₂n̥rés (talk) 17:18, 3 November 2020 (UTC)

Don't create entries with the intention of making work for other people. Either do a good job, or acknowledge that you don't have the resources and leave it for someone else. The WOLD data are actually perfectly reliable for Hausa; the problem is that you are ignorant of Hausa orthography and you don't know how to convert it. Again, the problem is with your actions. —Μετάknowledge^{discuss/deeds} 20:32, 4 November 2020 (UTC)

Unattested Tamazight words

@Fenakhay has noticed that words like ⵓⵙⵏⴰⴽ don't seem to be attested in Central Atlas Tamazight, and even as coinages intended for the artificial Moroccan standard, they aren't found in the IRCAM dictionary. Are you getting these from Wikipedia or a similar source that does not meet CFI? If so, these are protologisms that should be deleted. —Μετάknowledge^{discuss/deeds} 23:10, 28 December 2020 (UTC)

IPA modules

Hey Malku. I just wanted to let you know that if making modules like the Central Atlas Tamazight one is something you enjoy doing, then it might be a good place to focus your efforts. Personally, I haven't got the patience to make them, so I appreciate when someone else does. If you're interested, there's a long list of languages that could use it, but the highest priority on my list would be Hausa. —Μετάknowledge^{discuss/deeds} 17:05, 20 April 2021 (UTC)

Hi @Metaknowledge! Currently I'd like to focus on Central Atlas Tamazight, developing the infrastructure with templates for declension and conjugation and adding basic vocabulary which is lacking like pronouns for instance. I've found several sources and I hope to carry on other Berber languages (and Proto-Berber if serious reconstructions exist). I'm peculiarly interested in the Afroasiatic family, more than others, however I never tried to learn Hausa (or another Chadic language).

As for the module I created, the lack of true Regular Expressions (Regex) in Lua makes the task quite longer (but not that harder). I used to program in Javascript and use Regex before joining the dictionary (whence my "ability" to manage complex Lua patterns). What type of module would you like? If it's for conjugation, declension or pronunciation I would need the regular correspondences. Malku H₂n̥rés (talk) 17:25, 20 April 2021 (UTC)

There's a lot of work to do overall, and I'm sure you can do good work if you restrict yourself to good resources (like Penchoen). If you're curious about Proto-Berber or other Berber resources, I've got a good handle on what's out there, so send me a DM on Discord or an email. In terms of my requests, I would specifically love a Hausa IPA module (just phonemic). —Μετάknowledge^{discuss/deeds} 18:55, 20 April 2021 (UTC)

@Metaknowledge it's ok I can make this Hausa pronunciation module. Do you want me to see Hausa phonology or do you prefer to indicate the rules to follow? I need to know if it's rather regular from the way it's written (like IPA, 1 sign = 1 phoneme) or if there are complex rules occurring like in the Ayt Ndhir dialect of Central Atlas Tamazight. I'm very interested in the sources you can propose me because it seems there are quite few works about the Berber family (and Proto-Berber even less) on the Internet. Malku H₂n̥rés (talk) 19:16, 20 April 2021 (UTC)

Hausa phonology is very regular; you can get (almost?) everything you need from Wikipedia, I think, and I can check it (although if you want a more in-depth resource, nothing beats Newman's The Hausa Language: An Encyclopedic Reference Grammar. As for Berber stuff, send me a message. —Μετάknowledge^{discuss/deeds} 21:21, 20 April 2021 (UTC)

Module Error at ⵢⵓⵜ

This entry currently has a module error because you commented it out in Module:number list/data/tzm: "--don't add ⵢⵓⵜ now". Can you give me some idea of how and when this is going to be resolved? I know very little about Central Atlas Tamazight, so I can't do anything myself, but a module error is a useless eyesore that reflects poorly on the dictionary, and we need to keep CAT:E as clear as possible so any new errors can be easily spotted. Chuck Entz (talk) 18:04, 15 May 2021 (UTC)

Hello @Chuck Entz. ⵢⵓⵜ is the feminine form of ⵢⵓⵏ, which is a number and uses a module to generate the box. It's currently the sole entry for a feminine number in Central Atlas Tamazight. The best would be to delete for now, until I add the feminine forms in the module. The problem is that I don't know how to do so, therefore I'm just procrastinating the problem. In detail, Central Atlas Tamazight can't be managed with the classic number box parameter (cardinal, ordinal, etc.) because it has three sets of numerals: two native, one of both masculine and the other feminine, and the third is borrowed from Arabic. I remember that the template for number box can accept only one extra parameter, not six:

• native cardinal masculine
• native cardinal feminine
• native ordinal masculine
• native ordinal feminine
• Arabic-borrowed cardinal
• Arabic-borrowed ordinal

If you know how to solve this problem it'll be very helpful for me. Else just delete the page and when it will be the turn of numerals to be edited by me, I'll see how to handle this. Malku H₂n̥rés (talk) 18:27, 15 May 2021 (UTC)

The entry as it is now is rather misleading. You should just mark the feminine forms as {{feminine of}} the masculine forms, and leave those to be linked in the module. After all, IIRC only yut is irregular. —Μετάknowledge^{discuss/deeds} 18:43, 15 May 2021 (UTC)

(edit conflict) Well, we don't want to delete an entry that's attested. The simplest solution is to remove the template from ⵢⵓⵏ, which I have done.

As for masculine vs. feminine: there are lots of languages with grammatical gender. The general practice is to have one gender as the main one, and to put all the general information including the number box in that entry. The other genders have only the information specific to the gender, and link to the main gender. See the difference between Latin unus, una and unum, or Arabic واحد and Arabic واحدة for instance. That eliminates half of those six.

As for the native/borrowed dichotomy: the only language I can think of off the top of my head with a similar native/borrowed dichotomy is Japanese, and the Japanese editors seem to deal with it by only including the native numbers in the cardinal box. Of course, Japanese uses the same character for both the native and borrowed numbers and only the pronunciation changes, so it's not quite the same. For one thing, that means they're both on the same page. How you handle it also depends a lot on the patterns of usage and cultural implications of using one or the other in a given context. Chuck Entz (talk) 19:32, 15 May 2021 (UTC)

The Arabic forms should definitely be linked. Japanese isn't a great example; Module:number list/data/ko is much more germane. —Μετάknowledge^{discuss/deeds} 19:50, 15 May 2021 (UTC)

@Chuck Entz, Metaknowledge why removing the template from ⵢⵓⵏ ? This entry has no problem currently. One would think that the main numeral set is the native one, but the Arabic-borrowed one is widely used in everyday life. But this one begins at 4, because the numbers 1, 2 and 3 are native only. I agree for the gender, but the question is which set will we left unmarked ? That is, is it better "Cardinal: [...] Arabic-derived cardinal: [...]" or "Cardinal: [...] Native cardinal: [...]" ? The feminine spellings are regularly derived from the masculine ones (not in pronunciation, for assimilations reasons, but my template manage that, anyway) except for ⵢⵓⵜ, as Metaknowledge noticed, so I'll do like in other languages. But this way, there are still two sets of numerals in the number box, the native masculine and the Arabic borrowings, both having ordinal forms and more, so it would more than one extra parameter. However Module:number list/data/ko seems able to manage several non-classic parameters, so I'll see how it works. So the solution for ⵢⵓⵏ and ⵢⵓⵜ is to remove the latter from the module and to link both each other, so that the box link to ⵢⵓⵏ, which in turn links to its feminine counterpart. Malku H₂n̥rés (talk) 09:25, 16 May 2021 (UTC)

"So that I ensure both of you read it"

Except you didn't. Adding a ping to an existing edit accomplishes nothing, even if you sign it again. Feel free to ask @Metaknowledge and @Fay Freak if they got the ping in this edit. I would be quite surprised if they did. Don't feel bad- I've seen people who have been admins for years making the same mistake. Your edit comment was the main reason I posted WT:Beer parlour/2021/June#A Primer on Proper Pinging, though I've thought about doing it before. Chuck Entz (talk) 01:30, 28 June 2021 (UTC)

@Chuck Entz: Ok thank you very much for telling me this! But how can I ping them after realizing I haven't when I replied (that is, in the current situation, how can I get them pinged)? Malku H₂n̥rés (talk) 07:58, 28 June 2021 (UTC)

Category:Central Atlas Tamazight roots

I just noticed that these pages exist. I don't like the idea of creating root entries; unlike Semitic, Berber doesn't have a very productive system of roots (as you can see from how short the sections are in any Berber dictionary that organises lemmas that way). Most active derivational processes occur from the verb, and we can handle the rest in 'Related terms'. Do you have any arguments for keeping it? —Μετάknowledge^{discuss/deeds} 20:43, 25 August 2021 (UTC)

@Metaknowledge: Fenakhay told me the same thing, but the choice was between a template making a little box linking to the root (as he does for Maltese) and putting the root as an actual (synchronic) etymology just like Arabic, and I chose the second option because the system of roots is not neglectable in CAT at least, and roots behave also as verbs (when there's no verb section on a CAT page with a root section, it's just that I haven't added it). Moreover, roots are given in Penchoen (see pages 102-107 IIRC). Now I admit I would accept to lower the importance of roots to a simple box as in Maltese, but I'd oppose deleting the very root entries as is. Best I can do is accepting to merge root sections with verb sections and assume what is currently marked as terms derived from the root would be derived from the verb, but the {{rootsee}} template would fail. Malku H₂n̥rés (talk) 21:22, 25 August 2021 (UTC)

I don't think we should be using {{rootsee}} at all. We could use root boxes if you want, but I don't really see the point. From a lexicographical perspective, you will never find an Arabic dictionary that doesn't use roots — but you can easily find Berber ones. That says a lot about their utility. —Μετάknowledge^{discuss/deeds} 21:29, 25 August 2021 (UTC)

Ok for "Related terms" instead of "Derived terms", and turning roots into verbs. But PB entries are only words, not roots, so I wonder how to provide an entry for a common etymology to all words derived from an actual consonantal root. Unless you want PB verbs to indicate that other words derive from them. Malku H₂n̥rés (talk) 21:44, 25 August 2021 (UTC)

If they are derived from the verb, then yes! The whole idea that a root can replace an actual etymology isn't even true for Semitic, and the fact that you see it in Arabic entries is because people have been lazy or lacked the requisite resources. Are you able to convert the root entries to verb entries youself? —Μετάknowledge^{discuss/deeds} 22:16, 25 August 2021 (UTC)

I understand you're skeptical towards the root system since it's synchronic, like a surface analysis, which is chronologically wrong unlike inheritance from an older language. However you need to acknowledge (if you haven't already done) that there's a productive derivational system in CAT, peculiarly remarkable with Moroccan Arabic borrowings because words are derived from a common borrowed "root" which now we'll call "verb", not independently borrowed from Moroccan Arabic. Yes I can convert the root entries into verbs that'll be easy since the non-gloss definition is systematically "related to [verb]ing" so I replace with "to [verb]". Malku H₂n̥rés (talk) 07:35, 26 August 2021 (UTC)

@Metaknowledge: I forgot to ping you. Malku H₂n̥rés (talk) 16:21, 26 August 2021 (UTC)

I know, I just don't see a reason to treat them as roots and create entries for what are essentially fictions. There are other ways to handle languages that have active derivation from a verb — see {{sw-der}}, which we may want to adapt for CAT and other Berber languages. —Μετάknowledge^{discuss/deeds} 16:32, 26 August 2021 (UTC)

Roots are not words, and of course in a non-concatenative morphology they're fictions. Since any root is also a verb we can delete "root" as a valid POS for CAT. {{der}} for verbs is analogous to {{ar-verb forms}} for roots I guess, so yes such a template is required as an Ersatz to categories of root-derived terms. Malku H₂n̥rés (talk) 18:18, 26 August 2021 (UTC)

• {{sw-der}} Malku H₂n̥rés (talk) 18:18, 26 August 2021 (UTC)

@Metaknowledge: done. I've cleaned Category:Central Atlas Tamazight roots. Now there are the root pages to delete. Malku H₂n̥rés (talk) 22:07, 26 August 2021 (UTC)

The categories for pages by root need to be emptied first. —Μετάknowledge^{discuss/deeds} 23:40, 26 August 2021 (UTC)

@Metaknowledge: done, replaced {{tzm-root}} with {{m|tzm}}. Malku H₂n̥rés (talk) 06:45, 27 August 2021 (UTC)