Talk:皇

Also "emperor"?

Can this also mean "emperor"? 71.66.97.228 06:52, 19 January 2011 (UTC)

Wubi input

I am not sure what kind of material is applicable to this page, so I add the information here. The Wubi86 input code for this character is RGF.

Generally, it would be great to also directly link to the referenced dictionaries, and maybe to reference material about the input methods, too - but maybe in a separate part about languages (eg. one page per language which gives a good overview about relevant dictionaries etc.). 79.228.183.87 21:00, 8 December 2012 (UTC)

Kun'yomi

Hello @Poketalker, non-on readings listed in JA dictionaries, that apply in toto to a single kanji, are generally regarded as kun'yomi unless they're some kind of relatively recent innovation. Rather that すべ・すべら・すめ・すめら are quite ancient -- see also https://kotobank.jp/word/%E7%9A%87-448702 -- these would have to be kun'yomi. I have restored them as such in the entry. ‑‑ Eiríkr Útlendi │^{Tala við mig} 01:12, 22 December 2019 (UTC)

@Eirikr: the same link, Daijisen section has this: ［音］コウ（クヮウ）（漢）　オウ（ワウ）（呉）　［訓］きみ　すめらぎ　すべらぎ

Rather than creating kun readings from dictionary entries as you did, the kun readings should be based on the kanji entry in question; therefore my POV that sube and sume being irregular readings. Unless someone has a physical kanji dictionary which list all of the possible kun readings of a specific kanji (don't have any at the moment; the online Jigen probably isn't one)... ～ POKéTalker（═◉═） 01:26, 22 December 2019 (UTC)

To turn it around, we could say that, conversely, you're limiting the list of kun'yomi based on dictionary entries.  :)

Per w:ja:訓読み,

訓読み（くんよみ）とは、日本語において、個々の漢字をその意味に相当する和語（大和言葉、日本語の固有語）によって読む読み方が定着したもの。

Further down, the article lists a few specific cases of 訓読み. For 皇, the readings in question aren't 熟字訓, nor 義訓, nor yet 国訓. They might be viewed as "irregular", but again, these readings are very well established quite early in the historical record: 「読み方が定着したもの。」 See also MYS 77, among various other places, where 皇 appears with one of these readings. ‑‑ Eiríkr Útlendi │^{Tala við mig} 02:35, 22 December 2019 (UTC)

@Eirikr: first of all you're confusing the interpretation of the man'yogana in the poem with *the original poem*:

吾大王物莫御念須賣神乃嗣而賜流吾莫勿久爾

wago₂ opoki₁mi₁ mono₂ na-omoposi sume₁kami₂ no tugi₁te tamape₁ru ware nake₁naku ni

(please add an English translation of this usage example)

...that means there is no written 皇 (sume₁-) in the above poem. If I make an OJP entry for sume₁-, there might be a dilemma: should it be created at 皇 or すめ?

What I mean is, take 祈 for example. Its kun reading is 祈る (inoru). However, dictionaries list 祈ぐ (negu) with this kanji despite Daijisen not having this in its kanji entry; the ne- can be an irregular reading. That is at least my stance on the issue, unless another dictionary has ね-ぐ on its 祈 entry.

If you like though, if my single-kanji edits are in need of more readings, feel free. For now I follow the Daijisen rule of {{ja-readings}}. ～ POKéTalker（═◉═） 03:22, 22 December 2019 (UTC)

• Re: the MYS, I did not intend to do an exhaustive search. For the sume reading in the context of the compound term 皇神 (sumekami), where the kanji 皇 is used in the source rather than a phonetic equivalent, see MYS 5 #894. I'm sure we could ferret out other cases where 皇 is used in compounds in some poems, and the same compound with a phonetic spelling is used somewhere else as confirmation of the reading.
• Re: dictionaries, I did some poking around.

• Nelson's kanji dictionary for English readers includes the sumera reading as a kun'yomi. Google Books isn't cooperating (hiding the relevant page from preview), but I can upload a photo of my own dead-tree copy if you'd like. Nelson's reading listings are minimal rather than exhaustive, and it does not include any of the sube- readings, nor sume nor sumeragi.
• The monolingual Japanese SMK5's kanji lookup index in the mini paper version I have here indicates kun'yomi as 「△すめ（ら）△すめらぎ」. The triangle is used in the index to indicate non-Jōyō readings. I can't presently find an explanation for the SMK5's parenthesis notation here, but I do see that the main body of the dictionary includes an entry for 皇 (sume) as a prefix alongside a separate entry for noun 皇 (sumera), suggesting that the （ら） indicates a kind of branching, where readers can expect to find an entry with the listed additional kana, and an entry without. Both entries show a 《 above the readings (vertically oriented, as the text itself is so oriented), which is explained in the introductory material as indicating non-Jōyō readings. In addition to omitting rarer senses, forms, and readings, as befits a smaller dictionary, the Shinmeikai in general is known for being a bit odd, as dictionaries go; c.f. w:ja:Shin Meikai kokugo jiten.
• Daijisen's entry seems to use a similar notation, a downward pointing ▽ to show non-Jōyō readings, as explained in the 「2. 常用漢字」 section of their 凡例 page (can't link directly to that section). I do find it odd that Daijisen's ［漢字項目］ section for reading kō doesn't list all kun'yomi, and then the fuller entry below includes nothing for any of the three listed kun'yomi of kimi, sumeragi, or suberagi -- none of which are included in the Jōyō readings, either. In general, I've found that the DJS seems to omit information about rarer senses, forms, and readings.
• The Daijirin entry at Kotobank is missing such metadata about readings, but the Weblio version includes it. We again see a ▽ used to indicate non-Jōyō readings (see also their 凡例 page). The DJR also seems to include less information than the KDJ. Compare the DJS, DJR, and KDJ entries for 丸太, for instance -- the DJS and DJR are both missing the additional KDJ sense related to fish, and while the DJS and DJR provide brief definitions related to Edo-period prostitutes, we have to read the KDJ entry to get any additional information about why "debarked log" came to mean "unlicensed prostitute masquerading as a Buddhist nun".

I've also got a dead-tree KDJ somewhere, I'll see if I can find that one too. I've gravitated toward using the KDJ more and more over time, as it provides more information than many other JA dictionaries. I liken it to the OED as an exhaustive resource, as compared to the more "abridged" information provided by Merriam-Webster, Collins, etc. Similar to the OED, however, the physical KDJ is a bit of a monster, so I mostly use the electronic CD version, due to ease of use and portability. ^_^  :)

Cheers, ‑‑ Eiríkr Útlendi │^{Tala við mig} 20:02, 23 December 2019 (UTC)

@Eirikr: as long as the Shin Meikai (and any other dictionaries) have すべ(ら)/すめ(ら) as kun readings of 皇, that is all. By the way, does the Shin Meikai include ね-ぐ in the 祈 entry?

Side note: went to Kinokuniya Books in Little Tokyo a few days before Christmas, skimmed through Tuttle's large kanji dictionary in the language shelf, and it does not have すべ(ら)/すめ(ら) kun reading for 皇 nor ね-ぐ for 祈. Rare readings indeed. ～ POKéTalker（═◉═） 05:20, 27 December 2019 (UTC)

@Poketalker I'm resurrecting this discussion 5 years on to say that I agree with @Eirikr here, and I have two major problems with using the term "irregular" to refer to readings on single-kanji entries:

1. The definition of "kun'yomi" is not something that is prescribed by dictionaries or some authority, so we should not be limiting ourselves only to readings which we can find in other dictionaries, since that fundamentally misunderstands what it means for a reading to be kun'yomi in the first place and is pretty misleading to our readers. For example, you originally added 刺(いら) (ira) (etymology 3) as an "irregular" reading ([1]), but it's (a) a native Japonic word that (b) appears in multiple compound terms, and (c) hasn't undergone any unusual development. By any conventional understanding of the term "kun'yomi", 刺(いら) (ira) should be covered, but I assume you marked it as irregular because it didn't appear in the reference dictionary/dictionaries you consulted, since it's not a common reading in modern Japanese. However, that doesn't mean it's not kun'yomi - it just means it's rare.
2. More importantly, marking such readings as irregular is also a misuse of the word "irregular": to call something irregular, we have to have a "regular" standard to measure it against. Outside of defined categories like on'yomi etc, we don't have anything like a spelling standard to measure kanji readings against, so the only thing that can make a reading irregular is if it only occurs in certain fixed contexts, because otherwise there's nothing that makes them distinct. By definition, that means they can only occur in compound terms. Being rare or obsolete doesn't make a reading irregular, since there's still no contextual restriction on its usage; they just don't gets used very often, but that's irrelevant. This works the other way, too: the kana は (ha) has the irregular reading は (wa), which is used in a single word, but that one word occurs about 10 times more frequently than all uses of は (ha) combined. It's still irregular, though, becuase - unlike kanji - kana usually only have one reading. The point is that "irregular" does not mean "other" or "we don't know"; it refers to a specific phenomenon which cannot occur with single-kanji terms.

I really think we should amend all these entries, and I'd like to add a restriction to kanjitab to prevent it being added to single-kanji terms. I recently added a new restriction so that you can't specify one-kanji readings as being jukujikun (since they're multi-kanji by definition), and adding that threw around 100 entries into CAT:E; having been over quite a few of them, I got the distinct impression that a lot of the bad inputs were caused by people being confused by the base kanji entries calling such readings irregular, which has led to lots of nonsense. Theknightwho (talk) 06:48, 29 June 2024 (UTC)

From that reply alone, that totally defeats the purpose of said "prescribed dictionaries". Am going to stay out of this discussion until something like a "neutral irregular" close to "rare" is included as a reading type... which can never happen. Just a few kanji dictionaries including ira as a native (kun) reading is enough for me to warrant either that or not that. The more exacerbations on the code due to radical change, the better. ～ POKéTalker（＝◉＝） 07:45, 29 June 2024 (UTC)

And do not drag は (ha/wa) to this discussion; former depends on kanji used, and latter is a particle not a "word" per se like in English. It seems your user page lacks a Japanese level Babel template at the time of this reply; at least mention your proficiency level of it so we have equal ground... ～ POKéTalker（＝◉＝） 07:55, 29 June 2024 (UTC)

@Poketalker The point of bringing up は (ha/wa) was to illustrate why it's wrong to use "irregular" to mean "rare", because it's a good example of something that is irregular despite being more common than the expected regular reading. I picked it as an example because I thought you'd understand where I was coming from, but the specifics of when each reading get used or whether a particle is a word have nothing to do with what I was saying (though FWIW I disagree with you).

Could you please point me to something that says "kun'yomi" is defined by what's in dictionaries, or what your basis is for wanting a separate reading type (which is something I haven't seen any other dictionaries do)? That's not my understanding of the term at all, and it's not Eirikr's either, so I don't really understand why we can't just use the label "rare" on the definition line. Plus, rare readings may be semantically restricted or restricted to certain time periods, but that still doesn't make them irregular, because there is no frame of reference by which we can call it irregular when it's a single-kanji term, because irregularity is a linguistic question, not a frequency one. Theknightwho (talk) 15:55, 29 June 2024 (UTC)

@Poketalker @Eirikr I hesitate to suggest this as it feels like a big can of worms, but would some kind of distinction between 正訓(せいくん) (seikun) and 義訓(ぎくん) (gikun) work? Theknightwho (talk) Theknightwho (talk) 16:58, 29 June 2024 (UTC)

Am already having a bad feeling about rendering native historical/obsolete "readings" as kun outright in regards to inclusion to {{ja-readings}} and {{ja-kanjitab}}; this seems counter or to be a radical turn... would be a bigger can opening to much more confusion. My other rationale for remaining them as irregular, aside from explicitly excluding as such in said dictionaries and counterproductivity, is the possibility of becoming more common on the public consciousness; will cold water be called mo(h)i after your recent edit(s)? kimi/suberagi to address the Emperor with this 皇 kanji in writing? The more digital the more liberal, so they'd say...

The Daijisen defines kun'yomi as synonym to kundoku, in turn:

１ Reading a Chinese character (kanji) in correspondence to its meaning and Japanese "pronunciation". Examples: 花 is [read] hana and 草 kusa. Antonym is ondoku.
２ Reading classical Chinese texts (kanbun) applying Japanese grammar rules and annotation marks (kunten).

No explanation about rare and obsolete readings here, so would leave them labeled as irregular or through discernment.

Seikun simply means a kun reading with definition(s) conforming to its original Chinese character, while gikun is as straightforward as "meaning-reading".

Wish that water in general wouldn't be translated as moi in the future... ～ POKéTalker（＝◉＝） 01:43, 30 June 2024 (UTC)

@Poketalker Three things:

1. The definition provided by Daijisen does not exclude rare or obsolete readings. 刺(いら) (ira), for example, is read according to its meaning ("thorn") and its Japanese pronunciation (ira), as it is a native Japonic word. Nothing in the Daijisen definition says that kun is only limited to common readings, so by the very definition you are using, 刺(いら) (ira) is kun'yomi.
2. If I understand you correctly, you are concerned that labelling these readings as kun will lead to confusion. This is an issue that isn't unique to Japanese, and our Chinese entries deal with it quite well. For example, alongside common readings like Mandarin 參／参 (cān, “to join; to consult”) and 參／参 (shēn, “ginseng”), theres also 參／参 (sān, “three”, financial, obsolete). The solution is to label the senses appropriately, not to label the reading as irregular. In several cases of readings that you've labelled irregular, the reading has turned out to be entirely regular in derived compounds, so you actually caused confusion for me, instead of dispelling it.
3. Leading on from point 2, your approach would cause me genuine confusion if a rare kun reading that you've labelled irregular was used alongside a genuinely irregular reading (e.g. a contraction). Some good examples of this occur in terms that use the rare kun reading 省(つかさ) (tsukasa, “ministry”), which occurred in the names of various Japanese ministries from the middle ages (see here). In many cases, these were simply taken from (or modelled on) Chinese as 〜省 , with the Chinese word 部 (bù, “ministry”) remaining as a fossilised part of the term. For example, Chinese 兵部 (Bīngbù, “Ministry of War”) became 兵(ひょう)部(ぶ)省(しょう) (Hyōbushō), also read 兵部省(つわもののつかさ) (Tsuwamono no Tsukasa). In the second reading, we have 兵(つわもの) (tsuwamono, kun) + の (no) + 省(つかさ) (tsukasa, kun), where 部 has an irregular "reading" of not being read at all. If we were to take your approach, {{ja-kanjitab}} would say kun, irregular, irregular, since 省(つかさ) (tsukasa) is obsolete in the modern day, but this would be completely misleading in context, as the only truly irregular "reading" is that of 部. It's definitely not jukujikun, either, since 省(つかさ) (tsukasa) occurs in the names of other ministries that don't contain 部, and has the same reading in all of them: 中(なかの)務(まつりごとの)省(つかさ) (Naka no Matsurigoto no Tsukasa), 宮(みやの)内(うちの)省(つかさ) (Miya no Uchi no Tsukasa), 刑部(うったえただす)省(つかさ) (Uttae Tadasu Tsukasa) etc.

I simply don't see how this is a radical change. The reading is but one of many things in an entry, and if it's labelled "kun" but all the senses are labelled "obsolete", "dialectal", "rare" or whatever else, that tells the reader exactly what they need to know. Theknightwho (talk) 16:59, 30 June 2024 (UTC)

Until tsukasa is seen as an explicit kun-reading on a monolingual kanji dictionary, we are going nowhere; too many worms to can back. 刺 only has the kun readings sa-su, sa-saru, and toge in the online Daijisen entry, same with kaeri-miru and habu-ku for aformentioned. Any nandokus like 刺草 (irakusa) are up to my discernment, whether to be as kun or remain irregular in template labeling. Don't drag the Chinese joining/consulting, ginseng, and financial three to a Japanese sorting/categorizing issue; they bear little to no mutual intelligibility with each other. Don't think in my opinion a native Japanese would agree with your looser interpretation of "native (kun) reading" in this regard... ～ POKéTalker（＝◉＝） 05:10, 1 July 2024 (UTC)

@Poketalker You're avoiding the points I'm making, and the fact that you have actively caused me confusion as a user of the dictionary by misusing the term "irregular" as a catch-all. There is no "looser" interpretation of "kun" - you've just decided it excludes rare readings for some reason... Wiktionary is not bound by what other dictionaries say, and I don't understand why you seem to think we can only mark something as "kun" if other dictionaries say so, when we can simply follow the actual definition as given by Daijisen. I have tried to give analogies to explain why I think your approach is misguided, but you've just dismissed them out of hand without trying to understand them; the fact that Chinese isn't mutually intelligible with Japanese is completely irrelevant to the point I was making, just as your dismissal of the ha/wa issue was. You're better than that. Theknightwho (talk) 12:20, 1 July 2024 (UTC)

@Poketalker, it's worth pointing out that Daijisen's coverage of rare, archaic, obsolete, and dialectal terms is not very good. The fact that Daijisen doesn't include tsukasa as a kun'yomi for the 省 character should not be interpreted as meaning that tsukasa perforce cannot be a kun'yomi for this character. My dead-tree copy of Nelson's, for instance, does include tsukasa as a kun'yomi for the 省 character.

Daijisen's omission is simply because their goal is coverage of the middle of the bell curve of the Japanese language: a focus on modern, everyday use.

Meanwhile, our goal is the coverage of *all* of the Japanese language (as best we can, anyway), and that is consequently going to include a lot of things that aren't in more abridged resources like Daijisen.

----

I think I understand (some of?) what you're saying about potential for confusion — a big part of learning to read Japanese is learning the readings for each character, and if we include all of the rare and obsolete kun'yomi for every character, that task becomes much more daunting.

That said, we also deliberately include notes in the output of both {{ja-readings}} {{ja-kanjitab}} to indicate whether a given reading is part of the jōyō kanji, and for {{ja-kanjitab}}, we also show what grade that kanji + reading combination is taught in. For instance, our entry at 省#Japanese clearly shows that the shō, sei, kaeri-miru, and habu-ku readings are jōyō, while the others are not. And the instance of {{ja-kanjitab}} in the entry sub-section for the shō reading shows that this kanji + reading is taught in grade 4.

This kind of information should help learners in focusing their efforts, and I think this should also, in combination with sense labels like "archaic / obsolete / dialect" etc., help readers avoid confusion about how to use the different readings. For 水 (moi), for instance, we include an "obsolete" label, which should be a clear marker to readers that this will not be understood by modern speakers.

If I've misunderstood your concerns, please help me understand better. ‑‑ Eiríkr Útlendi │^{Tala við mig} 18:24, 1 July 2024 (UTC)

This is a hard porridge to swallow, really. All @Theknightwho has done is the reverse of "catch-all" to the kun-reading definition; and I don't know how a native Japanese reader/editor would react... is "irregular" reading parameter for {{ja-kanjitab}} being reinterpreted?

Am asking for a compromise now as their edits are about to be like "spread like wildfire"; no way I would follow the make the all native archaic, obsolete, dialect, etc. readings categorized into one unified Kun: line or kun in kanjitab. Said readings would be inconsistent to those seen in monolingual dictionaries, especially kanji ones (not talking about Nelson's by Tuttle as above). Until then, must "hew to Japanese dictionaries" on my edits as you said... who knows moi will be more common than mizu in referring to cold water.

...or stay out of editing modern Japanese for the time being, this return of "capture many native readings as you can like 'mons" seems radical to me. ～ POKéTalker（＝◉＝） 11:12, 2 July 2024 (UTC)

I think we should, no matter what we decide, take very careful stock of the extent of the impact it would have to entirely stray from what is published by other dictionaries. I'll maybe try to make a more compelling argument later, but especially reading some historical texts, which are way more liberal in assigning whatever kun readings they want to whatever kanji they want, I think the number of kun'yomi could blow up massively if we aren't careful. This would make a huge proportion of the listed readings pretty useless to the average reader since they're unlikely to ever actually occur, besides in a rare usage of gikun or something.

The argument in this case isn't based on frequency, just on utility; whilst including every kun'yomi that's ever been interpreted could satisfy the technical, dictionary definition of kun'yomi, it sort of misses the mark for the practical meaning of the word IRL, where FTR not all kun'yomi readings are even of native Japanese origin (just to point out the flaws in adhering only to a brief, dictionary definition of kun'yomi; we should try to create our own, more detailed policy, since evidently we are disagreeing on some of these points! :)).

As a remedy, how about we require a minimum attestation count for readings, just like the basic inclusion criteria for words in general? 3 uses quoted, or if it obviously exists. I think this would prevent annoying hapaxes or things that people would question from setting in, and also ensure that a reading isn't just favored by one author or source or whatever. What do you think? Kiril kovachev (talk・contribs) 11:56, 2 July 2024 (UTC)

@Kiril kovachev I don’t think I’m reinterpreting anything here - this seems to be down to Poketalker’s misunderstanding of what the word “irregular” means, as this is something they clearly disagreed with Eirikr on several years ago, and it’s not something I’ve seen coming from many, if any, other users..

I agree with you that rare readings should be subject to minimum attestation requirements, but what I don’t understand is all this hand-wringing about misleading people, when we can - as Eirikr and I have both pointed - use labels for this job, instead of misusing the reading parameter. We already give plenty of rare or obsolete readings as it is; people won’t suddenly be mislead if we label them “kun'yomi”, but they will be mislead, as I was, if we wrongly label them irregular when they’re actually rare, dialectal or archaic, because that’s way too vague and obscures actual irregularities. Theknightwho (talk) 15:34, 2 July 2024 (UTC)

@Theknightwho Yeah I agree with that. I think I misunderstood what we were talking about, lol. So, as long as the reading is used consistently (for the words that it's used for), but not necessarily commonly, it can be safely called a regular kun'yomi - right? Kiril kovachev (talk・contribs) 15:43, 2 July 2024 (UTC)

@Kiril kovachev Yes, precisely - that’s my understanding: if it’s not on'yomi and not ad hoc, it’s kun'yomi. We shouldn’t be adding ad hoc readings since they won’t pass CFI, so the only irregular readings should occur within compound terms, since they’re irregular by being limited to that compound.

What I find confusing is where Poketalker says they want us to be consistent with monolingual dictionaries but they‘re “not talking about Nelson's by Tuttle as above” (since these have been found to contain mum readings they claimed weren’t in monolingual dictionaries), which… seems a bit like trying to have their cake and eat it. I think it gets to the heart of the issue, though: they want a general-purpose dictionary, but Wiktionary has always been about being a comprehensive dictionary, and we need to be clear in how we label that. Theknightwho (talk) 15:54, 2 July 2024 (UTC)

I will leave my commentaries here then, as I don't think I have much more to add - I think we're already doing well with listing kanji readings, and I am of the opinion that we don't need to agree with other dictionaries, just that they *are* a reasonable standard for what a sensible notion of "having a reading" is; but if we're judicious in diverging from that, there should be no issue at all IMO. My goal ATM is to ensure that all readings in Kanjipedia / Kanken Kanji Jiten, are included, and we can continue to expand on that with more readings if we find them out there.

The thing we must avoid is the situation of Unihan, where lots of kanji have had ridiculous kun'yomi entered that there's no way that they really exist. This was a discussion we had a while ago, and in fact Wiktionary is currently still polluted by many of these readings. However, all of those were added here by a bot, so I trust we don't really consciously lean that way when editing by hand. Kiril kovachev (talk・contribs) 16:07, 2 July 2024 (UTC)

@Kiril kovachev Yeah, I agree. I know that Unihan are doing (have just completed?) a massive review of everything in Dai Kan-Wa Jiten, because they were very aware of how much crap was in their system, so it might be worth re-looking at their data in that light. Theknightwho (talk) 17:34, 2 July 2024 (UTC)

@Theknightwho That would be good, maybe if they've finished it we could try comparing what readings it now has compared to ours. Maybe that could show which were the bogus ones. I was meant to try that before already, but if they have indeed finished this would be good to make use of... Kiril kovachev (talk・contribs) 23:06, 2 July 2024 (UTC)