Wiktionary talk:About Japanese/Archive 2

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


From talk:かく[edit]

Please avoid the term "furigana" since it is more specialized than "romaji" which is also more widely understood. Also "information" should not be in headings. Information such as headwords, hiragana spellings, etc belongs in plain text below the headings. — Hippietrail 15:07, 11 Dec 2004 (UTC)

[The following post split into topic-specific sections by Rodasmith 20:19, 27 April 2006 (UTC)]
<Jun-Dai 16:19, 11 Dec 2004 (UTC)>I'm not sure what the standard format is for Japanese furigana entries. I was basing this off of what I saw in しゅ. Since this entry and others like it will be almost exclusively of interest to people that want to know what kanji are constructed from this pronunciation (most likely broken into on and kun readings), it made sense to create a furigana section. Also, the term "furigana" should be known by anyone looking up an entry like this one, since it is Japanese 101 terminology, and this entry (which is really just a word pronunciation, albeit in Japanese) will be of zero value to someone that doesn't know at least a tiny bit of Japanese.</Jun-Dai>


I've just been going through the categories trying to learn more about how things here are organized, and run into some serious concerns, as it seems like a good bit of earnest hard work has been in unfortunately misguided directions. The Furigana category is one such instance. Specifically, the header to that page describing what the category is for notes the following:

Furigana (振り仮名) refers to the hiragana (ひらがな) representation of a Japanese word for the indication of pronunciation. It is most frequently used to refer to hiragana written just above or below a sequence of kanji (漢字) characters, but it can also be used to refer to the hiragana representation of Japanese words as, for example, the entries of a dictionary, or in an index (this makes alphabetized indices of any sort much easier to use).

I've followed some of the discussion here, though I came in late and was unable to participate in a lot of it directly, and it seems people have (hopefully) come to the realization that furigana are not the same as hiragana. Nonetheless, I feel the need to clarify on the one hand, and ask the community what we can do to refactor the goofs on the other.

To clarify, furigana are not just the hiragana (ひらがな) representation of a Japanese word for the indication of pronunciation, they are only those kana placed next to or above the kanji (and occasionally rōmaji) to indicate pronuncation. This equates to the first clause of the second sentence in the quote above, minus the words most frequently. In the word 食べる, for instance, the only furigana this word could possibly have is , as we see below. The trailing hiragana べる are what is called okurigana (送り仮名, おくりがな).

Note: For those interested in how to make furigana, have a look at the source. But be warned -- it's not pretty, it doesn't scale across font sizes if the user changes them, and you have to manually figure out the values for "top" and "left" to make things look right. Unless you really need to use furigana right now, it's probably best to wait for CSS 3, which will include proper ruby (i.e. furigana) support.

The Category:Hiragana explanation is correct, and seems to fill the index role proposed on the Category:Furigana page.

The hiragana (平仮名, ひらがな) form of a Japanese word is a phonetic representation of that word.

Might I suggest that the Category:Furigana entries be changed over to Category:Hiragana? Category:Furigana itself should probably be removed once this is done. If there are Category:Furigana entries where the entry term itself is in fact not written in hiragana, I suggest the term be removed from Category:Furigana and not added to Category:Hiragana, but instead indexed under Category:Japanese language, as below for the term 音読み:

[[Category:Japanese language|おんよみ]]

This term entry is not written in hiragana, and thus should not be categorized under Category:Hiragana. However, adding the pipe and hiragana pronunciation to Category:Japanese language ensures that the term is still indexed properly.

(As a side note, it seems Category:Japanese nouns is not sorting properly, as 音読み is listed before お父さん... I'll see about fixing Template:janoun to add proper indexing, but just in case, are there any techies out there who can explain why this might be, and hopefully help fix this?)

If anyone has any strong arguments for keeping Category:Furigana, by all means let me know. If I see no replies to the contrary, I'll start refactoring the category to get it ready for deletion. Thank you, Eiríkr Útlendi | Tala við mig 20:34, 6 April 2006 (UTC)

Given the complete lack of comment, I've begun refactoring. For hiragana entries that are actually used as furigana for kango, I'm using a slight modification of the format seen on the くつ page. Cheers, Eiríkr Útlendi | Tala við mig 23:27, 13 April 2006 (UTC)


Hi all :) I'm pretty new so I'll probably be popping in here with questions as I figure out how this all works. As for Romaji entries, the current policy is that yes, they are desirable? I can say that I find them useful, as my Mozilla search bar doesn't seem to deal well with non-latin characters, and I think they're likely useful for those who can't read kana or kanji, or who don't have a kana/kanji input method installed at all. However, the links to Romaji entries are being consistantly removed by User:Gliorszio, even in cases where the Romaji entry exists. I guess I'm asking for some sort of confirmation of consensus on this. I'm going to ask him to come here to discuss via his talk page... Cruinne 14:56, 15 February 2006 (UTC)

I'd like to be able to search for everything using romaji, but it doesn't seem necessary to link to a romaji page from a kanji or hiragana page, since it wouldn't contain any new information. Kappa 23:22, 28 March 2006 (UTC)
I was also confused about whether to create and link to romaji entries as well. I think I understand the criteria for inclusion now, though, which suggests that we should only create romaji entries for romaji that are comonly used (i.e. not in word lists or teaching material when hiragana or kanji are also given). The "comonly" criteria is a bit fuzzy, but it's probably safe to say that we should include romaji entries for romaji that appear, for example, more than 100,000 times in a Google search that excludes the corresponding hiragana and kanji terms. Each kanji and hiragana entry should show romaji transliterations, though, so that searches for the romaji will work as expected. Rodasmith 05:26, 29 March 2006 (UTC)
Rats. Kappa 05:43, 29 March 2006 (UTC)
Point well taken, Kapa. :-) There are not yet any explicit guidelines regarding CFI for romaji and kana transliterations, despite the suggestion for such guidelines from the Wiktionary Transliteration Policy. So, I will open that discussion here. Rodasmith 17:55, 29 March 2006 (UTC)
What about simply having redirects from "uncommonly used" romaji entries to the corresponding kana or kanji entries? This would allow folks to search Wiktionary using just Latin characters, which, as some point out, is the only way certain users have of entering text. Eiríkr Útlendi | Tala við mig 17:33, 30 March 2006 (UTC)
For that matter, the electronic J dictionaries I've seen have often allowed lookup by either kana or kanji, and in some cases also by romaji -- i.e., in Wiktionary terms, we'd have one entry for each Japanese word, standardizing possibly on hiragana as it is indeed Japanese-only, with redirects from the other renderings (hiragana, katakana, romaji...?). This would certainly cut down on the amount of duplication -- do we really need full entries for each Japanese word, once for the kanji, once for the hiragana, once for the romaji? I've only been digging around here for a day and a half, and already I've seen significant format variation between different entries for the same word. Consolidating sure looks like it'd make things simpler. Cheers, Eiríkr Útlendi | Tala við mig 17:40, 30 March 2006 (UTC)
As noted below, see esa and から for examples of terms that prevent us from adopting the proposed #REDIRECT solution. Rodasmith 19:20, 30 March 2006 (UTC)

Criteria for inclusion of romaji and kana[edit]

A week has passed since my proposal to relax the CFI for romaji and kana entries. Finding no objections, I updated Considerations about Japanese language entries to stipulate that CFI relaxation. [Section refactored for brevity.] Rodasmith 04:42, 5 April 2006 (UTC)

Rōmaji entry format[edit]

I'm a little puzzled by the suggested format for rōmaji entries. The main article page shows the following:

 '''romaji version''' (hiragana version)
 # [[kanji form 1]]: short definition 1
 # [[kanji form 2]]: short definition 2

I see a number of problems here.

  1. We lose part-of-speech information. This is somewhat jarring in its inconsistency for those rōmaji entries that are homonyms of words in other languages, where the other languages follow the usual format of:
    ===Part of speech===
  2. If we were deliberatly looking up a rōmaji term, this subheading serves no function, as we already know it's rōmaji.
  3. To the uninitiated, Rōmaji might be taken as a category similar to a part of speech on the one hand, and the kanji forms might be mistakenly understood as aspects of rōmaji-ness, as it were, on the other.
  4. Kanji forms do not necessarily imply different meanings. As I note below in the Hiragana and romaji in entries subsection, the word tsuku has very many kanji forms, many of which are effectively synonymous. Organizing an entry's meanings by kanji forms is unproductive and potentially very confusing -- imagine an entry with multiple kanji forms, all with the same short definition.

As an alternate, I propose the following, just as an idea to start from:

 === part of speech ===
 # short definition 1
 # short definition 2
 === Written forms ===
 * [[Rōmaji]]: ''romaji form''
 * [[Hiragana]]: [[hiragana form]]
 * [[Kanji]]: [[kanji form 1]], [[kanji form 2]], etc...

Where kanji forms are specific to certain meanings, we could indicate that by a simple note. Another option is that a new Kanji forms fourth-level subheading could be added under each relevant definition.

What thoughts? Cheers, Eiríkr Útlendi | Tala við mig 23:36, 6 April 2006 (UTC)

Yes, it would be nice to stop misleadingly categorizing rōmaji as a part of speech. How would the casual reader who stumbles upon a romaji entry know where to find the main entries corresponding to that romaji? Perhaps the "Written forms" section should link only to main entries. I.e., if a romaji is the romanization of both a wago term (whose main entry are at the hiragana page) and two kango terms (whose main entries are at the kanji+okurigana pages), the hiragana and the kango terms should be linked, but not the kanji-compound for the wago definition. Rodasmith 16:50, 7 April 2006 (UTC)
Minor clarification -- if a word has okurigana, it is by definition a wago, and I think it has to be a verb too (or at least a verb form), as okurigana provide the conjugational ending. Kango have no conjugation and are almost always written exclusively in kanji, with the only exceptions being words where the kanji in question is written in kana instead as it is either too complicated or too unusual, such as 濾過 roka "filter", where the first kanji is often replaced, as in the form ろ過.
I'm confused about what you mean by but not the kanji-compound for the wago definition? Do you mean that a link to the wago 食べる (there's that kanji + okurigana by the way) should point to たべる rather than 食べる? If so, I agree, in principle. For entries with multiple overlapping kanji forms such as つく, however, this gets tricky. I still argue that wago definitions belong on the kana entry page, but if we have something like [[つく|就く]] it would be good for the user if we had some way of indicating which definition of つく we mean, such as [[つく]]: 就く, II.2., and this is difficult in a rapidly-changing project like this where we are not sure how long that "II.2." indication will be correct. Hmm, something else to ponder... :) Cheers, Eiríkr Útlendi | Tala við mig 20:16, 10 April 2006 (UTC)

Rōmaji transliterations[edit]

There was some question over on Talk:on'yomi about the use of apostrophes in the rōmaji transliteration of Japanese words. Upon consideration, the points brought up might be more useful here, so I am copying the relevant portion below. Comments are welcome. Eiríkr Útlendi | Tala við mig 00:43, 17 April 2006 (UTC)

I don't know whether Wiktionary follows the same rules as Wikipedia, but w:Wikipedia:Manual of Style (Japan-related articles)#Romanisation states, "Article titles should omit apostrophes after syllabic n." Rodasmith 03:45, 4 April 2006 (UTC)
After looking over what the w:Wikipedia_talk:Manual of Style (Japan-related articles) page had to say (searching for "apostrophe"), I could find no clear discussion of why apostrophes are to be avoided. There's a lot of talk about why macrons should be used, and the justifications given about distinguishing words, achieving a closer transliteration, and clarifying pronunciation, would seem to also apply to using apostrophes between ん and following vowels. The main point brought up for why not to use the apostrophe essentially comes down to cosmetics -- correct or no, people don't like how it looks -- which strikes me as a weak argument. I can understand that viewpoint somewhat for an encyclopedia focusing on a general English-reading audience, but it seems very inappropriate for a dictionary where part of the point is to show the correct rendering, even more so when it's a bilingual dictionary and as such should be more specific to the conventions of the non-English language. Yet again even more so once you start factoring in words that, without the apostrophe, would have the same rendering in rōmaji: one such instance would be 文案 (ぶんあん, "draft") and 無難 (ぶなん, "safe"), which should be bun'an and bunan, but with the no-apostrophe rule both would be listed under bunan.
Such confusion in a dictionary strikes me as unacceptable. Some homophones are inevitable, but bun'an and bunan aren't even homophones. Beginning studiers of Japanese might not know better, but this is easy enough to allow for by adding a simple line at the top pointing to the other rendering, as we see on Wikipedia at w:Mandoline, for instance. I am strongly in favor of being more specific where possible, and breaking ranks with Wikipedia to use the apostrophe in article titles looks very much like the right thing to do here. But if anyone has a strong argument otherwise, by all means let me know. Cheers, Eiríkr Útlendi | Tala við mig 17:59, 5 April 2006 (UTC)

Sheesh. Amazing what you can find when you look. Et voilà, Wiktionary:About Japanese/Transliteration, though still a draft, is a policy recommendation that looks by and large pretty good to me. The page discusses apostrophe use, and comes down on the side of using them between ん (moraic /n/) and vowels. For anyone working on rōmaji, please read this page. Incidentally, looks like I have some cleaning up of my own to do (that bit about the diaresis was news to me...).

Just for the record, I have some serious concerns about the recommended treatment of compound nouns and the nearly innumerable する verb combinations (eg the 県庁所在地 and 勉強する examples, each of which by my judgment should be two "words" when romanizing, and the use of hyphens), which I shall take up on Wiktionary talk:About Japanese/Transliteration. Those of you interested in such issues, please look there -- though not tonight, I'm signing off for now. :) Thank you, Eiríkr Útlendi | Tala við mig 03:40, 28 April 2006 (UTC)

Hiragana and romaji entry format[edit]

[This post split into topic-specific sections by Rodasmith 20:19, 27 April 2006 (UTC)]

I have serious reservations about how the proposed template suggests we deal with hiragana and romaji in entries. The article page suggests entering kanji under the hiragana or romaji heading, which frankly makes no sense to me. An example:

 '''hiragana form''' (''romaji form'')
 # kanji form: short definition
 # kanji form: short definition

My confusion is, if the entry is already the hiragana (or romaji) rendering,

  1. why do we need to point out in a subheader that it's hiragana (or romaji)? Isn't it already obvious, as that's what the entry heading is (i.e. that's what we typed in to get here)? Then,
  2. shouldn't the kanji be listed under a ===Kanji=== subheading? Otherwise it looks most illogically as if the kanji are supposed to be some component of hiragana (romaji). [...] Cheers, Eiríkr Útlendi | Tala við mig 18:14, 30 March 2006 (UTC)

[remainder of post moved to #Kango and wago entries, below by Rodasmith]

Location for main entry[edit]

From talk:かく[edit]

[The following post split into topic-specific sections by Rodasmith 20:19, 27 April 2006 (UTC)]
<Jun-Dai 16:19, 11 Dec 2004 (UTC)>I haven't seen any discussion for the formatting of entries like this one, but the format I constructed here is extensible and seems useful for furigana entries, whereas the one at くる will quickly become unwieldy once it is even moderately expanded upon. The furigana pages should really just be landing/reference pages for the kanji entries (though a rudimentary meaning should be next to the kanji in parentheses, because entries like , once flushed out, will contain references to dozens of kanji.</Jun-Dai>
Actually it is not so simple. As you know, many Japanese words have no kanji, especially particles, adverbs, and many "grammar" words. I have no idea if there are any such words spelled "かく" but it would be terribly unweildy to have one format for words which always map to kanji and a different one for words which map to kanji sometimes and not others.
Some of these topics have been discussed in random articles as they are edited, much as we're doing now. We should have a solid place to discuss these issues but such a thing has not evolved as yet.
I agree that the very terse format including a gloss is very good for words which are usually spelled in kanji. Words which can only be spelled in hiragana (note that these might be called okurigana but not furigana, whereas hiragana is always appropriate) will need a full regular entry. Words which are often spelled either way, or which were spelled pre-1947 with kanji and post-1947 with kana, ateji such as sushi, words with obsolete kanji spellings, all need to be taken into account so that we can come up with one format that fits all these oddball cases. Japanese is particularly challenging for us! — Hippietrail 16:34, 11 Dec 2004 (UTC)
<Jun-Dai 16:53, 11 Dec 2004 (UTC)>What you say is true, but I don't see why a Furigana section at the bottom is mutually exclusive of a definition section at the top. Here's my proposal:
  • If a hiragana word has no kanji, then it should have a definition at the top of the page.
  • If a hiragana word has kanji, but that kanji doesn't really exist in modern Japanese (e.g., しまう), then the primary entry should be on the hiragana page, and anything on the kanji page should indicate that it is archaic.
  • If a hiragana word has kanji, but is frequently expressed in hiragana, even in formal (technical, commercial, official) contexts (e.g., いる, する), then the primary entry should be on the hiragana page, and anything on the kanji page should indicate that the word is normally expressed in hiragana.
  • If a hiragana word has kanji and is expressed in formal writing in kanji, but is expressed in informal writing in hiragana (e.g., 醤油, 美味しい), then the primary entry should be on the kanji page, but the word should be given a line in the definition section of the hiragana page, which should include a reference to the kanji page.
  • If a hiragana word has kanji and is expressed in kanji in any context where kanji are used at all (e.g., , 書く), it should only receive a reference on the hiragana page in a "furigana" section, or any information in the definition section should contain a reference to the kanji section.
  • If a hiragana combination forms the onyomi or kunyomi reading of a kanji (or phrase with kanji, e.g., とびこむ -> 飛び込む), then the kanji should be given a listing in a furigana section, along with a very brief English-language definition or "general sense" of the meaning of the kanji (to make it easier to differentiate between the many kanji in the furigana section).
Let me know what you think, and we or someone else can put it on the main lang consideration page.</Jun-Dai>
I second Jun-Dai's proposal Millie 16:04, 6 December 2005 (UTC)
The above rules (with qualifiers "frequently" and "informal") seem too complex and subjective. #Hiragana and romaji in entries below proposes instead to define wago kango terms (which have kanji that are read as on'yomi) fully on their kanji page and kango wago terms (which either lack kanji or have kanji read using kun'yomi) fully on their kana page. Rodasmith 23:54, 5 April 2006 (UTC) Rodasmith 06:18, 6 April 2006 (UTC)
** Correction ** -- Switch the terms wago and kango in the paragraph above -- kango (漢語) are words of Chinese origin (or inspiration), and thus have kanji and are pronounced with the on'yomi (reading based on ancient Chinese pronunciation), whereas wago (和語) are words of ancient Japanese origin, and thus might not have kanji and are pronounced with the kun'yomi (native Japanese pronunciation). Cheers, Eiríkr Útlendi | Tala við mig 05:14, 6 April 2006 (UTC)
Oops. Switched. Rodasmith 06:18, 6 April 2006 (UTC)

Kango and wago entries[edit]

[This post split into separate topics by Rodasmith 20:19, 27 April 2006 (UTC)]

If we [...] consolidate the various Japanese word entries currently divided by rendering (i.e. with each single J word split up into one entry for kanji, one for hiragana, one for romaji) into one single entry, using redirects from the other renderings, much as is done in other online and otherwise electronic J dictionaries, some of this confusion is obviated.

I'm interested in anyone's thoughts here, particularly on the idea of consolidating under one Japanese rendering for each entry and redirecting from the others. Cheers, Eiríkr Útlendi | Tala við mig 18:14, 30 March 2006 (UTC)

I think the ===Hiraga=== or ===Romaji=== heading must appear because each Wiktionary entry shows details under a third-level heading for the speech described. In the kanji entry, we detail the term's meanings by separating it into ===Noun=== and ===Verb=== sections.
Regarding the redirects to consolidated kanji entries, I don't think that is possible in general. For one, a particular romaji term may correspond to a term in another language, so a redirect would not be appropriate. See esa, for example. Likewise most hiragana terms correspond to multiple kanji, so a redirect from hiragana to kanji is not possible. See から for an example of that. Rodasmith 19:17, 30 March 2006 (UTC)
Thanks for clarifying, Rodasmith. I understand that redirects from romaji might be problematic, but I overlooked that in my previous post. Above, I suggested consolidating on kana entries; would that work? Individual kanji and kanji compounds that are used in other languages would need to be their own thing, clearly, but kanji + okurigana combinations could be safely redirected as they would be Japanese-only, and thus we would not need separate entries for つかう, 使う, and 遣う, for instance. Cheers, Eiríkr Útlendi | Tala við mig 19:46, 30 March 2006 (UTC)
Refactored post no longer relevant
Hello Rodasmith --
I agree that this would be inconsistent with the policy for simple kanji terms, where the main entry must be the kanji page itself, as you put it, which is why I was thinking of consolidating つかう, 使う, and 遣う under the heading つかう, with subsections noting the two kanji renderings and linking through to the separate pages for 使 and 遣. This way, if someone searches for 遣う, they'd be redirected to the つかう page, but if they searched for 遣 alone, they'd get the kanji page, which would include 遣う as a Japanese usage linking back to つかう. Plus, inflection information is common to both 遣う and 使う, and thus would seem to belong more on the つかう page common to both.
I got the feeling we're maybe talking past each other; does this description make any more sense?
I agree that romaji entries should be kept simple. However, for some kana entries, I wonder how things would be organized? For つく, for instance, I count:
  • 付く・着く・就く・即く・憑く
  • 突く・衝く・撞く・搗く・舂く・築く・吐く
  • 尽く
  • 漬く
  • 点く
And then the つく entries without 送り仮名:
  • 銑・釻・柄
The items separated by nakaguro are all lumped together as variants (as far as meaning is concerned) in my electronic copy of Shogakukan's 国語大辞典 (1988). Clicking on one of the variant headings shows the shared meanings for all of them, with all the variants listed across the top. Separating all these out into individual pages loses that common relationship.
I'm wondering how we could go about organizing the entries to make it clear that separate kanji do not always denote cut-and-dried differences in meaning? Especially when the J word in question defies a simple one-to-one English correlation -- For the first kanji group, for example, do we render this as "stick to", "accompany", "arrive", etc...? If we make the kana entry the core, we could then break things down in more logical groupings, perhaps:
(Hiragana heading)
== Verb ==
(I) Def 1, 2, 3, ...
(kanji list)
(II) Def 1, 2, 3, ...
(kanji list)
== Noun ==
(I) Def 1, 2, 3, ...
(kanji list)
(II) Def 1, 2, 3, ...
Each logical group would go under the Roman numeral, with the individual meanings listed under the Arabic numerals. This would preserve synonymous relationships without requiring that we duplicate information across multiple entries. Any thoughts? Eiríkr Útlendi | Tala við mig 01:03, 31 March 2006 (UTC)
Refactored post no longer relevant
Regarding etymologies, it's important to realize that, for words of ancient Japanese origin (i.e. 和語), the kanji are purely secondary, something applied later. This is why some words have so many alternate kanji renderings, as the kanji were essentially "translations" of the Japanese concept into Chinese. As with much translation between highly dissimilar languages, one concept in the source language often matches multiple concepts in the target. One must be careful therefore not to be confused by alternate "spellings" of what might essentially be the same word, as we see with つく for instance.
For 漢語, however, the kanji are integral to the meaning, as the words themselves either came from Chinese or were coined along Chinese lines (c.f. 社会, 共和国, and the like).
With this in mind, my ¥2 contribution is that it would seem to make sense to have the full entry on the kanji page for on'yomi words, and on the hiragana page for kun'yomi words. Any other views out there? Cheers, Eiríkr Útlendi | Tala við mig 06:48, 31 March 2006 (UTC)

Good point, Eiríkr Útlendi. I agree with putting the main content for 和語 terms on the kana page and the main content for 漢語 terms on the kanji-compound page. Doing so reflects etymology and minimizes redundancy. So, following is the updated proposal for Japanese entry organization:

The kanji for words of ancient Japanese origin (i.e. 和語, wago, read as kun'yomi) were transliterations of the ancient Japanese terms into hanzi (kanji), so some 和語 (wago) have many alternate kanji renderings. Likewise, the kana for Japanese words of Chinese origin (i.e. 漢語, kango, read as on'yomi) are transliterations of the Chinese terms. In order to maximizing usability and minimize homonym redundancy, the full entry for each 漢語 (kango, read as on'yomi) should be on its kanji or kanji-compound page, with links to and from brief kana and romaji entries. Likewise, the full entry for each 和語 (wago, read as kun'yomi) should be on its hiragana page, with links to and from brief kanji-compound and romaji entries.

If there are no objections, I will update Wiktionary:About Japanese to reflect the above organization guidelines. Rodasmith 05:34, 5 April 2006 (UTC)

Would "putting the main content for 和語 terms on the kana page" minimize redundancy? 掛ける, 欠ける, 賭ける and 駆ける are different words, aren't they? They are colliding under the policy.
I'm one who took part in establishing the previous-generation practices now on the content page. What we thought was to scatter main entries by the most common written representations of the words. With such approach, 4 words mentioned above will have their own main entries. The minor variant 懸ける will belong to the common representation 掛ける, and 駈ける will belong to 駆ける likewise.
When it comes to つく, I think 着く, 付く and 就く can be separated into different pages as they are treated as different words in modern Japanese while they are certainly cognate/akin. 衝く would belong to the more common form 突く as they are interchangeable.
I'd like to consider approaches like this as the options to deal with the homophone-richness of Japanese. --Tohru 08:38, 9 April 2006 (UTC)
Part of the trouble appears to me to be that Japanese has not just many homophones, but many "words" that are in fact different shades of meaning of the same underlying word, as we see very clearly with 突く and 衝く, and less so with 着く, 付く and 就く.
One of my concerns is that focusing too much on the kanji obscures the underlying relationships between different words, and can thus impede learning. If I hear つく, but I don't know the word, which one do I look up? If つく only has a collection of links to 着く, 付く, 就く etc., it takes much more time and effort to find the one I want -- and that's assuming I don't get confused along the way and choose the wrong meaning, simply because it sort of fits and I can't be bothered to look through the other five entries, all on separate pages. This is even more of a problem in the significant portion of the world that does not have broadband.
Basically, from a useability standpoint, splitting out the meanings to separate pages for each "spelling" becomes quite cumbersome.
Whereas, if we split the meanings for each "spelling", but keep them on one page, we can make it clear that each is distinct without introducing unnecessary hurdles or obscuring the interrelationships, and also make clear which kanji belong to which meaning. Using つく as an example, here is an adapted excerpt from Shogakukan's 1988 edition of its Japanese-Japanese 国語大辞典 (my own rough and ready translations, and mostly my own sample sentences):
== Verb ==
(I) To become close to or inseparable from another thing. To stick to.
1. To contact, or to come close as if to contact.
Shushō ni wa goei ga tsuite ita.
The bodyguards kept close to the prime minister.
2. To stick to something and not come off. To cling or adhere.
Doro ga amagappa ni tsuita.
Mud got on / stuck to the raincoat.
付く, (other kanji)
(II) To obey or follow another person or thing.
1. To be sympathetic to or in love with someone. To obey someone, to learn from someone.
Yūmei na sensei ni tsuite mokuchō wo benkyō suru.
I will study woodcarving under a famous teacher.
2. To follow after. To stay with someone or something, to attend or accompany.
就く, (other kanji)
== Noun ==
(I) The moon.
1. Ancient eastern Japanese dialect version of つき tsuki, "the moon".
2. [Lexeme] An alternate or archaic form of つき tsuki, used before other nouns. Ex: 「つく夜」 tsuku yo "moon(lit) night", 「つくよみ」 tsuku yomi "the moon god" (lit., "moon reading"; originally meant a ceremony for "reading" the moon to divine how far until the next full or new moon).
If needed, the kanji list could go after each Arabic-numbered definition, rather than at the end of the major Roman-numbered definition group. Later in the same つく entry, Shogakukan uses the following format. Here they make it very clear which kanji goes with which meaning.
(5) ある地位、場所などに身を置く。
1 (着)進んで行ってある場所に至る。「成田に着く」*源氏‐須磨「申の時ばかりにかの浦につき給ひぬ」
2 (即)ある地位、特に、帝位にのぼる。*源氏‐若菜下「内のみかど、御くらゐにつかせ給ひて」
3 (着)座をしめる。ある場所にすわる。「会談の席につく」*枕‐一六一「上達部のつき給ふ倚子(いし)などに」
4 (就)ある役目や任務を負う。「任(職)につく」*今鏡‐一〇「出雲にて失せ給にし大将殿のつき給へりける年とかや」
I agree that Japanese poses an interesting organizational challenge with the many different forms of the words, but I feel rather strongly that splitting 和語 definitions across multiple pages is not the best way to go from a useability standpoint. If I look up homonyms in an English dictionary, for instance, I expect to find them all on the same page, ideally even in the same entry, such as moon, or record, with both noun and verb meanings. Splitting them across multiple entries works better on paper, when you can view the whole page with the words listed alphabetically one after the other, but it does not work so well when using a more visually limited web browser as our viewing portal, where we can only see one entry at a time.
I apologize for the length of this post, but I had a lot of ground to cover. I hope these examples provide food for thought. Thank you, Eiríkr Útlendi | Tala við mig 19:59, 10 April 2006 (UTC)

I think all the Japanese (except linguists) doesn't mind whether speaking word is 漢語 or 和語. I see that it is very annoying when you are adding a Japanese word, but the definition of a word doesn't matter whether you read kun'yomi or on'yomi for most of Japanese words. Therefore, no Japanese-Japanese dictionary mentions whether it is 漢語 or 和語, so I think this kind of separating is useless. If someone has editted the Hiragana page in proper format, you can see simple definitions in one sight. It may be a special ability which Japanese have, to read Hiragana in a right definition easily. However, also for Japanese, it is easy to understand Kanji words rather than Hiragana words, for every Kanji words usually have single definition. --M. Powel 00:44, 29 April 2006 (UTC)

Reading the above, I think that M. Powel is mostly correct. 漢語 and 和語, while useful information, is not relevant enough to dictate policy. Secondly, the difference between 漢語 and 和語 is usually pretty obvious to people who know enough Japanese to care at all about the difference. Thirdly, there are a lot of edge cases that would have to be defined. But the first point really covers it.
As for the organization of these articles, well, I'm semi-retired from working on this, so my opinions not very important, but I feel that the most logical approach is to have unique articles for each "term" (which of course would have to be determined on a case-by-case basis, involving precedents by other dictionaries). Each "term" would be written in kanji, if there is a single standard way of writing it, will all the variants prominently displayed. If the term is almost exclusively written in kana, that should be the article page, with all of the kanji (and variant spellings) prominently displayed. If the term has more than one way of being commonly written, which is common for onyomi words (e.g., 柔らかい and 軟らかい), then it should be at the article with kana and kanji listed in parentheses (in order of commonness, perhaps?). Ex.: きる (切る, 斬る). Of course 斬る and 切る would redirect to this article. This would best match how Japanese dictionaries are formatted, as far as I've seen, and it resembles the Wikipedia's approach to dealing with ambiguous entry names.
Following Japanese etymologies is hard, but I think it would be valuable information if anyone can dig them up. For example, I strongly suspect that 初めて (first) and 始める (to begin) are closely related words. Much less certain is 捨てる (to throw away) and 廃る (to go obsolete, to fall into ruination), but since they have similar meanings and follow the Japanese pattern of e<->a transitive/intransitive transformations, I suspect there is some etymological connection.
Following Chinese etymologies is easier--I know of a number of dictionaries that include this information (particularly Kanji dictionaries). While the amount of effort involved is vast and the usefulness limited, it would be nice to have the practice identified in our styleguide.
Another thing I'd like to see that I haven't so far is the distinction between the various Chinese readings of Japanese characters. When I was involved here last, the kanji entries were a mess, but because they were being closely connected to the Chinese and Korean entries for the characters by people that were doing a lot of work and had some ideas of their own, I left them alone. What I am speaking of is 漢音, 呉音, and so forth.
Lastly, but not least, it would be nice to have a page that identified a lot of things that people want from a Japanese-English dictionary that have been brought up in the talks, so that anyone working on policy or reformatting entries can look through the list and try to hit them all as best they can. Many people use dictionaries for quite different purposes (etymology, pronunciation, example usages, learning a language, etc.) and come from very different backgrounds. This is even more true for foreign-language dictionaries. It's important that we try to address as many of these backgrounds and purposes as possible, which means making it easy to look up words by romaji, kana, kanji, etc. We should also include etymology, variant pronunciations, regional information, example sentences, usage information, etc. An example of useful information that you don't often see in any dictionary is how much a term is used in speech vs. writing as well as what sorts of people use the term. How many dictionaries would be so helpful as to tell me that ambulate is a term only frequently found in relation to medicine, and is otherwise uncommon in daily conversation (used in literature occasionally) and probably unknown by 70% of the population? Yet if I'm learning a language, this is extremely useful information. Obviously coming up with precise data is going to be tough, but we can at least start with impressions (which is basically what we're doing with the rest of this dictionary), which will be useful. I'm starting to wander off into things that are not specific to Japanese dictionaries, and should probably find another soapbox for it. Incidentally, I think example sentences and citations/quotations are both very useful and serve quite different purposes. For Japanese we should have both, though I'm not sure how they should be handled. Citations are useful, because they give a sense of broadness and context to the term (i.e., that Basho, Murakami, and NHK used a term tells us something about how old and still-used it is). Example sentences are useful because they can be optimized to show different ways in which the term is usable. (end rambling)
Jun-Dai 19:37, 22 June 2006 (UTC)

Particle positions[edit]

Your input is requested regarding whether and how to document positions of Japanese particles. The discussion is at Category talk:Japanese particles#Particle positions. Rodasmith 06:10, 26 April 2006 (UTC)

Keiyōdōshi declension[edit]

Please see Template talk:ja-na, where I propose to change the structure of {{ja-na}} to show the declension of 形容動詞 (keiyōdōshi) more accurately. Rod (☎ Smith) 06:12, 11 May 2006 (UTC)

janoun template[edit]

Can I use janoun template for explanation of reading for other parts of speech (verbs, adjectives, etc.)? --M. Powel 12:05, 15 May 2006 (UTC)

It would be better to use a separate template for each part of speech. See Category:Japanese inflection templates. Verbs and "quasi-adjectives" (形容動詞) are already there. What should we call the template for i-adjectives (形容詞)? What else should be there? Adverbs (副詞)? Particles (助詞)? Rod (A. Smith) 18:59, 15 May 2006 (UTC)
We have a new Template:ja-noun, I will be working on the other parts of speech. We do need to get POS headings and templates into romaji and hiragana/katakana entries; there is no reason why there needs to be an exception to WT:ELE (or to WT:CFI for that matter... romaji and hiragana entries are fine if formatted by WT:ELE.) Robert Ullmann 14:33, 22 August 2006 (UTC)


I just fixed the article text -- the Conjunctions header listed a couple "words" that are neither single words nor really conjunctions, specifically だから (should be から, as だ is the copula) and それなのに (should be のに, as それ is "that", and な is a form of the copula). -- Eiríkr Útlendi | Tala við mig 18:00, 16 May 2006 (UTC)

added radical/stroke index to Template:kanji[edit]

Since the template adds the kanji to Category:Japanese kanji it needs the radical/stroke sort index as defined at the end of WT:AJ. See the document and the template. Entries without the index will sort under ? in the category to indicate that they need attention.

moving Category:romaji to Japanese romaji, same for hirigana[edit]

Category:Romaji is being moved to Category:Japanese romaji for consistancy with the other languages as well as with the other Japanese category names (e.g. Japanese kanji). Most of this is happening inside templates, so it doesn't require manual fixes.

Similarily, Category:Hiragana to Category:Japanese hiragana.

We should end up with the category references all within templates, so we can make whatever changes are desired later. (E.g. if desired, we can move them to a new name with a one-line change.)

Question: why shouldn't the individual hiragana characters be in the hiragana category? They would appear very conveniently at the beginning of each sort section. Very easy to find with the TOC. Robert Ullmann 12:31, 24 August 2006 (UTC)

I think some individual katakana entries redirect to the corresponding hiragana. If that's still the case, it should be corrected and each individual kana entry should be categorized as you suggest. Rod (A. Smith) 14:25, 24 August 2006 (UTC)
Yes, I've noticed that. (Confused me a bit!) They need their own entries. Robert Ullmann 15:53, 24 August 2006 (UTC)

Template:ja-verb and Template:ja-adj[edit]

We now have proper templates for each as with ja-noun.

In addition to the common parameters, ja-verb takes a type= (1, 2, or 3) parameter to put the entry in the correct category. The ja-adj template takes a q= (い or な) parameter, again to categorize properly. Robert Ullmann 15:53, 24 August 2006 (UTC)

Templates for hiragana and romaji modified to present in the same style. Robert Ullmann 21:25, 24 August 2006 (UTC)

non-POS headers[edit]

E.g. ===Romaji===, ===Hiragana=== (not to mention Furigana) and Kanji when not for a single kanji. Can we just remove these? They aren't POS headers even though they are pretending to be.

Just use ==Japanese== and then the romaji or hirigana templates, or the proper POS headers and the POS templates? Robert Ullmann 21:47, 24 August 2006 (UTC)

We now have proper templates for all of the POS; I've edited to show how these are to be used, replacing the hiragana and romaji headers, and identifying the script forms. They also have the inflection information not in the separate conjugation/declension table templates. Robert Ullmann 15:40, 3 September 2006 (UTC)

suru form and quasi-adjectives[edit]

Edited to remove references to redirecting pages; the wiktionary does not redirect forms. Also each form is its own entry, with references as appropriate to the others. 平安な should not appear on the same page as 平安, it is a different entry.

(It should be noted that the wiktionary was redirecting forms in at least some cases at the time this was originally written; this is not a criticism of the author(s), quite the contrary!) Robert Ullmann 14:00, 26 September 2006 (UTC)


Four questions about romanization/transliteration:

  • The strict Japanese transliteration rules state that -oう where the う is the non-past indicative verb inflection should not be romanized as ō (e.g. in 誘う sasou), but that -iい must always be romanized as ī, even if it is the non-past indicative 形容詞 adjective inflection (e.g. in いい, 難しい). Was this an oversight or is this intentional?
    As you might already know, domestic standards about Japanese romanization such as American ANSI Z39.11-1972 [1] and British BS 4812 : 1972 [2] say that a macron should not be used spanning a morpheme boundary. If it is the rule to obey, 難しい needs to be transliterated as muzukashii just like it is now in the article. The rule don't connect morphemes with macron sounds reasonable, so I believe we should employ it and keep いい and 難しい as they are. (I'll go through these items one by one, as writing in English is fun but considerably time-consuming for me :) --Tohru 03:06, 5 January 2007 (UTC)
    Ah, that seems like a good idea. So by that rule, for 場合 we should write baai, because 場 + 合 is ba + ai? Cynewulf 01:55, 21 January 2007 (UTC)
  • For 形容動詞 na-adjectives mainly, and similar things (-に, etc) by extension, does the standard romanization include a space between the root word and the type conversion particle? For example, is 静かな shizukana or shizuka na?
  • It is implied via wikipedia that ぢ (e.g., in ちぢれる) is ji, and づ (e.g., in つづく) is zu. Is this correct? Should ぢ be differentiated from じ ji, and づ from ず zu?

Cynewulf 09:14, 4 January 2007 (UTC)


Am I correct in understanding that redirects are inappropriate for ALL cases? Or is it SOME but not ALL cases? I mean, I know there are some fantastic examples out there where a redirect would be problematic. —Tokek 11:27, 2 February 2007 (UTC)

Right. On the English Wiktionary, all redirects are discouraged in the main namespace. The term esa is an example of why. --Connel MacKenzie 11:53, 2 February 2007 (UTC)
Thanks for the clarification. I know there are some fantastic examples, like esa, where redirects would be problematic for SOME (but not ALL) cases. --Tokek 08:25, 3 February 2007 (UTC)

Kanji Entries[edit]

Template:Kanji has been deleted. How should kanji entires be formmated now? --Balloonguy 21:27, 25 March 2007 (UTC)

With template ja-kanji, a ====Readings==== headers, and template ja-readings. You do have a point, the doc needs to be updated. I'll get to it soon. (you have seen this in a lot of entries, right? ;-) Robert Ullmann 01:15, 26 March 2007 (UTC)


There are a couple things about the 形容動詞(-related) entries that I find bemusing.

Why is the -な form used as the main form (i.e. used in article names)? Isn't the basic form -だ, e.g. 静かだ? Japanese dictionaries such as 大辞林 and 広辞苑 only use 静か, but in many Japanese grammar discussions I've seen 静かだ used as the non-inflected form. Sometimes this results in such unsightly article names as 本当な - even the article itself(!) recommends against using this form. With -だ there wouldn't be such a problem.

Why are there both forms with -な and without as full articles, e.g. 静か and 静かな? This would make sense if 静か were a 名詞 and 静かな were a 形容動詞, but there is no such 名詞 as 静か (静か is linked to in this policy article so it's not by accident). The article openly admits this by describing a 形容動詞 only. It is also said that "Note that the "plain form", e.g. 平安, will always be a noun in English: peace." Thus 静か should also be a noun (名詞), yet it is not. Depending on how you understand grammar, you might not even consider 静か to be a word at all, but only a part of one (静かだ). -- Coffee2theorems 04:38, 30 April 2007 (UTC)

Come to think of it, why not have adjectives be in the dictionary form articles, like other dictionaries have? Then 本当 would have both noun and adjective sections and 本当な would be gone. The adverb 本当に would stay where it is, as it does not inflect. -- Coffee2theorems 12:34, 30 April 2007 (UTC)

The link to 静か is left over from an old version (and the entry should be fixed, or as you observe, removed, it is not really a word, it would be "quiet" in English as a noun, if it was a noun ...), it should be 平安な, see the more expansive section later.
Understand that the en.wikt has entries for all forms of a word used, not just for the "dictionary form". (This is is endlessly confusing to people who want to do things "the way Japanese dictionaries do"; but this is an English dictionary of Japanese ;-). And the Attributive form is the "normal" form for English, which is (to oversimplify a good bit) why these adjectives are described in English as な-na adjectives.
So as to the last comment, we don't use the "dictionary form"; entries are under all the inflections with the "main" entry under the chosen lemma form. Robert Ullmann 13:16, 30 April 2007 (UTC)
Yes, I understand that Wiktionary has articles for many forms that are not present in traditional dictionaries. However, it's not true that it has entries for all forms for a word used. There's only 食べる, no 食べられる, 食べれる, 食べない (! actually this one exists!), 食べられなかった, 食べられたくなかった, 食べられなくなった, 食べられたくて, 食べられながら, regardless of whether each of these is used much or not. These are regular and could all be added by a bot if it's deemed that such articles should exist. It doesn't make much sense unless the meaning differs from what's expected using the standard rules, though. Indeed, you can (and at least some grammars do) consider these to be the word 食べ that is used with various auxiliary words such as る, ない and ながら. And from WT:CFI, "An expression is “idiomatic” if its full meaning cannot be easily derived from the meaning of its separate components." The same for adjectives.
In any case, you pick one form (say, 静かな) and use that either as the main article or the only article. It is only the choice of this form that I was talking about. You could pick any of 静かな、静かだ、静か (or 静かに、 静かさ or maybe something else, but I shall ignore these as nobody is proposing them). Of these to me 静かだ looks like the best, because it is not an unnatural form for any such adjective, whereas both 静か and 本当な are strange. If you have many articles that point at a main article, it would be odd to have them all say that "this is a form of <link to word in a strange form>", and then the main article say that "the preferred form is actually <link to the preferred form>, but most of the explanation is in this article for consistency".
Of the other two, 静か has the merit of being the form that is used by other dictionaries, which I assume to be based on some form of accepted grammar (as far as I know there are many grammars for Japanese). You can consider 静か to be a word that is mostly used with a form of the word だ (like you can consider 食べる to be 食べ+る). It's also not always used in the declensions given here, e.g. 静かね。 is OK but the ね is not a declension.
静かだ has the advantage of being 終止形 just like e.g. 食べる and 美しい, so it's consistent (consider e.g. 夕日が綺麗だ、夕日が美しい、夕日が見える). Of course 綺麗な (連体形) is consistent with the others too, as they are also 連体形 (consider e.g. 綺麗な夕日、美しい夕日、見える夕日), but the 本当な problem persists and besides, if you have to pick between two forms the more self-contained (instead of being just a modifier) sounds better for an article title ("The sunset is beautiful."-beautiful or "Beautiful!"-beautiful rather than "Beautiful sunset."-beautiful).
The article says "Thus, the parts of speech should adhere to the common Japanese way of teaching grammar in modern 国語 kokugo texts, the method with the widest agreement among Japanese scholars." I don't have any such texts at hand, but at least jp Wikipedia says that 綺麗だ is a 形容動詞: "さらに、形容動詞の「綺麗だ」には、[...]", which reinforces my understanding that the -だ form is indeed the basic form in school (国語) grammar. Unfortunately I don't have any such texts, so I can't verify this. I'm pretty sure they don't use 綺麗な as the basic form, however. -- Coffee2theorems 18:08, 1 May 2007 (UTC)

Distinguishing between hiragana and katakana[edit]

Why is a distinction made between hiragana and katakana in templates etc.? Because of this e.g. the rōmaji article refers to a non-existent "ろうまじ" article. This kana usage is non-standard to say the least. If you read by kana in a text that has furigana, what you get is ローマじ. This is also what's used by dictionaries such as 大辞林 and 広辞苑. -- Coffee2theorems 11:28, 30 April 2007 (UTC)

It makes the distinction precisely for this reason. But that entry didn't use the template quite as well as it could, fixed. There should be an entry at ろうまじ, it is written that way sometimes. Robert Ullmann 13:22, 30 April 2007 (UTC)
What do you mean by "this reason"? I don't understand your change of the rōmaji article to say that ローマ字 is katakana either, as the last character is not. While it looks like ろうまじ will indeed have the three usages in permanent media that are required by WT:CFI, I doubt this is true of all words the template is/will be used for. FWIW, I'll mark it non-standard like jist is due to its extreme rarity. -- Coffee2theorems 15:36, 1 May 2007 (UTC)
Other things that are essentially never written in hiragana use kata= and hidx= for the indexing. Robert Ullmann 15:45, 1 May 2007 (UTC)
I'm afraid I don't understand how that works. Could you add a clear explanation to Template talk:ja-noun or is it explained somewhere else? (I could add a link on that page, it was the first place where I went looking) In any case, it would be clearer to change the visible expansion of the template which claims that ローマ字 is katakana, because it's obviously not. Perhaps "usual spelling"? Or "non-hiragana"? Or "katakana and kanji" (how about numbers in things like 911?). Or something else which describes whatever is meant. -- Coffee2theorems 16:04, 1 May 2007 (UTC)


How should numbers be formatted?--Balloonguy 21:51, 1 May 2007 (UTC)


For the compounds section for single kanji entries, should compounds be listed as:

  • compound linked

or should it be

  • compound linked (hiragana linked, romaji linked)?--Balloonguy 21:53, 1 May 2007 (UTC)
As far as I know, the first style is the usual, original practice though there would be no explicit standard about it. Gradually it is evolving into the second one; as you can imagine, after someone once added hiragana and romaji information to a compound link, no one will dare to delete it. The first one is sufficient, while you can choose the second style if you like. --Tohru 05:20, 6 May 2007 (UTC)
When I add Chinese, Japanese, or Korean compounds, I always try too add the romanization (and kana for Japanese) for the same reason I always try to include genders in the synonyms, related, derived, and see also sections for european languages. This stuff is intrinsic to the words and a user would often have to look it up anyway. I know I do when I use print dictionaries! — Hippietrail 22:47, 6 May 2007 (UTC)
There does need to be some sort of standard. A lot of kanji pages have the second one, plus a short definition. You get more information, but it clutters the page up. I brought this up in a roundabout way in the beer parlour at Wiktionary:Beer parlour#Japanese Proper Names on Kanji Pages. The point was made that each page is for the entry itself, not every other entry it's part of. The only place on Wiktionary that talks about this is at Wiktionary:About Japanese#Compounds and the section directly under it , Kana, Romaji, English translation. The Compounds sections says "list every compound," and the section under it says every Japanese term should be given as kanji (formal), kana, romaji, and English (hence the title). The question is whether this rule applies to every example of every Japanese word in every article, or to just each entry. And even if it does, it may not necessarily be the best format. Personally, I do favor having the compound then kana, romaji, and a short definition. However, some people have expressed feeling against this, as it does add clutter and affect readability of the page. Thoughts anyone?--Hikui87 02:21, 1 June 2007 (UTC)

Kanji Readings[edit]

Shouldn't the on-readings for the individual kanji be in katakana? Most dictionaries I use do this or make the distinction in some way. I'm not saying the hiragana for compounds should be changed, just the list of on-readings. For example, 大, on-reading: ダイ; compound: 大学 (だいがく).


I'm now considering converting Template:japdef to something named in the proper and correct manner. Apparently, the language should be represented with "ja" or at least "jpn", not with "jap"... "ja-def" would be just fine. At the same time, I'd like to change the way in which the template is applied into one like the following, allowing several spelling variants to be put in a line and excluding the explanation from the template arguments:

 *{{ja-def|勝ち|勝}} a [[victory]]
 *{{ja-def|価値}} [[value]]
 *{{ja-def|徒歩}} [[walking]]

And the result will look like this one as usual. What do you think about this change? After consensus reached, or no one is opposing, I will put it into effect with the standard pywikipedia framework, maybe without requesting botflag. The number of the entries to be handled is about 1,200. --Tohru 10:33, 23 July 2007 (UTC)

Sounds fine to me. The name isn't that important, but moving the definition text outside/after the template is a good idea. If you run the automated replacement, make sure you restrict it to the main namespace ( -namespace:0 in replace.py), people forget this, and it is annoying to have talk and user pages modified. Robert Ullmann 05:48, 27 July 2007 (UTC)
Thank you for the feedback. I will watch it out in the preparation of a wikilink list that will be feeded to replace.py. --Tohru 05:56, 27 July 2007 (UTC)
-ref:Template:japdef should be all you need.
python replace.py -ref:Template:japdef -namespace:0 -regex "{{japdef\\|(.*?)\\|(.*?)}}" "{{ja-def|\\1}} \\2"
is pretty close I think. (assuming I got the right level of escapes on the |'s ;-) Try it, then run it with -always if it is working okay ... Robert Ullmann 06:07, 27 July 2007 (UTC)
Thanks again! I will employ the ref option. I've already tested the regex on my local Wikt environment on Fedora Core. Single \ should be ok for both escaping the pipe and referencing groups :). --Tohru 06:34, 27 July 2007 (UTC)
Almost one week passed after I posted this first, and there seems to be no objection. I'll start the test run on several entries soon... --Tohru 08:20, 30 July 2007 (UTC)
Done without any particular problems. The diffs are here: [3], [4], [5], [6], and [7]. --Tohru 09:07, 30 July 2007 (UTC)

Example sentence format[edit]

As noted here, the recommendation in WT:AJ to format example sentences with a leading bullet ("#* 例...") was developed before the modern WT:ELE. We should probably align that recommendation with WT:ELE (i.e. "#: 例..."). That means we must also remove the recommendation to provide brief usage notes as "#: note...", and usage notes should be moved into the standard "====Usage notes====" section. To do all this, it seems we need a vote to approve the change, a bot to move usage notes into "====Usage notes====", and a bot to convert existing examples to the new format. Thoughts? Rod (A. Smith) 04:45, 24 July 2007 (UTC)

The following is from Wiktionary:Votes/2007-08/Layout of example sentences for Japanese entries:

Sorry for not saying anything before; I didn't read this vote page until now. :-/   The problems I see are: (1) the lack of mention of quotations; (2) the insistence on a hiragana-only version rather than a kanji-free version (I don't actually know Japanese, but if an example sentence contains, say, katakana, or Hindu-Arabic numerals, I don't see why those would need to be translated into hiragana); (3) the inconsistency between the examples — some put quotation marks around the English translation, and some don't. (Also, it would be nice for there to be some guidance on example sentences themselves, particularly as regards choice of scripts, but that's not a sticking-point for me.) —RuakhTALK 03:20, 18 August 2007 (UTC)
Thanks for the input. I should have known better than to expect any vote to run without dissent. ;-) Seriously, though, your points are valid and this is an excellent opportunity to address them. I don't have the energy right now, so for now I will just freeze this vote to give me time to write up the new text. I will also copy this conversation to Wiktionary talk:About Japanese#Example sentence format, where we can continue this conversation. Rod (A. Smith) 03:31, 18 August 2007 (UTC)

Please add any additional suggestions here. Rod (A. Smith) 03:43, 18 August 2007 (UTC)

Since each major change requires a vote, and since this latest vote is intended to align WT:AJ with WT:ELE, I drafted a new version at Wiktionary:About Japanese/proposal. Please help me refine that draft. Rod (A. Smith) 06:33, 18 August 2007 (UTC)

Any suggestions on a better wording for “hiragana version”? The idea is to translate all kanji from the example or quotation into hiragana, but to leave katakana, romaji, etc. as is. Too bad we cannot use Ruby here and just get rid of that line. Or can we? Rod (A. Smith) 03:43, 18 August 2007 (UTC)

I slightly prefer the way Ruakh suggested. It sounds like a good idea to always present an example sentence with the variant in the language's most fundamental script. For example, telling how to pronounce "36" like in the following would be helpful for many people though not essential.
Sanjū-rokkai renzoku de jettokōsutā ni norimashita.
I tried the roller-coaster thirty-six times consecutively.
While it is apparent that Ruby can be a useful tool here, we still need more discussions and experiments about it to get consensus on the definite formatting conventions. I think we should stick to the original plan and put off introducing Ruby for now, keeping an open attitude to furhter experiments.
And about Ruakh's first point: Quotations in Japanese pages are quite rare at the moment, and so it seems too early to formulate the local policy. Maybe we'd better simply exclude it from the scope of this amendment. ―Tohru 14:02, 18 August 2007 (UTC)
Your note about using hiragana as a canonical transcription makes sense. (Note, though, that it actually Ruakh's initial suggestion to change the hiragana line.) So, I removed the strike-out around “hiragana” in Wiktionary:About Japanese/proposal and invited Ruakh to reply here. Regarding your uncertainty about quotations, I don't know of any reason to deviate from the current standard from WT:ELE. If we exclude the quotation format, it would seem to imply that we do not want quotations in Japanese entries. Do you think the quotation format proposed at Wiktionary:About Japanese/proposal is questionable?
By the way, to see the full set of changes I am now proposing, see [8] and any subsequent difference from [9]. Rod (A. Smith) 17:46, 18 August 2007 (UTC)
On second thought, I realized that there is no uncertainty left about quotations in effect... And on a different matter, I revised the corresponding part of the draft importing a few elements from WT:QUOTE. ―Tohru 08:14, 19 August 2007 (UTC)
Re: hiragana-only vs. kanji-free: I'll definitely defer on this to editors who actually speak the language. To me it seemed that the point of the hiragana transcription was simply to address the problems inherent in using kanji; for comparison (choosing French, a language that I do know and that y'all can probably catch the drift of), it seems silly to me to have:
Le premier dauphin naquit en 1095.
Le premier dauphin naquit en mille quatre vingt quinze.
The first dauphin was born in 1095.
but if y'all disagree, or if there's something that makes Japanese different in this regard, I'm quite willing to give up on this point.
By the way, another thing: in English example sentences, we bold the occurrence of the headword (or form thereof), and in Hebrew example sentences I've been doing this in both the Hebrew and the English translation (in part because a single Hebrew word might correspond to several words in the English translation — e.g., Template:HEchar (uvayom) "and on the day"); I don't know if bolding is feasible in Japanese (I seem to recall reading that the Japanese equivalent of bolding or italics is simply to use katakana instead of hiragana, which obviously wouldn't serve our purpose), but if so, I think you should use it, and if not, I think you should find an alternative that serves the same purpose.
RuakhTALK 19:17, 18 August 2007 (UTC)
As for transliterating numbers into hiragana, Japanese numbers are read differently depending on context, so it can be helpful to show the hiragana as a guide. As for how to embolden the headword in the example, katakana seems strange to me for that purpose, but considering my poor Japanese proficiency, I'll defer to Tohru. Rod (A. Smith) 19:40, 18 August 2007 (UTC)
I personally don't see any problem with embolding the headword in the example sentence regardless of the script, while we're usually avoiding embolding kanji especially in inflection lines. (I edited the example above.) Sometimes readers might not be able to see the details of a boldfaced kanji of a large sroke-number, but the kanjitab and inflection line can serve for the purpose and complement it. ―Tohru 07:26, 19 August 2007 (UTC)
OK. (unindenting...)

Thanks for the help with this update. As you probably have noticed, the proposed changes include the following:

  • The format given for examples and quotations is now matches that of WT:ELE and WT:QUOTE.
  • The introduction now contains the text, “This policy explains considerations for Japanese entries that are not covered by WT:ELE and other general policies.”
  • Entry layout headings now match those of WT:ELE, e.g. “===A very simple example===”, “===Headings before the definitions===”, and “===The article core===”.
  • The new section “==Lemma entries==” describes lemma entries in a fairly non-restrictive way.
  • The “==== Etymology ====” section now has an example for entries with multiple etymologies.
  • “== Hiragana and Romaji ==” is now called “== Non-lemma forms ==”
  • The POS headlines are now described as valid at either L3 or L4, in line with how we format entries with multiple etymologies.
  • Formatting examples for English translations now are enclosed in directed quotes (i.e. “English translation”)
    Note: This change reflects recent thoughts I have had for consistently differentiating between English text, non-roman script examples, transliterations, and translations. However, this convention has not actually been discussed by the community, so if there is reasonable opposition to it, I would be happy to remove it in order to reduce obstacles to this update.
  • Minor tweaks were also made (e.g. adding transliteration for non-roman terms, updating links, linking to templates using {{temp}}, etc.).

Unless objections arise for the proposed changes, I will open the vote soon. Rod (A. Smith) 20:02, 19 August 2007 (UTC)

Looks fine as far as example sentences go, although I don't really feel the need to put translations in quotes -- they're indented from the definitions, and the romaji will be in italics.
As for various other things (since the changes aren't strictly limited to examples), I'd like to raise an issue about the placement of kanjitab -- I've come to prefer having it always be just below the ==Japanese== line, whether there's an Etymology section there or not. The component kanji don't necessarily have anything to do with the etymology -- not all words are Sino-Japanese compounds.
Some things I'd like to see clarified: It's implied that 平安な is romanized as "heian na". How are we supposed to determine whether to put a space between a word and an adjoining particle? Is "捨てる" "sute ru"?
As another clarification, I'd like it explicitly stated whether the romaji version of (say) 希望 goes at kibo or kibō (or both) -- and if not kibō then all the templates and their usages need to be changed. Though these things probably belong in a different vote... Cynewulf 22:39, 19 August 2007 (UTC)
Thank you for that good feedback, Cynewulf. We should certainly address those issues. In order to maximize the speed and likelihood of approving this much overdue alignment of WT:AJ with WT:ELE and other existing practices, I dropped the recommendation to designate translations with quote marks. (I actually hope to introduce that standard much more broadly, especially in places that use non-English terms followed inline by translations, e.g. etymology sections, derived terms, and single-line translations. This, however, is not the appropriate place for me to try to introduce a new convention.) I don't have any opinion for or against moving {{ja-kanjitab}} out of the etymology section or about potential updates to Wiktionary:About Japanese/Transliteration, but as you say, it would be best to split those into another discussion and corresponding vote. Does the current version now seem ready for a vote? Rod (A. Smith) 23:19, 19 August 2007 (UTC)
Yeah, it's probably better to do things one at a time (I'm not necessarily hostile to quotes, just don't see the need right now). I just thought of another thing: can we add "words in example sentences may be wikilinked"? I really like what User:A-cai is doing, adding quotes with everything wikilinked (see 始終), and would like to experiment with doing the same thing in Japanese. People who don't know much Japanese might have difficulty breaking sentences into words to look up, and are probably going to want to look up words in examples/quotes unless we make them outrageously simple. Or we can leave this for later too. Cynewulf 23:38, 19 August 2007 (UTC)
Hmm. Maybe it needs clarification for determining the lemma of 一寸/ちょっと, where it technically does have a jōyō kanji form but nobody uses it. Other than this one thing, looks good to go. Cynewulf 23:38, 19 August 2007 (UTC)
(er, it has a form that contains only jōyō kanji, but the form isn't 常用 (commonly used), I have now given myself a headache) Cynewulf 23:57, 19 August 2007 (UTC)
LOL. Fair enough. Should the lemma section just say that the jōyō set is a good rule of thumb or is there a better way to describe the appropriate lemma form? Rod (A. Smith) 05:51, 20 August 2007 (UTC)
It's hard to cover all possibilities. I think just "The kanji form is usually the lemma. When there is no kanji (as with つるつる), or it is rarely used compared to another form (as with ちょっと), the most common form will be the lemma. When the situation is unclear, editors are advised to use their best judgment on a case by case basis." -- give some examples, and leave the rest to common sense. Feel free to rewrite this. Cynewulf 19:59, 20 August 2007 (UTC)
Done. Does it seem ready for a vote yet? Rod (A. Smith) 22:05, 20 August 2007 (UTC)
See current version, proposed changes, and proposed version. Rod (A. Smith) 22:17, 20 August 2007 (UTC)
Looks like completely ready for a vote. Though there are some more possible modifications mentioned above which I can fully support, it's probably better to handle them in the next batch as already noted. ―Tohru 01:41, 22 August 2007 (UTC)

Reciprocal (for verbs)[edit]

I removed the following recent addition to WT:AJ in the interest of due process. Should we have a WT:VOTE to discuss the proposed addition?

  • If a verb is part of a reciprocal pair, where two verbs share the same kanji but have different endings to determine activeness/passiveness- as for example 見える and 見る- you may indicate here its reciprocal (linked).

Rod (A. Smith) 15:50, 25 July 2007 (UTC)

There is no regular rule, thus it's very useful to be able to see the reciprocals in wiktionary. It's also a way for us to compete against other japanese-english dictionaries, who do not list the reciprocals. Now, a common rejoinder might be, "why not put the reciprocal under 'see also'?" The situation is analogous to the synonyms/antonyms headers in English. We could just as well list synonyms under 'see also', but we don't. Because there's a specific, concrete, and consistent association involved. See also is very vague. Language Lover 00:42, 26 July 2007 (UTC)

We are trying to sort out transitive-intransitive pairs like this [10], aren't we? If so, this looks like where "related terms" header is applicable due to the strong etymologycal connection of the pairs, though exactly speaking those intransitive verbs can be considered as derived terms of the corresponding transitive verbs. I think it is a good idea to neglect this dependency so as to allow cross-referencing. Also, giving a label that explains the relation would be desirable. Please see 脱ぐ and 脱げる for illustration. ―Tohru 04:24, 27 July 2007 (UTC)
Those are some nice examples, the only problem is the same I said about "see also". If we put OTHER things in "related terms", then it's not so clear what the reciprocal is. Eg (yes, this example is contrived, I'm just making up some other things that someone might list under related terms):
====Related terms====
Language Lover 04:35, 27 July 2007 (UTC)
As 脱げた and 脱げます are just verb forms, they would not be put in under the related term header. Are there any good examples of other related terms? I suppose they are not so plentiful if exist. the example was said as "contrived" so my saying was off the point. --05:36, 27 July 2007 (UTC)
I put in some slightly less contrived examples
By the way, I will appreciate if you could provide some pointers to good references in which the term "reciprocal" is being used in this context? I'm wondering if this term is common and appropriate enough for us to use in this purpose. I'm searching on the Web and Japanese textbooks in a big bookstore for the attestation in this two days but still not successful... ―Tohru 05:07, 27 July 2007 (UTC)
I made the term up, and am quite open to better suggestions :) Language Lover 06:16, 27 July 2007 (UTC)
The term is used, but uncommonly, see [11] for example, but does not mean what Language Lover thinks it does. It has its usual meaning: a reciprocal of (e.g.) "close(s)", would be "is closed by". This should be Related terms anyway. In any case, all of Language Lover's examples are wrong. Robert Ullmann 05:28, 27 July 2007 (UTC)
Thanks, that's a pretty cool paper. It's talking about something totally different though, not really related. It's talking more about the "eachother" type of reciprocal. Language Lover 06:16, 27 July 2007 (UTC)
From the examples provided so far, this seems similar to etymologically related causative (사동사, sadongsa)/passive (피동사, pidongsa) verb pairs in Korean. Although some common suffixes are used to form such related terms, which suffix to use for any given verb is unpredictable, so it is helpful for Korean dictionaries to list those related terms. Similarly, many slavic languages (e.g. Russian and Polish) have perfective and imperfective verb pairs, one of which is often formed by prepending one of a few not necessarily predictable prefixes. Rather than add new headings for ====Perfective====, ====Imperfective====, ====Causative====, ====Passive====, ====Reciprocal====, and whatever other related sets of words occur in other languages, perhaps we should just consistently apply tags like "(causative)", "(passive)", "(perfective)", etc. to such terms within the ====Related terms==== section. Rod (A. Smith) 05:53, 27 July 2007 (UTC)

To both Rod and Tohru: don't be confused by looking at WT:ELE's definition of "related terms", it is overruled by WT:AJ's definition of same, which is much more threadbare and itself needs revision. Language Lover 06:16, 27 July 2007 (UTC)

Pitch accent[edit]

How are we supposed to indicate it in pronunciations? --Ptcamn 17:52, 19 November 2007 (UTC)

ja.wikt uses diagonal arrows ↘ for the pitch contours (see e.g. ja:日本), en.pedia uses vertical ↑↓ in w:Japanese_phonology. I'd be fine with either. I've never seen a source for this, and for the few cases that I've been able to work out, I just marked vowels as unvoiced (due to low-pitch vowel between consonants): 捨てる. Do we need to mark anything other than the falling pitch contour? Cynewulf 02:38, 20 November 2007 (UTC)
For standard Japanese, the fall in pitch is all you need to mark, but some other dialects have more complex patterns. Osaka Japanese for example has accentless high-tone words, accentless low-tone words, high-tone words with a falling acent and low-tone words with a raised accent. --Ptcamn 13:41, 20 November 2007 (UTC)
For standard Japanese you can use ˺ (U+02FA, MODIFIER LETTER END HIGH TONE) and ˼ (U+02FC, MODIFIER LETTER END LOW TONE) which are AFAIK more suited for this case. IHMO, にほ˺ん is easier to read than にほ↘ん --Aleχ 06:54, 15 February 2011 (UTC)
How about using a template? Personally I like the number system. The syntax is easy, copy-paste friendly and allows for a link with an explanation. E.g. "にほん [2]" The "overline system" allows you to see at a glance were the down-step happens, which I think is valuable too. I'm not sure how to render this in wiktionary though, but this could be added to the template eventually. Showing both would be nice.
As to what's used: 新明解 (Shinmeikai) uses the number system (not the backwards one). It seems that Chinese students of Japanese unlike western ones actually are taught pitch accent, and have an easier time hearing it (since Chinese is tonal I guess). I've heard they typically use the number system, with words marked in the vocab list. The NHK accent dictionary uses a line above (overlines?) the "high" kanas, that ends with a downwards tip where the downstep occurs (if any). The downstep is a bit similar to that Unicode "˺ (U+02FA, MODIFIER LETTER END HIGH TONE)" mentioned above. Japanese wiktionary is expectedly somewhat inconsistent, but mostly uses e.g. "コ↗ーヒ↘ー". A few pages uses "く↗び→", "わ↑かやま↓けん", or "第2モーラのみ高(_ ̄__ ____ _)", and quite a few mark it in IPA too. I haven't seen this notation used in other places.
I'm planning on batch-importing some of the pitch information from the Japanese wiktionary once I have cleaned it up and parsed it, and I'm wondering about which format I should use. A template would be a good idea, right? There are at least a couple of thousands entries with tone-information in Japanese wiktionary, though I'm unsure of how many I can use. --Vaste 09:44, 21 May 2011 (UTC)

Reading entries[edit]

How should something like れつ be handled? WT:AJ#Hiragana entries seems to say that there should be a "Hiragana" level 3 header, but then which template should be used for the inflection line since the readings are a mix of nouns and adjectives? Should this entry really split the kanji by POS and have "Noun" and "Adjective" sections (with ja-noun and ja-adj respectively)? Mike Dillon 03:31, 21 February 2008 (UTC)

See cleaned up entry. There's only one noun I'm aware of and no adjectives. If it's only readings, see にゃく for an example. "Hiragana" is no longer an acceptable header. Cynewulf 04:23, 21 February 2008 (UTC)
Hmm. I think that goes beyond the semi-informed changes I've been doing... I guess I'll just leave those or add {{ja-attention}}. Mike Dillon 04:48, 21 February 2008 (UTC)

Were these changes to きり accurate? Mike Dillon 05:07, 21 February 2008 (UTC)

Yes, from what I've seen you're handling things pretty well. If you're uncertain, tag it {{ja-attention}} and somebody will get to it. Cynewulf 20:25, 23 February 2008 (UTC)

Quasi-adjective categories[edit]

I've got another question for the Nipponophiles: Are the categories Category:Japanese quasi adjectives and Category:Japanese な-na adjectives redundant? Mike Dillon 04:26, 23 February 2008 (UTC)

Yes. I think we've had trouble determining the best name to use for these words. Rod (A. Smith) 04:37, 23 February 2008 (UTC)
Yes, just let {{ja-adj}} handle it. When we decide on a name, we can then change the template rather than ten million words. Cynewulf 20:23, 23 February 2008 (UTC)
Ok. Currently, the ones done by {{ja-adj}} with decl=な or decl=na end up in Category:Japanese な-na adjectives. The 22 remaining entries in Category:Japanese quasi adjectives are all manual. I've removed the three where the page name actually ends in "na". Since I don't know enough about Japanese, I've only been dealing with the ones that should have "な" added when there is also a "Noun" sense listed. In cases where all the content would be moved to the "な" page name, I don't have the knowledge of Japanese necessary to know what to leave in its place if there isn't already something there. Mike Dillon 21:33, 23 February 2008 (UTC)
I was working on removing the term "quasi-adjective", both as header and category; it only has a quasi-meaning (;-). The term is used for other unrelated languages to refer to things that are sort-of adjectives, but the various authors couldn't figure out what to call them. Not a well-defined term. If we get them all templated (ja-adj) and get rid of the header (should just be adjective), we can lose the Category:Japanese quasi adjectives category. Robert Ullmann 12:40, 24 February 2008 (UTC)
Note that this document specifies the use of "Adjective" as the header, and {ja-adj}, and does not list the "quasi adjectives" category. Robert Ullmann 12:44, 24 February 2008 (UTC)

Additional help[edit]

I have added an "Additional help" section to the end of WT:AJA. Please review and correct this section. --EncycloPetey 00:50, 17 March 2008 (UTC)

Radical characters[edit]

I've noticed that Category:Japanese kanji is using regular kanji in place of Unicode radical characters (e.g., 乙 U+4E59 instead of ⼄ U+2F04). I assume that's because most people don't have the right font(s) for displaying them (on this computer I'm on now, I see the second example character in my previous parenthetical remark as a "?")... right? - dcljr 01:00, 24 March 2008 (UTC)

This is very common; the Unicode radicals block is essentially unused; very nearly complete deprecation. The regular Han characters (Kanji, Kangxi Han, whatever) are used. The radicals block turns out to be well-intentioned, but useless. The lack of font support is effect, not cause. Robert Ullmann 22:58, 24 March 2008 (UTC)

Alternate forms[edit]

Shouldn't the header be "Alternative forms" to match the spelling used for all other languages? --EncycloPetey 12:15, 11 April 2008 (UTC)

It depends on what you consider a "spelling." If it applies to all written language, then sure. If it applies only to alphabets and not to syllabaries or collections of idiographs, then no. It does seem odd to say a Japanese word has a "spelling," especially since there is more than one way to write a word correctly (this applies to formal language, too). It is not really the same as saying harbour is an alternative spelling of harbor. --Hikui87 14:48, 11 April 2008 (UTC)
Pardon me, but EP is referring to the spelling of the header, and noting that "alternate" is incorrect (see alternate#Adjective, sense 3). It should be "Alternative forms". Fixed, I should have noticed that long ago. Robert Ullmann 14:58, 11 April 2008 (UTC)
Yeah, I misread that. Some people have been using "Alternative spellings," though, as in 腹切り.

'see' formating[edit]

The question was raised in a discussion on User_talk:Connel_MacKenzie#JA_formating whether the 'see' template should be used before or after the language headings. As of now, this template calls for 'see' after the language heading, but other languages require it at the very top of the page.--Hikui87 02:40, 23 May 2008 (UTC)

The documentation for {{see}} calls for its placement above the language headers, but I think the restriction is more a matter of convention than of policy. It's a good convention, though, so unless anyone thinks {{see}} should be allowed under language headers, WT:AJ should be changed (with a vote, of course). Rod (A. Smith) 16:09, 23 May 2008 (UTC)
I think it's a good idea to place it below the language header. Terms in kanji often belong to several other languages, but not all forms of the term. 安心 may belong to Japanese and Chinese, but 安心する can belong only to Japanese.--Hikui87 22:41, 28 May 2008 (UTC)

Kanji in headword lines or definition lines[edit]

I was under the mistaken assumption that WT:AJ called for kana entries to show the corresponding kanji in the headword lines, so I made this edit. I now see that the recommendation is for kanji to appear in definition lines instead. Is either format preferred? Rod (A. Smith) 20:41, 10 June 2008 (UTC)

The thing is there are tons of homonyms in Japanese and one hiragana entry can have lots of kanji with that reading Language Lover 23:00, 10 June 2008 (UTC)
Indeed. Likewise with Korean, but the solution we adopted there is consistent with the rest of the English Wiktionary. I.e., such homophones almost always have different etymologies (as in Japanese), so they get different etymology sections, and hence, different headword lines. Rod (A. Smith) 23:07, 10 June 2008 (UTC)
But the "etymology" of hiragana entries is almost always the same, namely, "Reading of the kanji". Eg., we shouldn't have etymological info at あらわれる, we should have it at 現れる. Otherwise it would be like putting etymological info about cat in cats. Duplication of effort and a nightmare to sync. With words that usually *aren't* written in Hiragana, the biggest use of having the entry in the first place, is to find what the kanji is. Having a bunch of separated headwords with the same POS just because of etymology, would be confusing, people would just look at the first one and assume that's what the hiragana means. Language Lover 01:31, 11 June 2008 (UTC)
I suppose that makes sense in entries for hiragana words that are usually written in kanji. Thanks for reminding me that the Japanese portion of this project actually does distinguish between full lemma entries (here, the kanji entry) and intentionally abbreviated non-lemma entries (the hiragana equivalent), despite the unfortunate results of Wiktionary:Votes/2007-10/Lemma entries. I've undone my edit [12]. Rod (A. Smith) 02:12, 11 June 2008 (UTC)
That is not a distinction between "lemma" and "non-lemma" forms, that is a distinction between script forms. This says that full entries usually occur at the kanji script form (or the kana form normally written). "non-lemma" entries (inflections) should still have full definitions and usage examples to be as useful as possible to readers. Robert Ullmann 12:48, 11 June 2008 (UTC)
I have a hard time thinking of something sillier and more unnecessarily-confusing than example sentences at かく. Language Lover 16:52, 11 June 2008 (UTC)
Robert, if you're under the impression that I or anyone else ever advocated eliminating example sentences from non-lemma entries, then you're mistaken. Regardless, I'm quite pleased that WT:AJ makes a voter-approved distinction between full entries and intentionally abbreviated secondary entries. Rod (A. Smith) 14:55, 11 June 2008 (UTC)

Declension of 辛い[edit]

At 辛い we see a flaw in our declension system. A reader with a small monitor could easily totally miss the fact that there's a whole separate sense hidden below that massive declension table. Language Lover 17:16, 12 July 2008 (UTC)

Category for Four Kanji Idiomatic Compounds[edit]

I believe we need a category for yojijukugo akin to Category:四字熟語 on the Japanese Wiktionary, or at very least a list of them akin to Wiktionary:四字熟語の一覧. They're certainly unique enough to warrant identification. It would either be a subsection of Japanese idioms or a section in its own right. The naming is up to question (I prefer Japanese Four Character Idioms or simply Yojijukugo). Any opinions on such a category? Estemi 00:09, 1 September 2008 (UTC)

Why are adjectives called "declined"?[edit]

To me, it's rather clear that adjectives are verbs and should be treated as such in the terminology. But just 1/50 of an euro. (Lâche anonyme) 20:04, 24 September 2008 (UTC)

In traditional Japanese grammar, both verbs and adjectives are classified together as 用言 (yōgen). Yōgen are words that conjugate (or decline) and can stand as a predicate. They contrast with 体言 (taigen), which nouns are a part of, which do no not conjugate (or decline) and that can stand as a subject. Bendono 08:12, 3 September 2009 (UTC)
A late 2p, but when it comes to w:Inflection, the terms conjugate and decline are distinct, and refer obliquely to the part of speech that undergoes inflection: decline is generally used for nouns or pronouns (or even adjectives, for certain langauges) that have case endings that change depending on grammatical function (number, subject, object, indirect object, etc.), whereas conjugate is generally used for verbs. Latin and German adjectives decline because these words are true adjectives (purely modifiers for nouns, and that require a verb to be used as the predicate of a sentence) and their endings change to match the noun they modify. Japanese adjectives conjugate because they are also verbs (they can be used as the complete predicate of a sentence) and their endings change to show tense (or, technically speaking for Japanese, w:Grammatical aspect rather than w:Grammatical tense).
I'm guessing that this might be what prompted the IP user's comment. -- Eiríkr ÚtlendiTala við mig 16:05, 31 October 2011 (UTC)

-な, -の adjectives -する verbs[edit]


IMO, when adding translations, it makes sense to use a display name (alt=) with -な, -の and -する but the actual link to have without it. Otherwise each entry for adjective and verbs should have a redirect or an entry with these endings. See my analyse translation into Japanese to see what I mean and suggest.

* Japanese: {{t|ja|分析|alt=分析する|tr=ぶんせきする, bunseki-suru|sc=Jpan}} gives:

The alternative would be to supply an additional entry for 分析する and for each verb, formed this way. Anatoli 01:06, 31 July 2009 (UTC)

(I corrected the wikitext of your {{t}} above so it shows your "alt", "tr", and "sc" parameters.)
Linking to the main entry seems good, but I'm not sure about using "tr" to show transliterations in two different scripts. —Rod (A. Smith) 17:46, 31 July 2009 (UTC)
(I corrected it again using <nowiki>.)
分析 and 分析する are different words, and they should have separate entries. Note that 宿題する is just an ellipse of 宿題を する, which contains a direct object and a verb. — TAKASUGI Shinji (talk) 05:01, 7 December 2011 (UTC)
It's understandable they are different words and even different parts of speech but perhaps the entry for 分析 could include 分析する section. I'm not saying it should, it's just a thought, する is just too productive to make verbs and user may not necesarily know the lemma form of する when they see 分析します. It makes much more sense for -, - adjectives (- adverbs) to have the entries without the suffixes but displaying them in the header and advising the user that it is a -/- adjective. I think we don't need an entry for 病気な if we have 病気 (currently doesn't have an adjective section but can be added). --Anatoli (обсудить) 05:25, 7 December 2011 (UTC)
Just added an adjectival noun section in 病気. I was suggesting a similar apporoach to -する verbs derived from nouns. --Anatoli (обсудить) 05:39, 7 December 2011 (UTC)
I have corrected the adjectival noun entry of 病気. — TAKASUGI Shinji (talk) 06:35, 7 December 2011 (UTC)
Speaking of 分析する (analyze), it is different from 分析を する (do an analysis). The former can have a direct object. — TAKASUGI Shinji (talk) 06:40, 7 December 2011 (UTC)
Thank you, so あの人は病気なんです means "he/she is abnormal" only (not 気分が悪い)? What about 病気になりました? isn't it "fell ill"? Perhaps another sense is warranted with a description.
I understand the difference between 分析する and 分析をする. The rationale for merging verbs/adjectives and nouns is just to make is easier for user. A Japanese-English dictionary will have 分析する and 分析 in one entry, 分析をする could be listed as "see also" or in "usage notes". --Anatoli (обсудить) 06:52, 7 December 2011 (UTC)
あの人は病気なんです means "it is that he is sick." The copula だ always becomes な before の or ん, whether it is attached to a noun or an adjectival noun. 病気な人 only means an abnormal person, while 病気の人 means a sick person. — TAKASUGI Shinji (talk) 07:09, 7 December 2011 (UTC)
Yes, I know this. It's actually a -no adjective, not -na, I have just changed the entry but I can't make the no parameter work the same way as na. See Template talk:ja-adj. In あの人は病気です (without なん or なの) 病気 is still a -no adjective in a predicative form. --Anatoli (обсудить) 07:15, 7 December 2011 (UTC)
My point is, both 病気の and 病気な can go into 病気. --Anatoli (обсудить) 07:19, 7 December 2011 (UTC)
I know the term -no adjective, but I just don’t believe its existence. It is used only for explanation for non-native speakers, and for Japanese, 病気の人 (a person of disease) is syntactically nothing different from 病気の診断 (a diagnosis of disease). — TAKASUGI Shinji (talk) 07:57, 7 December 2011 (UTC)

Adjective etymologies[edit]

Japanese linguistics as well as lexicography classify regular adjectives into two categories: -ku and -shiku.


  • 暑し: atsuku, atsushi, atsuki
  • 高し: takaku, takashi, takaki
  • 新し: atarashiku, atarashi, atarashiki
  • 美し: utsukushiku, utsukushi, utsukushiki

The nomenclature is defined by the whether the first form, known as adverbial, ends in -ku or -shiku. Dictionaries list adjectives, just like verbs, in the second form, the conclusive, at which point the distinction can not be made. The third form is called attributive and ends in either -ki or -shiki. This medial -k- regularly drops out resulting in endings of either -i or -shii. Over time, the original conclusive form was overtaken by the attributive, essentially merging or replacing it. Thus, with the drop of the medial -k-, the new "conclusive" now ends in either -i or -shii. With this merger the two conjugation classes have now merged into one. This is the colloquial adjectival system that many of us are familiar with. Note that while adjectives ending in -shi or -(shi)ki is rather formal and stiff, it is far from rare in the right context.

The etymology of such adjectives is quite regular, so I have created the following two templates:

Please apply them where appropriate. Please note that since the two categories merged together, you can not blindly apply them to every adjective listed at Category:Japanese い-i adjectives. However, any adjective ending in -shii will be a -shiku adjective, and those ending in -i will be -ku adjectives. The production of new adjectives ending in -(shi)i is quite rare so should not be a major problem. When in doubt, ask for help. Regards, Bendono 07:48, 3 September 2009 (UTC)

They were two different groups of adjectives in Early Middle Japanese, but the difference has already disappeared. — TAKASUGI Shinji (talk) 06:59, 3 November 2011 (UTC)

Translingualness of Hiragana[edit]

How are/should w:Ryukyuan languages be handled here on wiktionary. The only discussion I can find is [[Wiktionary:Information desk/Archive 2008/January-June#how to list Okinawan words]] which seemed to be only informed by the ISO 639-3 codes. This decision could affect whether or not Hiragana entries are considered Translingual. Since Hiragana is used to write Japanese and some Ryukyuan languages if they are considered separate on Wiktionary, then Hiragana characters could be considered Translingual. If the Ryukyuan languages are all considered "Japanese" then the Hiragana entries should just be Japanese and not Translingual. --Bequw¢τ 19:41, 15 October 2009 (UTC)

At least katakana must be translingual, because Ainu uses them, including special katakana not used in Japanese, such as and セ゜. There is also a hiragana used in Chinese: . — TAKASUGI Shinji (talk) 05:30, 12 December 2011 (UTC)

Kanji Tab Redirect[edit]

Since Template:ja-kanjitab is a Japanese template, wouldn't it make sense to automatically redirect to the Japanese section of each entry, similar to the way Template:term works? --Hikui87 21:18, 15 October 2009 (UTC)

Using infl[edit]

AFAICT any Romaji script entry using {{infl|ja|pos}} should have sc=Latn in it, as the default for ja is Jpan. Perhaps we would read the latest dump to find the ones that use 'infl'. Mglovesfun (talk) 12:15, 30 December 2009 (UTC)

Does anyone know if this lack of correct script code causes problems? --Bequw¢τ 06:04, 18 January 2010 (UTC)

Character references, dictionaries, and entry methods[edit]

I posted this at About Chinese characters and am reposting here.What is the guideline for including cross references other than those found in {{Han ref}}? Since character entries are targeted at an English-speaking audience, my thought is more information can be provided, particularly for those that are looking at these from a Japanese perspective. Right now the entry methods (other than the Unicode listing) are limited to Chinese-language entry methods. Shouldn't we also include Kuten, Shift-JIS, and JIS codes as appropriate? Would these go under Transligual or Japanese? THe Japanese wiktionary lists all relevant entry codes for its character articles. There is {{ja-kref}} used in a handful of articles that does this (under the Japanese heading), which could probably be merged with {{Han ref}} or remain as a stand-alone template.

Also, what about references to dictionaries/indexes other than the major one's listed. I doubt your average English student of chinese characters is going to have access to the Morohashi dictionary or the Dae Jaewon. But he or she may have one of the other character dictionaries on the market, like the Classic or New Nelson's, One of Halpern's dictionaries, or Spahn's Kanji and Kana dictionary. Several of the more common ones are listed at this website. We wouldn't use them all, obviously, but the main 4 or 5 common ones could be included. Also, we could also addSKIP Patterns to the entries, since that is also a common indexing format in addition to the traditional radical method. Again, some of these are Japanese centric and probably belong under that language heading.

Just a few comments to see what other folks on here think about this proposal before I start making some additions on my own.Dcmacnut 03:02, 16 February 2010 (UTC)

な-adjective translations[edit]

Should the な-adjective translations look like this?

退屈な (たいくつな, taikutsu-na)

{{t|ja|退屈|tr=たいくつな, taikutsu-na|alt=退屈な|sc=Jpan}}

--Anatoli 05:38, 5 July 2010 (UTC)