Wiktionary talk:About Korean/2007-2013

Standard romanisation

I was writing here when all of a sudden all three browser windows disappeared even though I only X-ed one of them via the taskbar. I feel sad and angry and shall attempt to write my text again this weekend. It will be about what's good and bad about the romanisation systems (RevRom transcription, RevRom transliteration, Yale, McC-Rsr) available for Wiktionary to use, and about which of them is the coolest that kicks the others' butts.

It is RR transliteration.

Dustsucker

Breaking up articles by etymology

Why is it that Korean is the only language on Wiktionary where all homophones are grouped together in a section regardless of differing etymologies? Shouldn't we have standard article formats as much as possible? — Hippietrail 10:38, 22 February 2007 (UTC)

Deletion discussions

Korean related entries that are being considered for deletion may be listed here:

치는 사람

(discussion)

雪濃湯

a very obscure hanja of 설렁탕, only the last caracter '湯' is correct.

really? It is used 13000+ times on the web. Mostly in Chinese text of course, but it is a Korean word! Robert Ullmann 18:16, 18 January 2007 (UTC)

Would you please transliterate '雪濃湯' in Korean, if you can? And you can't maintain that some words you find in web are all correct.--아흔(A-heun) 18:19, 18 January 2007 (UTC)

• I think we should give this as the hanja for 설농탕, and explain 설농탕 as an (obsolete?) variant form of 설렁탕. Kappa 18:28, 18 January 2007 (UTC)

Etymologically, that version is somehow reasonable according to a phonetical change. However, it's also possible that the word has something to do with '설렁하다'. I am not sure...--아흔(A-heun) 18:43, 18 January 2007 (UTC)

Although I would hesitate to call these authoritative, here are a couple of online etymology pages that do support this as an archaic form of 설렁탕 (which apparently was originally 선농탕 / 先農湯): [1] [2]. Be that as it may, I suggest that the entry should emphasize the current overwhelming use of this word as the Chinese/Japanese "translation" of seolleongtang. -- Visviva 16:54, 19 January 2007 (UTC)

There are some hypotheses:

1. it might come from hanja.eo 선농단(先農壇), which had been served for agricultural ceremony.
   • 선농단 → 선농탕 → 선롱탕 → 설롱탕 → 설렁탕: It is of cource a hard phonetical change, and it misses a semantical component.
2. it might come from a mongolian word, transliterated into Korean '슐루', which means broth of meat.
   • 슐루탕 → 설렁탕: Explaining a phonetical problem is not so easy.

Whether of these hypotheses is plausible is open. I've been thinking about '설렁하다' since this adj. could explain the signifié of 설렁 (+ 탕). I share with Visviva's opinion in respect to currently usage. --아흔(A-heun) 19:52, 7 February 2007 (UTC)

Categories

I think it would be helpful if someone other than me commented at Wiktionary:Requests_for_deletion/Others#Category:Native_Korean_words et infra. -- Visviva 07:26, 4 June 2007 (UTC)

Parts of speech

I don't have the official ROK school grammar mentioned by User:KYPark on his talk page, but do have a copy of 서정수's (Seo Jeong-su's) compendious 국어문법 (Gugeo munbeop, "Korean grammar"), rev. ed. 2006. This work includes a digest of the part-of-speech schemes used in the South and North, and confirms that the 9-part scheme given here does correctly represent the standard South Korean school grammar as of 1990 (I don't think it's been revised since then). A very similar scheme is apparently in use in the North, although it contains only 8 parts of speech; 토, roughly equivalent to affixes and particles, are not considered a distinct POS in the North Korean system. The comparison chart is on p. 131, for those following along at home.  ;-) -- Visviva 17:03, 20 May 2007 (UTC)

Excellent. Thanks, Visviva! How does that grammar classify the copulas (이다 and 아니다) and the existence verbs (있다 and 없다)? Are they verbs (동사), adjectives (형용사), or something else? Rod (A. Smith) 19:39, 20 May 2007 (UTC)

Although I'm not certain -- the book is a recent acquisition, and at 1500 pages it will take me a while to get fully conversant with it -- Seo appears to do an end-run around this problem by using a functional grammar in the body of the work, thus classifying all of these as 용언.

Other sources, however, are fairly unanimous in treating the copula as a particle (조사) and the negative copula as an adjective.

Koh & Nam (표준 국어 문법, rev. ed. 2002) go into some depth on the 있다/없다 question; the upshot being that it's the general modern consensus that these are adjectives with certain verbal properties. Koh & Nam also note that in certain contexts, such as "거기 있어라," 있다 can function entirely as a verb. This is supported by the 표준국어사전, which has both "verb" and "adjective" headings for 있다. The same appears to be true for 없다, for which the Pyojun gives an archaic verbal meaning (=없어지다), and Naver notes a contemporary North Korean verbal meaning (=죽다)... So it would seem best to list these as both adjectives and verbs. Cheers, -- Visviva 02:21, 22 May 2007 (UTC)

Within the noun class (명사 myeongsa), we should probably distinguish dependent nouns (의존 명사, uijon myeongsa), since they have their own grammatical properties. Dependent nouns would include counters (단위 명사, dan-wi myeongsa). I suppose a {{ko-dependent noun}} tag would be in order, but should such a tag go in the entry headline (a.k.a. the "inflection"/"part of speech" line) or on the definition line? Rod (A. Smith) 20:21, 21 May 2007 (UTC)

Not opposed... but the more granularity we have, the more hairy decisions have to be made. -- Visviva 02:21, 22 May 2007 (UTC)

"Hanja with this reading" section in Korean syllable entries

Thoughts on this? I thought it was a good idea, and so have copied it in various entries. However, it is currently non-standard, so should be addressed on this page if we're going to have it. -- Visviva 02:26, 22 May 2007 (UTC)

For clarification: I thought it was a good idea when I saw it already in use. -- Visviva 12:07, 22 May 2007 (UTC)

I don't know. The hanja reading information is useful, but it may be just as legible to list the hanja in a ===Syllable=== section, e.g.:

==Korean==
===Syllable===
{{ko-pos|syllable|rv=wae|hidx=ㅇ}}

# Reading of hanja [[倭]], [[媧]], [[歪]], and [[蛙]]

Is that any better or worse than ===Hanja reading=== or ===Hanja with this reading=== ? Rod (A. Smith) 04:05, 22 May 2007 (UTC)

Hmm... It's shorter, but I don't know if we can consider "Syllable" a valid part of speech... and for a monosyllabic word, aren't all uses of the word syllables? Actually, looking at WT:POS, it appears that ===Hanja=== would be the preferred heading; and since that is (allegedly) a Policy here we should probably follow it. -- Visviva 12:07, 22 May 2007 (UTC)

"Hanja" is the POS header for the character itself, it its entry. This should be "Syllable", which we have been using for Pinyin and such. (Only for CJKV languages for the syllable represented by a single character.) The above is fine, but for more consistency with (e.g.) Mandarin it should be:

==Korean==
===Syllable===
{{ko-pos|syllable|rv=wae|hidx=ㅇ}}

# [[倭]]:
# [[媧]]:
# [[歪]]:
# [[蛙]]: frog

if short meanings (like glosses) can be provided for each (I'm only really sure of one for Korean ;-). See yí for a extended example (and don't worry about the "Pinyin" header there; it is going to end up "Syllable" ;-) Robert Ullmann 12:22, 22 May 2007 (UTC)

OK, many thanks for the info. I will use "Syllable" in the future -- although I have to admit it still seems like a funny name for a part of speech. -- Visviva 09:05, 23 May 2007 (UTC)

Principal parts

The Pyojun, like most other Korean-Korean dictionaries, gives three principal parts for most verbs and "adjectives" -- viz., the dictionary form (하다), the polite stem (하여/해), and the ni form (하니). (See for example the mirror of the 베끼다 entry on Naver: [3]). This seems like the bare minimum that we should provide in these adjective/verb entries... some dictionaries also give the formal declarative (합니다), which might not be a bad idea either.

A mostly unrelated point: we need to be careful not to mix adnominal/attributives (관형사), i.e. the words that behave like actual English adjectives, with actual Korean "adjectives" (형용사). Attributives cannot be conjugated, and will presumably be needing a template of their own. -- Visviva 13:22, 23 May 2007 (UTC)

Thanks for that link. Yes, we should give readers at least the basic forms of Korean verbs and adjectives. Besides your additional parameters on the headword/POS/inflection line, I created {{ko-conj-verb}}. I'm not yet sure which forms it should include but added it to a couple of verbs to illustrate how it might help. See 가다 and 먹다.

Also, note that 관형사 (gwanhyeongsa) has several different English translations: adnominals/attributives/determiners/pre-nouns/unconjugated-adjectives, so I'm not sure what to call our part of speech (POS) category and headline/POS/inflection template for 관형사 entries. The existing template names and categories are based on English names of parts of speech, but I don't have any books that discuss 관형사 and Google shows no English books containing "관형사". When we find a good name, we should merge Category:Korean adnominals and Category:Korean attributives into that name. Let me know if you have an opinion regarding how we should name them. For now, I've created {{ko-attrib}}, which we can just rename if we find an authority on the matter. Rod (A. Smith) 07:40, 24 May 2007 (UTC)

Seo 2006 gives "determinative" for 관형사; he's the most authoritative source I'm aware of. At the time I created the "Adnominals" cat, the only source I had at hand used that translation, but I don't think it was very authoritative... on the other hand, "Korean adnominal" has 24 mostly-somewhat-relevant Google hits, compared to zero relevant hits for "Korean determiner" and "Korean determinative," and only 2 relevant hits for "Korean attributive." Anyway, I don't have any strong preference, and agree that the categories should be merged to whichever term we choose to use. -- Visviva 15:45, 24 May 2007 (UTC)

24 Google hits isn't terribly compelling, so assuming nobody objects to "determinative" as our English translation for 관형사, we'll deprecate Category:Korean adnominals and Category:Korean attributives, create Category:Korean determinatives, rename {{ko-attrib}} to {{ko-determ}}, and have that template categorize entries as Category:Korean determinatives. Would you also think a "===Determinative===" header for these entries is appropriate? If so, I'll announce the intention to use "===Determinative===" in WT:BP. Rod (A. Smith) 19:12, 24 May 2007 (UTC)

That seems wise; not all determinatives are even descended from adjectives or verbs, so they do need their own heading sooner or later. -- Visviva 01:34, 25 May 2007 (UTC)

The conversation at WT:BP#Determinative indicates we should choose the more common synonym: "determiner". So, I'll have WT:AK, Category:Korean language, and {{ko-determ}} adopt that term unless anyone objects either here or at WT:BP#Determinative. Rod (A. Smith) 18:44, 28 May 2007 (UTC)

Terminology

We really need a set of standard terms for all relevant grammatical concepts (of which there are A Lot). I suppose such a list is best placed in Appendix: namespace? -- Visviva 05:22, 24 May 2007 (UTC)

Well, I started the first appendix. Please review Appendix:Korean verb inflections. Maybe 하다 (hada, “hada”) wasn't the best verb to show there because of its stem change, but the main point is to establish a baseline of verb inflections. Our per-entry templates can show a subset of the inflections and point to the appendix for the full set, right? Rod (A. Smith) 04:02, 30 May 2007 (UTC)

I think it would help if neighboring columns had a slight shift in background color. It would make the columns easier to parse visually. --EncycloPetey 04:05, 30 May 2007 (UTC)

Yeah. It's a beast to read now, but I'm pretty sure it's only around half complete in terms of the forms it should show. When we determine the set of forms to show, we may find that we need to rearrange it significantly. So, we should probably wait until it is (nearly) complete to make it pretty. Rod (A. Smith) 04:23, 30 May 2007 (UTC)

Looks promising; suggest we limit that table to conjugations by speech level/honorific, and develop other appendices for nominal/conjunctive/etc. forms which do not (usually) vary by speech level. -- Visviva 09:34, 1 June 2007 (UTC)

Actually what I was specifically thinking of was something like Appendix:Korean grammatical terms, to which improvements are eagerly sought. -- Visviva 09:34, 1 June 2007 (UTC)

Nice. I linked the Korean terms to the corresponding Wiktionary entries and added "Notes" columns for examples and brief explanations. The reader can click through to the Korean terms' entries for more details.

Regarding the speech levels, I italicized some "Wiktionary translations" that I thought were ambiguous and hence questionable. I also suggested some translations. We should probably discuss translation options and choose an official one for each term.

As for the parts of speech, I listed 느린 (neurin, “neorin”) as an example of an attributive because that's how we're currently classifying the -ㄴ/-은/-는 forms of 형용사 (hyeong-yeongsa, “adjectives”), but I'm not confident that those forms are actually 관형사. They grammatically act as 관형사, but since they are actually inflected forms of the underlying adjective, they're probably more accurately listed as non-lemma entries for the corresponding 형용사. It would be nice to have that formally decided one way or the other.

Also, 체언 (cheeon, “che-eon”) and 용언 (yong'eon, “yong-eon”) are umbrella terms that probably shouldn't be listed as POS themselves, but I don't know where better to list them in the appendix. Rod (A. Smith) 18:41, 1 June 2007 (UTC)

Quite so... those terms could perhaps go under "Other" for now. Thanks for your work on that page!

I've put in my 2 cents on the adjective/determiner matter below (vide #Determiners and adjective forms), in hopes of keeping this conversation from getting too trackless. ;-) -- Visviva 09:06, 2 June 2007 (UTC)

Template:ko-pron

Curious to get some feedback on this... It isn't really orthogonal to the current approach, but perhaps some things can be adapted.

1. Is the display method something that we might adapt for the POS templates, or is it too idiosyncratic?
2. In entries generally, is it OK/wise to include Yale romanizations? (my feeling is yes, if only because WT is a linguistic resource and Korean linguistics still uses YR quite extensively)
3. In entries generally, is it OK/wise to include phonetic hangul? (this is arguably redundant with Yale, except that Yale roms cannot generally be verified directly)
4. Given that Yale is almost entirely phonetic, and MR is avowedly (though not entirely) phonetic, where do we draw the line between information that belongs under "Pronunciation" and information that belongs in the POS heading? -- Visviva 17:47, 24 May 2007 (UTC)

My preference would be to show just one romanization in the headline/POS/inflection area and to put the others in a separate section. {{ko-pron}} does the latter part of that nicely when it's in a ===Pronunciation=== section. I cannot remember where, but somewhere in the Wiktionary namespace it says that RR is our official romanization system, so I'd be in favor of eventually having a bot move all the Yale and MR romanizations into ===Pronunciation===/{{ko-pron}}. Yes for Yale for the reason you mention. And yes for phonetic hangeul since it's so easily verified in Korean references and seems to be used by native speakers of Korean. (When I talk with my native Seoul friend, he corrects my pronunciation by writing words in phonetic hangeul, except that he doesn't mark the long vowels as such.) Rod (A. Smith) 18:13, 24 May 2007 (UTC)

Groovy, seems like we're on the same page then. I'll start implementing this more widely, and we'll see if anyone screams. -- Visviva 04:59, 25 May 2007 (UTC)

Well, calling a romanization a pronunciation is an utter travesty ... I think we should drop revised (rr=) from the ko-pron template, it should always be on the inflection line, and is the recommended SK written romaja form. (whether used or not is a separate issue ;-) I think the Yale and MR should only display if specified? Robert Ullmann 09:47, 1 June 2007 (UTC)

Yes, this has bothered me. Still, the romanizations do each provide pronunciation information, each in its own way, and for that matter IPA can be regarded as a romanization of sorts. I wouldn't be averse to a separate "Romanizations" heading, if that were approved at WT:ELE, but it seems unnecessary. Floating boxes are ugly, and putting the romanizations all in a row just creates an unparseable mess... so I'm not sure what would be a better solution. At any rate, I have made the rr, mr, and yr parameters optional, per your suggestion. -- Visviva 10:50, 1 June 2007 (UTC)

Category sorting

WT:AK needs a note to describe the DEFAULTSORT recommendation. Something this:

DEFAULTSORT (e.g. "{{DEFAULTSORT:ㅁ말하다}}") helps sort Korean language categories. The parameter should be the jamo for the initial consonant (including ㅇ, ieung if applicable, and not doubled, so use ㄱ instead of ㄲ) immediately followed by the headword. Doing so causes the headword to appear where it is expected in any manually set categories (e.g. "Category:Native Korean words"). Note: The expected category collation is similar but not identical to that of most Korean dictionaries because MediaWiki categories sort with binary collation. For example, most Korean dictionaries sort 가까이 (gakkai, “gakkai”) before 가감 (gagam, “gagam”), but MediaWiki sorts "ㄱ가감" before "ㄱ가까이".

If we want to sort our Korean categories even more accurately, we'd need developer support to install a Korean collation feature or we must change the category index parameters to something like "ㄱㅏ*가ㄱㅏㅁ감" (for 가감) and "ㄱㅏ*가ㄱㅏ*까ㅇㅣ*이" (for 가까이). That seems pretty cumbersome, though, so perhaps properly sorting Korean categories is a job best left to bots or put off until MediaWiki can sort hangeul properly. Thoughts? Rod (A. Smith) 23:02, 28 May 2007 (UTC)

irrelevancies stricken from above Rod (A. Smith) 21:32, 1 June 2007 (UTC)

I'd prefer to keep things as natural as possible and trust that the software will eventually catch up. (In view of which I'm not entirely happy with indexing ㄲ as ㄱ, though I can live with it). -- Visviva 01:06, 29 May 2007 (UTC)

Update: Actually ㄲ seems to be indexed appropriately by the software, so I'm not sure why we can't use doubled consonants for the sortkey. -- Visviva 07:37, 1 June 2007 (UTC)

There is a similar problem with Chinese, where we use rad/stroke of the first character to get it mostly right, but then we are basically stuck with Unicode collating order, which is not bad. Robert Ullmann 01:43, 29 May 2007 (UTC)

Note: the dictionaries I have at hand sort 가까이 after 가감. -- Visviva 13:26, 30 May 2007 (UTC)

Interesting. My Korean dictionary is probably an anomaly, then. It has this order:

• ㄱ
• 가
• 가가 대소
• 까까머리
• ...
• 가까이
• ...
• 가가 하하
• 가각
• ...
• 가감

Since that seems not to be the normal Korean order, it would be interesting to know whether the normal order matches Unicode sort order. Rod (A. Smith) 16:08, 30 May 2007 (UTC)

As far as I can tell, that seems to be the case; I wouldn't be surprised if there are some inconsistencies between standard SK order and Unicode order, but for the most part they seem to coincide. The sorting issues that I'm aware of affecting our work here are:

1. Categories are sorted by jamo rather than letter by default. This is a bug (or rather, the absence of a feature) in MW, and might be worth making a patch for, since it affects not only every Wiktionary but every KO project.

2. Sort order in most categories is very messy, due to several mutually-incompatible schemes being in use. This is a mess that we can (and must) clean up fairly quickly.

Am I missing anything? -- Visviva 07:37, 1 June 2007 (UTC)

When you say "Categories are sorted by jamo rather than letter by default" above, do you mean, "Categories are sorted by hangeul character rather than jamo by default"? Rod (A. Smith) 18:43, 1 June 2007 (UTC)

Well, no, by default in MediaWiki the entries are sorted by jamo, for example 각기 goes under 각 rather than ㄱ... which is why we have to fiddle around with sortkeys this way. -- Visviva 05:48, 2 June 2007 (UTC)

My understanding of "jamo" apparently differs from yours. The 24 jamo include ㄱ but not the hangeul character 각. Rod (A. Smith) 16:56, 2 June 2007 (UTC)

Er... that's right, of course. How embarrassing. :-o -- Visviva 06:10, 3 June 2007 (UTC)

Unfortunately, I just discovered that MW only uses the last DEFAULTSORT directive on a page. That makes DEFAULTSORT inappropriate for non-hangeul entries. We can still use it in hangeul entries, but hanja and romaja entries will typically have non-Korean language sections, whose categories need their non-Korean category index. Rod (A. Smith) 00:35, 2 June 2007 (UTC)

Good point. -- Visviva 05:48, 2 June 2007 (UTC)

Actually it had never occurred to me to use hangul sortkeys for romaja. Not sure if that would be wise... -- Visviva 13:51, 2 June 2007 (UTC)

Since the North and South Korean jamo orders differ, there seems to be no perfect way to index our hangeul entries without aligning with either side. E.g., in South Korean order, ㄲ (kk, “kk”) is treated as a sequence of two ㄱ (g, “g”) jamo and alphabetized accordingly, but in North Korean order ㄲ (kk, “kk”) is a separate jamo, alphabetized much later. So, if we want to maintain jamo-based category indices, we need to choose the North Korean (indexing words that begin with doubled jamo under that doubled jamo) or South Korean (indexing such words under the corresponding single jamo) collation system.

Perhaps we should skirt this issue by assuming that MW will eventually have proper hangeul collation features. If so, we could remove the DEFAULTSORT, hidx, and [[Category:...|?...]] recommendations for the hangeul entries. The resulting Unicode sort seems at least as good as aligning with the North or South sort order. (We could easily alter the hangeul category TOC index to jump to the hangeul syllable starting with the clicked jamo.) Rod (A. Smith) 19:36, 16 June 2007 (UTC)

Well, when we use DEFAULTSORT, we aren't strictly aligning with North or South, but with the Unicode order of jamo (which treats ㄲ as a separate character immediately following ㄱ, and is arguably something of a middle ground). I'm not really holding my breath for any improvements to Mediawiki in terms of hangul collation (unless one of us is going to write the patch), so I'd prefer to stick with DEFAULTSORT as the least-messy solution. I don't consider the default by-syllable sorting to be adequate, though you're right that the category indices do get around the problems somewhat. -- Visviva 09:28, 27 June 2007 (UTC)

“[I]n South Korean order, ㄲ (kk, “kk”) is treated as a sequence of two ㄱ (g, “g”) jamo and alphabetized accordingly” – You mean SK order would have 어따 (eotta, “etta”) and 얻다 (eotda, “etta”) next to each other? Where does it put words like 얹다 (eonda, “entta”) that would otherwise go in between? I would love to see a source for that. – Dustsucker 05:44, 31 July 2008 (UTC)

Determiners and adjective forms

I think we need to keep in mind that determiner/관형사 is the actual part of speech for words like 느린. These forms should also be categorized in Category:Korean adjective forms, since they are adjective forms; however, "adjective form" isn't really a POS in itself (neither is "verb form," come to think of it). The case is arguably similar to verbal nouns. An entry for a verbal noun such as 가르침 (gareuchim, “teaching”) would still have "noun" as its POS heading/category, but would also be filed under Category:Korean verb forms or a more specific subcategory.

The ideal solution, methinks, is to make like Category:English present participles, and have subcategories for each kind of verb/adjective form. Then words like 느린 could just go in Category:Korean determinative adjective forms (although that in turn might need eventual subdivision, considering 느릴, 느리던, 느리다는, etc.), and that category would be a subcat of both Category:Korean determiners and Category:Korean adjective forms. -- Visviva 09:07, 2 June 2007 (UTC)

That sounds quite sensible. Rod (A. Smith) 16:56, 2 June 2007 (UTC)

I think this needs to be moved again. Determiners isn't right as what has been labeled "determiners" in Korean doesn't behave like English determiners and would only serve to confuse english readers who are familiar with the technical aspects of determiners. [4]. This academic paper has a break down on why they're not determiners and makes a case for Korean being a determiner-less language. I've seen reliable sources refer to these words as "pre-nouns" and I think that is a much better description of them. [5] See page 19 and 21.--Crossmr 01:54, 4 January 2010 (UTC)

Yes, "determiner" has always been a very awkward fit. The constraining factor is WT:POS. For various mostly-good reasons, Wiktionary frowns upon innovation in POS headers.

I would prefer "adnominal" to "pre-noun", as it is the term used by Martin and others such as Lee & Ramsey to refer to the full class of words/forms in question. Whatever header we use will have to apply not only to true 관형사 like 저, 이 &c., but also to 관형어 like 작은, 큰, &c. "Adnominal" also has the advantage that some editors have used it already, such as for Japanese under 東#Adnominal; so it is not as much of an innovation. -- Visviva 06:01, 4 January 2010 (UTC)

The link you just gave shows the section using the title "prenoun" not adnominal. The book you gave lists the same parts of speech as the link I gave. Nouns, Pronouns, (2 are skipped as part of the preview), prenouns, adverbs, and then a few more skipped pages. In fact that book describes an adnominal as something entirely different. If you search it for adnominal you'll see the definition in the preview is specifically "A sentence that is embedded in another sentence and is used to as an element modifying a following noun phrase is an adnominal". That isn't what we're talking about here at all. That book uses prenoun to describe exactly what we've labeled as determiners. An adnominal seems to be similar to an adjective clause.--Crossmr 01:54, 5 January 2010 (UTC)

No, that book uses "prenoun" to refer to the small closed set of words like 이 and 저. From the paragraph just below that heading: "Prenouns are [...] few in number. In fact they form the smallest, and therefore most unusual, word class in Korean." (Note that this usage is at variance with Martin's, who equates "adnoun" and "prenoun" and uses the term to cover all adnominal modifiers.) The present discussion is/was about how to handle adnominal verb/adjective forms like 가는 and 할.

That said, given the above-mentioned strictures regarding headings on Wiktionary, I think that the pragmatic argument for treating "prenouns" (in the limited sense of Lee & Ramsey) as "determiners" is rather strong. For a multilingual translating dictionary, consistency across languages is key. It makes intuitive sense that the word that translates this#Determiner would be filed under the same POS heading.

On the other hand, I think my position at the start of this discussion was a bridge too far. Using "===Verb===" and "===Adjective===" headings for all verb and adjective forms is better, simpler, and consistent with broad Wiktionary practice. As a bonus, it neatly limits the POS header issue to a small closed set. -- Visviva 05:06, 5 January 2010 (UTC)

Unfortunately these words in Korean don't behave as determiners, so while it may be quaint to refer to them as determiners, it is misleading and provides false information about how these words work and what they are. It seems more like an attempt to shoe horn a language into an already existing box than properly define and represent it. Abnominals is also out as the definition from the book clearly isn't it. The paper which talks about korean being a determiner-less language seems to indicate that those don't function the same, so labeling them as such doesn't make sense. This book [6] only mentions prenouns a couple of times, but the first mention seems to be put them in the class or equal to adnouns. Adnouns are later described as including unconjugated adjectives, much like Chul Young Lee's korean grammar book describes them [7]. This book separates prenouns and adjectives [8] which doesn't match up with the first text book which refers to adjectives solely as those with are used as predicates. I think there are two issues. 1. What is the name of this class, whatever is settled out, I completely disagree with determiner as I think its misleading and 2. what words are included. A couple sources seem to indicate that adjectives are included or related and a couple sources seem to separate them.--Crossmr 05:58, 5 January 2010 (UTC)

Romanization (oh dear god)

OK, I don't want to open the which-romanization-is-better debate (ever!), but here are a couple of things on my mind:

A. Without disrupting the canonical use of RR, we need to sort out exactly where alternative romanizations should go, whether they should be repeated within an entry, etc. As I've expressed above, I think Template:ko-pron is the least bad solution so far, but I'm not married to it.

B. I just noticed that Chapter 3 Paragraph 8 of the RR edict, aka 국어의 로마자 표기법, actually describes a system of RR transliteration for academic use very similar to what I think Dustsucker was suggesting at the top of this page. I'd opposed this idea because it would be introducing a novel and unfamiliar system; however, inasmuch as this is a system with a formal description, and in some respects more transparent than standard RR, I'm curious as to the thoughts of others on this. Some examples of differences:

좋다: standard RR jota (or johta?), transliteration johda

집: standard RR jip, transliteration jib

독립: standard RR dongnip, transliteration doglib

... again, I'd like to avoid the broader romanization debate, but am very curious as to what people think of these matters. Happiness, -- Visviva 13:39, 2 June 2007 (UTC)

Adopting tansliteration sure would reduce the number of errors I make. A complete Wiktionary entry should certainly have the RR transcription wherever it has MR or Yale forms, but maybe the transliteration system makes more sense for the canonical romanization following each use of hangeul. I.e., {{ko-inline}}, {{ko-pos}}, etc. could use transliteration but the {{ko-pron}} (or whatever we end up calling it) could use transcription. Rod (A. Smith) 16:56, 2 June 2007 (UTC)

OK, well, guess we should mention this at the BP and see what folks think (since it would be a fairly wide-ranging change). It's worth noting that about half of our current entries are using something like RR transliteration, or were until I went around obsessively changing them. -- Visviva 06:20, 3 June 2007 (UTC)

Yes, that's it – I thought it would be clear I was referring to § 8. This RR transliteration (RR-lit) seems to be South Korea's official standard for situations that demand disambiguity, it is documented, and so much easier to learn than Yale and McC-Rsr, especially for people familiar with RR transcription (RR-scr). Unlike RR-scr and McC-Rsr, RR-lit unambiguously represents Korean orthography, which is good for users who cannot install a legible hangul font for whatever reason. On the other hand, it doesn't really do anything else, so RR-lit would be less of a pronunciation aid than RR-scr (let alone Yale and McC-Rsr) is. That being said, I have never seen RR-lit in use anywhere, not even in official publications. – Dustsucker 06:27, 31 July 2008 (UTC)

Middle Korean

Two issues that have been bugging me with Middle Korean in Etymology sections:

• Yethangeul: Simple MK characters, for example syllables which are unusual only in their inclusion of arae-a, seem to be problem-free. But more arcane characters like  are problematic.  shows up as a nonexistent Chinese character in my browser, but becomes the syllable ㅅㄱㅗ when pasted into (Korean) MS Word. Not sure if it's a proprietary codepoint or what. It would be a good idea to have some general guideline for handling this.
• Romanization: The only romanization that is actually used for Middle Korean is Yale, specifically the Middle Korean-specialized version of Yale. So this is presumably what we should be using; it will, of course, be somewhat disorienting to the reader. However, there doesn't seem to be any alternative. -- Visviva 06:20, 3 June 2007 (UTC)

And another:

• Language code: There is an ISO code for Middle Korean (okm) and also one for Old Korean (oko), so two questions arise: 1. should templates for citing Middle Korean, such as {{ko-etym-native}}, use {{lang|okm|foo}} instead of {{KOchar|foo}}? 2. Should we, in the future, use "Middle Korean" as a language header where appropriate? It does seem to be the most common term for referring to 14th-16th century materials. -- Visviva 09:11, 28 June 2007 (UTC)

Thanks to User:Robert Ullmann for clearing up some of the Yethangeul issues (see User_talk:Visviva#.EF.94.BA.EE.9E.B4). Characters like  which use the Private Use Area of Unicode are clearly deprecated, and for good reason... Although I'm still not quite sure how best to proceed; even with the recommended fonts installed I can't get the hangeul at wikisource:ko:훈민정음언해 to display properly, and there isn't much to be said for replacing a bad system with one that no user or editor will be able to view properly. -- Visviva 17:10, 30 June 2007 (UTC)

wikisource:ko:훈민정음언해 works for me. A few years ago, Jungshik Shin (신정식) has published a comparison of methods to display old hangeul in browsers, and AFAIR the only free Unicode font to display decently was UnJamoBatang (은 자모 바탕). Aesthetically, UnJamoBatang compares rather unfavourably against any non-옛한글 fonts, but unlike even Microsoft's fonts at that time, it did (and does) handle bangjeom. Would it be odd to use Yale romanisations throughout in place of actual Unicode yethangeul, instead of confining them to a “Romanizations” section? – Dustsucker 2008-07-31

Korean parts of speech

I notice that Appendix:Korean grammatical terms lists 조사 (josa) and defines it as "particle" and "postposition". Pardon my ignorance, but I'm not sure how this fits in with our usual POS scheme. If a word is a Postposition, I would expect that header just as we use Preposition as a header (Hungarian uses postpositions rather than prepositions, for example). Particles are usually short words with a specific (yet hard to define) grammatical function. That is, if it's a particle, then it doesn't fit among the usual parts of speech. So, is a josa a particle or a postposition? --EncycloPetey 06:26, 3 June 2007 (UTC)

"Particle" is definitely the most common translation, for what that's worth. These are most commonly grammatical markers of sorts. -- Visviva 07:29, 4 June 2007 (UTC) signed later

I don't think postposition would work; it's just noted in the table as an alternate translation that some have used. These are fairly similar to particles in Japanese (and probably other languages), in terms of their general grammatical function, so I assume if it is correct to call the Japanese morphemes "particles" then this can be applied to the Korean case as well. (Disclaimer: my Japanese is rudimentary; I could just be confused.) -- Visviva 07:29, 4 June 2007 (UTC)

Coverage status

Of some interest may be User:Visviva/List and User:Visviva/1000. The former is drawn from ko:Special:Allpages, and the second from the (first 1000 words of) the 한국어 학습용 어휘목록 by NIKL. I believe this last to be free of copyright concerns, as it is part of an official bulletin of a government agency, and such bulletins are not eligible for copyright under the Korean Copyright Act. I might of course be mistaken. -- Visviva 09:15, 28 June 2007 (UTC)

Update: the first (most important?) 100-some are all blue now. -- Visviva 17:04, 30 June 2007 (UTC)

Yet another template

Created {{ko-sort}}. This simply adds the DEFAULTSORT magic word in the way we've been doing. I propose that we use this instead so that if we ever decide that the DEFAULTSORT approach has become obsolete (because of changes in MediaWiki or because we change our minds about sortkeys), the change can be made with a single edit, and we can then remove the obsoleted templates at our leisure. Otherwise all the sortkey cleanup we do now is likely to just end up being part of a larger mess in the future... Also, this is a little easier to type ({{ko-sort|ㅎ}} instead of {{DEFAULTSORT:ㅎ하다}}), although I wouldn't have proposed it for that reason. -- Visviva 15:27, 29 June 2007 (UTC)

Good step. I was thinking of creating something like {{sortkey}} that would be useful in a number of cases; remembering that it is a default that applies to the whole page. But there are a number of complications; I wish someone had put more thought intt DEFAULTSORT. (It is generally useful in the 'pedias, but gives us (wikts) troubles.) Robert Ullmann 15:35, 29 June 2007 (UTC)

Yeah, thanks. Rod (A. Smith) 15:45, 29 June 2007 (UTC)

This & that...

Hi,

Here are some useful/interesting things I've discovered lately:

• It is possible to link directly to both search results and individual entries in the 표준국어사전. Woot! Search results can be linked in the form "http://www.korean.go.kr/06_new/dic/search_input.jsp?att1=(urlencoded page name)", and individual entries in the form "http://www.korean.go.kr/06_new/dic/View.jsp?idx=(page ID)". See for example [9] and [10]. ... Those URLs don't look very stable, but hopefully it will be a while before the backend changes.

• I've been building a series of preload templates to ease the process of new article creation. These are all in my userspace for now, but I'm thinking they're about ready to go live. The end notion is to use something like User:Visviva/Meta as a base template, to which more specialized templates like User:Visviva/Noun-h (for nouns composed of 2 hanja) can pass parameters. Wordlists like User:Visviva/1000 could then use a template (maybe a substed one) based on User:Visviva/Tracker to send us directly to preloaded pages with useful information above the edit box: example (still a ways to go, but interested in feedback on the idea)

• I recently acquired a pronouncing dictionary which contains a reprint of the 1999 IPA Handbook section on Korean... but, um, there is quite a gulf between the system actually used there and the one used in the w:Korean language article which I had been following heretofore. Not sure where to go from here, as the system actually used in the Handbook (or at least in the reprint) does all sorts of weird things like using z/s to represent ㅅ/ㅆ ... On the other hand, it's difficult to justify not following authority in an area as hopelessly tangled and trackless as phonology. -- Visviva 08:01, 7 July 2007 (UTC)

Nice. Do promote your Korean preload templates into production. Also, if either of your wordlists is in the public domain, maybe we should move it to Appendix:1000 Korean basic words (ala Appendix:1000 Japanese basic words) that would use your preload templates for non-existent entries and maybe also create Category:1000 Korean basic words (ala Category:1000 Japanese basic words).

Regarding the ㅅ/ㅆ as z/s issue, regional differences may be involved. I don't know enough to have a strong recommendation. Rod (A. Smith) 18:30, 7 July 2007 (UTC)

Thanks! Your URI seems to work if we use “08” instead of “06”. You might want to avoid copyright infringement à la NanshuBot's copying of the Unihan DB, though. – Dustsucker 2008-07-31

Verb forms of nouns

Based on WT:AJ#Verb forms of nouns, I included the following instructions in WT:AK:

Many nouns in Korean have a verb form that is created by simply appending 하다 (hada). Because this type of verb is so closely associated with the noun stem, each form should refer to the other with the {{see}} template used at the top of the Korean section, before the first listed part of speech.

I'm not convinced that's any better than adding the -하다 derivative under ====Derived terms====, so perhaps that exception should be removed. Thoughts? Rod (A. Smith) 19:10, 7 July 2007 (UTC)

This seems like a good idea, on the whole. But I wonder if we should extend it to cover 되다 and 적 derivatives as well; these are similarly tightly-bound to the noun stem, and like the 하다 forms are often rolled into the root entry in Korean dictionaries. -- Visviva 10:35, 9 July 2007 (UTC)

I tooled up a little thing at Template:ko-see which takes 하다, 되다, and 적 in any order as unlabeled parameters. Just to have it handy... Template:see seems to disagree with Template:ko-inline, and as usual with such things, if we change our minds it would be helpful to be able to just check Special:Whatlinkshere/Template:ko-see in order to see what needs to be cleaned up. I put in a test implementation at 시간, but don't plan to use it more widely just yet...

It seems that the ongoing parlour chat on see-alsos for plurals &c. is considered relevant to this. -- Visviva 16:28, 9 July 2007 (UTC)

Example sentence format

As noted here, the recommendation in WT:AK to format example sentences with a leading bullet ("#* 보기...") was inherited from WT:AJ, which was developed before the modern WT:ELE. We should probably align that recommendation with WT:ELE (i.e. "#: 보기..."). 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 bot to move usage notes into "====Usage notes====" and a bot to convert existing examples to the new format. Thoughts? Rod (A. Smith) 04:47, 24 July 2007 (UTC)

I wouldn't have thought there would be so many as to require a bot; but otherwise, sounds good. -- Visviva 01:56, 25 July 2007 (UTC)

Hanja entries

The following, up to 19:07, 14 September 2007 (UTC), moved here from User talk:Connel MacKenzie#about Hanja. Rod (A. Smith) 19:07, 14 September 2007 (UTC)

I'm really sorry to discover repeatedly that the en or some users regard Hanja as Korean. There had been discussions on this issue (Beer parlour archive/2007/February, Robert Ullmann/2007, A-heun). I don't know whether users who Hanja (like 建物, 競馬, 怪物, 九月, 木曜日, 大學校, 飜譯, 木曜日 ...) regarded as Korean have ever been informed about the Korean language. I hope the en won't intend to publish wrong informations about the Korean language. --아흔(A-heun) 09:10, 14 September 2007 (UTC)

Yes, "hanja" means "Chinese character(s)." When a Korean word is written in Chinese script, it is still a Korean language word. Yes, the script is hanja. If one writes "geonmul" the script is Latin; the word is still Korean. At 建物 we have a proper entry for a Korean word written in Chinese script. The heading used is the language, not the script. Robert Ullmann 11:47, 14 September 2007 (UTC)

I think 아흔 (A-heun) is concerned that readers of such entries might get the wrong impression about how modern Korean is written. In our properly formatted hanja entries, we indicate hanja on the headword line and show the equivalent hangeul spelling, but that does not seem very enlightening. Perhaps we would better serve readers to add usage notes or to link that hanja indicator to an appendix that explains how hanja has limited use in modern Korean (mostly for people's names and to disambiguate homophones in certain government documents and legal contracts). Whatever we end up doing (or not doing), though, we should probably discuss this somewhere more public than Connel's talk page, e.g. WT:BP. Rod (A. Smith) 16:45, 14 September 2007 (UTC)

Thank you (all) for your comments. Wiktionary talk:About Korean seems like the right place to me. My guess is that A-heun is looking for support; that is, we probably should not recommend spellings that are wildly obsolete, nor should we portray them as current language elements. List them; yes, but only with an indication that they are "classical" only. Our Korean entries are much better than they used to be, but still have a long way to go. I don't see how I personally can help, though. --Connel MacKenzie 18:57, 14 September 2007 (UTC)

Note: I expanded Appendix:Korean script with more information, mostly from English Wikipedia articles. Any help that anyone can provide with that appendix would be much appreciated. If there are no objections, I'll have {{ko-noun}} et al. link the hanja tag it outputs for hanja entries to that appendix. Rod (A. Smith) 20:06, 14 September 2007 (UTC)

One option is to create a bot that adds something like

====Usage notes====
*{{ko-hanja-message}}

to all Korean entries in the Hanja script, where Template:ko-hanja-message can be anything the Korean-speakers here deem appropriate (including a link to Appendix:Korean script). I don't think we've ever tried having such widely duplicated usage notes, but if they work out and don't turn out to be a pain, I can see other uses for them; indeed, EncycloPetey recently suggested one at Wiktionary:Beer parlour#Multiple context tags.

—Ruakh_TALK 23:12, 14 September 2007 (UTC)

Yes, I very much like that idea. 아흔 (A-heun), what do you think about that approach? Rod (A. Smith) 23:29, 14 September 2007 (UTC)

Thanks for the attention of all.

• I find Appendix:Korean script is a very good idea for information about hanja (thanks Rodasmith for editing)
• The template:ko-noun can't assign hanja entries to the Category:Korean Hanja, which is curiously a clean-up candidate. All hanja entries should be classed under that sub-category of Korean. Instead of template:ko-noun one should always use Template:ko-hanja in the case of hanja. Above listed hanja used template:ko-noun.
• The headline for hanja entries should refer precisely to its identification to clear up a misunderstanding. Hanja is not Korean per se !
• I think, the possible Template:ko-hanja-message can be pasted directly after the headline. Or more simply in the headline ==Korean Hanja {{ko-hanja-message}}==.

current manual (e.g.)	proposal
英語 ==Korean== ===Noun=== {{ko-noun|hj|hangeul=영어|rv=yeong-eo|hidx=ㅇㅕㅇㅇㅓ}}	英語 ==Korean Hanja == {{ko-hanja-message}} ===Noun=== {{ko-hanja|hangeul=[[영어]]|eumhun=[[영]], [[어]]|rv=yeongeo|mr=yŏngŏ|y=yenge}}

--아흔(A-heun) 09:32, 15 September 2007 (UTC)

If hanja is not Korean, I have a lot of very confused books in my library. Oh, look, what's this? 위례성, copyright 2003. Let's open it to a random page, shall we? "내용은 사가의 기록에서 빠졌거나 자세히 드러나지 않은 것을 드러내 표현한 것으로 神異한 史話가 많다는 특징이 있다. <sarcasm> Oh my god! I must have accidentally purchased a book in Chinese! How foolish of me.</sarcasm> Granted, that was an archeological text, prone to archaisms. Let's try another; ah, here's the 北韓總覽, the official South Korean guide to all things North Korean. Well, the title is in hanja, but surely that's just a strange quirk of design. I open it to a random page, and what do I see? Table after table, labeled in hanja: "<表 8>", says one, "北韓의 國民總生産." And in the accompanying text, one paragraph after another beginning "이를 美貨로 표시하면..." My textbook of law presents the entire 1987 South Korean constitution in mixed script (I am not lawyer enough to say if the binding version of the constitution contains hanja, but it would be an odd thing to do otherwise).

Walking on an unremarkable Busan earlier street today, my eyes were assaulted by signs advertising "食品", "玉", "美容室." In the local shopper, eager companies advertise "사원 大모집!" In the dailies, sage tongues wag over the latest gaffes of President 盧 and the fickle fortunes of the 新黨.

Are hanja forms as common as their hangul equivalents? No; in most cases they are used only for disambiguation and precision (which is why every Korean dictionary includes them). Or they are used for impact (as in the signs and advertisements) or to show the erudition of the author (as in the scholarly texts cited). But are they in current, Korean use? Obviously. And are they Korean? Obviously. Any assertion to the contrary is political and tendentious at best.-- Visviva 10:27, 15 September 2007 (UTC)

The above is unnecessarily snarky. I apologize for this. -- Visviva 17:26, 15 September 2007 (UTC)

These seems a little excessive; IMO all of these entries should simply be soft redirects to the hangul entry using {{hanja form of|}}. I don't see any need for a "Usage notes" section at all, although a little superscript "?" or something might be reasonable. Neither do I see the need to treat the hanja forms of nouns as if they were not nouns (as proposed by A-heun). Having a Category:Korean hanja nouns (or "hanja forms of Korean nouns" if that's preferred) should be quite sufficient. -- Visviva 10:27, 15 September 2007 (UTC)

Well, you say above that hanja forms are sometimes used "for disambiguation and precision"; wouldn't simply defining hanja forms in terms of hangul forms remove that benefit? (I see that {{hanja form of}} has a slot for a gloss, which is great, but the way it's written, it makes it sound like the gloss is for the hangul form, not for the hanja form, and like the hanja is exactly synonymous with the hangul.) Or am I misunderstanding something? This seems especially to be a problem because, from what I understand, hanja forms sometimes appear in parentheses after hangul forms, which means that someone looking up a hanja form might already know what the corresponding hangul form is, no? (Actually, maybe that's not true, since in that case the easiest way to get to the hanja page would be to visit the hangul page and click the link.) That said, this seems like this can easily be solved by rearranging {{hanja form of}} slightly, so that the gloss comes first, and therefore looks like a gloss for the hanja. Would that work? (N.B. I don't know Korean, so if something seems obviously stupid with what I've said, don't bother reading too deeply to figure out what I must "actually" mean; I probably just mean the obviously stupid thing.) —Ruakh_TALK 17:11, 15 September 2007 (UTC)

That's a very good point... and yes, I think that would work adequately. The alternative would be to have it link to a section of the hangul entry, but since our current section-labelling system does not really allow this to be done properly, it's probably better to do as you suggest. -- Visviva 17:26, 15 September 2007 (UTC)

• I'm just wondering about that Visviva discovered only the everyday use of hanja in Korea. One can find, indeed, also written English and Korean side by side in some books, advertising pamphlets, papers, websites etc in Korea. What is such English used in Korea or elsewhere? If this is not English, I have no idea, what this is.
• "Hanja is obviously Korean." That's an opinion without warranty. One can ask at the same time, whether that sort of assertion is not a political motivated and tendentious one, too. But what kind of political and tendentious? I have no idea. I've tried and try to call attention to wrong supposition.
• But what is hanja? I say, hanja is a Chinese character that is called so and used as an auxiliary means, according to circumstances and custom in Korea, but not as the Korean language. I don't know any official or private institutions which would have confirmed that hanja was Korean.
• To the templates: The reason I've made a proposal above is not to overlook hanja NOUN (see other forms of hanja on the ko), but to allude to hanja entries like 建物, 競馬, 怪物, 九月, 木曜日 ...that are classed in the category:Korean nouns automatically by the template:ko-noun. They are not Korean noun. By using the template:ko-hanja it doesn't happen. --아흔(A-heun) 08:44, 16 September 2007 (UTC)

  아흔 (A-heun) makes an excellent point. We only categorize the lemma forms of English nouns or verbs in Category:English nouns and Category:English verbs. It makes sense, then, to use a category other than Category:Korean nouns for words written in this alternative (non-canonical) orthography.

  This is the English Wiktionary, though. Here, language headers identify the language of the term or phrase, not the script in which the expression is written. So, we will still need to use "==Korean==" for hanja entries. (Notice "==Japanese==" in sayonara?) Rod (A. Smith) 09:38, 16 September 2007 (UTC)

OK, I've reworked {{ko-noun}} to automatically place all "hj"-tagged entries in Category:Korean hanja nouns, which is a category we were due for in any case. -- Visviva 11:33, 16 September 2007 (UTC)

Excellent. Thanks, Visviva! Rod (A. Smith) 18:45, 16 September 2007 (UTC)

I update {{hanja form of}} to show a superscripted "?" that links to Appendix:Korean script. The entry 建物 now uses that template. Comments? Rod (A. Smith) 21:23, 16 September 2007 (UTC)

Looks good to me, thanks. -- Visviva 08:57, 18 September 2007 (UTC)

If the romanized forms of Korean words come into general use, so that they are used in normal running text the way hanja often are -- if, that is, instead of phrases like this: "建物이 所在하는 土地" [11], we have phrases like this: "geonmul이 sojae하는 toji", then certainly it would be just as appropriate to have entries for romanized Korean words. But I have never seen Korean written like that; have you? -- Visviva 11:33, 16 September 2007 (UTC)

Nesting problems in Syllable entries

What is the correct format for an entry such as 미, where one hanja reading also functions as a freestanding noun? Should it be Etymology 1> Syllable, Etymology 2> Noun, even though the first etymology is superordinate? If not -- if something like the current Etymology > Syllable-Noun format is OK -- where should the pronunciation section go? It can't go in the usual place under Etymology, because it would apply only to 美 and other 미 hanja which are also read with a long vowel. Would it ever be permissible to have a Etymology > Noun > Pronunciation structure, with pronunciation at L4 or L5? -- Visviva 08:56, 18 September 2007 (UTC)

Good question. Wiktionary:Votes/2007-06/Level of basic headings attempted to answer this situation. If you can think of a something, reasonable, take it to WT:BP. In the meantime, I think the convention is any of the following:

• to group etymology by pronunciation: ===Pronunciation 1===/====Etymology 1====/=====Noun=====/====Etymology 2====/Various/=====Syllable=====/===Pronunciation 2===/====Etymology====/Various/====Syllable====. This layout seems problematic because some speakers may have (or develop) different pronunciations.
• give an etymology header (even without etym text) for each syllable: ===Etymology 1===/====Pronunciation====/====Noun====/===Etymology 2===/====Pronunciation====/====Syllable====/===Etymology 3===/====Pronunciation====/====Syllable====/===Etymology 4===/.... This layout seems overly verbose.
• for a more abbreviated variation of above: ===Etymology 1===/====Pronunciation====/====Noun====/===Etymology 2===/Various/====Pronunciation====/[long vowel]/====Syllable====/===Etymology 3===/Various/====Pronunciation====/[short vowel]/====Syllable====.

None seem ideal to me, but the last is perhaps the least poor. Rod (A. Smith) 16:29, 18 September 2007 (UTC)

Improving conjugation

I'm thinking it makes the most sense to move the nominal, adverbial, determinative, and other non-verbal forms of verbs and adjectives into ====Derived terms====, and leave only the sentence-final forms under "Conjugation." This seems to accord with normal practice for entries in other languages, and has the added bonus of cutting the conjugation table down to a mere screenful or so. If there are no objections, I will change {{ko-conj-verb}} and {{ko-conj-adj}} to the format shown at {{ko-conj-verb/draft}}. It will probably be sensible to create a new template or templates for the Derived terms section, basically copying the material cut out of the conjugation section, perhaps with some additions. -- Visviva 03:04, 7 February 2008 (UTC)

I don't understand why you'd want to remove those forms, nor why you think their equivalents are absent in most other languages. To me, they are useful and seem similar to participles, which most of our conjugation templates include. Rod (A. Smith) 03:41, 7 February 2008 (UTC)

Huh. Well, okay; somehow that's not the response I expected.  :-) I'll have to devote some further thought to this. Maybe we could break the honorific and non-honorific forms into separate boxes? -- Visviva 07:45, 7 February 2008 (UTC)

If a goal is to reduce the height of the conjugation box, splitting it up seems helpful. We could split out the forms that are not sentence-final, in line with your original thinking, or split out the honorific forms, as you mention above. Each format has its own advantages. If it's not so much about box height but about visual clarity, another option is to apply some more shades or colors. Rod (A. Smith) 07:33, 8 February 2008 (UTC)

Jamo

Seeing the organization at w:List of hangul jamo, I have a concern about how we handle the jamo variants in Unicode. The Halfwidth Jamo are all soft redirects to the Hangul Compatibility Jamo which seems correct and in keeping with practices for half/full-width variants. All other forms of jamo (in either Hangul Jamo, Hangul Jamo Extended-A, or Hangul Jamo Extended-B), however are being hard redirected to these compatibility "letter" jamo when it exists. This was talked about in the BP before with some saying "there are two codepoints for every jamo so we should just use the compatibility jamo" and others saying "we should have a page for each codepoint". The situation is slightly more complicated so I'd like to rediscuss the current practice. First off, it's not always that there are two identical codepoints for the same jamo. There are many cases where there is no compatibility jamo codepoint including cases where a consonant has both initial & final forms as well cases where a jamo has only an initial or only a final form. This is poorly documented currently on our entries and is confusing (at least to me).

I'd like to suggest an approach that Visviva appears to have tried at ᄫ . Assuming all the codepoints exist in documents, each would get its own page (no hard-redirects though it could soft-redirect to a more common form) and the appropriate {{character info}} templates are used to show encoding details and link to the other forms. If one form is to be considered the "lemma" that is fine, but we'd have to be able to handle the case where a consonant has an initial and final form but no compatibility form. An ancillary benefit would be that this bring practice into line with WT:REDIR. Does this seem reasonable? --Bequw → ¢ • τ 19:46, 16 December 2009 (UTC)

This is the approach I would prefer (unsurprisingly). Redirects confuse and annoy, especially when people want to know which specific form they have ended up with (e.g. in a document file), or validate that they are using the correct form (e.g. when building Middle Korean syllabic blocks by hand).

As far as lemma entries, I would suggest that these be at the compatibility jamo where available (since that is the form most likely to be found in the wild), and otherwise at the syllable-initial form. -- Visviva 05:16, 5 January 2010 (UTC)

RL circumstances have massively stalled me on this, but if there are no objections I'll go through and soften the redirects within the month. -- Visviva 05:16, 5 January 2010 (UTC)

Hanja entries and Northern vs. Southern spelling

When a single string of hanja has different hangeul spellings in North vs. South Korea, is there any specific template I should use, or should I just put two entries on separate lines? (If my question doesn't make sense, see 規律#Korean for an example of what I did). Thanks, Quant18 (talk) 14:01, 8 April 2012 (UTC)

Sound changes reflected in the romanisation (RRK)

Copying this in full from Wikipedia for consideration. Please pay special attention to jamo "ㄹ (l)" and "ㅎ (h)".

Consonant letters

Hangul		ㄱ	ㄲ	ㄴ	ㄷ	ㄸ	ㄹ	ㅁ	ㅂ	ㅃ	ㅅ	ㅆ	ㅇ	ㅈ	ㅉ	ㅊ	ㅋ	ㅌ	ㅍ	ㅎ
Romanization	Initial	g	kk	n	d	tt	r	m	b	pp	s	ss	-	j	jj	ch	k	t	p	h
	Final	k	k	n	t	-	l	m	p	-	t	t	ng	t	-	t	k	t	p	t

The revised romanization transcribes certain phonetic changes that occur with combinations of the ending consonant of a character and the initial consonant of the next, for example Hanguk → Hangug-eo. The significant changes are highlighted:

	next initial →	ㅇ	ㄱ	ㄴ	ㄷ	ㄹ	ㅁ	ㅂ	ㅅ	ㅈ	ㅊ	ㅋ	ㅌ	ㅍ	ㅎ
previous ending ↓			g	n	d	r	m	b	s	j	ch	k	t	p	h
ㄱ	k	g	kg	ngn	kd	ngn	ngm	kb	ks	kj	kch	k-k	kt	kp	kh, k
ㄴ	n	n	n-g	nn	nd	ll, nn	nm	nb	ns	nj	nch	nk	nt	np	nh
ㄷ	t	d, j	tg	nn	td	nn	nm	tb	ts	tj	tch	tk	t-t	tp	th, t, ch
ㄹ	l	r	lg	ll, nn	ld	ll	lm	lb	ls	lj	lch	lk	lt	lp	lh
ㅁ	m	m	mg	mn	md	mn	mm	mb	ms	mj	mch	mk	mt	mp	mh
ㅂ	p	b	pg	mn	pd	mn	mm	pb	ps	pj	pch	pk	pt	p-p	ph, p
ㅇ	ng	ng-	ngg	ngn	ngd	ngn	ngm	ngb	ngs	ngj	ngch	ngk	ngt	ngp	ngh

For consistency and simplicity, I suggest to remove comma-separated values and leave just the first value, like ㄷ + ㅎ = th (not, t or ch).

I suggest to review this section Wiktionary:About_Korean#Han-geul.2C_Romanization.2C_and_English_translation and RKK (2000) a more strict standard.

In addition, posting the vowel transliteration table, also from Wikipedia:

Vowel letters

Hangul	ㅏ	ㅐ	ㅑ	ㅒ	ㅓ	ㅔ	ㅕ	ㅖ	ㅗ	ㅘ	ㅙ	ㅚ	ㅛ	ㅜ	ㅝ	ㅞ	ㅟ	ㅠ	ㅡ	ㅢ	ㅣ
Romanization	a	ae	ya	yae	eo	e	yeo	ye	o	wa	wae	oe	yo	u	wo	we	wi	yu	eu	ui	i

--Anatoli ^{(обсудить}/^вклад) 00:59, 8 May 2013 (UTC)