Wiktionary talk:Idioms

First response

 * This all seems more theoretical than practical, probably because it dates to the early days before we had much practical experience. In practice, if an entry becomes encyclopedic (or more commonly, when someone dumps in an encyclopedic entry in lieu of a dictionary entry), the encycopedic material moves to Wikipedia.  Detailed lexicographical discussion ends up, unsurprisingly, on the discussion page.  We now have over 50,000 entries, and relatively few have links to Wikipedia.


 * There does seem to be quite a bit of dictionary material on Wikipedia which needs to move over eventually. Idioms are a good example.  Wikipedia currently has a fairly long list of idioms, most of which aren't in wiktionary yet.  Besides potentially duplicating effort on Wiktionary, a single listing page forgoes all the advantages of the dictionary format.  Fortunately, we now have Category:Idioms, as well as Category:English idioms and Category:French idioms, which are starting to fill in as more idioms get entered and more existing idioms get tagged accordingly.


 * The striking thing about this is that the Idioms category integrates the "Idiom Dictionary" aspect of Wiktionary smoothly with the rest of Wiktionary. I'm becoming more and more convinced that the full power of the Wikimedia category machinery is only starting to be tapped here.


 * The "defining vocabulary" issue seems particularly esoteric. I'm not aware that any of the well-known print dictionaries has an explicit defining vocabulary, and certainly the lack of such a vocabulary hasn't stopped us from acquiring over 50,000 entries.  I don't think trying to establish a defining vocabulary early would have worked very well anyway.  It would be instructive to look at the whole database now to see which words are defined in terms of which others.  This could well suggest targeted changes to improve consistency.


 * One would expect to see a core group of words that tend to turn up in definitions, and one would expect to see some cycles that will be difficult if not impossible to break up, but who knows. I think the database is only just getting big enough to do this sort of data mining meaningfully.


 * It might also be interesting to establish a "Core Vocabulary" category and tag entries with it, and then see how well that lines up with the results of the data-mining approach. I'm personally not sure what to expect, except that there are almost certain to be some surprises. -dmh 21:30, 13 Jan 2005 (UTC)

Translation of idioms
I may well be preaching to the converted here, but a little plea...

When translating idioms, it is essential to provide idioms in the target language. Anyone who knows anything about translation will tell you that translating word for word is simply wrong. I find it unlikely that both French and Spanish should have idioms that match English's the grass is always greener on the other side verbatim, and so I have commented them out. I might be wrong, but idioms very rarely translate word for word into other languages. The point of providing translations is, of course, for the user to make himself or herself understood in a foreign language, and verbatim translations do the opposite.

So please provide equivalent idioms and do not translate word for word. This applies to other multiple-word entries as well, of course. &mdash; Paul G 16:44, 3 August 2005 (UTC)


 * A further request: when defining idioms across languages, please provide both a literal translation (the grass is greener) and the meaning of the idiom (the stuff further away tends to look better from here). --Dvortygirl 02:04, 4 August 2005 (UTC)


 * A literal translation is very close to a translation word by word so which convention should we follow ? Should the literal translation be given in the English and/or the non-English entries ? Should we copy the idiomatic meaning from the English entry in the non-English entries ? See this discussion about Czech proverbs Thanks in advance. ThomasWasHere 18:02, 25 March 2008 (UTC)

Debate on "fictional character" moved from rfd
No more than the sum of its parts: a character who is fictional. &mdash; Hippietrail 18:34, 25 September 2005 (UTC)
 * Does not mean 'a fictional distinguishing feature', 'fictional moral strength', 'a fictional written symbol' (such as those in Dr. Seuss's On Beyond Zebra), etc. but only 'a fictional individual appearing in a literary work.' So it doesn't mean any character that is fictional. Less than the sum of its parts, thus idiomatic.  —Muke Tever 19:55, 25 September 2005 (UTC)


 * Muke, I don't think you know what "idiomatic" means. If a combination of any of the senses of the component words gives the meaning, then it is not idiomatic. For almost every phrase in every language there can be more than one combination of senses yet only a small minority of phrases are idiomatic.
 * Non-idiomatic phrases in your previous post which can be interpreted in multiple ways: not mean, moral strength, written symbol, literary work, doesn't mean, any character, that is, the sum, sum of, its parts, thus idiomatic. Do you really believe Wiktionary will be better if you include "definitions" for all of these phrases just because in your post you intended only one meaning for each?
 * Somebody lying about their non-existant moral strength, an invented kanji in a work of fiction, and every other combination of fictional and character are completely valid uses of those two words. Find a print dictionary of idioms that includes "fictional character" or do you believe its only opaque when the amount of paper used by a dictionary has no limit? &mdash; Hippietrail 21:00, 25 September 2005 (UTC)


 * I said it was idiomatic, not that it was an idiom. If anything I would call it a compound word.  The first attestation of “fictional character” in a different meaning from that given that I was able to find after several pages of Google Print hits: “In describing the fictional character of everyday world phenomena, including persons, Tsongkhapa gives two interpretations of the phrase ‘illusion-like.’” in a book by Thupten Jinpa, a Tibetan.  (Of course, I'm not sure how 'illusion-like' is a phrase...)


 * That’s interesting. Are you saying that idiomatic itself has an idiomatic sense besides the literal sense “of/relating to/pertaining to an idiom”?


 * Look at idiomatic's first sense—no, I didn't add it—or Webster 1913's only sense (compare it to his 'idiom' — our usual sense of idiom is #2, while the sense of 'idiomatic' derives from #1.) —Muke Tever 00:11, 27 September 2005 (UTC)

Anyway: “The object is to describe what it takes to use a language properly as a member of society. Part of this is knowing what things to say, when to say them and how to say them in conventional ways. [...] Instead of striving to keep the lexicon small we need to enrich it. In fact we apply the terms ‘lexicon’, ‘lexeme’ (or ‘lexical item’) and ‘lexicalized’ in ways quite different from the grammarian. Now these terms are defined with respect to cultural facts as well as with respect to purely structural criteria. Complex words and compounds, and perhaps phrases, are considered part of the speaker's cultural lexicon if we can show that they have entered the social tradition, that they have attained the status of social institutions, being recognized as conventional ‘names of things’, as ‘terms’ in a set or terminology, as ‘set phrases’, and perhaps as ‘appropriate things to say’. All grammatical strings are not socially equal. We award special status to those strings that are culturally significant, even though they may also be perfectly grammatical. The upshot is an enormous increase in the number of lexemes compared to the ideal grammarian’s dictionary.” Andrew Pawley, as quoted in Making Dictionaries In the same source is quoted his list of criteria for lexeme/headworthiness, which I have beforehand shared with the IRC channel: Generally the only criterion I see used here for multi-word entries is #21. I do know that you are personally against some of these, but as I've mentioned before I still don't agree that being an idiom or not is the best criterion to fixate on. —Muke Tever 03:01, 26 September 2005 (UTC)
 * 1) The naming test: Can the candidate for a lexeme be referred to in questions or statements such as the following: ‘What is it called?’ ‘It is called X.’ ‘We call it X, but they call it Y.’
 * 2) Membership in a terminological system: [...] Does X encompass other terms; can one say ‘it (dog) is a kind of X (animal)’ (=generic)? Is it a member of a set of similar things; can one say ‘X (a chair) is a kind of Y (furniture)’ (=specific)? Can it be used to show contrast; ‘is it a kind of X (fruit), but not a Y (vegetable)’? Does it have synonyms or antonyms?
 * 3) Customary status: Does the use of the phrase imply certain behavior patterns, values, or sequences of activities that are known by society at large? They represent conventionalized knowledge. For example, expected behavior at the front door is different from at the back door (besides their participation in idioms), indicating that these function as cultural units (lexemes) that are more significant than the sum of the parts. Consider go to the mosque, get off work, take a vacation.
 * 4) Legal status: Some phrases have such status that they are codified in legal usage: driving under the influence, breaking and entering, assault and battery, justifiable homicide. Even so-called ‘primitive’ societies with unwritten languages have categories of this sort for dealing with things like marriage negotiations and litigations over land, property, and adultery.
 * 5) Speech act formulas: Every language has some formulas “which carry out conversational moves” (Pawley 1986:106). For example, excuse me, how are you, y'all have a nice day, etc.
 * 6) Use of acronyms: This is often proof that a multi-word phrase represents concepts that have attained conventionalized or institutionalized status. Consider: VIP, DWI/DUI, IQ, RBI, SAT, ASAP, PTO, PTL, AWOL, BS, RSVP, R and R; in Indonesia: KB, DKI, KK, ABRI, DPRD, GBHN, etc.
 * 7) Single-word synonyms: the only one of its kind ↔ unique.
 * 8) Belonging to a terminological set: This is similar to (2), but focuses more on a pair of antonyms. Consider: tell the truth ↔ tell a lie, take care of ↔ neglect.
 * 9) Base for inflected or derived forms: short temper → short-tempered; ooh and ah → oohing and ahing, Indonesian ke mana → dikemanakannya (‘to where’ → ‘wind up where’).
 * 10) Internal pause unacceptable: The unacceptability of inserting a pause in the middle of clichés, idioms, and compounds is partial indication of their functioning as a unit. Consider the functional differences between bunch of baloney vs. bunch of bananas. One can say two bunches of bananas, but cannot do the same with the figurative sense of bunch of baloney.
 * 11) Inseparability of constituents: Insertion of other material changes the unity or naturalness of a phrasal lexeme. Consider: lead up the garden path. Saying lead up the beautiful garden path shifts it from a figurative to a literal interpretation. This is similar to (10) above.
 * 12) Ambiguity as to whether it should be written as a single word: whatchamacallit, thingamajig, man-in-the-street, oneupmanship.
 * 13) Conventionally reduced pronunciation: bosun (boatswain), won't, can't, o'clock, Newfoundland, Christmas, Worchestershire, thruppence (three pence) etc.
 * 14) Conventionally truncated forms: Widespread occurrence of shortened forms often indicate their role as a lexeme in the language: exam(ination), rad(ical), ex-con(vict), con(vict), con(fidence man), con(fidence trick), ex(-husband/-wife), pro and con, etc.
 * 15) Omission of headword: The modifier stands metonymically for the whole: She had an oral (examination), He had a physical (examination), A short (circuit) cut off the (electrical) power.
 * 16) Omission of final constituents: This often implies conventionalized knowledge: If you can’t beat ’em..., A stitch in time..., I haven’t the faintest (idea). These elided forms are often marked by peculiar intonation.
 * 17) Stress and intonation patterns: Different languages give different phonological clues for what is seen to function as a unit. English often uses stress and intonation. Government jargon is often coined through these means. Consider political matters memorandum.
 * 18) Invariable constituents or grammatical frame: The demanding and rhetorical Who do you think you are? does not have the same impact in the future. Kick the bucket does not mean the same when put in the passive. The thought had crossed my mind, and he took the law into his own hands are unnatural in the passive. Compare also stripped down formulaic sentences easier said than done, spoken like a man! There are also syntactically irregular or archaic idioms like easy does it, no go, no way, be that as it may, (she) wants in, once upon a time.
 * 19) Use of definite article on first mention: In English this can indicate the conventionalized nature of the ‘object’, showing the speaker assumes the identity is understood by the addressee: the fire department, the foreign legion, the eight ball.
 * 20) Writing conventions: Where there is a written tradition these may provide clues to perceived status as a unit. Capitals may indicate lexemes that are not typical proper nouns: Third World, Big Bang, Inner City. Beware that where a society has the luxury of supporting a literary community, some writers manipulate the use of capitals for unconventional purposes. Quotation marks may also indicate unitary status: he was considered a ‘bad boy’. Orally, some speakers use so-called or a preceding pause to mark an equivalent to quote marks.
 * 21) Unpredictability of form-meaning relation in semantic idioms: kick the bucket, chew the fat, shoot the breeze.
 * 22) Arbitrary selection of one meaning: Notice that button hole is a hole FOR putting buttons THROUGH, whereas bullet hole is a hole MADE BY bullets, post hole is a hole FOR setting posts IN, etc.
 * 23) Use in ritual language of parallelism: This is a special case of (2) and (8). Ritual language in parallelisms is widespread. It is found, for example, in Biblical Hebrew and many Austronesian languages, particularly in eastern Indonesia (Fox 1988). Existence as a paired entity in this context is sufficient for justifying its status as a conventionalized unit, and hence a lexeme.
 * I agree with everything Pawley says here and still believe phrases such as fictional character, Egyptian pyramid, radioactive material, item of furniture, leaf storm, and whichever others I've nominated on this basis to be unsuitable for a dictionary. Perhaps you should contact him for his views on a list of our most controversial terms for clarification. I think this is one of the most important issues for Wiktionary to decide. How inclusive should it be? I truly believe that large numbers of such terms will mislead ad confuse people trying to improve their understanding of English, particularly children and speakers of other languages.
 * While it seems that I'm sometimes seen as a radical on Wiktionary I think that embracing such terms is far too radical a break from traditional dictionaries. We need to decide whether we're to be a traditional dictionary or one of these new-fangled all-embracing dictionaries some contributors seem to want. If people choose the latter I believe those entries should be marked in some way so as not to bewilder people who are looking for a traditional dictionary. &mdash; Hippietrail 16:36, 26 September 2005 (UTC)


 * A big problem is there doesnt seem to be any consistent idea of which out of several pretty-much contradictory kinds of dictionary Wikt wants to be. I know from studying other languages that I would much prefer a dictionary that tells me the idiomatic way to say things in a language; suppose I'm a Spanish speaker, I would want it to give me  the set term "fictional character" and not leave me in the lurch to invent something non-'idiomatic like "personage of fiction" (after the Spanish equivalent personaje de ficción.)  —Muke Tever 00:11, 27 September 2005 (UTC)


 * Thank you, yes! I guess I have the same perspective from the opposite side, teaching English as a second language and wondering what some of my students were thinking in their essays. I'm not going to write down 20+ points, but basically there are certain combinations that just click, fictional character and financial stability being two examples. On the flip side, I don't suppose someone has already deleted bookshop as nothing more than a sum of its parts? Davilla 21:16, 27 September 2005 (UTC)


 * Compounds which are written as one word are counted as one word, same goes for hyphenated compounds. This is of course the best sign that a term has entered the lexicon and is beyond dispute by any of our contributors. Two-word spelling variants of such terms are of course also always allowed. &mdash; Hippietrail 14:28, 29 September 2005 (UTC)


 * Sufficient but not necessary, as you well know. Placing some of my earlier sarcasm aside, the boundary of what constitutes a word is of morphic and cyclic nature from what I understand of linguistics. It's quite variable between languages (please Germanthinken) and in my opinion more an artifact of language as it is written. Insomuch as "insomuch" is a word, along with "nonetheless" and "notwithstanding", the spacing doesn't add as much weight to the argument. Alright? ;-) Davilla 18:36, 2 October 2005 (UTC)


 * I don't think that it's accurate to say that #21 is the only one of the criteria that we would apply. The list is nevertheless a useful set of guidelines, but few of those criteria can act alone in determining.  I certainly don't think that it's enough to say that a particular combination of words occurs frequently.  There needs to be more than that.  Living examples will be the basis for determining whether a lexeme should be recognized.


 * I said that #21, being an idiom, is the only one that is applied, because otherwise it would be entirely insensible to nominate it for deletion solely on the grounds that it is not an idiom.


 * A dictionary that tells you the idiomatic way to say things is being prescriptive. How we handle personaje de ficción is interesting.  The Spanish speaker's intuitive solution would sound strange to the English speaker, but he would understand it.  Nevertheless, the issue there probably has more to do with the structure of the two languages, and how generally de in Romance languages is translated.  "Of" is technically correct, but there is a broader patter that applies which is also linked to the English practice of having the adjective precede the noun.  I would, however, find "fictional personage" a perfectly acceptable alternative that would break stylistic monotony in someone's writing.


 * A dictionary that tells you the way things are said by native speakers of English is being descriptive. That is what the whole conflict between description and prescription is, the difference between everyday, idiomatic language, and language augmented by rules imposed from without.  "Fictional personage" is acceptable English but is not a set term in the way that 'fictional character' is, and someone, say, who wrote literary criticism only saying "fictional personage" and never "fictional character" would be most unusual.  (Incidentally, "fictional personage" is even rarer in raw google hits than the whole phrase "has a fictional character" discussed below.)


 * What kind of dictionary we want is an important question. There are some major questionss that influence that.  How prescriptive can we be?  What are the proportional roles of a translating and an own language dictionary?  What is the relative importance of the historical language, and modern developments in the use of the language?  In the interpretation of text, how much weight to we put on context and connotation, much of which can only be translated with great difficulty?  A reasonably competent writer will give new meanings to words without the need to have them appear in a dictionary, and his readers will often understand the implicit subtleties.  There would be a tremendous difference between the sentences, "John is a fictional character," and "John has a fictional character."


 * Is that claim verifiable, or is it Original Research based on intuition? Querying the Google corpus for 'has a fictional character' (where 'has' is a content verb, i.e. unlike "not since X has a fictional character done Y", which appears to be unusually common fsr):
 * It is documentary in style, but it has a fictional character at its center.
 * ...l'Acadie, the virtual nation that has a fictional character as national icon...
 * I'm sorry honey, but surly almost everyone on this forum has a fictional character... el gilko? Either that or some very cruel parents to give their children such strange names.
 * Except it has a fictional character in it.
 * Perhaps he has a fictional character called Roy G Biv?
 * Each wing of this alliance has a fictional character with his own elaborate plot line devoted to it:
 * I am writing a children's book that has a fictional character who wanted an American Flyer...
 * Chances are that every family has a fictional character or two to grow up with and through the years some will make their way to the written page.
 * ...the fictional drama has a fictional character facing a moral dilemma...
 * HSTP as its integral compnent has a fictional character called ÔSawaliramÕ to whom children are encouraged to send their queries...
 * ...he even has a fictional character (Grandfather Twilight) endorsing him in The form of a press release on his website...
 * In "The Greening of Mars" (1984), with Micheal Allaby, he has a fictional character, Travers Foxe, launching rockets to Mars...
 * ...the list, released in this week's magazine, also has a fictional character, Arli$$ Michaels, agent, at No. 100.
 * A fictional movie that makes all the scientific sense of a Wile E. Coyote battle against the roadrunner has a fictional character that looks like Cheney, and that fictional character doesn't believe in something.
 * In "The Poisonwood Bible," novelist Barbara Kingsolver has a fictional character in 1960 overhear official US plans...
 * ...that biography was just called Dutch, has a fictional character in it who actually is Morris...
 * I imagine when one has a fictional character as a hero, what one is actually doing is admiring The possibility of what "could be,"...
 * The violence in this movie actually happened to a person. Any other movie has a fictional character, and the violence doesnt serve a purporse, except to be entertaining.
 * Each player has a fictional character, called a Player Character (or “PC”)


 * Now, it is true that *occasionally* the word is used in (only) one other meaning:
 * Showing that his game has a fictional character, his Q's couldn't reign over Lindy's J 9, which hit for two pair.
 * This information, which is not found elsewhere, has a fictional character and seems to have derived from a distortion of the traditions on the Jewish services on the Temple Mount. (in a translation from a Hebrew book, may be representing Hebrew turns of speech)
 * In 1 C. 7:7 it has a fictional character. (in a translation from a German book, same caveat as for the previous)
 * The writer is usually forced to be highly selective in the presentation of the story. Even the most complete ethnography still has a fictional character,...


 * If the Ultimate Wiktionary supporters ever stop trying to be everything to everybody, and present some working software it will go a long way toward dealing with the translations issue. This would allow us to devote more of our energy to the other questions.


 * Amen!


 * I think that evidence remains the most important feature to a credible dictionary. I believe that everything here should be verifiable.  For common words a reference to a well-known dictionary may be enough.  For a composite term like "fictional character"  we really should have evidence that would establish what makes it more special than its component words.  Eclecticology 20:05, 29 September 2005 (UTC)


 * As far as well-known dictionaries, fictional character is in WordNet. Anyway, as for the headworthiness criteria posted, it meets #1 (what's it called? a fictional character), #2 (according to wordnet, a fictional character is a kind of imaginary being, a protagonist is a kind of fictional character, etc.), and #22 (the meaning arbitrarily selected being: an imaginary person such as may appear in a story). —Muke Tever 04:17, 30 September 2005 (UTC)


 * Despite the heading for this section, this discussion is not really about fictional character. Though I would prefer deleting it, it is still a marginal matter for me.  Be that as it may, a number of other disputed terms have been mentioned in the course of this discussion, and that suggests that a much wider issue is at stake.  That wider discussion should probably take place on an other page, The issue would be less about what the uncontroversial meaning of these expressions, than about where they merit being a headword.


 * I saw prescriptivism embodied in your comment, "I would much prefer a dictionary that tells me the idiomatic way to say things in a language.&quot; This implieas that there is a unique way to deal with the translation.  "The idiomatic way" suggests a unique solution that gives no choice where in fact there are alternatives such as the one that I suggested.  Prescribing specific terminology can be just as prescriptive as putting it in terms of some set of inviolable rules. -Ec

Please, everybody sign your comments. I cannot see who holds which position! For me the proof that fictional character is not idiomatic is that even though it may be the most common way (I haven't researched it), there are certainly other ways to say it which I also do not count as warranting dictionary articles: fictitious character (530,000 Google s/pl/misspellings) - which actually has an entry in thefreedictionary.com, made-up character (44,300 Google s/pl), and invented character (25,400 Google s/pl). Another is that the meaning is not hurt by inserting a word between fictional and character. Some I found: fictional Balkan characters, fictional holiday character, fictional, paranormal character, fictional, animated character, fictional composite character, fictional, unrealistic character, fictional, mustachioed character, fictional, archetypal character, fictional Cowboy character, fictional Kilburn character. These all break the "Inseparability of constituents" rule above plus those with commas also break the "Internal pause unacceptable" rule.

While the fact that fictional character is usual may be interesting, it is specialist information which would be more at home in a dictionary of collocations. These actually exist and I have seen several marketed to Japanese learners of English. I am in favour of enhancing Wiktionary with collocation information, but creation of headwords which mislead the casual reader into thinking such phrases are set in stone is not a good way to do it. &mdash; Hippietrail 19:28, 1 October 2005 (UTC)


 * Sorry, about the missing signature. Some of us who do regularly sign our posts will still occasionally and inintentionally miss doing so.  That was a strong argument.  One more term that could be on your list is literary character. Eclecticology 05:01, 2 October 2005 (UTC)


 * There's some great insight. These common phrases seem to be of interest to a number of people, so let's approach this a different way by rephrasing the question. How would one go about creating a wikified dictionary of collocations? By including synonyms, translations and so forth, Wiktionary is already a number of other things. I'm sure if we work this out we could find a way to include common phrases in some fashion. The objection seems to be that "fictional character" doesn't deserve a dictionary entry, one that defines the term. Well, maybe it doesn't, but that wouldn't mean it should be ignored altogether. These and other common phrases are, in the opinion expressed, integral to language in the same way as, or even to a greater extent than, some of the sayings.
 * Here's a better example, taken straight from my Chinese students' work. Not knowing how to express themselves in English, several apparently entered a common Chinese word into their translators and came up with "conflagration". Now, conflagration&mdash;that is, its translation&mdash;may be quite common in Chinese, but this poor English speaker had to look it up. Turns out what they meant was forest fire, a pretty obvious sum of words to us but, despite having both words in their vocabulary, not to them. And I don't think it's the fault of the translator either, unless anyone would suggest that Dante should have written about forest fires instead of his inferno.
 * Here's the dilemma. If someone were to look up "conflagration" from the translation... or however they get there, there's got to be a page, and on that page, if Wiktionary is worth its weight in kilobytes, are the synonyms "inferno" and "forest fire". Under strict rules, the latter will always be dimmed red because there can be no "forest fire" page because it's too obvious. Now I don't mind a work in progress, but I have considerable objection to the idea that scattered links should deliberately be permanently broken. Either it shouldn't be linked at all, which sounds like a really bad idea in terms of maintaining what should or shouldn't be linked, or there should be something there. Now whatever's there doesn't have to be an entry. It could be a meta command letting the server know that the double-bracket designation for this string isn't really a link. It could be a link to the separate words in some sort of skeleton entry. But it all comes back to that question of a dictionary of collocations and how best to (externalize or) integrate that, which until someone frames a better question is what's really at the heart of this. Davilla 18:09, 2 October 2005 (UTC)
 * I think best practice for collocations is to look at some actual print collocation dictionaries. Unfortunately I do not currently have access to any. One thing we might think of doing is expanding our types of entries. For terms which would belong in a translating dictionary, rhyming dictionary, thesaurus, phrasebook, or collocation dictionary - but not in a "regular" dictionary, we would need to leave out the defintion section of course but also we would need to mark it in some way as being a "specialist" entry - or whatever we might like to call such things. We might even think about changing our headings structure with one for each type of dictionary - but that would be an awful lot of work. But we need to think more about our headings and their structure anyway, since the Flemish "I love you" (below) also warrants coverage of some kind but doesn't need a definition and doesn't fit into our current system of headings.
 * Having said that, there could be a case for forest fire since we don't really use that term in Australia and our transparent term bushfire gets entries in the dictionaries of other countries, sometimes being marked as Australian. We might also list brushfire / brush fire as a synonym for bushfire but not for forest fire.
 * Another thing re the Chinese learners of English. The problem could be due to the dictionaries they're currently using. I don't know much about Chinese but Japanese bilingual dictionaries have a poor reputation and are famous for just this type of thing. One example is the often-heard "very terrible". Also English teaching practices in Japan are not good with total emphasis on passing examinations and zero emphasis on conversational skills.
 * PS I'm happy for this conversation to be moved to a better place... &mdash; Hippietrail 20:58, 3 October 2005 (UTC)

DELETE -- Dipping my oar into the fire... (Yes, I just made that phrase up. Please don't add it as a word.) If you can look up fictional and character and understand what fictional character means, then you don't need a separate definition. --shark 23:54, 27 November 2005 (UTC)

Delete - not more than sum of parts Παρατηρητής 11:34, 1 January 2006 (UTC)

Word of Mouth
another idiom
 * Yup, and it's in Wiktionary: word of mouth!
 * Nbarth 01:34, 27 January 2008 (UTC)

Alternative forms?
How should we deal with alternative forms of idioms?

I raised this question also over at: Wiktionary talk:Alternative spellings


 * OIC: use
 * Nbarth 01:55, 27 January 2008 (UTC)

Form of idioms
I've listed a recommendation to use "one" or "one's", and to not use infinitives; this is based on common practice AFAICT; I can't find any discussion or policy to this effect.

Nbarth (email) (talk) 22:58, 19 February 2008 (UTC)

Header
What h3 header to use? I see some use ===Verb===. --Nemo 07:44, 22 August 2015 (UTC)
 * Usually you should use the POS of the phrase (i.e. Verb if it is a verb phrase, Noun if it is a noun phrase, etc.) When the phrase doesn’t have a specific POS, you can use Phrase. Some Chinese entries use Idiom; I don’t know if that’s the accepted practice or just older formatting. — Ungoliant (falai) 13:50, 22 August 2015 (UTC)