Wiktionary:Beer parlour/2008/July

Etymology trees

It would be nice to have a system in place to show the development of word stems to their present forms in modern language, and a section to show language branching. Perhaps with a legend for the differentiation of meanings across languages. Just a language tree would be nice to have. Thoughts?

For reconstructible PIE nouns there are some fairly complete entries in the appendix (*ph₂tḗr, *méh₂tēr, *bʰréh₂tēr, *swésōr, *gʷḗn, wĺ̥kʷos, *ḱḗr, *h₁nḗh₃mn̥ etc.). Semantic shifts and not so evident meanings are usually indicated with glosses when they occur. For the PIE roots - not so much complete entries unfortunately, other than for very basic ones as *bʰer-, *steh₂-, *h₁es-, *deh₃- which haven't changed much in reflexes neither in meaning nor shape even till today. You are welcome to contribute to any language family of your preference. --Ivan Štambuk 05:28, 1 July 2008 (UTC)

Scope of related terms

I am wondering about the proper scope of the section Related terms. I have added philosophy of science to the section Related terms of epistemology, as the distinction proper between the two terms is unclear to me; for instance, I am at loss whether Popper's falsification theory is a contribution to the former, to the latter, or both. WikiPedant has removed philosophy of science from epistemology per WT:ELE's definition of what counts as related term: only etymologically related terms count as related. Per WT:ELE his removal is right AFAICS.

However, I have been treating Related terms more broadly, and as I now see not in align with WT:ELE. For instance, I have added "computer science" and "information theory" as related terms of information science, based on the felt risk of confusion of the terms, not based on their etymological relationship.

So I wonder: where should I put terms that I want to have contradistinguished from the term of the entry? Or should they be not there at all?

Also, I have been moving terms from See also to Related terms, which too seems to be wrong.

--Daniel Polansky 07:39, 1 July 2008 (UTC)

If the additional terms are of real value for understanding the entry, and cannot be included either as part of the definition or as etymologically realted terms, then a ===See also=== section may be used. However, it should be used only sparingly, and some community members dislike seeing that section at all. --EncycloPetey 07:42, 1 July 2008 (UTC)

Thanks. I have now noted that See also is mentioned at WT:ELE and that it could be used for the purpose. I have now also noticed the existence of Wiktionary:Semantic relations, which however is not a policy and does not seem to be endorsed. It seems to me that the relationships between terms that I wanted to enter do not help understanding the entry alone; they help to understand a group of terms within a domain of discourse, and in other cases, they help to remind the reader that two or more terms have vastly different meaning, in spite of their syntactic similarity, such as information theory and information science. In any case, they are not about single terms; they are always about two or more terms. --Daniel Polansky 08:06, 1 July 2008 (UTC)

There is a limit to what a general dictionary can do. The semantic relations headers other than Synonym can be difficult to apply and can be difficult to understand (even Antonym in many cases). I sometimes amuse myself with such things, but I don't understand how they could be very useful to users. Often the best "See also" is just one or more {{pedialite}}. DCDuring TALK 00:00, 2 July 2008 (UTC)

I find synonyms, antonyms, derived terms, and related terms very useful, especially synonyms and antonyms. My perspective is one of a non-native speaker. I have spoken in favor of derived terms in a recent discussion in Beer parlour.

Specifically on antonyms: Often, I want to express an idea but only know its antonym, in whose entry I find candidate terms for my idea under the head of Antonyms. Also, there are pairs of words that syntactically appear as antonyms, but semantically are not; I've got no example right now. Antonym forming in English is not altogether regular, featuring the prefixes "non-", "non", "un", "im" and the like, unlike in my mother tongue where it is rather regular, so knowing which prefix applies to which particular word is useful. Also, when different antonyms are given to more senses, my confidence in the meaning often increases, again without examples.--Daniel Polansky 08:05, 2 July 2008 (UTC)

To clarify, "derived terms" and "related terms" aren't semantic relations per se. —Ruakh_TALK 18:04, 2 July 2008 (UTC)

Correct, thanks. I've mixed them up in my reply. --Daniel Polansky 09:36, 3 July 2008 (UTC)

Let me see if I understand:

1. Related terms = defined list of Semantic relations + words morphologically connected to head word that are not derived terms;
2. Derived terms are those derived from the word (compounds and other morphological derivatives using the unaltered headword or inflected forms)
3. See also = Semantic relations not on the defined list. DCDuring TALK 10:58, 3 July 2008 (UTC)

Yes regarding "See also" and "Derived terms", but no regarding "Related terms". All items in the "Related terms" sections should be etymologically related to the head word. If a term is semantically related to the head word but shares no history with it, it should not be included in the "Related terms" section, but perhaps in one of the sections devoted to semantic relations. I also use the "Related terms" as a repository for terms that clearly share some history with the head word, even if I don't know whether they were derived directly from it, hoping that another editor who knows more of the history of the words can sort it out. Rod (A. Smith) 15:59, 3 July 2008 (UTC)

If that is the case than the original question is that "surjection", "injection", and "bijection" are "Related terms" mutually. My association of Wiktionary:Semantic relations with "Related terms" was wrong, although based on the actual practice of using "Related terms" as a holding pen for terms related eihter semantically (same PoS, usually no direct etymological relationship) or etymologically (usually different PoS).

I would expect that "Related terms" should not include terms that share only minor morphemes like prefix and suffix, that at least one stem should be involved. DCDuring TALK 16:34, 3 July 2008 (UTC)

Related: Do injection and bijection qualify as related terms of surjection, per sharing a grammatical root, even if not a prefix? It seems to me that the definition of "related term" for the purpose of the Wiktionary sections deserves a more detailed elaboration that the one currently found at WT:ELE, including some borderline counterexamples, or even obvious counterexamples. --Daniel Polansky 15:30, 1 July 2008 (UTC)

Using Wiktionary:Semantic relations (which hasn't been rejected) and if I get these terms properly:

IF "jection" were a valid name OR if "mapping" were suitably defined, THEN

they would all three be "Coordinate terms" and could appear as "Related terms" in each other's entries ELSE

"bijection" is hyponym of both "surjection" and "injection", which are then, by definition, hypernyms of "bijection", but, not having one of the named semantic relationships, "surjection" and "injection" can only appear in each other's "See also" sections. DCDuring TALK 00:00, 2 July 2008 (UTC)

With my question, I intended to head in a slightly different direction: does "having a shared grammatical stem and differing only in prefix", such as "to project" and "to inject", qualify as "having strong etymological connections"? If not, I think "to project" and "to inject" should better be mentioned as counterexamples in WT:ELE#Related terms. --Daniel Polansky 08:05, 2 July 2008 (UTC)

I think that kind of relationship should fit into Etymology or Derived terms. To facilitate pursuing those relationships Etymology can and should include morphology for constructed words so that links are available to go to the entries that have the constituents of the word. Those constituents should have Derived terms that show other words derived from them. I don't think there is a good home for all aspects of the mixed semantic-morphological relationship that exists in the case of the mathematical "-jections". The semantic relationships, at least, can be presented at the level of an individual definition. Related terms and Etymology do not penetrate to the definition level. In the case of the "-jections", there is almost certainly a case to be made for creating an entry for at least one of "jection" or "-jection", which would then allow separate etymology for the various "jections" and allow the use of the existing etymological and semantic headings to display the relationships more clearly. DCDuring TALK 10:52, 2 July 2008 (UTC)

I support creating an entry for (deprecated template usage) -jection. (But I must say that I'm surprised to see you advocating it, as you requested that our entry for (deprecated template usage) -scribe be deleted, and I don't see how that case is different?) —Ruakh_TALK 18:04, 2 July 2008 (UTC)

I came up short of advocating it. I was just trying to advance the discussion. I am interested in how far the existing structures go. I don't like -jection (because it is like -scribe), but I haven't found any evidence that jection exists. In English, we have no provision for "Stems" as a part of speech heading, though we do have "Prefix" and "Suffix". It would be interesting to have a morpheme namespace to and from which we could have various mappings from and to principal namespace. This would give a home for "infixes" and other troublesome entries, possibly Symbols. DCDuring TALK 18:28, 2 July 2008 (UTC)

Yet Another Interminable Discussion about Wikisaurus

Hello everyone, your favorite mosquito here. I've got yet another question about wikisaurus. Now I know you have gotten to the point of ignoring me and letting me do my thing, but I have a question about the Wiktionary:Wikisaurus page. On it there is, of course, the list of words created to date. So far so good. Then we have this:

   * Help:Creating a Wikisaurus entry
   * Wiktionary:Thesaurus considerations - Original discussion about the project.
   * /purpose - Centre for discussion of the purpose of Wikisaurus
   * /improvements - Discussion about the direction and overall project.
   * /criteria - Discussion about the content and criteria for inclusion within the Wikisaurus project.
   * /format - Discussion about the formatting and general contents of a Wikisaurus entry.
   * /requested entries
   * /templates - short list

My question is: just how much of this is still:

• 1. valid
  2. needed
  3. confusing

I plan on doing the job for quite some time yet, but it wouldn't be impossible that someone would come along wanting to help. Is there a point when we can start doing cleanup of the resource pages so that people entering don't get hopelessly lost in the clutter? Amina (sack36) 21:46, 1 July 2008 (UTC)

And while we're on the subject of wikisaurus clutter, I'd like to point out a tiny glitch in something mentioned on the policy of Wikisaurus. It says to put in a section for linking non-synonymous non-antonymous related words. I have updated one record to follow that strategy. This one was not picked on purpose, guys, it was just the one that was mentioned as an example! May I point your attention to this page: http://en.wiktionary.org/wiki/User:Sack36/sandbox and then look at the other linkages I found that should be added to that page. I found them here: http://en.wiktionary.org/wiki/Wikisaurus:beer I haven't looked anywhere else and yet It still starts looking like a three book tome. What would you suggest? Amina (sack36) 00:17, 2 July 2008 (UTC)

It looks like Wikisaurus is gaining a little momentum, maybe enough that some of these issues can be discussed, instead of brought up and unanswered. In my opinion none of it is completely valid, especially since any of it is subject to change.

Format: Wikisaurus should be more than just a list of synonyms. Ideally it would actually explain when you would use what word under what circumstances, to the extent that we can generalize it. Somehow I don't think we can jump in with the experience of American Heritage though. Maybe a good starting point are collocations as illustation.

Criteria: Exact same criteria as for dictionary entries. Why duplicate the process? One of the advantages of linking to the main space is that it's easy to check. Having criteria is important because without it the crud builds up very quickly.

Templates: The ones I've made express the ideas that I've had about titling pages. Wikisaurus has an entry on vomit in the sense of "regurgitate". I don't think any single word can pin the concept down precisely. Both vomit and regurgitate have other meanings, but the intersection is a clearer, single concept. DAVilla 07:35, 16 July 2008 (UTC)

Wikibooks

The lead featured book over on Wikibooks is the introduction to Spanish. We've been talking about increasing our internet profile, and so I was dismayed to find that not only does this book not link to Wiktionary, but there isn't even a template for linking to Wiktionary over there! --EncycloPetey 19:10, 2 July 2008 (UTC)

I added a few wikt links to that book back in November ([1]), but it would certainly be nice to have an easier way to make those links. Rod (A. Smith) 19:35, 2 July 2008 (UTC)

I reckon it would help if some people just went through it and placed wikt links on all important words. I've started already :) --EivindJ 23:03, 2 July 2008 (UTC)

There really ought to be a prominent link from the first page too, don't you think? Something to let readers know that there is an available companion resource. --EncycloPetey 06:28, 3 July 2008 (UTC)

[2]—msh210℠ 17:56, 3 July 2008 (UTC)

Also started added a few links [3]. & [4] Mutante 09:45, 3 July 2008 (UTC)

Redirect from "article + noun"

I reckon it might be ok to make redirects from "article + noun" to noun. In English it might not be necessary to redirect "an apple" to apple, but the articles clearly plays a greater role in other language, like e.g. Norwegian or Spanish. Should we, when we stumble upon them, redirect la lunatica to lunatica and et tre to tre, or should we simply mark them for speedy deletion if people create them? I think it is plausible entries and I reckon many people enter the article when they search. --EivindJ 16:11, 3 July 2008 (UTC)

Articles are prefixes or enclitics in a number of languages, including the Semitic languages, the Scandinavian languages, Romanian, and Bulgarian. Certainly these should be kept either as redirects or as entries similar to "form of". In some cases, the form with the article is different from the base noun (e.g., Den Haag, 's-Gravenhage, La Haya, A Haia) and deserves a full entry. —Stephen 16:35, 3 July 2008 (UTC)

An old BP discussion decided (against my better judgement) that we should not have (most) Hebrew words that include clitics, not even as form-of entries. (Hard redirection was not discussed.) For some reason, that discussion was about conjunction and preposition clitics, but not the definite-article clitic.—msh210℠ 17:48, 3 July 2008 (UTC)

Here I was basically thinking about articles that indicates the gender of a word, like in Spanish and Scandinavian language. I cannot see that they deserve a full entry or "form of" entry, but at least I reckon they should be redirected directly to the noun. --EivindJ 18:15, 3 July 2008 (UTC)

The Hebrew definite article is different from the examples discussed there in that it's less clearly a clitic: it mostly attaches to individual nouns and adjectives, not to full nominals. For example, as I'm sure you know, "the two children" is not *"ha-sh'nei y'ladim" but rather "sh'nei ha-y'ladim", and "the tall children" is not "ha-y'ladim g'vohim" (which means "the children are tall") but rather "ha-y'ladim ha-g'vohim". Traditional Hebrew analysis has viewed nouns as having three "states" — indefinite (status absolutus: bayit = house), definite (status emphaticus/determinatus: ha-bayit = the house), and construct (status constructus: beit = house-of) — and likewise for adjectives but without a construct state. I don't advocate following traditional analysis on all points, but this strikes me as a sound way to look at it, until such time as we find that modern linguists with Hebrew expertise have adopted a better analysis (and even then we have to ask what forms of Hebrew are covered by said better analysis). There are some cases in colloquial Modern Israeli Hebrew where it looks more clitic-like (e.g. "ha-beit sefer", which gets 122,000 hits — nothing compared to "beit ha-sefer"'s 2,470,000, but still nothing to sneeze at), but even for those, I'm not sure what the best analysis is: is it ha-{beit sefer} because ha- is a clitic for those speakers, or because {beit sefer} is a single word? ("Beit-sefer-im" gets essentially no hits compared to "batei-sefer", which seems to rule out the latter interpretation, but it could be somewhere in between.) If we do follow the traditional analysis, I believe definite nouns and adjectives would have normal form-of entries ("singular/plural definite form of ____"). —Ruakh_TALK 22:06, 3 July 2008 (UTC)

If there's no language ambiguity then this sounds fine. We already do hard redirects on phrases (with placeholders etc.) because a collision is so unlikely. I'd want to be really sure about the no ambiguity part though. DAVilla 09:27, 15 July 2008 (UTC)

trans-top

Does the sense given in template:trans-top make it seem like the translations given are of that term? For example, slide rule gives as its definition "an analog calculator consisting of three interlocking strips marked with...", so has {{trans-top|analog calculator}}, including, inter alia, Czech: logaritmické pravítko. Would the reader, not acquainted with Wiktionary conventions, read that as meaning that logaritmické pravítko means "analog calculator" (rather than as meaning that logaritmické pravítko means "slide rule")? I fear so. Perhaps template:trans-top, instead of <div class="NavHead" align="left">{{{1|Translations}}}</div>, should have something like <div class="NavHead" align="left">{{#if:{{{1|}}}|''In the sense of:'' {{{1}}}|Translations}}</div>.—msh210℠ 17:42, 3 July 2008 (UTC)

That seems like a possible concern for some newer users and even newer translator-contributors. But the gloss is supposed to be no longer than the definition and the extra 1.5 inches of repetitive text on a series of collapsed translation bars might look silly. Does or could the template easily provide for optionally suppressing the "in the sense of 'headword'" for those cases where a knowledgeable editor had prepared a good shorter gloss that didn't need the phrase? Perhaps the translation header is where the repetition should be for the benefit of users before expansion, with the headword only appearing in the gloss after expansion, if indeed it were still needed. All of this might be annoying for veteran users so that having the option of suppressing if would be nice. I fear that all of this conditionality could make the design of the template complicated, so just take it as idle wishful thinking. DCDuring TALK 18:25, 3 July 2008 (UTC)

I'm sorry for not having been clear. The {{{1}}} referred to in trans-top is not the headword but a rewrite of the individual sense; so, in the example I gave, {{{1}}} is analog calculator. So I was suggesting that the header of the drop-down box read "In the sense of: analog calculator" instead of merely "analog calculator". To your points, though: I don't see why we couldn't suppress "In the sense of" in cases where {{{1}}} is a good gloss of the headword (so that the translations given are translations of {{{1}}}); we could use something along the lines of {{#if:{{{1|}}}|{{#if:{{{goodgloss|}}}||''In the sense of:'' }}{{{1}}}|Translations}}.—msh210℠ 18:47, 3 July 2008 (UTC)

I still can't read template code, though I'm working on it. With the tech details, this discussion seems more GPish than BPish. In any event, the option for suppressing the "in the sense of: " would leave us with a visible marker for indicating glosses that haven't been looked at, assuming nobody runs a bot to automatically insert what turns that text off. DCDuring TALK 19:27, 3 July 2008 (UTC)

That haven't been looked at, yes, or that are not good glosses of the headword. They needn't be, need they? I mean, "analog calculator" is fine for the translation table (as long as, as noted above, indication is given that that's not what' sbeing translated) even though it's not a good gloss. No?—msh210℠ 19:50, 3 July 2008 (UTC)

The purpose of the gloss is to disambiguate individual senses. I don't see the point of repeating part of a definition when there is only one sense. Better to repeat the headword “slide rule,” to reinforce exactly what is being translated.

When there are multiple senses, it would be clearer to use the headword, and add the disambiguating gloss to reinforce its supplementary function, as “slide rule (analog calculator).” —Michael Z. 2008-07-03 19:46 z

I always add the gloss, even if there is only one sense. There is no guarantee that a new sense will never be added. --Panda10 22:10, 3 July 2008 (UTC)

When another sense is added, a gloss should be added to each as a matter of course. But there is no purpose to embellishing the table header with an unnecessary gloss—useless elements of an interface necessarily make it worse. —Michael Z. 2008-07-04 00:06 z

Michael, I agree. When another sense is added, a gloss should be added. But not every editor will follow this rule. So I am trying to prevent a lot of additional work. Not to mention, that I am monitoring Category:Translation table header lacks gloss and if a trans table is created without gloss, this category will get a new member and it will grow quickly. --Panda10 02:39, 4 July 2008 (UTC)

This seems like a very good idea and very easy to implement. How much harm can that little extra sentence do compared to what gain it might give us. The way I see it, be bold and do it ... --EivindJ 22:44, 3 July 2008 (UTC)

For entries with a single sense, why not do both? That is, for slide rule, the trans-top gloss would read "slide rule - an analog calculator", or some such. That way, we don't confuse new users by having only the definition, but we also don't limit ourselves to forever having just one sense for the entry. --EncycloPetey 00:16, 4 July 2008 (UTC)

Can't this be done in the template, adding {HEADWORD}: (colon)? edit: Wait... I think I meant PAGENAME. As must be obvious, I really know my template talk... %-) -- Thisis0 19:20, 4 July 2008 (UTC)

Alternatively, we could write "in the sense of X". Whatever we decide to to, it looks like a bot might be required... — Paul G 16:22, 28 August 2008 (UTC)

If space is at a premium, it could be reduced to just “sense of analogue calculator”, or “sense: analogue calculator”. —Michael Z. 2008-08-28 17:47 z

Lack of consistency

After some contribution to this project I've reached the conclusion that this is a very nice project with many nice contributors with a very nice final product, but I've also understood that some basic (and sometimes more than basic) understanding of wiki syntax is needed. The extensive use of templates probably makes this project one of the hardest places for newbies to start and contribute, but then again it might be that Wiktionary rely on already experienced Wikipedians from other projects. Anyways, I was thinking about a more spesific problem, and that is the lack of consistency when it comes to the names on the parametres on many of the templates. When you have an entry with more than one word it is often preferrable to link to the different words inside the three '''. The {{infl}} then have the parametre "head=", {{en-noun}} has "sg=" and {{en-adj}} has "pos" ... Would it be an idea to make this kind of parametres (having the same function on more or less all the templates) have the same name on all templates. I reckon "head=" would be a nice parameter name which I would prefer if we could have on all templates of that kind. Any reasons why we shouldn't? --EivindJ 13:55, 4 July 2008 (UTC)

+1 —Ruakh_TALK 14:22, 4 July 2008 (UTC)

I'd also strongly support an effort to make template parameters consistent. Rod (A. Smith) 16:26, 4 July 2008 (UTC)

The only reason not to do it is that the current crew is used to this way. But we need more contributors and any reasonable standardization would help in getting them up to speed.

There is some logic to the existing system, because the parameters have somewhat different functions. In {{en-verb}} "inf" automatically adds "to", which might otherwise be forgotten. "pos" can used for all(?) of the en- templates (at least) other than {{en-verb}}, I think. {{infl}} can be used for any language, which puts much more burden on the person entering.

Maybe a place to start is to document the named parameters used in classes of templates. They may be more standardized than appears at first glance. It might also be useful to replace some of the less-used templates with the best modern forms in entries so that the obsolete templates can be eliminated. DCDuring TALK 16:31, 4 July 2008 (UTC)

The "pos" parameter in {{en-adj}} stands for "positive", as opposed to "comparative" and "superlative". To me, "head" seems a better name for the general parameter because it cannot be confused with "part of speech", "positive", or {{pos}} (the ISO 639-3 code template for "Sayula Popoluca").  :-) Rod (A. Smith) 19:38, 4 July 2008 (UTC)

I confess to having assumed it referred to "Part of Speech", which didn't keep me from using it correctly, I think. Your reasons seem fine, but, in my limited understanding, "head" in infl works differently than "pos" in the en- templates. "pos" also appears in other en- templates where I didn't read it as meaning "positive", {{en-pron}}, {{en-intj}}, {{en-prep}}, {{en-con}}, {{en-det}}, (not the most commonly used); as well as {{comparative of}}, {{superlative of}}, where positive makes more sense. It would be tough if we let ISO 639-3 deprive us of many desirable three-letter abbreviations for our templates and for parameter names. DCDuring TALK 20:01, 4 July 2008 (UTC)

There are some good reasons for the inconsistencies. Now, for {{en-noun}}, it might be fine, and the template might be set up to accept either head or sg for the parameter name; I think it already has a couple of built-in options. It would also have the benefit of allowing the template to more freely be used on plurale tantum nouns, where "sg=" makes no sense. For {{en-verb}}, I think the suggestion would also work well. Butw hile uniformly using "head=" might seem like a good idea, there are some situations where it creates problems. For adjectives and adverbs, I'm skeptical. If someone could generate by bot a short list of entries that use either of those templates with an explicit pos=, it would either allay my fears as unfounded, or else reveal a problem that needs to be addressed first.

For foreign language entries, the use of "head=" isn't always the best choice. For some of us, the "usual" parameter is "alt=", so there will be some inconsistency even if English can be made uniform. Additionally, some templates work entirely on the basis of roots or stems, such as the Esperanto templates, so head= would be meaningless in those cases. --EncycloPetey 04:12, 7 July 2008 (UTC)

Not to change the subject, but another inconsistency that bugs me is with language parameters. {{attention}} and {{infl}} use {{{1}}} and the ISO code; {{rfc}} uses {{{lang}}} and the ISO code; {{etyl}} uses {{{2}}} and the ISO code; {{abbreviation}} and {{acronym}} use {{{1}}} and the spelled-out name of the language. In my opinion, which you can take or leave, all templates that call a language parameter should allow users to enter the language code or the spelled-out language; and all should allow users to use {{{lang}}} (in addition to whatever numbered parameter they use now, if any).—msh210℠ 17:53, 4 July 2008 (UTC)

That problem has been noted and raised by many of us. Some folks have made some real progress towards solving the issue, and ao some templates already have this problem corrected. However, there are some cases where the user must be forced to use one or the other, especially in category names where some require ISO and others require the full name, or we will get incorrectly named categories requiring mass cleanup. now, {{rfc}} requires use of a named parameter because that parameter is optional, and {{{1}}} is the reason/note regarding cleanup. But {{infl}} must have a language, so it's silly to require the user to always have "lang=". Now, you're mistaken slightly about {{etyl}}, because it uses both {{{1}}} and {{{2}}}, and interprets one as the language name and the other is ISO code. As a result, it can't use "lang=", because it may require two such values in a single use. You are correct about {{acronym}} and {{abbreviation}}; both of those ought to accept the ISO code with lang=, because the code should always be interpreted the same way. In those cases, the feature has simply not yet been added, because the templates predate {{lang}} and other ISO interpreting code. --EncycloPetey 04:12, 7 July 2008 (UTC)

I agree with nearly all the comments made here so far (with the exception of msh210's critique of the deftly written {{etyl}}, which EP seems to have rebutted nicely). I think that a greater deal of uniformity is an excellent idea, and we have a great deal of room to improve on this. However, as EP notes, there are a few places where an editor will simply have to know the quirks of a specific template, with etyl being an excellent example of a template which can't take a lang parameter. A good start would be to make many of the templates take multiple parameters to mean the same thing (e.g. let {{en-adj}} take pos or head, and have them both work). I bet if we asked nicely, Robert Ullmann would be more than capable of both writing these in and/or siccing AutoFormat on the parameters we would like to deprecate, so that they can leisurely be exterminated, and the parameter redundancy be lifted. -Atelaes λάλει ἐμοί 08:10, 7 July 2008 (UTC)

Well, {{etyl}} can have {{{lang}}} s a synonym of {{{2}}}, much s other templates have {{{lang}}} to indicate the language of the word being defined on that page. {{etyl}}'s having {{{lang}}} as a synonym of {{{1}}} would not be in line with other templates.—msh210℠ 16:47, 15 July 2008 (UTC)

While it can have that, it would be terribly confusing for editors to have a template that accepts "lang=" for only one of its two ISO slots. --EncycloPetey 17:45, 15 July 2008 (UTC)

Quotations from newspapers / magazines

Hi, I was wondering how to format quotations from newspapers and / or magazines. Please see Wiktionary talk:Quotations#Quoting of newspapers/magazines H. (talk) 15:06, 4 July 2008 (UTC)

Community censure of Robert Ullmann

I have spoken to Robert, and a couple of others have chimed in about his poor behavior, all founded on assumptions of the reasons for a single pair of reversions [5] [6] of an anon contributor. The anon had moved a quote off an entry's main page to another page; I moved it back. The anon has since returned and decided to again remove the quotation [7], because he/she feels that (contrary to prior community discussion) the forms do not match exactly and so should not appear on the page [8]. That is the background.

Here is Robert's initial response, in which he charges me with vandalism for making the revert: [9] The ensuing conversation is at [User_talk:Robert_Ullmann#Vandalism charge], in which I chastise Robert for making such ugly accusations, and in which Robert is than chided by two other community members. Robert holds firmly and vocally in the discussion to his charge of vandalism, and proceeds to ascribe many motives to my actions which are clearly assumed. He has refused to retract his serious allegations that my reverts were motivated and carried out as deliberate vandalism.

If Robert had a charge to level against me, he should have come to me (as I went first to him). As it stands, Robert has yet to ask me for my side of things. (It was a simple pair of reverts.) Instead, he has woven an extensive fabrication of untruths and half-truths to claim that I have acted in vandalism.

I request that the coommunity make it clear that such serious charges are unwarranted, and that the community make a plain statement that such attacks should never be posted on third-party pages. Robert's actions are unwarranted and are a blot on the reputation of Wiktionary. --EncycloPetey 16:35, 6 July 2008 (UTC)

I removed the comment from the IP anon talk page. I do stand by my primary complaint that removing content from non-lemma pages is very wrong, but this has been discussed before and will be again. I do want to apologise to EP. My explanation, while not an excuse, is that I have been having a desperately bad day; normally the wikt is a relief, requiring a lot of intellectual focus and thus distraction. This has not been working today. (If you want the gory details of the bad day, send me email, I'm certainly not going to splat it here! ;-) May I ask you to please stop escalating this; it was only on my talk page (after I took it off the anon, who has since replied there.) Robert Ullmann 16:45, 6 July 2008 (UTC)

First of all, yes, charging another admin with vandalism is a bad idea, and I believe Robert was in the wrong to do so. However, he has apologized (in bold no less), and retracted his statement. In my opinion that closes the matter. We all have our off days and make imprudent actions from time to time. -Atelaes λάλει ἐμοί 17:42, 6 July 2008 (UTC)

(after edit conflict) I object to the use of the word (deprecated template usage) vandalism, as to me that word implies bad faith, but I think it's important that administrators be honest with new users, rather than pretending that administrators all always do the right thing and always agree with each other. Our policies are poorly documented, and we frequently disagree on whether a given thing has been decided or not; so, it's not like we can just hide from newbies the fact that we lack agreement about even very basic things (e.g. what kind of entry is optimal for non-lemma forms). R.U. having now apologized (or having now expressed an affirmative desire to apologize, which is pretty much the same thing), I see no need to censure him. Everyone makes mistakes, everyone feels strongly about random things that others find unimportant, everyone sees mastodons occasionally, everyone learns to forgive and forget. —Ruakh_TALK 17:53, 6 July 2008 (UTC)

Mistakes happen. Conflict happens. Please let it go. DCDuring TALK 18:04, 6 July 2008 (UTC)

As Robert has been willing to apologize, I accept his apology. Please note, Robert, that this was restoration of the quote to where it started only, and nothing more. If you believe that the quote should be duplicated on the page for the matching inflected/plural form, I won't disagree with that. I merely wanted to return the quote to the "primary" page to ensure we had a citation there. That (and that alone) was my motivation in making the reversions. The comment I posted to the anon reflects this, as it only metions removal of the quote from the lemma page, and does not discuss the issue of adding it to the plural page. --EncycloPetey 20:23, 6 July 2008 (UTC)

I'm not going to comment on any of the charges as they appear to have been resolved. I would like to say, though, that I completely understand Robert's point about duplicated content, which is not something that others have backed him on. The frustration I think comes as a result of fully reverted changes that admins label flatly as wrong without asking if any part of the change might be productive. No, I wouldn't call it by the V-word at all since in this case the removed content still exists in an appropriate place on Wiktionary, but there's nothing inherently wrong with having it in both places. I say this despite disagreement on precisely this point about the titles under which Citations should be combined or not, but my position is not inconsistent. I see Citations as a space for potentially holding all quotations, not every one of which could be so illustrative as to be included in a dictionary entry somewhere. Thus, anything listed in the main space as illustration would be, potentially, a duplicate anyway. If it's helpful in more than one location, then why force a choice between them? This is a fairly minor point of course, relative to the escalation. DAVilla 09:11, 15 July 2008 (UTC)

Wikipedia/Wiktionary

Quick question please -- can a word be defined both an here ond on Wiktipedia?? Smith Jones 00:49, 7 July 2008 (UTC)

Wikipedia avoids definitions, but a Wikipedia article might have the same name as an entry here; for example, our horse defines the word (deprecated template usage) horse, while Wikipedia's Horse discusses horses. —Ruakh_TALK 00:59, 7 July 2008 (UTC)

I was refering more to the phrase detente balam, which is more or less defined her and goes into very little more detail on Wikipeda. it seems redundnat to me so I wanted to make sur thta tit was ckosehr. Thanks for replaying? Smith Jones 01:02, 7 July 2008 (UTC)

Ah, I see. We decide whether to include an entry based on Wiktionary:Criteria for inclusion, and Wikipedia decides whether to include an article based on w:Wikipedia:Notability and other policies. Since a dictionary and an encyclopedia tend to include different things, an entry here will generally differ significantly from a corresponding article there; but in some cases they'll be quite similar, and there's nothing wrong with that in itself. —Ruakh_TALK 02:33, 7 July 2008 (UTC)

For what it's worth, a common reason given for deletion on Wikipedia is (or at least used to be) "dicdef": that the article is merely a "dictionary definition" and unlikely to be more than that. Likewise, a common reason here is "encyclopedic" or "not dictionary material": that the entry is not a word but subject for an encyclopedia article (like Columbia-Presbyterian; cf. w:Columbia-Presbyterian).—msh210℠ 16:43, 9 July 2008 (UTC)

Merriam-Webster copying us?

Is it just me, or has Merriam-Webster flat-out copied our first two examples for mondegreen? bd2412 T 09:10, 7 July 2008 (UTC)

Is there a problem with this? Are we really worried about copyright infringement on our GNU license? According to that, "...this means that the entries will remain free forever and can be used by anybody." Amina (sack36) 16:49, 7 July 2008 (UTC)

It's good to know and to be able to cite chapter and verse when discussing the relative quality of on-line dictionaries. In this case it's a little hard to tell because there are only about 10-20 commonly cited mondegreens. DCDuring TALK 17:36, 7 July 2008 (UTC)

It's a bit flattering actually, although even the GNU technically requires that we be credited. I suppose they could argue that it's a coincidence, but if it is, it is certainly quite the coincidence. bd2412 T 21:52, 7 July 2008 (UTC)

Perhaps they did, but we didn't invent those examples ourselves, so to me it sounds a little strange to say we have a copyright on them. Consider, for example, how we've copied quotations directly out of the OED. And numbering two, it's hardly a collection. DAVilla 08:33, 15 July 2008 (UTC)

Placing of Etymology

Is there a reason why etymology preceeds definition? I have found words(though at the moment I can't remember them) that have the same spelling but different etymologies, and the second most frequent use of a dictionary is for definition of a word, not etymology. (for some wrong-headed reason spelling is number one. Go figure.) Amina (sack36) 16:45, 7 July 2008 (UTC)

Etymologies are placed first because a word's origin is the only reasonable way to sort it. While this is less important for the average word, for words that have four completely different words sharing the same spelling, it becomes highly important to be able to sort out what is actually related to each other, and what is related simply by happenstance. 'Tis a lesson learned from biology. -Atelaes λάλει ἐμοί 17:08, 7 July 2008 (UTC)

We already have two entry structures: one where Etymology comes first and at the same level as POS sections, and one where POS sections are grouped under Etymology. Even if we don't want to change the second structure, that doesn't prevent us from changing the first one and putting Etymology last and at the same level as POS sections. —Ruakh_TALK 18:44, 7 July 2008 (UTC)

Another take: for a given entry, there can be multiple parts of speech and multiple etymologies (as many as 10!: kaeru#Japanese). For a given etymology-part of speech combination, there can be multiple definitions. Grouping the parts of speech that share the same etymology can help someone understand the range of meanings the word has more easily. It is even more of a help to someone trying to prepare an entry. In a wiki the later consideration counts for more than it would in an on-line dictionary with a paid professional staff. If we could efficiently present differently to "normal users" (say, all unregistered users and registered users who opted to be treated as unregistered users with respect to presentation), then we could present one way to those with more sophisticated tastes in language and another way to "normal users". Under the new WMF budget there is supposed to be much more spending for technology, so it may become possible to contemplate more developer-, server-, and bandwidth-intensive possibilities. In the meantime, portions of excessively long etymology and pronunciation sections could be placed under a show/hide bar using {{rel-top}} and long "Alternative forms" or "Alternative spellings" lists can be laid out as horizontal lists instead of vertical ones. DCDuring TALK 17:32, 7 July 2008 (UTC)

Further, consider placing the etymology if it did not precede the definitions. Because there can be multiple etymologies, we would have to mark each and every definition line as to its etymology if we did that. To avoid needless repitition and screen clutter, it is better to simply use the Etymology as a grouping mechanism. (Note to Atelaes: Although it is not incorrect to say we learned this lesson from biology, the biologists got the idea from Hennig, who got the idea from the historical linguists. And so we come full circle.)--EncycloPetey 18:04, 7 July 2008 (UTC)

The Merriam-Webster on-line solution is to present a single etymology-PoS combination at a time, which easily allows the etymology to be presented after the definitions. Their solution is not perfect by any means: for words that have multiple etymologies they may present two entries for the same part of speech on a disambiguation page, without even giving a gloss to guide the user's choice.

There are design tradeoffs if one only has one way of presenting to the user. For us, registration provides a convenient point for differentiating types of users. Users who register, learn about preferences, and then log in, can have custom profiles which are limited by the capabilities of CSS etc., developer time and capability, and the willingness of WMF to tolerate the resource load. DCDuring TALK 19:03, 7 July 2008 (UTC)

M-W is also presenting words in only a single language, and so has far less of the overarching data structure we must contend with. --EncycloPetey 23:27, 7 July 2008 (UTC)

Category:Wikisaurus

Should there be a [[:Category:Wikisaurus]]? I ask because there’s recently been some traffic on my talk page on exactly this point.

See:

• User talk:Nbarth#Category:Wikisaurus
• User talk:EncycloPetey#Category:Wikisaurus

Summary:

“I removed “Category:Wikisaurus” from the few WS pages that had it, since it seemed to be deprecated, both:

• Because the category itself says that it is deprecated:

  This category has been superceded, since the introduction of the Wikisaurus namespace, by Special:Allpages/Wikisaurus:

• ..and ostensibly, since few WS pages had it.”

EncycloPetey wrote:

Actually, there is a need for the category. All Wii pages are expected to be categorized, or they clog up the list of uncategorized pages. It may seem superfluous, since there is a separate namespace, but it is technically required.

Re-adding the category would be best. We categorize items in all namespaces. All Citations: pages are categorized in Category:Citations; etc.

…and thus I dutifully added [[Category:Wikisaurus|{{PAGENAME}}]] to {{saurus-head}}, and refreshed all the Wikisaurus pages.

Then Robert Ullmann wrote:

…

Having a category which replicates Special:AllPages/Wikisaurus: may be useful, or may not.

In more detail, he wrote:

Note that we categorize things in some even-numbered namespaces (not Mediawiki, and I'm not sure what with Citations), the odd-numbered namespaces (talk pages) are not. But it isn't about Special:UncategorizedPages, which is only NS:0; there are others for other namespaces (Special:UncategorizedTemplates, Special:UncategorizedCategories, Special:UncategorizedImages). And there is no "Special:UncategorizedWikisaurus" (which would require writing or modifying a code module for the MW software).

…so I figured I’d bring it to the beer parlour – what do people think, and, concretely, should there be a catch-all Wikisaurus category?

Nbarth (email) (talk) 21:18, 7 July 2008 (UTC)

I'd just as soon have the category. As my father would say, it can't hurt and it might help. bd2412 T 21:50, 7 July 2008 (UTC)

See also WT:RFDO#Category:Citations.—msh210℠ 21:56, 7 July 2008 (UTC)

I'd just as soon remove the category. As BD2412's father might say on a more pessimistic day, it can't help and it might hurt. —Ruakh_TALK 21:57, 7 July 2008 (UTC)

Hurt how, exactly? I can't conceive of any possible harm. bd2412 T 09:04, 8 July 2008 (UTC)

I think an incomplete [[:Category:Wikisaurus]] could be actively detrimental. We can do various things to mitigate that risk, such as task a bot with populating the category, but I don't see the point. If we want to invest effort in it, I think we'd be better off using JavaScript to link from Wikisaurus entries to Special:PrefixIndex/Wikisaurus:. —Ruakh_TALK 23:00, 8 July 2008 (UTC)

I'd like to see us keep and expand it. If Wikisaurus is expected to (finally) grow, then a category structure would enable us to group listings by themes, the same way the Roget's hierarchy does. --EncycloPetey 23:25, 7 July 2008 (UTC)

With or without Wikisaurus, I've been thinking, we can have Category:Roget's 123 (etc.) on entries.—msh210℠ 23:30, 7 July 2008 (UTC)

Possibly, but only if that information is not proprietary. --EncycloPetey 01:51, 8 July 2008 (UTC)

The original Roget's is from the 1850s and from the U.S., if I recall correctly. If that's right, then there's no copyright on it (IANAL). I don't know about patents, though. (And note that the numbering system has changed since the first edition.)—msh210℠ 16:36, 9 July 2008 (UTC)

Idea for a new bot

Greetings. I run bots on Wikipedia, Wikisource, and Commons, and I have an idea for a bot that might be useful here. I've read Wiktionary's Bot policy, and so I wanted to bring the idea here to see if it would be worth writing. Here is my idea.

Nutshell: This bot would help users easily add citations on the citation pages for words, when the source of the citation is Wikisource.

Justification: The "citation" pages are a great idea, but they are underutilized. It seems like a lot of work to add them, involving cross-referencing with the definition page, as well as pages for the found quote. A bot could make this easier.

Detailed description: This would be a tool running on the toolserver. It would provide a user with a textbox for a word, a textbox for a quote, and a textbox for a URL to Wikisource. For instance, I might type in "delicacy", "Between me and the other world there is ever an unasked question: unasked by some through feelings of delicacy; by others through the difficulty of rightly framing it." and "http://en.wikisource.org/wiki/The_Souls_of_Black_Folk/I". If the URL is valid and the word is found in the quote, the bot will look up the word in Wiktionary and return a list of definitions for the user to choose between. In this case, the choices would be:

• Noun: The quality of being delicate.
• Noun: Something appealing, usually a pleasing food, especially a choice dish of a certain culture suggesting rarity and refinement -a Chinese delicacy
• Noun: Fineness or elegance of construction or appearance.
• Noun: Frailty of health or fitness.
• Noun: Refinement in taste or discrimination.
• Noun: Tact and propriety; the need for such tact.

Let's say the user chooses #6 -- a sensible choice. The bot would look up the URL in Wikisource to see if it can get the author, the name of the work, and the year. In this case it can. By going up a level in the URL, it can get the author and title from the "header" template, and the year from the category. Having all this information, the bot could create or edit Citations:delicacy with the following:

===Noun: ''Tact and propriety; the need for such tact.''===
{{timeline|
1900s=1903}}
*'''1903''' - [[w:W. E. B. Du Bois|W. E. B. Du Bois]], ''[[s:The Souls of Black Folk/I|The Souls of Black Folk]]''
*:Between me and the other world there is ever an unasked question: unasked by some through feelings of '''delicacy'''; by others through the difficulty of rightly framing it.

Built-in limitations:

1. This bot will not make any decisions about whether a quote is appropriate -- that's up to the user, just as it would be if the user didn't use this tool. The tool only makes the job easier.
2. If the Wiktionary definition is not formatted correctly, the bot will not be able to correctly list the definitions for the user to choose from.
3. If the Wiktionary citations page is not formatted correctly, the bot may insert the new quote in an unexpected part of the page.
4. If the Wikisource article is missing information, then the bot can't get it. For instance, s:The Two Gentlemen of Verona does not give the year.

This is just the planning phase, and I haven't even started on the bot yet. I wanted to bring it up here and get some feedback to see whether this function would be useful and encouraged or not. All the best, Quadell 19:24, 9 July 2008 (UTC)

Intriguing. Note that the year is often missing from the texts pages themselves, but are (more) often included in the list of author's works on the author page. If the bot could locate the line on the author page that links back to the text examined (a little tricky, yes), then it could also get the date that way. --EncycloPetey 22:50, 9 July 2008 (UTC)

Additional thought: It would be nice if the bot could be customized to search a particular language of Wikisource. The Latin Wikisource has a huge corpus of Classical literature (in part because of fewer copyright problems, I suppose). --EncycloPetey 00:06, 10 July 2008 (UTC)

You may wish to talk to TheDaveRoss who was also writing a similar bot (though I think he was using google books). I don't think that this would be that useful without a human glance at each quote, as although lots of quotes is adequate, a few good quotes is better. Conrad.Irwin 23:02, 9 July 2008 (UTC)

I like this idea a lot. I think that if Wikisource doesn't have a date, or other bibliographic info, the script can insert some template like {{Quadell-bot-missing-info|date}}. Note that this is not really a bot in the usual sense; every edit would be made by a human contributor, as I understand it (right?). Yielding more results than having the user type in a word, a quotation, and a URL, wouldbe for the user to type in only the word, and have the script search Wikisource; that's similar to what TDR is doing (as noted above by Conrad).—msh210℠ 23:29, 9 July 2008 (UTC)

Good feedback. I hadn't thought of using Wikiquote, although that might (might) be possible as well, if the necessary data is easily harvestable. I like the idea of entering a word and searching Wikisource (or Wikiquote), but that would be tougher. I would have to either integrate with Wikisource's search, or Google's API, and I don't have any experience with either. Still, it certainly sounds like a useful idea. (And yes, to confirm, every addition would be initiated and confirmed by a human.) I'll give TheDaveRoss a ring to see what his thoughts are. Quadell 01:31, 10 July 2008 (UTC)

I was/am part way through a bot with the same intent, I was using texts from the Gutenberg Project. My project is on hold for the summer while I spend time doing other things. I was going at it a bit differently, I was simply going to have the bot add the citations directly to the page based on some rules for what qualifies as a good cite, your bot seems like it would be a great, but different, tool. - [The]DaveRoss 19:53, 11 July 2008 (UTC)

Examples

One of the reasonably common "illegal" L4 headings is "Examples". These are not usage examples, but rather illustrations of the concept. I have been working on the entry for prolepsis, which included in-line examples of the rhetorical and grammatical device involved. Arguably these do not belong in a dictionary entry because they do not illustrate the usage of the word. Presumably, we are referring users to articles in WP that actually illustrate the phenomenon.

Is it ever appropriate for us to illustrate the concept in words (by an example), rather than define the word and show its usage? (Obviously we "illustrate" by the use of images, BTW). If so, how ought it be presented ? The parallel case of pictorial illustration suggests that a sidebar "box" of some kind might be appropriate to mark the pattern-breaking nature of such examples. DCDuring TALK 20:02, 9 July 2008 (UTC)

It's sometimes appropriate to illustrate the use of a grammar term in an example sentence with the following technique from the entry for (deprecated template usage) synonym:

“Happy” is a synonym of “glad”.

That is, an fitting example sentence can usually be made by mentioning some examples of the concept. Does that help? Rod (A. Smith) 21:18, 9 July 2008 (UTC)

Judicious and, dare I say, witty selection of usage examples has been fun and has enabled me to finesse the general problem on many occasions. But it doesn't always work. (As in the more complex definitions of prolepsis.) It can be hard to find a passage that briefly illustrates the definition. In addition, our use of graphics is usually to illustrate concepts rather than decorate entries. It is the parallel to those graphics that struck me. Also the abundant use of boxed mini-cases and sidebars in textbooks and some other non-fiction struck me as a useful format for texty material that could illustrate, say, rhetorical devices, grammatical constructions, etc. Just a thought. DCDuring TALK 22:10, 9 July 2008 (UTC)

Yeah, that's a good idea. Make it into a "picture", and then it "illustrates" the concept. DAVilla 07:17, 16 July 2008 (UTC)

Bot Request for QualiaBot

I ask for a bot flag for QualiaBot. The half automated bot will work interwiki (the main aim is to integrate upper sorbian wiktionary in the interwiki and to make interwiki in the hsb wiktionary). To be effective with ressources it is senseful that the bot writes his findings also in the pt dictionary. The bot uses pywikipedia. His wikipedia sister has bot flag in 26 wikipedias (e.g. en, pt, es, de, cs, pl, sk, ru, uk, bs, sr, hr, eo, el, it, no, da, ro, sl, lt, lb, tr, nds, nds-nl, be-x-old, cv) Thank you! Qualia 17:12, 10 July 2008 (UTC)

See above, Wiktionary:Beer parlour#botflag_for_CarsracBot and Wiktionary:Beer parlour#Bot_flag_request_for_User:EivindBot. In what sense is your bot "half[-]automated"?—msh210℠ 19:24, 10 July 2008 (UTC)

Okay, I see there would be only be a small time benefit if QualiaBot would write here. If you don't like this benefit, not my problem. By the way I would think it would be - with respect to the effectiveness - better if Interwicket would write its findings also in other wiktionaries. Because not all wiktionaries have such a powerful bot and it would be more effective for the wictionary projects at all. Qualia 12:28, 11 July 2008 (UTC)

Wikisaurus at cross purposes

We seem to be working at cross purposes with wikisaurus and it's wasting everybodys time because we are overlaying each others work. We need to stop the entry work there and clearly define how the page is going to work.

I have been listing all the synonyms I can come up with for a given word within that words headpage. Thus, noun, adjective, adverb, and verb forms of the same word all have their place on that page. Each synonym is linked to their own headpage where wiktionary can find them. I was told to take a page to the sandbox to alter it and when I went to put it back, the page had been altered beyond recognition. It won't let me archive and overlay any more.

The altered page has broken out all of the list of words into different groupings (I guess) and sent them to their own page. A great deal more information has been added to the page (most of it I have no problem with) that wasn't there before and all of the words link back to wiktionary instead of wikisaurus.

I've been putting in a lot of hours on this and I want to know that my hard work won't be wiped away just because of our lack of organization. The casual editor isn't the problem here, it's the bunch of us who are on this thing all the time and now that I've breathed new life into the project, people are no longer eager to have it demolished. That's good, but it is also a plan for disaster if the heavy hitters all pounce on the wagon at once without getting up to speed on where things sit.

I appreciate the Category:Wikisaurus. It was a great improvement, but I wasn't told it was happening and found it by accident after all was accomplished. I was reading someplace that courtesy dictates we notify the people who seem to be most involved in a certain area. Everyone here knows that I am most involved in the Wikisaurus project. Where was i notified? Amina (sack36) 20:59, 10 July 2008 (UTC)

Who did what to whom? Have the changes been rolled back? Were you working in a public sandbox or on your own user page? On a wiki, if it is public, it is open for editing by all. We have civility rules and we also things that are hard to find (one's own user pages, if cunningly named. DCDuring TALK 21:40, 10 July 2008 (UTC)

I'd rather not spend time pointing fingers. The page that I prepared is at sack36/sandbox. I was asked earlier to move things to either the public sandbox or one at my page. I chose my page. the other item is at Wikisaurus:drunk.

The point is that the pages look HUGELY different. Words that I had painstakingly transferred to point to wikisaurus are now all pointing to wiktionary. Lord only knows why. they can type the word in search and be there as fast or they can click the wikisaurus pointer and go to where it is a headword, click again and you're there. Not every word is a clear equal to the headword. You have to have synonyms set up for each word because of that.Amina (sack36) 22:34, 10 July 2008 (UTC)

The Wikisaurus pages must link to the dictionary pages, because that's where the meaning of words is stored. I can see why you would want to redirect from the other Wikisaurus pages to the main one, but can't work out why you would want to link from Wikisaurus to Wikisaurus on every item in the list. Is there a page like Wiktionary:Wikisaurus layout explained that I should read/someone should write? Conrad.Irwin 22:42, 10 July 2008 (UTC)

Hiyas, There are a number of issues being discussed here, and some evident frustration, at least some of which relate to me, so I thought I’d add my two cents.

• WikiSaurus format

There seems to be a lack of documentation (and perhaps consensus) on format for WS entries. I find Wiktionary:Wikisaurus/format and Help:Creating a Wikisaurus entry#Formatting (which I’ve linked to each other).

Regarding linking, the Help page says:

“Please be sure to wikify all words and idioms listed in Wikisaurus, and in applicable places link to Wikisaurus pages.”

Other than “applicable places”, I can’t find a mention of where to link to on WS pages – perhaps this could be clarified?

• Category:Wikisaurus

This is discussed above at: #Category:Wikisaurus – briefly, I found few pages with the apparently deprecated category, removed it, then at EncycloPetey’s suggestion, added it – I didn’t know that anyone wanted to be appraised, and it seemed uncontroversial. Sorry if I stepped on anyone’s (Amina’s?) toes.

• Altered page?

I don’t follow Amina’s question – Amina? The page currently at User:Sack36/sandbox (current revision) is on “alcoholic”, and Wikisaurus:alcoholic has not been edited in over 9 months (history), so I’m not sure what the question is.

In case it adds useful context, I was the person who suggested using a sandbox when ones edits break a page, at: User talk:Sack36#Wikisaurus:gigantic.

Suggestions:

• WikiSaurus layout

As Conrad suggested, having WS layout laid out somewhere would at least give us somewhere to point, be it at a new Wiktionary:Wikisaurus layout explained (WS:ELE?) or an existing Wiktionary:Wikisaurus/format or Help:Creating a Wikisaurus entry#Formatting.

• WikiProject pages

If people want to broadcast that they’re quite involved with (for instance, Amina is very involved in Wikisaurus, User:A-cai is clearly very active in Chinese (see: recent contributions), lately I’ve been doing some work in etymology, Japanese, and Chinese characters, etc.), may I suggest WikiProjects?

For instance, Wiktionary:WikiProject WikiSaurus, etc.?

Among other things, this means that someone interested in WikiSaurus (say) can find who’s currently involved?

Amina says “Everyone here knows that I am most involved in the Wikisaurus project.”

I didn’t. I now do. Were there a WikiProject stating: “These people are very involved in WS”, future people would learn much faster.

Nils (Nbarth) (email) (talk) 00:03, 11 July 2008 (UTC)

Added given name at 00:22, 11 July 2008 (UTC)

I'm sorry Nbarth, if you hadn't heard. I joined the Wikisaurus project a few months ago. There was at that time an informal vote going on as to whether the project should get folded into wiktionary and archived. In other words, since nobody had worked on it in "quite some time" everyone was saying it was a dead project. I volunteered to work on it so it wouldn't be a dead project. I was told that since no one else was doing anything I could have at it and don't worry about painting with broad strokes.

So I had at it. I explained everything that I was doing as I did it, dutifully recording it here in the Beer parlour. I asked for feedback numerous times with DC being the biggest contributor to the discussions. I had assumed (I usually know better) that with the amount of chatter I was generating, that people would notice things were happening with wikisaurus and who was making it happen. I know I have gotten a fairly good idea of people's interests based on what they comment on in the discussion rooms.

Now about the linking of non-headword words to wiktionary. Why is it necessary for the wikisaurus synonyms to be directly linked to find out their meaning? If you click on them when linked to wikisaurus, it takes you to their headword document where the abbreviated meaning is the second thing listed after part of speech. It they need to have something more detailed after that, clicking on the headword takes you to their wiktionary page. Why clutter the page with links that are unnecessary? There is no main page. This was explained earlier in the Beer parlour.

Nils, I think the idea of setting up a project similar to the Wikipedia projects is a wonderful idea. In that, one person designates the project and invites anyone and everyone who wants to join in the discussion, right? Then different segments of the project are farmed out to different people to do the preliminary work. I've been working on wikipedia and it's children for around 4 years. For a short time I was project leader on one of the Wikipedia projects. That was until I located Wikisaurus and got so excited about it. I was really impressed with the designation of projects and I kinda wondered why it didn't happen here.

Finally, I'd like everyone to take a look at my user page User:Sack36, please? I talk at length about how I feel the wikisaurus should be laid out. Does this clear up all the questions etc. that people wanted? Amina (sack36) 08:38, 11 July 2008 (UTC)

Ok, I finally see what you are driving at, what a brilliant idea! Can I suggest that you keep your posts slightly shorter, then more people might read them :). Your user page describes something very interesting, but I think we need to constrain it slightly to fit within the bounds of wiki-ness, (though maybe you've just worked out Web 3.0!)

Wikisaurus 2: Instead of thinking of Wikisaurus as a place to dump synonyms when main space entries get too long, Wikisaurus is a big map describing the network of words. There are lots of connections between words, in the grammatical sense the simplest being synonyms and antonyms, the more complex ones including hyponyms and hypernyms. Thus each word should have a Wikisaurus page that describes each of its connections. The interesting bit is that each of the connections is not a simple link, it can (and should) be described too, probably using tags like {context}.

Issues: it will be necessary to give a Wikisaurus entry to each definition of a word, which can get messy for words with lots of definitions. Laying out a multi-dimensional web on a two dimensional page. Conrad.Irwin 23:46, 12 July 2008 (UTC)

My idea: see Wikisaurus:lucky for a simple, reasonably clean format that can be extended to include whatever relationships we want. I consciously chose to imitate your layout == Part of speech == / === gloss of definition === as that makes it very easy to navigate, I also chose to link to both wiktionary and wikisaurus, though it is easy to change the format of each line by editing {{wsword}}. Not sure whether using ;synonyms is better than ==== synonyms ====, but we can always customise appearance with CSS. Conrad.Irwin 00:47, 13 July 2008 (UTC)

Wikisaurus is fundamentally flawed

Interlinking Wikisaurus with itself is great when it's appropriate, but there's no reason to give every word an entry in Wikisaurus as it has in Wiktionary. If that were the case, then why not combine every PAGENAME with its corresponding Wikisaurus:PAGENAME via the synonyms and antonyms sections?

The fundamental flaw is one that has existed since the beginning. A thesaurus entry is not supposed to list every meaning of a string of characters. A thesaurus entry doesn't even have to correspond with a single word. It's best if it corresponds with a concept, which is why titles like fast (quickly) and fast (speedy) work better in my opinion. Most of those have been deleted because of strong objection from a contributor to titles with parentheses, although in my view nothing about Wikisaurus can be said to be all that rigid even today.

That's the reason why I didn't revert any of the changes that I've seen being made, although in earnest I don't always agree with them. I'd rather give some wiggle room. But since it's now an issue, I'm laying my opinion out there. DAVilla 08:16, 15 July 2008 (UTC)

no word of the day

there is no word of the day today. in its place was a template to let the beer parlour know about this fact. Special:Contributions/70.231.238.8c70.231.238.8 19:12, 11 July 2008 (UTC)

OK, I don't get it. What am I supposed to be seeing here? All I get with your or my id's is no contributions. Is that right?! Amina (sack36) 08:05, 12 July 2008 (UTC)

The Word of the Day today is laissez faire. You can see this on the Main Page. --EncycloPetey 22:10, 13 July 2008 (UTC)

Phrasebook

While I agree that a phrasebook can be useful, it's more Wikibooks material. Would it be possible to transwiki all these phrases, then delete the entries here on Wiktionary?--♠TBC♠ 05:12, 12 July 2008 (UTC)

There's also the issue of determining how common a phrase has to be to merit an inclusion into the phrasebook. But that's an issue the Wikibooks editors have to resolve (if these phrases do get transwiki'd).--♠TBC♠ 05:16, 12 July 2008 (UTC)

No, we want to keep them for the sake of the translations. Wikibooks is not about maintaining the Translations like we do, and usually resents becoming a dumping ground for other WM projects. --EncycloPetey 22:08, 13 July 2008 (UTC)

The one problem I see with having a phrasebook, is that it sets a bad precedent for other non-dictionary entries. For example, having Is it going to rain? might merit the creation of Is it going to snow?, Is it going to be a cloudy day?, and so on.--TBC 04:12, 14 July 2008 (UTC)

Fortunately, that particular issue has never been a problem here. --EncycloPetey 17:58, 14 July 2008 (UTC)

Wouldn't it possible to list variations together, as is usual in actual phrasebooks:

• Is it going to rain/snow/hail/be windy/be sunny ?

Circeus 23:51, 5 August 2008 (UTC)

I think common phrases are helpful to include for the sake of translations, but I see no reason in trying to be complete in all variations there may be. What might be good could be to list the "accepted phrases" somewhere to keep a balance of which phrases are common enough, and which phrases are not, also a balance of phrases used in different situations, grouped like "At the restaurant", "Questions about time" etc. Nothing I will put time on personally probably, just a suggestion. ~ Dodde 00:40, 10 August 2008 (UTC)

Length

The length of the RfD page is way too long, over 400 KB, making it very cumbersome to scroll through. I propose we split it by months, like Wiktionary talk:Requests for deletion/Log/2008 June and Wiktionary talk:Requests for deletion/Log/2008 July. Any objections?--♠TBC♠ 04:04, 12 July 2008 (UTC)

I don't like that idea now (though I did support it last time it was proposed I think), if we split up that page it will never get processed - it just needs someone to go through it every month or so and close the old debates, if it is split into months then the necessary work won't be apparent. I would suppose splitting discussion pages like this into topic sections, and having the main page only as a place-holder (as we do for WT:VOTE) as that would make archiving them much easier. Conrad.Irwin 12:04, 12 July 2008 (UTC)

That sounds okay. A problem is that, unless we date subpage titles as we do for Votes, they won't be unique. (If we ever decide to do similarly for WT:TR, by the way, we can instead of subpage transclude ns:1 pages into the TR, and make Rfdresult (or whatever that template is called) noincluded.)—msh210℠ 23:16, 14 July 2008 (UTC)

I, too, object, for the record.—msh210℠ 22:32, 14 July 2008 (UTC)

Well, what we could do is start doing all the big discussion pages like we do WT:VOTE, with threads being contained in subpages which are transcluded. This would confer a number of benefits: First, archiving is a snap, and retains a faithful edit history. A word passes rfv? It's discussion is simply removed from the rfv page and transcluded into the entry's talk page, taking the whole edit history along with it. This would make the BP much more efficient as well. People can watchlist topics of interest to them, and keep track of what people are saying. Conrad.Irwin says that it would be rather easy to write a bot which would archive discussion topics which have had no comments for a month. That way, short discussions are moved out of the way, and long discussions can stay as long as they need to. Even better, an archived discussion could be back onto the BP if it starts up again. Anyone remember an occasion where they were trying to find an old, unresolved discussion to bring context into a new discussion? Just revive it! Obviously, the archiving policies would have to be different for the rf pages, as they shouldn't be archived until resolved, but I imagine this would be a simple matter. Thoughts? -Atelaes λάλει ἐμοί 23:28, 14 July 2008 (UTC)

I like this idea. It certainly would make things a lot easier and more convenient.--TBC 23:53, 14 July 2008 (UTC)

I've no objections either way. We have discussed this idea before, so I know there are some proponents out there. --EncycloPetey 03:39, 15 July 2008 (UTC)

Starting a vote. Any other additions before this goes up?--TBC 10:52, 17 July 2008 (UTC)

I like this idea, but am not sure exactly how it would work when creating a new section. Would we expect new users to be able to create appropriately-named discussion pages from scratch, and then transclude them successfully into RFD? Also, it's currently a bit of a pain to watchlist votes; I don't mind it there, because there aren't very many, but if we're going to do this for RFD, I think we need a better approach, likely using JavaScript. —Ruakh_TALK 17:23, 17 July 2008 (UTC)

Concerning creation, I imagine the process would be similar to that of WT:VOTE. To tell the truth, I'm not completely sure of how that works, but I have to imagine it could be duplicated. If I'm not mistaken, the rfv/rfd templates currently do everything for you simply by pressing the little + button, and I would hope we can duplicate that with the new system. I must admit, I hadn't thought about how watchlisting would work for rfv/d. Yes, that would be a pain in the ass. I wonder if there's some magic that can automatically watch included pages on a page or something. That would have to be worked out. -Atelaes λάλει ἐμοί 20:15, 18 July 2008 (UTC)

We can definitely duplicate the process of starting a vote, but to start a vote, one must first create the vote subpage, and then manually transclude it into Wiktionary:Votes. I realize this is just a two-step process, but I think that that two-step process is enough to discourage many newcomers (maybe even non-newcomers). —Ruakh_TALK 21:22, 18 July 2008 (UTC)

My idea would be to have User:Conrad.Bot (or another) sit on recent changes and control the beer parlour, it would automatically add new discussions, and could (optionally) remove topics once they had become stale (after a month or so of no reply). Conrad.Irwin 23:49, 20 July 2008 (UTC)

Eep. I have to say, I find a bit frightening the idea of being completely dependent on a bot's catching every single relevant edit as it goes through recent changes. Before Interwicket, one of the problems with some of the interwiki-link-adding bots was that they only tracked recent changes, so when they missed an edit (as they often did), that was that. Maybe, in addition to what you describe, we could have a cronbot go through (e.g.) Special:PrefixIndex/Wiktionary:Beer parlour/2008-07/ once a day? —Ruakh_TALK 01:29, 21 July 2008 (UTC)

I had taken that as a given :). Yes, of course. Conrad.Irwin 09:53, 21 July 2008 (UTC)

Modernism

Many prominent modernist writers, like T. S. Eliot and F. Scott Fitzgerald, use nouns as adjectives, ones that are typically not considered as such. Fitzgerald, for example, uses the word "cocktail" in the Great Gatsby to describe a festive song. So then, should we count these words as adjectives?--♠TBC♠ 19:41, 12 July 2008 (UTC)

The decision has been no if the noun is only used as "attributively". For example, it is good English in anyone's book to say "cocktail party" or "cocktail dress". One does not normally say "That party (or dress) is cocktail", which would be a predicative use. One also cannot normally grade such a word when it used adjectivally: "That dress is very cocktail". Or compare: "That dress is less cocktail than the other." The use could be analyzed as a compound noun "cocktaildress", but not spelled solid. OTOH, note the weasel word "normally". I am sure that thousands of people have said: "That dress is so party." I am sure that thousands of people have said: "That dress is so party." Some nouns originally only used attributively do become more truly adjective-like over time.

Omitting the Adjective Part of Speech section, however, invites contributors to add such a section. That is why I have advocated having some kind of filler note under an Adjective header for such a noun saying that the noun is often used attributively as an adjective (or to form compound nouns) and providing links to an appendix or WP article providing a fuller explanation. The response here has not been enthusiastic because this is deemed to violate the metaphysically non-adjectival nature of such words and to mislead users. DCDuring TALK 00:03, 13 July 2008 (UTC)

Perhaps instead we add a note under the Noun section, indicating that the noun can be used as an attributive adjective with a different meaning than simply "of or relating to the noun". For example, "cocktail" in "cocktail dress" means "informal".--♠TBC♠ 04:23, 13 July 2008 (UTC)

I think that information belongs in the definition of cocktail dress, because in many circumstances it does mean “of or relating to:” cocktail shaker, cocktail party, cocktail lounge, cocktail wiener, cocktail hour, etc.

This seems to be a case where we're over-specifying to the point of being prescriptive. Other dictionaries mention attributive nouns, or assume that the reader can figure it out. Our convention of mentioning every possible usage leaves no room for the grey area: if we leave out the adjective it implies that this is not “normally” possible, whatever that means. But although “that dress is cocktail” is non-standard, it is not uncommon nor “incorrect” in colloquial speech.

Perhaps an adjective heading is a good idea, or an “attributive” label for the noun's inflection line. —Michael Z. 2008-07-13 05:55 z

We'd then be adding "empty" information to just about every noun in the English language. I'd hazard a guess that more than 90% of all English nouns can be used attributively, including many proper nouns (e.g. "Paris skyline", "United States government). Creating an "adjective" definition for all of these nouns would be a waste of our time and take up "valuable real estate"^tm on the entry pages. (Note use of both adjective and entry as attributive nouns in that last sentence.) --EncycloPetey 10:03, 13 July 2008 (UTC)

I wouldn't dispute the repetitiveness of the information, only that it might serve to prevent the addition of adjective senses by contributors while remaining at least as much an open Wiki as we are. As to "valuable", I would have assumed that any attributive-use Adjective header and line of linked text would follow the Noun and therefore fall into less valuable below-the-fold real estate, except for very short entries (one screen), where no real estate is especially valuable. DCDuring TALK 16:03, 13 July 2008 (UTC)

Let me clarify, this is not a request to add in all attributive adjectives, just the nouns that don't simply mean "of or relating to said noun" when used as such. For instance, the word "coffee", which can be used as an attributive adjective to mean "pale brown".--♠TBC♠ 10:27, 13 July 2008 (UTC)

Coffee and most other entries for colors have adjective PoS sections even though there is a universal rule that "allows" such words to be used as adjectives. DCDuring TALK 20:50, 13 July 2008 (UTC)

Actually, the rule runs the other way. Most color words are (originally) adjectives, but they are permitted to be used as nouns. For the historical reason, then, new color names acquire a sense of being an adjective. Also note that some color words also function as a verb meaning "to turn that color" (The page yellowed; The lawn began to green), although these are more aften formed by addition of a suffix to the color name. --EncycloPetey 22:06, 13 July 2008 (UTC)

Some print dictionaries contain a notation on some noun entries that indicates frequent attributive use. DCDuring TALK 11:53, 13 July 2008 (UTC)

What nouns are not used as adjectives? Would it make sense to mark these? —Michael Z. 2008-07-13 18:28 z

That would be worth noting, if we could think to ask the question when looking at the noun entry. DCDuring TALK 20:50, 13 July 2008 (UTC)

Another example, "spot" is an attributive-only adjective that can be defined as to be "made, paid, delivered, etc., at once" (as in a spot sale, which is obviously not a sale of spots).--♠TBC♠ 20:02, 13 July 2008 (UTC)

This is a fairly practical question. What combination of characteristics of word-as-noun, word-used-attributively, and Wiktionary-entry-for-word might warrant some kind of Adjective PoS or other indication? The practical benefit for us is to prevent needless entry of adjective PoS sections that we believe would be misleading. It also gives us the chance to explain a bit about attributive use of nouns, thereby fulfilling a modest educational function. The cost is about 3/4" of usually below-the-fold vertical screen space.

One possible limited application of the stub Adjective PoS section would be those words for which a contributor entered an Adjective PoS section that we subsequently determined to be merely attributive use. If there were no indication that contributors found the absence of an Adjective PoS confusing, then there would be little reason for us to add it.

Another application would be cases where the noun is attestable (and etymologically prior ?), but most of the usage is actually attributive. The justification would be that such instances are highly likely to lead to a contributor seeking to enter the Adjective PoS section. DCDuring TALK 20:50, 13 July 2008 (UTC)

Is this the same as in spot check? Would this be a derivative of on the spot?

Possibly, but that's more of an etymology concern.--TBC 14:44, 14 July 2008 (UTC)

Idioms vs Chengyu

Are the entries in "Category:Mandarin idioms" idioms in the English sense of the word, or in the sense of the Chinese equivalent of an idiom, the chengyu (using "chengyu" since I'm too lazy to type "four character idiom")? Chengyus are a lot more specific than idioms, as they require four characters and they usually come with a background story. For example, this issue is evident with 黑馬, which means dark horse, or an unexpected success. English speakers would immediately identify this as an idiom, since it can't be understood simply by knowing the meanings of each individual character. Chinese speakers, on the other hand, would not consider this to be a chengyu; in fact, they would likely just consider it to a normal compound and not idiomatic at all. So then, any thoughts?--♠TBC♠ 11:20, 13 July 2008 (UTC)

One possible solution to differentiate the two is to create a subcategory in Category:Mandarin idioms for chengyus. But then, what should we call the category? Category:Mandarin four-character idioms?--♠TBC♠ 11:23, 13 July 2008 (UTC)

For purposes of Wiktionary, Category:Mandarin idioms refers to idioms in the English sense of the word. As you mentioned, most people associate the term chéngyǔ with the numerous four-character expressions found in Chinese. However, Category:Mandarin idioms is not limited to just chéngyǔ. You could perhaps create a subcategory for chéngyǔ. However, I'm not sure how useful such a category would be. -- A-cai 12:26, 13 July 2008 (UTC)

I think it's important that we distinguish chéngyǔ from idioms, so that readers (especially those starting out in Chinese) can understand the culture significance of four character idioms in comparison to normal idioms. A subcategory certainly won't do any harm.--TBC 14:11, 14 July 2008 (UTC)

Category created.--TBC 14:15, 14 July 2008 (UTC)

Oscar the Grouch, Cookie Monster, and others

(I believe this has been brought up before, but I'd like to add in my two cents)

I'm fine with keeping entries like Nintendo or Tupperware, companies so ubiquitous that they've become synonymous with their products. But I draw the line at fictional characters. I know there's been some discussion of this before (a year ago I believe, with Care Bears and Teletubbies), but keeping these entries sets up a god awful precedent for keeping any pop cultural icon associated with a certain personality, nuance, or quirk. This opens up a Pandora's Box for entries like Bart Simpson (associated with adolescent rebelliousness), Eric Cartman (associated with excessive profanity and manipulation), Cheech and Chong (associated with "stoned slackers" and buddy flicks), and so on. I'm tempted to take this to RfD, but it's more of a policy issue. --TBC 13:05, 14 July 2008 (UTC)

Shouldn't the test be whether the phrase is used sufficiently to generate citations conveying meaning independent of the work with which the character is primarily associated? bd2412 T 13:09, 14 July 2008 (UTC)

But the problem is, virtually any pop cultural icon can be applied in such a way. Bart Simpson, Indiana Jones, the list is endless.--TBC 13:14, 14 July 2008 (UTC)

Concerning the Oscar the Grouch entry, I'm particularly concerned about citations like "I'd have looked a lot like Oscar the Grouch", "a pop-up “Oscar the Grouch” toy", "Oscar the Grouch flip-flops", where the quote is clearly referring to "Oscar the Grouch" as in the character from Sesame Street. For the other quotes, my comment above on pop cultural icons applies.--TBC 13:24, 14 July 2008 (UTC)

Also, regarding pop cultural icons, this applies to real people (and entries such as The Beatles and Abbott and Costello) as well: "He's a Jimmy Hendrix on the guitar", "He's the da Vinci of our time", "He's an Orson Orwell when it comes to directing", etc. See how this can be applied to nearly anyone in pop culture with an identifiable style, motif, or personality?--TBC 13:32, 14 July 2008 (UTC)

Adding on, the "out-of-context" rationale used to defend entries like these is extremely weak. Anything that's relatively famous can (and probably has) been used out-of-context. More examples (this time with cites): "the original TLJ was the Lord of the Rings of the gaming world" [10], "the Jewish people need a Ronald Reagan", [11], "the Bill Gates of Artificial Life" [12], "don't have a Jesse Jackson to speak for us" [13]--TBC 13:45, 14 July 2008 (UTC)

First of all, the pop-up toy and flip-flops do not necessarily support an out-of-context rationale. They are there to support an attributive-use rationale that worked its way into CFI and that I'd just as soon see work its way out. Out-of-context is superior in my view, and completely defendable. If you ask anyone on the street the who the following are:

• Bart Simpson
• Bill Gates
• Cheech and Chong
• da Vinci
• Eric Cartman
• Jesse Jackson
• Jimmy Hendrix
• Ronald Reagan
• The Beatles

then I'd bet the majority would know every one of them, if they don't nail, right on the head, the significance of every single one of them. What you're saying is that you think a word, the metaphor of an idea, that every English speaker knows, that we can use to convey a concept in conversation, that as you demonstrated we do in fact use without any additional references to its meaning or origin, should not because of some technicality be part of a corpus of knowledge that defines terms. A younger person today might not understand "Abbott and Costello routine" any more than they understand "dance routine". Why should we define one and not the other? Fifty years from now Bart Simpson might not be known either, nor I'm guessing are Cheech and Chong by many non-native speakers. It's not that the Orson Wells of directing is difficult to presume correctly, rather "the Orson Welles of the hand-held camera" or "the Orson Welles of software" or even "the Orson Welles of mice". If someone is likely to run across it and want to know what it means, then we should provide them with a brief explanation. DAVilla 21:01, 14 July 2008 (UTC)

The problem here is that, this logic would merit the inclusion of virtually any famous figure or pop cultural icon. I wouldn't opposed a soft redirect to Wikipedia, though.--TBC 21:32, 14 July 2008 (UTC)

I don't see how that's a problem. If someone is likely to run across the figurative use of any famous figure or pop cultural icon, and needed to know what is meant, then include them all, I say. This line is not exactly how CFI reads, by the way, which is part of the reason there are still red links. But we don't do soft links to pedia, maybe because we're just too proud.

I don't blame you for bringing it up as there will always be more debate. It's not so easy to reach consensus on some terms, and harder to generalize what the community opinion is more generally, and close to impossible to pin a rationale. DAVilla 07:49, 15 July 2008 (UTC)

That sounds reasonable. I personally don't believe that such entries belong in a dictionary, but then again, Wiktionary isn't really your traditional paper dictionary.--TBC 08:23, 15 July 2008 (UTC)

Regarding the "'attributive-use" rationale; couldn't any major brand or series name be used attributively? As in, "a Hot Wheels car", a "G.I. Joe figurine", a "Hitchhiker's Guide to the Galaxy t-shirt".--TBC 21:48, 14 July 2008 (UTC)

Yes, and that's why we went through a lengthy discussion and vote to set policy for brand names. Brand names now have their own amended requirements under WT:CFI. --EncycloPetey 03:32, 15 July 2008 (UTC)

That was a wording mistake, I meant brands as in franchises and pop cultural characters (like Indiana Jones), not products and trademarks like Nintendo and Band-aid, which CFI does cover.--TBC 08:23, 15 July 2008 (UTC)

You could have saved yourself a lot of work by reading the Talk page for Oscar the Grouch. That entry has already passed RfD, and was kept. Yes, pop culture icons can end up surviving CFI and RFD. Some cultural icons have penetrated everyday speech to an extent that they should be included in a reasonable dictionary, so that they can be explained to persons learning the language. Having an entry for Oscar the Grouch is no stranger than having entries for Hercules, Cinderella, or Romeo. These are characters in stories that mean something to the culture that shares them. Note that this is not the same as the full name of a real person, as that is a separate issue. --EncycloPetey 17:56, 14 July 2008 (UTC)

I've read through the discussion, and I found the rationale for keeping the entry extremely unsatisfactory. I realize that this has already survived RfD, but how does that bar us from discussing it? Consensus is never set in stone (I know this is from Wikipedia, but the concept still applies) and decisions can change; after all, this is a wiki. Again, allowing these entries to be kept would set a precedent meriting the inclusion of virtually any famous person, pop cultural icon, character, book, and so on.--TBC 21:32, 14 July 2008 (UTC)

Please do not cite Wikipedia policies or guidelines to support actions here; they do not apply. We don't want to keep hashing over the same RfD discussions over and over, which is why we archive them on the talk pages of the entries. Once a decision has been reached, an argument of "I don't agree" isn't sufficient to reopen the discussion. Yes, we have set a precedent with that, but not the one you think; please re-read the end of the discussion. The item must still have citations suitable to pass the requirements of WT:CFI. --EncycloPetey 22:08, 14 July 2008 (UTC)

I understand, but the concept that decisions are never binding and that consensus can always change on a wiki still applies to Wiktionary. And regarding citations, my point is that such citations can (and likely will) be found for any famous person, book, pop cultural icon, etc.--TBC 22:13, 14 July 2008 (UTC)

But I do agree with you on one point: We need a comprehensive policy for proper nouns.--TBC 22:20, 14 July 2008 (UTC)

You mean additional, beyond what WT:CFI already says? It already says that Thomas Jefferson will not have an entry because it is a specific person's name that is not used attributively. Yes, citations can be found for the things you mentioned, but in most cases those citations will not be attributive use. Please review CFI carefully before you continue this discussion. You're not saying anything that hasn't already been said here many times before. --EncycloPetey 03:37, 15 July 2008 (UTC)

Thomas Jefferson's full name is not used attributively because Jefferson is a distinctive last name. But for famous figures with a common last name (like Jackson), full names are used attributively. Using Jackson as an example, there's: "he's a Jesse Jackson Democrat" (New York Times), "wearing a Michael Jackson hat" (New York Times), "fixed him with the Samuel Jackson stare" [14]. Either way, although I think attributive use is a good litmus test for determining an entry's inclusion, I just don't believe that it should be the only requirement for inclusion. Also, WT:CFI covers trademarks and products, but not fictional characters like Oscar the Grouch and Cookie Monster (which was the original point of this discussion).--TBC 08:23, 15 July 2008 (UTC)

Yes, it does. It has an entire section on "Names of specific entities" that does not limit itself to "real" versus "fictional". The few examples may not be fictional, but that doesn't mean that it doesn't apply to all names of specific entities. The Oscar the Grouch previous discussion specifically tied it to this same section. Again, I say please review CFI carefully before you continue this discussion. You're not saying anything that hasn't already been said here many times before. --EncycloPetey 19:49, 15 July 2008 (UTC)

Not trying to fight for the last word, but I did read the section carefully, and the section does not specifically talk about fictional characters. Although fictional characters might fall under specific entities, Oscar the Grouch is a lot different than New York. It doesn't matter either way; I can see the point to having such entries. --TBC 13:25, 16 July 2008 (UTC)

I like the examples you give. I wonder if, contrary to what CFI claims, Thomas Jefferson is used attributatively, somewheres out theres. DAVilla 06:55, 16 July 2008 (UTC)

Not sure about Jefferson (since he's usually referred to attributively by his last name), but George Bush has certainly been used attributively and out-of-context. Attributive examples: "He should take a George Bush approach" New York Times, "a George Bush compromise" [15], "Toevs calls Kolbe a George Bush Republican" [16] Out-of-context examples: "Has Campbell pulled a George Bush?" [17], "hes too old, a flip flopper, and a George Bush lite" Washington Post--TBC 13:25, 16 July 2008 (UTC)

I think the attributive-use rule is being interpreted too loosely. “... with a widely understood meaning” should be better defined.

“Oscar the Grouch toy,” “Oscar the Grouch suit,” and “Oscar the Grouch flip-flops” are not using his name attributively to invoke some widely-understood characteristic of said Grouch. These are merely the names of products. Likewise, I think “a Jesse Jackson democrat” is referring to someone's specific political affiliation with Mr Jackson, and not invoking him as an adjective with an understood meaning. “A Samuel L. Jackson stare” relies on context and the word stare for its meaning, and makes perfect sense to someone unfamiliar with this Mr. J. All of these uses refer to specific associations with the people, and none of them relies on a generalized understanding of who they are or some quality of theirs. (What the heck is a “Michael Jackson hat?”)

“A George Bush approach” refers Bush's specific actions in office, not generally to some understood Bush quality—it has no meaning out of its context. The same goes for every other of these Bush citations.

A more illustrative example: 2-term Winnipeg Mayor Sam Katz: “a Sam Katz party,” “a Sam Katz Party-dominated Council,” “a Sam Katz quote,” “a Sam Katz business investment,” “a Sam Katz campaign sign,” “a Sam Katz re-election campaign promise.” I would argue that despite these citations, Katz doesn't belong in Wiktionary.

Supporting citations should have meanings which are widely understood without reference to a specific context or circumstance.  —Michael Z. 2008-07-16 15:34 z

I agree. Part of the problem is that "attributive" has two senses: a technical sense that covers "George Bush" in "George Bush ally" but not in "ally of George Bush", and a non-technical sense that covers "George Bush ally" but not "George Bush" in "he's a George Bush ally". CFI aren't clear about which sense is meant, and (as you might expected) editors' ad-hoc interpretations vary significantly. —Ruakh_TALK 20:51, 16 July 2008 (UTC)

Can anyone suggest a revision of WT:CFI#Names of specific entities?

I also think that Empire State Building is a poor example, because it is so familiar and its special quality is understood worldwide. Here are some examples of its use attributively, or metaphorically (allegorically?):

1. “His ignorance was an Empire State Building of ignorance. You had to admire it for its size.” —Dorothy Parker (1893–1967)
2. “A lecherous and ruthless real estate developer, who built an Empire State Building-like Tower of Babel, corrupts an innocent girl in this film teeming ...”[18]
3. “... Cheerleaders but also the deadly Ass-teroids and King Dong, who has a unique way of fending off attackers from his Empire State Building-like perch.”[19]
4. “Canada, for instance, doesn't have an Empire State Building, and so makes do with the not-even-real-sounding CN Tower.”[20]
5. “ Unlike the Empire State-like Ukraine, where the rest of our group were staying, or the new mammoth Rossia, the National is a pre-Revolutionary hotel, and, compared with them, built on modest . . .”[21]
6. “... a musical experience any more than use of matter and windows will produce an Empire State Building unless one actually builds an Empire State Building.”[22]
7. “But more than anything else, Allie had pointed up, like an Empire State Building in a row of Neissen huts, that even Lepke—even the cool, calculating king—could make one mistake.”[23]
8. “Most people, in their drive to get rich, are trying to build an Empire State Building on a 6-inch slab.”[24]
9. “But, shit, when they kill the spirit of your better half, then you are left to do the dishes yourself and rekonstruct your Life like an Empire State Building out of toothpicks.”[25]
10. “He contended that establishing claims to knowledge was akin to building an Empire State building out of toothpicks, ‘most of which we haven't got and cannot be given.’”[26]

 —Michael Z. 2008-07-17 21:42 z

Awesome! DAVilla 04:42, 25 July 2008 (UTC)

Past decisions don't have any authority over present consensus. They can set precedent, affect or obviate new discussions, but anything can be reopened, especially if there is some point or information that hasn't been covered. We all signed the GFDL, so we shouldn't get too attached to our contributions. —Michael Z. 2008-07-14 23:24 z

Appendix:Mythological characters

It's insensitive to refer to God, Jesus, Muhammad, Buddha etc. as “characters.” Any objections to changing the title and references on the page to “Mythological figures?”. —Michael Z. 2008-07-14 19:58 z

"Mythological figures?"? I object. How about, instead, "Mythological figures"?  :-) —msh210℠ 20:38, 14 July 2008 (UTC)

Perhaps a separate appendix for Appendix:Deities? That way, God and the others won't directly be under mythological characters.--TBC 22:18, 14 July 2008 (UTC)

Careful now, there are still people who worship Thor - and consider him to be as real as any Christian, Muslim, or Buddhist counts their deity. bd2412 T 22:28, 14 July 2008 (UTC)

Another possibility is "Beings in X," thereby eliminating the word mythology. Wakablogger 23:09, 14 July 2008 (UTC)Wakablogger

There are atheists who faithfully consider any religion a fantasy, and there are probably adherents of just about any marginalized or even fictional religion. The dictionary shouldn't make judgments, and trying to categorize would be a can of worms. The simplest solution is to treat the subject with basic respect.

At first I also thought mythology sounded wrong, but it doesn't necessarily disparage or discount in the way tha myth or legend might, and is not uncommon. Cf. w:Christian mythology, w:Islamic mythology, w:Jewish mythology, w:Slavic mythology, w:Religion and mythology. —Michael Z. 2008-07-14 23:18 z

The Appendix in question is nothing more than a list of linked words, so why have this appenidx at all? We already have Category:Mythology (and subcats). Why not just use categories? --EncycloPetey 03:29, 15 July 2008 (UTC)

Agreed, delete. This appendix is a stone's throw from worthless, and we already have the cats. These sorts of things are just itching to push peoples' buttons. I'm an atheist, and even I think it a bit insensitive to put this all under mythology. -Atelaes λάλει ἐμοί 05:41, 15 July 2008 (UTC)

Good point, EP, Atelaes: delete. (Well, bring to RFDO, then delete.)—msh210℠ 17:58, 15 July 2008 (UTC)

I disagree with that - the list includes (and should include) redlinks which point us towards entries that we are missing. A category wouldn't. bd2412 T 19:21, 15 July 2008 (UTC)

A number of categories include, on the category page itself, a list of words, including redlinks, that someone ought to (write, if necessary, and) categorize in that same category. See, e.g., Category:English three letter words. Same can easily be done here, and imho it'd be more appropriate.—msh210℠ 19:25, 15 July 2008 (UTC)

That has got to be the worst category. It has like 17 entries in the category, and the list of words not in the category is simultaneously about twelve times as long as the words that are in it, and yet woefully short of the total count of three letter words (and some of those appear to be acronyms, rather than words anyway). Besides, are we going to have subcategories for the particular branch of mythology that each character comes from? bd2412 T 00:55, 16 July 2008 (UTC)

I agree the three-letter-word category needs help; I was just using it as an example. As to your final question, whether we need subcats for various mythologies, I think not. Frankly, I'm not that fond of topic categories; alas, they are, apparently, here to stay. But we definitely don't need very fine ones. (Note incidentally that my dislike of topic categories does not apply to lexicons, including lexicons of jargon used only in certain fields, as marked by context tags.)—msh210℠ 17:00, 16 July 2008 (UTC)

I agree that we ought not have categories specifying the mythology of a certian figure - but if we don't have them, is that a point in favor of an appendix making that sort of separation? Especially in a dictionary, where lexical origin of the set of words might be important? bd2412 T 20:32, 16 July 2008 (UTC)

Yes, but not a strong one, in my opinion.—msh210℠ 20:13, 18 July 2008 (UTC)

Would the same apply to Appendix:Fictional characters? Do we have a guideline setting out what kinds of lists are desirable, or how to obviate them with categories? —Michael Z. 2008-07-16 00:50 z

The ultimate goal of any category is to have all the red links blue (or removed because they fail CFI). At that point the list of items can usually be removed from the category because the items exist. One has to be careful though, because a blue link does not necessarily mean the entry exists in the desired language. --EncycloPetey 01:02, 16 July 2008 (UTC)

Rename it to "deities and mythological character". Does the trick without needlessly separating two tightly related topics where the breaking line can be difficult to establish. Circeus 23:46, 5 August 2008 (UTC)

Alt spellings

An idea: what if Autoformat used the information on one of the pages (say color) and duplicated it on the other page where translations, etc. might be listed but might be inconsistent between the two pages? This wouldn't be for all alternative spellings, just the ones where there are 2 complete entries. And the pages would have to be marked by hand beforehand to make sure that they really are identical. Nadando 04:54, 16 July 2008 (UTC)

But not all alternative spellings always apply to all senses (see Makemake). AF would have to know which senses paired with either other, and any edit made to clarify definitions would make the two different and stymie AF. The caveat of having to mark the pages beforehand is a big caveat. And there are many things that will have to be different betwen the pages, such as quotations and example sentences. --EncycloPetey 17:39, 17 July 2008 (UTC)

Fictional brand names in citations?

Hello. This edit and edit summary bring up a very interesting question. If an author uses a word as part of a fake brand name in a fictional work, does it count as a cite of the word? Language Lover 05:09, 16 July 2008 (UTC)

If it conveys the meaning that the definition posits, I don't see why not. But how often is that going to happen? bd2412 T 05:13, 16 July 2008 (UTC)

I've seen stuff like this now and again. It may not contribute directly to any of the definitions, or marginally so, but that doesn't mean it isn't worth keeping. DAVilla 06:47, 16 July 2008 (UTC)

I don't think a citation like the one removed in the edit linked to above is useful as an example or should count as attestation: there's no meaning conveyed by the word there, so not, in particular, the meaning given in our definition.—msh210℠ 16:40, 16 July 2008 (UTC)

Sure there is - have you heard the legal maxim, inclusio unius est exclusio alterius ("inclusion of one is exclusion of another")? When, in describing a weapon, you see the phrase "MegaHurt InstaKill Thunderbomb", you know right away that "InstaKill" is the kind of word that falls into the same category of meaning as MegaHurt and Thunderbomb", i.e. it has something to do with inflicting extremely large amounts of damage on something else. bd2412 T 20:29, 16 July 2008 (UTC)

Wiktionary:Internal links

Wiktionary:Internal links was written with the new or newish user in mind; please add to it (or blast it, as appropriate).—msh210℠ 17:06, 17 July 2008 (UTC)

It is a useful page indeed, particularly the links to the obsurish pages. Would this be better placed at Help:Internal links, seeing as it is written to aid the newish user and is not policy, but an amalgamation thereof? See #Spam Policy below for the same debate. Conrad.Irwin 22:27, 19 July 2008 (UTC)

Moved.—msh210℠ 21:20, 27 August 2008 (UTC)

Broken superlative template?

When I went to use the superlative template I got this in the edit box:

"#REDIRECT Template:new en superl bot"

dougher 02:16, 18 July 2008 (UTC)

It seems as if someone wants it work the same way that the new comparative template works, showing "most X" in the sense line. It would be desirable to have said appearance, but have the option of switching it off. DCDuring TALK 02:26, 18 July 2008 (UTC)

Well, exist that that isn't the case in all languages. In Latin the superlative can mean "most X", but it can also mean "extremely X" or "very much X" without comparison to other X's. If this template does it, it should be something that must be turned on, not the default. --EncycloPetey 06:24, 18 July 2008 (UTC)

Then why not {{en-superlative-of}}? DCDuring TALK 10:43, 18 July 2008 (UTC)

I guess I was looking for someone to fix this as I don't know how to. I've taken to using the comparative template and substituting the word "superlative" for "comparative". The templates I'm talking about are those provided by the links in the table below the following text: "You can create a new entry with one of the following preloaded entry templates: " - dougher 05:51, 25 July 2008 (UTC)

Phrasebook Criteria

We need a separate section on CFI detailing the criteria for inclusion of phrasebook entries. WT:CFI mentions phrasebook entries very briefly under the idioms section, but the wording is extremely vague (entries must be "very common" and "useful to non-native speakers"). Namely, we need to determine what qualifies as a common phrase, what qualifies as a useful phrase, and how phrasebook entries should be worded (grammatically, that is) and formatted. There should also be a guideline for writing phrasebook entries. At the very least, something more extensive than "See Wiktionary talk:Phrasebook for now".--TBC 09:39, 18 July 2008 (UTC)

Well this hasn't been a problem thus far, but I guess we could try to flesh it out a little bit more. It's not time to create completely objective criteria, however. I would want to work backwords from the consensus on multiple entries in RFD before trying a major overhaul. So far I don't know that any entries have failed.

To me, useful phrases are correlated to common or frequently occurring scenarios. For instance, being in a restaurant and needing to use the bathroom is a common scenario. Counting goldfish in a bowl is not. It's certainly conceivable that someone would count goldfish in a bowl, and an il-y-a construction might be "useful to non-native speakers" learning the structure of a language, but it's not something that happens frequently enough to warrant mention. It's not the type of entry you're normally think of being in a phrasebook.

The "very common" criterion isn't so important to me. What's more important is to say that this is the way that native speakers would tend to phrase a concept. There may be other ways of saying it that are both grammatically correct and attestable, but they're not always the most immediate or obvious ways of addressing the choice of wording. In some cases there may be more than one way of expressing an idea, and that's fine too. But they'd all have to be very common, first of all, and it's worth pointing out that the more informal terms are not something a non-native speaker would need to be introduced to right away. "Where's the little boy's room?" is a very common phrase, but it's not what you would teach someone who has trouble communicating. DAVilla 06:25, 19 July 2008 (UTC)

Spam Policy

Instead of using an interwiki link to the Wikipedia policy on spam (which might not completely apply with Wiktionary), I think we should have our own comprehensive policy on spam. Currently working on a draft.--TBC 10:20, 18 July 2008 (UTC)

Started up Wiktionary:Spam. Any other Wiktionary-specific additions?--TBC 10:50, 18 July 2008 (UTC)

I think this is a bad idea, the Wikipedia policy covers everything and our version will just go stale, a bit like that Vandalism page and all the other policies that are merely copies of Wikipedia. (see also Wiktionary:Three-revert rule Wiktionary:Be bold). I agree that Wiktionary could do with some explanatory pages, but copying from Wikipedia does not seem to work. Conrad.Irwin 20:51, 19 July 2008 (UTC)

I had a second thought! These pages should not be policy, I think that is what annoys me. Instead we should have lots of pages like Help:Spam which help newbies understand what the community thinks of spam. What does anyone else think? Conrad.Irwin 22:14, 19 July 2008 (UTC)

You mean like essentially a guideline?--TBC 21:16, 20 July 2008 (UTC)

"Guideline" to me implies a set of instructions, this isn't a set of instructions it's just a definition of what spam is, or have I misunderstood? I don't really see what the purpose of such a page can be except to help people understand what spam is, and so it seems logical (to me) that this should be in the Help: namespace. Conrad.Irwin 23:33, 20 July 2008 (UTC)

Could help. Can't hurt much, even if completely neglected. Doesn't require a vote. Why not? The approach has lots of applications. It's a good way of seeing what we really do think as a group and might lead to some policy decisions. Almost anything that would explain us to newbies would be very good. It is a actually a good task for a senior newbie or not so veteran contributor, like me. DCDuring TALK 23:15, 19 July 2008 (UTC)

Well ignoring WP:BEANS, no it doesn't hurt at all. However I can't see how this page functions as a policy. Conrad.Irwin 23:33, 20 July 2008 (UTC)

The "help" given could be wrong when written (because help is a low-priority/low-fun task) or become obsolete and therefore somewhat contradict to whatever policy might be. Definitely not policy itself, but could lead to policy if discussion is active and productive. DCDuring TALK 23:55, 20 July 2008 (UTC)

Right-hand ToC interactions with other templates

I really like having the ToC on the right so that I can both see much content and jump to content by clicking on the ToC. From time to time there are issues with other templates, which force the content (or the ToC ?) off the first page, defeating the purpose of having it. An example is {{rfe}}. In contrast {{etystub}} does not. Is it worth fixing {{rfe}} or could/should it be deprecated? How hard is it to avoid bad interactions with the right-hand ToC? Can the right-hand ToC code or CSS be tweaked to force proper display? DCDuring TALK 21:10, 18 July 2008 (UTC)

Hear, here. DAVilla 06:01, 19 July 2008 (UTC)

I take it that the source of the problem is the width of the pseudo-graphic. Templates with poor interaction with rh ToC: {{rfe}}, {{rfp}} {{rfap}} DCDuring TALK 10:37, 19 July 2008 (UTC)

I have reduced the width of the graphics to 50% in all 3 of these templates. They might look better not centered, but with a small left-hand margin. DCDuring TALK 10:53, 19 July 2008 (UTC)

I've now fixed them using the same CSS that is used on {{rfv}} etc. I have created the meta template {{request box}} for there, that mimicks {{maintenance box}}. I've put the width to match maintenance box, but that can easily be changed by editing Template:request_box. Conrad.Irwin 20:45, 19 July 2008 (UTC)

Your way is definitely better and would be my model, but the new meta template seems to obivate the need. DCDuring TALK 21:12, 19 July 2008 (UTC)

Are there other remaining barriers to making the right-hand ToC the default? Should we give more time for identifying problems? Should any that are discovered be reported here? or elsewhere? DCDuring TALK 21:15, 19 July 2008 (UTC)

There are a few issues with some of the CJK entries I think, but as I don't spend much time there I'm not sure what they are or what causes them. I definitely think we should activate this for a trial period of a few days, and then if things look good we might need to VOTE before enabling it permanently. Conrad.Irwin 22:18, 19 July 2008 (UTC)

RU seems to be aware of CJK issues. What are his thoughts? DCDuring TALK 23:18, 19 July 2008 (UTC)

Oh! Bless you and bless the hand that raised you! Having Wiktionary move to the right hand table of contents is going to make my stuff so much easier to do! How soon can we begin the test? — This unsigned comment was added by Sack36 (talk • contribs) at 14:58, 28 July 2008 (UTC).

Another observed template problem with right-hand ToC: {{rel-top5}} forced itself and everything following to below the ToC. I found it at -a and edited it to {{rel-top}}, which does not have the problem. I did not experimentation with similar templates to check the extent of the problem. DCDuring TALK 19:52, 23 August 2008 (UTC)

I have also noticed the interference with other show/hide bars for declensions and conjugations that occupy only the right-hand side of the screen. I dimly associate them with some CJK entries. I'll try to find some instances. DCDuring TALK 19:56, 23 August 2008 (UTC)

See {{ko-pron}}, as used in 직장. DCDuring TALK 20:02, 23 August 2008 (UTC)

Norwegian – again

Well, after some discussion on my talk page and the creation of Wiktionary:About Norwegian, I am to raise this question again. How should we handle Norwegian entries? Today we handle all words that is the same in Bokmål and Nynorsk as Norwegian, while we differ if a word only exists in one of the languages. But, as Kåre-Olav says, almost every word (nouns, adjectives and verbs) are inflected (slightly) different. We will renew the discussion on no.wikt, and see if we maybe should work differently there, but for now we do it the same way as here ... but I'm not sure if it's the best way. Any thoughts? Can we go for totally differing between the two languages (will give quite a lot of double entries, as it already is with Danish, Swedish and Norwegian in general)? --Eivind (t) 16:12, 20 July 2008 (UTC)

Previous discussion here. I'm going to stay out of this conversation, as I did so much waffling on the last one that I think I would be of little help. I would like to see Meco's input, as they are the only one who has stayed and consistently worked on Norwegian here. Not only are they going to be most affected by any decision, but they are in a better position to assert what will work. -Atelaes λάλει ἐμοί 19:29, 20 July 2008 (UTC)

I think that this is a complex issue that is not yet ripe for a comprehensive solution. I have for my two year tenure here mostly stayed away from the problem by sticking exclusively to Bokmål entries, basically ignoring the mere existence of Nynorsk, not out of disregard for Nynorsk (disdain for Nynorsk is prevalent among Norwegians), but because I have felt uneasy and unsure about how to deal with the Norwegian problem of two separate, however quite similar, languages. Only very recently have I begun fiddling with dual entries (partly inspired by the entrance of User:Kåre-Olav).

There are inconsistencies in the Norwegian connjugation and declension templates that are tentatively dealt with in a piecemeal way, and I propose that the strategy to be adopted for solving this problem is that of continual reconstruction and reconfigurations until we perceive that we are on a converging track. Having a separate Norwegian language forum is a good initiative, I think. I don't sense, though, that we are yet at the point where a comprehensive discussion is likely to provide clarity and solutions with regard to coming up with guidelines for how to deal with Norwegian entries. I think we should keep tabs on one another, taking pointers and giving feedback to fellow Norwegian entry contributors, and having a dedicated forum would be a useful adjunct in this respect.

Finally, I will mention two things. Firstly, the problems of the Chinese languages which does lead to a similar problem complex of not being able to pin down comprehensive, integral guidelines for that are only partly distinguishable. Secondly, that I don't think we should hope to find a solution by gleaning from the Norwegian Bokmål Wiktionary, as it's still too much in its infancy. We should keep lines of communication with it open though, as the problem is a common one, and a sustainable solution found either place would quite possibly be adoptable in the other. __meco 09:44, 21 July 2008 (UTC)

In my opinion, the mere existence of Wiktionary:About Norwegian is probably useful. I know that Wiktionary talk:About Ancient Greek has provided a nice centralized place for discussion, a grc:Beer parlor if you will. And yes, it is probably best to come to such a conclusion after a lot of editing and discussion, instead of simple chatter. -Atelaes λάλει ἐμοί 00:13, 22 July 2008 (UTC)

Since Norwegian is considered a macro language by SIL we have on svwikt already devided Bokmål and Nynorsk to be treated as two different languages. The same goes for all other kinds of macro languages according to SIL. For Norwegian also inflections are different between Bokmål and Nynorsk so that's yet another reason. I know deviding entries is quite some work, so I am not going to tell it is worth the effort, but my opinion is that it is much better with separated entries for the two languages. ~ Dodde 13:35, 23 July 2008 (UTC)

Googleability

Good news: Google now includes Wiktionary in its search for definitions using "define:" keyword. (E.g. [27]) --Dan Polansky 10:58, 21 July 2008 (UTC)

Callooh! Callay! DCDuring TALK 11:57, 21 July 2008 (UTC)

That's great to hear.--TBC 08:40, 22 July 2008 (UTC)

w00t! (About time too :p) On a related note, following Wiktionary:Votes/2008-06/Install_MetaKeywords_Extension I have filed a request to install the extension. Conrad.Irwin 08:56, 22 July 2008 (UTC)

However - if you try a search for define:andare - it finds it in Italian Wiktionary, but not in the English version. SemperBlotto 09:06, 22 July 2008 (UTC)

And it doesn't find recent additions - nothing for define:thermophotometry (added in mid June). SemperBlotto 09:26, 22 July 2008 (UTC)

Adding pastable examples to WT:ELE's "Example sentences" section

I hereby suggest that we add pastable wikicode examples of the format recommended in WT:ELE's "Example sentences" section, to that section. They could look like this:

#Definition of example
#:''This is a '''example''' sentence.''

#Definition of word-in-non-Latin-script
#: ''This is an example sentence using '''word''' in non-Latin script.''
#:: This is a transcription of the above example sentence.
#:: This is a English translation of the above example sentence.

EncycloPetey asked that I bring this up through channels, rather than just making the change, so I am doing so. Comments, suggestions, criticisms? JesseW 01:33, 22 July 2008 (UTC)

Conceptually I think examples are good — indeed, that's the principle underlying that section — but I worry about bloat. That page is already quite long. Might it be better to link to some well-formatted entries that have (among other things) example sentences? (Doing so would also resolve my other concern, which is that your examples only demonstrate the format of example sentences — the stuff in the bulleted list at WT:ELE#Example sentences — and not their contents, uses, etc. — the stuff in the numbered list below it. Arguably the latter is more important.) —Ruakh_TALK 02:43, 22 July 2008 (UTC)

I sympathize with the concern about bloat. Let me modify the proposal. Create a new page Wiktionary:Example sentences, and put the examples, and some of the textual material currently in the Example Sentences section, and make the ELE section just a summary (and link to) the outside page. If this seems like a good idea, I'll work it up in more details. JesseW 07:54, 22 July 2008 (UTC)

How about Help:Example sentences to expand what is currently in WT:ELE, I'm not sure that we wish to regulate the content of example sentences any more than is already done there. (In case people haven't noticed yet, I want to use the Help: namespace more!) Conrad.Irwin 08:45, 22 July 2008 (UTC)

We could do both. The Wiktionary: namespace is where we want style guides, and the Help: namespace is where we want helpful advice and basic "how-to" info. --EncycloPetey 01:01, 23 July 2008 (UTC)

I have no interest in changing or adding to policy at all. All I am suggesting is adding copyable examples of existing policy on example sentences to somewhere easily findable from WP:ELE. It doesn't particularly matter much to me where it's put. I suppose the simplest change would be, can we add a link at the bottom of the WP:ELE "Example sentences" section to Wiktionary:Example sentences, and put the copyable examples there? No other change, no modification to policy, no expansion of any sort, no danger of bloat. Just making it easier for people to format example sentences according to existing policy. (Please pardon the grumpy tone, I'm having a rather trying day right now.) JesseW 23:16, 23 July 2008 (UTC)

Yes, do it. The page is long, but it has to be too long before we can split it, or we'll just go in circles griping about how to split something that isn't there yet. DAVilla 04:37, 25 July 2008 (UTC)

Image count

Many modern print dictionaries brag about the number of illustratios they have. Is there a way to count the number of Image: calls we currently have? --EncycloPetey 01:59, 22 July 2008 (UTC)

We can do cat enwiktionary-20080613-pages-articles.xml | grep '\[\[Image:' | wc -l on the XML dump, which implies that we had approximately 9784 images on the 13th of June. Conrad.Irwin 08:42, 22 July 2008 (UTC)

Thanks. A ballpark figure was all I wanted. Compare that to the American Heritage Dictionary (4th ed.) which boasts 4000 images. --EncycloPetey 00:58, 23 July 2008 (UTC)

And I somehow doubt its images are full-color graphics. :-) —Ruakh_TALK 02:08, 23 July 2008 (UTC)

Actually, many of them are. Some are line drawings or diagrams, but the latest edition went to using many full-color photos. --EncycloPetey 18:09, 23 July 2008 (UTC)

<propaganda> Let me put in a good word for drawings and diagrams (with or without color) vs. photographs. Both can play a role. But the supposed "realism" and detail offered by photographs come at a price. What is relevant or important can be lost. I believe that the finest bird book for North American birds is Sibley's. It contains no photographs, but many drawings, highlighting distinctive features and eliminating background and foreground clutter. Color is often an irrelevant distraction from underlying structure. We could probably use more drawings for some entries for which there is no photograph that can convey the meaning. If you see such a possibility, please insert {{rfdrawing}} instead of {{rfphoto}} or {{rfimage}}. </propaganda> DCDuring TALK 18:47, 23 July 2008 (UTC)

<promotion> I can say from experience, it is quite a nice change of pace to just hit "random English" over and over looking for pages with no image, then hunting around commons for a good example photo or drawing. If anyone is looking for something to do, or has spent hours on a laborious task and needs something different, it doesn't take too much effort and improves the look and usefulness of Wiktionary quite a lot. - [The]DaveRoss 19:24, 23 July 2008 (UTC)

Good idea. Did you notice any systematic pattern? What are the kinds of things that needed images? DCDuring TALK 20:59, 23 July 2008 (UTC)

I'm guessing mostly animal and plant entries. They also need separate plural entries as well.--TBC 17:43, 24 July 2008 (UTC)

Lots of geographic locations (such as we have) could also use maps showing where they are. Many anatomical and architectural entries also lack images, or could use better ones. And there are still plenty of basic concrete nouns that lack images, beyond just the plants and animals. --EncycloPetey 21:34, 24 July 2008 (UTC)

Talk in foreign languages

I was under the impression that all talk pages had to be in English - so that we can all join in. Is that not the case? SemperBlotto 15:59, 24 July 2008 (UTC)

I haven't seen the past discussions on this, but my own opinion is that aside from [[User talk:*]], discussions really should be in English except in extraordinary circumstances (for exactly the reason you mention); and even on [[User talk:*]], it's preferable that they be in English (likewise). —Ruakh_TALK 16:52, 24 July 2008 (UTC)

Agreed. Excluding User Talk, talk pages should generally be in English. But, out of curiosity, what incident made you bring this up? --TBC 17:08, 24 July 2008 (UTC)

Jeez. Some IP asked an innocent question in Portuguese on Talk:waltz and SB deleted it while I was answering. Yes, in general discussion should be in English, but we have plenty of language knowledge here, minor questions and such should be fine. (In particular, explaining things in other languages for non-English speakers is helpful when possible, eh?) Robert Ullmann 17:21, 24 July 2008 (UTC)

I think a good middle ground is: If you see someone asked a question in a foeign language, translate it into English if you can; answer it in the FL if you can; and answer it in English if you can. As many of those as possible. Also, if you're dong any of those, you might also mention that discussion should be in English. And if a question posted in a FL goes unanswered and untranslated for some time, it can be deleted (as, if no one can translate it, then presumably no one can say whether it's irrelevant, spam, indecent, or cetera). (I do not intend any of this to be applicable to usertalkspace.) Does this seem reasonable?—msh210℠ 18:26, 24 July 2008 (UTC)

+1. It's worth pointing out that a foreign-language discussion means few of us can join in, while a deleted discussion means none of us can. —Ruakh_TALK 19:27, 24 July 2008 (UTC)

We want to encourage interaction in English, as many folks here communicate only in that language. And we shouldn't penalize or turn people away simply because their English communication skills are limited. However, often, a non-English comment or question cannot be read by most users here, and it can be difficult to judge whether such a posting is a comment, a question, personal information, vandlism, or random nonsense. --EncycloPetey 19:36, 24 July 2008 (UTC)

What about User: pages?—msh210℠ 17:09, 24 July 2008 (UTC)

As long as the user isn't using the page as a personal page (without anything informative or dictionary-related), he should be free to chose whatever language he wants (though English is still preferred).--TBC 17:17, 24 July 2008 (UTC)

Dictionary-related, you mean? Anyway, I disagree. Since our policy pages are in English only, it is expected all editors can understand English to some degree, and I see no reason their userpages should not afford the widest readership among editors (who are the only ones who are typically reading userpages, I imagine). (For this reason, too, I think our Babel boxes' texts should be in English.)—msh210℠ 18:26, 24 July 2008 (UTC)

Whoops, minor fluke. Comes from having to edit different wiki discussions simultaneously, I suppose. Anyhow, as long as the userpage isn't offensive or off-topic, honestly there shouldn't be much of a problem. --TBC 06:46, 25 July 2008 (UTC)

According to the draft proposal Wiktionary:Usernames and user pages, "User pages […] should be primarily in English." That's not the result of any discussion — I added it unilaterally — and I can't really imagine enforcing — but it's been there for eight months without any comments or objections that I can recall, despite some reason to believe that a fair number of editors have read that page. I agree that the text of Babel templates should be in English, but I don't mind if they're in both English and the mentioned language. —Ruakh_TALK 19:27, 24 July 2008 (UTC)

Then what about user talk pages? I've seen some that are not entirely in English, which neither stirred any trouble. I'm no more offended by that than I am by the fact that not all of our content is in English. DAVilla 04:16, 25 July 2008 (UTC)

OK - hornet's nests stirred up while you wait. What I should probably have done is marked the page with {{notenglish}}; the question could then have been answered, but would have been deleted after a month or so. SemperBlotto 21:27, 24 July 2008 (UTC)

Trivial words in English

The addition of nickel-rich suggests the need for some 110+ other words for materials that are rich in an element, some of which will be attestable. Further, other chemical nouns can probably form adjectives with -rich. If we like such words, why not have a bot create them? I personally think they add virtually nothing to Wiktionary, except to inflate our statistics. Do compounds formed by postfixing an adjective merit inclusion though virtual synonyms like high-nickel do not?

There are many trivial words, some formed by prefixes and suffixes, some by compounding. Sometimes the formations of adverbs and adjectives from nouns is not obvious and the meanings even less so. Many fairly trivial derivations should be entered. Even the formation of antonyms (by a-, un-, in-, non- (or not)) is worth being explicit about, especially since there are real differences.

I am less sure about words formed with other suffixes like -less, which can be applied to virtually any common noun.

I have serious doubts about many other compounds, including those on some of the "wanted" lists. The most relevant policy material seems to be WT:CFI#Idiomaticity and WT:CFI#Attestation vs. the slippery slope. Have there been discussions of special criteria for including or excluding classes of English compounds? For example, I would assume that any compound that is attested spelled solid merits inclusion. Is it worthwhile or even possible to have any more specific guidelines in this area? DCDuring TALK 16:20, 24 July 2008 (UTC)

As I understand it, we do not allow separate entries for lexemes used to create hyphenated compounds (like -rich) nor do we indicate at rich that the word is used to form such compounds, except by example in derived or related terms. DCDuring TALK 16:31, 24 July 2008 (UTC)

We don't? Do [[-collar]] and [[induced#Usage notes]] need to be deleted, then? —Ruakh_TALK 16:55, 24 July 2008 (UTC)

This would be why I test my understanding. It's hard to tell what we do based on what one happens to see in less than a year of editing. I certainly doubt that we are approaching any kind of consistency in or comprehensive coverage of such things. Are those excellent examples of how it ought to be done? Who says? Is it worth a guideline? DCDuring TALK 17:18, 24 July 2008 (UTC)

Is everyone agreed that -collar is a Noun? Are our users trained to look for this kind of thing in usage notes? Would a non-gloss definition be better? Shouldn't we have a fuller list of derived terms? Should they be shown as redlinks to suggest that contributors add the entries? In all cases? DCDuring TALK 17:29, 24 July 2008 (UTC)

-collar would be a special exception, since most of the -collar compounds have little do with the word collar (-collar compounds generally refer to job occupations).--TBC 17:20, 24 July 2008 (UTC)

It is etymologically derived from collar, The tenuous connection with the meaning of the non-combining form may not be atypical. For example, we did not have a sense of rich that corresponded well to nickel-rich. DCDuring TALK 17:29, 24 July 2008 (UTC)

"Abundant, plentiful, or ample" [28] is a sense for rich, and a commonly used one at that (as in, "a mountain rich in resources", "a vegetable rich with nutrients"). I'm surprised that our Wiktionary entry doesn't have the sense; I'll add it.--TBC 17:37, 24 July 2008 (UTC)

We don't have any policy, official or unofficial, that I can recall, nor any discussion I can remember that addressed this issue. My own feeling is that the construction Noun-Adjective or Noun-Noun is rather standard in English. This isn't a case of a suffix, so if we allow for entries for the components, they should not include hyphens. We've tended to include the hyphen only for affixes like -wise, where the separate word (e.g. (deprecated template usage) wise) does not carry the meaning that is found in the suffix (deprecated template usage) -wise. The two have separate meanings, and in this case it results from separate etymologies. In the case of dog-collar, however, the meaning of the second component is identical to the usual meaning of (deprecated template usage) collar, so (deprecated template usage) -collar would be a needless duplication. I think such terms are merely sum-of-parts and wouldn't bother including them, though admittedly the only "harm" in including them is that it might drain our limited resources from more useful additions, or set a precedent of including some kinds of strictly sum-of-parts entries. I have no strong feelings based on the conversation so far, but that's how I see the issue at this point. --EncycloPetey 18:38, 24 July 2008 (UTC)

For dog-collar, perhaps, but the meaning isn't obvious for components like "white-collar", "scarlet-collar", or "blue-collar."--TBC 06:30, 25 July 2008 (UTC)

I have assumed that solid-spelled compounds and prefix- and suffix-constructed words are to be presumed included no matter how SoP they may be. I have assumed that spelled-solid forms are main entries and that the other form (hyphenated or multi-word) are alternative spellings, except in the case of adjectives (hyphenated is main for adjective if most use is attributive) or contrary evidence. I will assume that we should apply the idiomaticity criteria to hyphenated compounds as well as to multi-word compounds. I intend to add non-gloss definitions or usage notes for words that frequently are commonly used in compounds when I notice and believe that there is potential for confusion. I'm not sure that these are really well-formed enough to constitute guidelines, but I will attempt to see how often they fail by noting the cases that don't fit. I'd welcome any thoughts on the adequacy of these simple guidelines and any cases for which the guidelines would not provide the right answer by more fundamental criteria. DCDuring TALK 20:31, 24 July 2008 (UTC)

If it meets the core CFI (three independant, non mention, cites over three years) we should include it. We include all plural forms even though they are trivially derived, there is no reason not to include these things - though I wouldn't ascribe a high priority to creating them. Conrad.Irwin 20:43, 24 July 2008 (UTC)

A comparison to plurals is a strawman argument. Plurals often have pronunciation changes, some are irregular, and even the regular ones require knowledge of the rules for -s vs. -es. They are inflectional forms of the same word, not a combination of words into a new compound. A better analogy would be the discussions we had that disfavored inclusion of possessives as separate entries, as we are really talking about enclitics. --EncycloPetey 21:01, 24 July 2008 (UTC)

I favor including many things, but compounds introduce the possibility of combinatorial explosion. In English plurals mean a doubling (roughly) and inflected forms of verbs less than a quadrupling of entry count. With plurals there are sometimes questions as to how to form the plural or whether a plural exists. Also sometimes users (me!) type in the plural in the search box. With the verb form, the participles are often adjectives or nouns as well. With some combining forms, the possibilities are hundreds or even thousands of words per productive compoundable word. (If nickel-rich is an entry, why not plutonium-rich, phosphate-rich, and PETE-rich?) I'd just like to know what are good reasons to include compounds but stop short of introducing many thousands of entries that make little contribution. I would not want to have such items clogging up RfV either.

If we are to exclude some of them, it would be vastly more efficient to do it by rule than based on an attestation test. We could simply put the burden on proof on those adding or defending such entries by requiring that they be attested within 3 (or 1, 6, or 12) months of entry, for example. If we are to include them, it would be more efficient to add them by bot and add them as derived terms to their stem words by bot as well.

I note that the on-line dictionaries we compete with rarely include compounds, but seem to have many redirects and suggested alternatives instead. DCDuring TALK 21:56, 24 July 2008 (UTC)

Generally, I believe that compounds should only be kept if the combination forms a new meaning, as with the -collar examples. Since the meaning of -rich (as part of a compound) is identical with a sense of rich (by itself), then I agree with you that an entry like nickel-rich should not merit inclusion, regardless of the cites.--TBC 06:36, 25 July 2008 (UTC)

First of all, I think I might have been responsible for the entry nickel-rich, or at least for a link to it from "nickel" when I researched terms derived from the names of the chemical elements. I would imagine "nickel-rich" is here because it is listed in the OED (I don't have it to hand to check right now).

The OED lists compounds either in bold or in italics. The ones in bold are usually idiomatic, and so likely to pass our CFI, while those in italics are usually non-idiomatic uses, listed only as examples of the sorts of compounds that the word can form. If "nickel-rich" is bold in the OED, then it can be argued it passes our CFI. If it is in italics, then it can probably be deleted. But we need to check first and research the existence of the term before any action is taken.

And naturally the slippery slope argument does not apply - the inclusion of a term does not mean that we automatically allow or require all similarly formed terms to be included. Terms are individually judged by our CFI, as indeed is happening here with this particular term. — Paul G 15:00, 19 August 2008 (UTC)

It's not in the OED Online; I mean, it appears in three quotations there (in the entries for (deprecated template usage) nimite, (deprecated template usage) vaesite, (deprecated template usage) westerveldite), but the dictionary itself never mentions it, if you see what I mean. I added the entry, because Zigzig20s requested it at (what is now) Wiktionary:Requested entries:English, with the thought that it might mean “pseudo-rich”; see Wiktionary:Requested entries:English?diff=1908711. On failing to find evidence for that sense, I added the straightforward sense. Nowise do I object to deleting it if others feel that appropriate; I really wasn't sure if it warranted an entry, but figured it was easier to just create it than to start a discussion at the requests page. —Ruakh_TALK 15:31, 19 August 2008 (UTC)

Thanks for checking, Ruakh. I checked too (in the paper version, 1989 - clearly you have access to the online version to have been able to find it in the full body of the text), and it's not there under "nickel". I could not have entered anyway it as it is not among the compounds listed at nickel. Was there any particular reason you created this entry? The prevailing view above seems to be that it doesn't pass CFI. — Paul G 09:44, 20 August 2008 (UTC)

Re: "Was there any particular reason you created this entry?": As I said above, it was listed at WT:RAE with a question if it meant “pseudo-rich”, so I added the entry with the correct definition. I really wasn't sure if it warranted an entry, but figured it was easiest to just create it and let DCDuring worry about whether I should have. ;-)   —Ruakh_TALK 23:30, 20 August 2008 (UTC)

As I understand this discussion, it is agreed that a hyphenated compound like nickel-rich ought to be more than SoP to be included. That is a "part" in "Sum of Parts" could be linked by hyphen to the other parts of the headword. It then has to meet CFI. In the case of nickel-rich, it would have been an entry showing both the SoP and non-SoP meaning. As it stands it ought to be deleted and care should be taken to avoid SoP compounds. Unless we want to stay ahead of fr.wikt.

A further issue is whether we should allow SoP alternative spellings for solid-spelled compounds as entries. They clearly are worth having as alternative spellings on the solid-spelled page, in part to help users find the right page. DCDuring TALK 00:07, 21 August 2008 (UTC)

Interwiki on Watchlist

Hey! I would very much fancy having interwikis on Special:Watchlist. Is there any negative consequences with giving Special:Watchlist interwikis to de:Special:Watchlist, sv:Special:Watchlist, fr:Special:Watchlist and so on? It is very easily done; MediaWiki:Watchlist-details. --Eivind (t) 12:17, 26 July 2008 (UTC)

Done. This is the kind of thing that should be in WT:GP in future. Conrad.Irwin 00:42, 31 July 2008 (UTC)

Ok, thanks! --Eivind (t) 19:37, 1 August 2008 (UTC)

Translation-table redlinks.

As of this writing:

At [[fuzzy logic#Translations]], we have a bunch of redlinks, because we don't yet have entries for various foreign-language translations of the phrase, and the target-language Wiktionaries don't have these entries, either. What's unfortunate is, we do have entries for a lot of the components of these redlinked translations — but they're not linked from that table. (For Finnish they were, but I “fixed” that when I changed it to use {{t-}}.) Likewise, while none of the target-language Wiktionaries has these entries, the corresponding Wikipedias all do — but except where we or the target-language Wiktionary has an entry, there's no link path to the Wikipedia article.

This seems unfortunate to me: we have lots of relevant information, but because the translations tables are only set up to include two specific links (our entry for the full translation, and the target-language Wiktionary's), none of that information is linked.

If y'all agree that that's a problem, do you have any thoughts for how to improve the situation?

—Ruakh_TALK 20:49, 26 July 2008 (UTC)

Interesting- maybe some kind of javascript magic could be set up to click on something in the output of the t template and split the word into its component parts. The parts would have to be set using "sg" or something. Nadando 20:53, 26 July 2008 (UTC)

That may not be desirable. Even when the components exist on the target wiki, we can't guarantee they'll be for the right language. You might end up following a link to the Catalan wiktionary and find a definition only of a Spanish word. --EncycloPetey 23:14, 26 July 2008 (UTC)

Broken RSS feed

Word of the day RSS feed is broken.

Link goes to this page: http://toolserver.org/~cmackenzie/wotd-rss.php

which currently displays an error message.

Thanks for letting us know, I'll see if I can get in touch with Connel, either directly or indirectly. Conrad.Irwin 22:51, 26 July 2008 (UTC)

This got "fixed", either by Connel or automagically. Conrad.Irwin 23:51, 29 July 2008 (UTC)

Retired langauge codes

SIL's list of retired ISO 639-3 codes show 5 that we're currently using. Most only have one usage here.

1. aiz (Aari) was split into Aari [aiw] (new identifier) and Gayil [gyl]

   Usage: Just one entry, isimanna

2. azr (Adzera) was split into three languages: Adzera [adz] (new identifier), Sukurum [zsu] and Sarasira [zsa]

   One entry , hai whose author User:Ptcamn was just notified.

3. bsd (Sarawak Bisaya) was merged into [bsb] Brunei Bisaya

   Just a red-link translation at iron

4. cit was split into Rohingya [rhg], and Chittagonian (new identifier [ctg])

   Many entries, but this was talked about in April

5. prv (Provençal) was merge into [oci] Occitan.

   One entry, Catalan

I reckon we'd want to address these in some way or another. #3 and #5 on the surface present little problem, #4 seems to be in-progress, #2 can hopefully by handled by Ptcamn, but #1 appears baffling as the author of the lone entry hasn't contributed in years. What are other people's thoughts? --Bequw → ¢ • τ 09:29, 27 July 2008 (UTC)

I've noted #5 to Medellia. -Atelaes λάλει ἐμοί 20:32, 1 August 2008 (UTC)

Section links for inflected form templates

My interpretation of a feedback item (See Wiktionary:Feedback#what is anchored is that the user couldn't find what he wanted after click through from anchored to anchor. It would increase the chances that a user could find what they want if the link directly to anchor#Verb. If this is desirable, it would have application to many inflected form templates that link to other pages. Is there a way this could be included in the templates reliably? DCDuring TALK 13:11, 27 July 2008 (UTC)

Unfortunately not, because we ask editors to manually linkify the lemma (as a simple way of ensuring that the entry shows up in our statistics). This means that any target-linking would have to be done manually — or botically. —Ruakh_TALK 23:10, 27 July 2008 (UTC)

This is one excellent reason why we should not be doing that, page count is not that important (and should form of entries even count towards it...). All the templates should be updated to include the relevant section link, and people should please just use the default. It would be quite easy to bot-fix all of the current entries if we wanted to. (If you want your page to count, add <!-- [[Making this page count]] --> to the entry... ) — This unsigned comment was added by Conrad.Irwin (talk • contribs) at 00:46, 31 July 2008 (UTC).

Indeed, the whole thing is silly, because the one kind of entry that we don't want to count — the much-reviled misspelling entry — still counts, because of the prolific editors at vi.wikt who seem to have every entry we do (or maybe they watch our create-log and keep pace?). —Ruakh_TALK 03:05, 31 July 2008 (UTC)

Agreed. I don't think I could possibly care less about entry count. I would completely support setting our egos aside in favour of better usability for the users. -Atelaes λάλει ἐμοί 07:11, 31 July 2008 (UTC)

Entry count seems to be a statistic of diminishing relevance as Wiktionary develops. Number of "good entries", of lemmas and of PoSs are arguably more relevant than raw entries. Something more reflective of our success in helping users would be nice. Number of visits, number of registered users, number of contributors, edits by type of contributors are all useful indicators. Google ranking is also useful. Whatever we measure is likely to influence us. It helps us set targets that can be converted into motivating objectives for some of the scut work. DCDuring TALK 07:59, 31 July 2008 (UTC)

Proposal to disable hotlinking

For all those interested, I have made a proposal to disable hotlinking on all Wikimedia projects. Please join the discussion at Meta. JohnnyMrNinja 20:52, 27 July 2008 (UTC)

As we have a whopping 9 images uploaded here, this won't affect us. Thanks for the notice, it's certainly an interesting question. Conrad.Irwin 20:56, 27 July 2008 (UTC)

Header font sizes

Do we really need for the header fonts to be as large as they are? Headers on the first screen that have fonts larger than our standard text are using up Precious Above-the-fold Real Estate™ and are using up screen space wherever they appear. The headers are: 1., large, 2., bold, 3., prominently located, 4., set off by white space, distinguished by the numbering. L2 headers are also set off by the lines. Reducing the amount of vertical space would be a way of increasing the information conveyed by a single screen without materially adversely effecting users, I think. DCDuring TALK 14:59, 29 July 2008 (UTC)

Yes, the defaults are appropriate. Some recent feedback has indicated that we need all the visual help for our users as they can get, since it is already confusing on large pages. Reducing the font sixe for top headers or eliminating the horizontal rules between L2 headers would only increase the difficulties. --EncycloPetey 16:34, 29 July 2008 (UTC)

Don't be so sure. Are you familiar with the concept of "banner blindness"? A reader who's looking for content will focus on contentful parts of a page — paragraphs and lists, I think — rather than on things like headings. (For an example, see the images at http://www.useit.com/alertbox/banner-blindness.html, showing what people looked at on content pages, and notice that the headers were basically ignored — even headlines on news articles!) —Ruakh_TALK 17:01, 29 July 2008 (UTC)

As the heading clearly indicated (!!!), I was only questioning the height of headers, not the language section dividers, which have virtually no effect on initial screen information content anyway. This is going to be a trade-off between the possible help in orienting a naive first-time user and the need to get as much information on a screen for all users. I would guess that the portion of users naive in this way, who don't have a reasonable ability to infer structure from layout, is declining. DCDuring TALK 17:58, 29 July 2008 (UTC)

What if we indented the section headers by level? It would make the layout a lot easier to parse and quicker to read. Nadando 23:42, 29 July 2008 (UTC)

That's a good idea, particularly for pages with lots of etymology sections, indenting the 3/4/5 level headings would make it much clearer to the eye as to which sections belong where. It wouldn't make any difference to machine parsing as it would be done with site css. I would also be in favour of reducing heading font size or 1/2/3 slightly, maybe leaving 3/4/5 the same size but indented differently? Conrad.Irwin 23:50, 29 July 2008 (UTC)

Interesting. Might be worthwhile and almost free. How would the text under each level of header be indented? I would guess that we would gain about two lines worth of space by font-size and (perhaps) white-space reduction. I would hate to "give back" some of the space gained by having one or two bits of text take an extra line because they were indented. I don't have an intuitive feel for how often that might happen (between different entry-layout characteristics, display device characteristics, etc.). DCDuring TALK 23:54, 29 July 2008 (UTC)

I hadn't thought about indenting the text itself, only the headings. Conrad.Irwin 00:12, 30 July 2008 (UTC)

I guess that is the way text is handled under headings in normal text, so it shouldn't look bad. An additional possibility is the use of some kind of extra bold wide font for L2 and PoS headers. As our layout works now, we can't readily differentiate between PoS headers (which might be either L3 or L4 (L5 under exceptional circumstances?) and other headers at those levels. If we could differentiate, it would be worthwhile to place more visual emphasis (like extra bold or size) on PoS rather than Alternative Forms and Spellings, Pronunciation, and Etymology, or the low-down L3 headings for Anagrams and References. DCDuring TALK 00:49, 30 July 2008 (UTC)

• There is definitely rom for improvement with it all. I think indentation of some kind is a good idea for L3s (perhaps some sort of bulleting?), and for L2s I'd quite like to see some kind of shading or colouring behind the text. Ƿidsiþ 16:02, 31 July 2008 (UTC)

Category:Wiktionary pages that don't exist

The name's misleading (and sort of paradoxical, reminiscent of these scenarios) especially for describing pages that "don't meet our criteria for inclusion but that are still useful to document." Also, wouldn't it be a lot more useful if these pages were included as soft redirects (especially for a page like Isaac Newton)?--TBC 00:30, 30 July 2008 (UTC)

I thought they were soft redirects (""The English Wikipedia has information at Isaac Newton""), the template was deliberately designed to resemble MediaWiki:Noarticletext as the words it is used on do not meet our Criteria for inclusion and thus we don't have articles for them. I agree with you that the category name is strange, do you have a better suggestion? I chose it to imply that the words it contains are not real words. Conrad.Irwin 00:39, 30 July 2008 (UTC)

Whoops, overlooked that part. :) Anyhow, how about Category:Tentatively-included Wiktionary pages?--TBC 00:48, 30 July 2008 (UTC)

Category:Soft redirects to other Wikimedia projects would seem like the most self-explanatory naming. If that is too unwieldy the "Wikimedia" could be dropped. Thryduulf 14:01, 31 July 2008 (UTC)

I think strongly that this category should have HIDDENCAT; the pages in this category should have no categories (shown), no interwiki links, etc. They are not entries.—msh210℠ 20:43, 20 August 2008 (UTC)

gotterdammerung

my grandmother told me that gotterdammerung meant "God rings his bells" describing the ultimate end of the world through a day of reckoning. Can anyone expand on this? thank you.

She was pulling your leg. Götterdämmerung means literally "gods’ twilight" (i.e., "gods’ dimming"). The "-rung" on the end is like English -ing (-ung): Götter + Dämmer + -ung. Thus it means "twilight of the gods". —Stephen 04:56, 30 July 2008 (UTC)

However, it is not "rung" which is similar to the eng. ing, but "-ung", in Götterdämmerung the second R appertains to the root Dämmer which means crepuscule as zou have finally written. Bogorm 13:24, 7 August 2008 (UTC)

Alerts for Keywords and New Members of Categories

To help distribute some of the patrolling workload and for other purposes, I was wondering whether we might have the ability to periodically alert registered users to the use of specific text anywhere in Wiktionary, everywhere except user pages, or in selected spaces. This would not have to be as soon as entered; it could be periodic, but probably at least weekly to be useful. If we had reliable weekly dumps it could be done off line. Would Google Alerts work for that purpose?

Also, is it possible to be alerted to new entries in a category? Once an item is meaningfully categorized it can pay for someone with specific knowledge in the category (eg, language, PoS, context) to review the entry. To have category membership alerts or a "recently added" view of a category can also be helpful in staying on top of edits. It would probably lead to improvement in the quality of some of the category memberships. Could Google Alerts work for this? DCDuring TALK 15:09, 30 July 2008 (UTC)

By language would also be useful. Personally, I'd love to be notified of (alerted to?) unpatrolled Hebrew entries and translations, and I'd love it if editors working in the less commonly known languages were notified similarly. (One thing I hate about patrolling is all the edits that I can't tell if they're good because I can't even read the script they're in, but if I just leave them be, they'll probably never get seen by anyone more knowledgeable.) —Ruakh_TALK 23:17, 30 July 2008 (UTC)

The language categories aren't good enough, are they? You would need L2 header alerts? I wonder if it is possible to get alerts for the use of certain character sets. Any component of this that was easy to do would be a worthwhile experiment unless there is some show-stopping consideration. DCDuring TALK 23:53, 30 July 2008 (UTC)

We can do basic categories with <dynamicpagelist>, - other stuff will require some hackering. An example of how to use this (very underused) extension follows, you can add as any categories as you are interested in. This won't, sadly, filter on patrolled/not-patrolled, but it's a start. Conrad.Irwin 00:39, 31 July 2008 (UTC)

<DynamicPageList> category=Hebrew nouns category=Hebrew verbs category=Hebrew proper nouns category=Hebrew numbers addfirstcategorydate = true count=10 </DynamicPageList>

19 September 2024: זב 15 August 2024: לבן 15 August 2024: לובן 7 August 2024: תמך 28 June 2024: אשם 5 June 2024: ביזה 5 June 2024: לוה 28 May 2024: לקט 21 May 2024: טבל 19 May 2024: חס

I thought dynamic page lists only worked on Wikinews. I'll have to make a ticker for myself. Circeus 23:21, 5 August 2008 (UTC)

Signature templates

There are at least two users here who are currently transcluding signature templates into talk pages, rather than using the raw signature option in special:preferences to create a custom signature.

On the English Wikipedia, there is a specific policy gainst this - w:Wikipedia:Signatures#Transclusion_of_templates. I quote from there the reasons for it:

1. Signature templates are vandalism targets, and will be forever, even if the user leaves the project.
2. Certain automated scripts (bots) are used to automatically archive particularly active talk pages. These bots read the source of the talk page, but don't transclude templates, and so don't recognize the template as a signature.
3. Signature templates are a small but unnecessary drain on the servers. Transcluded signatures require extra processing--whenever you change your signature source, all talk pages you have posted on must be re-cached.
4. Substitution of templates in signatures is also not allowed, as any such template either will violate the reasonable length restriction of 255 characters or will be redundant to using the same content as a raw signature.
5. Simple text signatures, which are stored along with the page content and use no more resources than the comments themselves, avoid these problems.

[bullet points changed to numbers (and the "Simple text signatures" line moved into the list) for ease of reference and links pointed to Wikipedia]

Points 1, 3 and 5 are identical to all projects. Regarding points 2 and 4, I do not know whether bots here work the same as on Wikipedia, and our draft Wiktionary:Usernames and user pages policy does not include any length restrictions on signatures, but I would like to add it here (see below).

For the reasons above I would like to introduce the above as a policy here. Additionally, I would also like to introduce the 255 character length restriction on signatures for the reasons below (copied verbatim (excluding formatting) from w:Wikipedia:Signatures#Length)

Keep signatures short, both in display and markup.

Extremely long signatures with a lot of HTML/wiki markup make page editing and discussion more difficult for the following reasons:

• signatures that take up more than two or three lines in the edit window clutter the page and make it harder to distinguish posts from signatures,
• long signatures give undue prominence to a given user's contribution,
• signatures which have long HTML/wiki markup and contain no spaces cause other editors' edit boxes to show unnecessary horizontal scrollbars (such signatures may have spaces added to them by any editor),
• signatures that occupy more space than necessary in the edit box displace meaningful comments, thus forcing the editor to scroll when writing his reply, and
• the presence of such long signatures in the discussion also disrupts the reading of comments when an editor is formulating his reply

The software will automatically truncate both plain and raw signatures to 255 characters (characters used for HTML/wiki markup are included!).

Thryduulf 15:24, 31 July 2008 (UTC)

Your points about not having long messy code in pages, making it difficult to edit, and difficult for 'bots to recognize signatures argue quite strongly for using templates. Indeed we might very well require templates for messy signature code. I, for one, am much happier with {{User:Nwspel/sig code}} in a page than the contents thereof. (And the page cache/whatever overhead is not a very serious problem.) Robert Ullmann 15:39, 31 July 2008 (UTC)

Note, this change would also affect User:TheDaveRoss. --nwspel ^tork _kontribz 15:50, 31 July 2008 (UTC)

Why not disallow all messy signature codes and be done with it? Long sig codes tend to produce visually distracting sigs, after all.--TBC 15:51, 31 July 2008 (UTC)

I'm with TBC on this one. If the code for a signature is not able to fit in the 255 character limit for raw signatures in special:preferences then it should not be allowed. Thryduulf 15:58, 31 July 2008 (UTC)

Well if anything like what you're saying is to be enforced, it needs to become policy on Wiktionary first. Wiktionary isn't Wikipedia, and rules on Wikipedia aren't automatically those on Wiktionary unless brought over. --nwspel ^tork _kontribz 16:08, 31 July 2008 (UTC)

Which is exactly why I am seeing if there is support to make it policy here. Thryduulf 16:33, 31 July 2008 (UTC)

I strongly support the cap on signature length, and the ban on templates which evade that cap. We are here to write a dictionary, and whatever dictionary-writing purpose is served by having a distinctive signature can be fully served by the wide variety of customizations that fit easily within the 255 character limit. bd2412 T 16:51, 31 July 2008 (UTC)

I also support a cap and disallowing of template signatures. Users who make use of templates must subst them. Otherwise, such a user who edits their signature produces an edit change for the server in every single page they've ever signed with a template. --EncycloPetey 17:50, 31 July 2008 (UTC)

"such a user who edits their signature produces an edit change for the server in every single page they've ever signed with a template." - Incorrect. If I were to update my signature right now, there would be no new "edit" to the pages I have left my signature on. --nwspel ^tork _kontribz 18:04, 31 July 2008 (UTC)

Not as such, but the job queue has to go through and re parse all the pages as though an edit had been made. I am personally for stopping this practice, not because of the server (which can easily cope), but because they are just a way to build more garish signatures. While there's nothing too wrong with such signatures, I find them unaesthetic to the point of irritating (particularly when they mess up the line-heights). I know this is an argument against garish signatures, not template signatures, but I feel that the two are reasonably synonymous. What's wrong with good ol' 80% sans-serif anyway... Conrad.Irwin 18:54, 31 July 2008 (UTC)

Template:ud Surely your point is with Signatures interfering with line-heights? You can have a signature that interferes with them, but it isn't neccessarily a template sig. 78.149.215.98 19:10, 31 July 2008 (UTC) (Sorry I got logged out. That IP is me)

I agree, disallow template signatures, garish signatures, and signatures with too many characters. —Stephen 19:13, 31 July 2008 (UTC)

My reasons for using a template are numerous, but the major one was that it allows styling to be hidden, without imposing on future editors. It isn't a vandalism target if it is protected, and even if it isn't protected, ALL templates are vandalism targets, using one more doesn't change that. As for the "resource" drain...that isn't a concern. Wikipedia has the policy, we shouldn't. - [The]DaveRoss 19:29, 31 July 2008 (UTC)

But why do you need such a complex signature? How does that advance the creation of a dictionary? bd2412 T 20:06, 31 July 2008 (UTC)

How does a simple signature advance the creation of a dictionary? My view is essentially that a user's private signature is their own business, and although I also dislike these long colourful graphical ones, I don't see that they impede what we're trying to do. Ƿidsiþ 20:09, 31 July 2008 (UTC)

A simple signature advances the creation of a dictionary by keeping the focus on the creation of the dictionary! bd2412 T 23:01, 31 July 2008 (UTC)

The point regarding vandalism is that yes, all templates are vandalism targets but edits to ones that are actively used will be noticed. Suppose that a user who signed talk page discussions with a template leaves the project for whatever reason, most/all the talk pages they commented on will get archived sooner or later. Suppose then two years later a vandal edits that users' template and replaces it (or insterts into it) libellous information, or changes the name from "User:Example" to "User:I dislike George Bush".

Every talk page they have ever signed with that template gets updated (no new edit, but the contents of the template is changed to match the current content on the template page) with the libellous information or incorrect attribution for the GFDL. It takes someone to either spot the edit to the users' template (and how many people watch other users' signature templates?) or someone looking at the old discussion who notices the incorrect attribution/libelous material (unless they were active at the time of, and actively remember, the discussion, the first of these is very very unlikely and the second easily overlooked) before the vandalism is spotted. Why leave yourself open to this? BD2412 shows that you can create a distinctive signature without resorting to a template. Thryduulf 22:44, 31 July 2008 (UTC)

It is not hard to have a signature protected. And to be honest, due to the way in which templated signatures work, it requires two templates templated from eachother - a vandal is not likely going to understand this, for even if they reach the first one, it won't make a difference, since changing that first one won't change the signature on every page edited by the user account so far. The only way that can happen is if the first template stays the same, and the second template is changed, and to be honest, a vandal is hardly going to figure out that he/she must do that - let's be realistic.

Despite my opposition to the ban on template sigs, I do actually think I will support some limitation on the height of the sig, so that it does not interfere with the lines - note, this will require me myself to change my sig anyway. --nwspel ^tork _kontribz 08:07, 1 August 2008 (UTC)

Agree that it would be good to implement the limits suggested by Thryduulf. With all due respect for the (in many cases excellent) contributions of folks with exotic sigs, please express yourself through what you say, not how many colors or fonts you can put in your signature. JesseW 19:21, 1 August 2008 (UTC)

If we choose to allow signature templates, we can solve most of the problems with a few simple rules:

• requiring their owners to fully protect them (or request their full protection, then wait till the request is honored) before beginning to use them.
• forbidding their owners, and everyone else, from modifying them. (Signature templates shouldn't be a way to modify past signatures; an editor who wishes to change his/her signature should create a new template.)
• requiring template signatures to follow a certain naming convention; for example, requiring that signatures of user Foo have names matching ^User:Foo/sig\b. (In the past we've already enforced something like this, e.g. by deleting [[User:Nwspel sig]] in favor of [[User:Nwspel/sig]].)

However, I have to admit that I'm no fan of garish signatures or signature templates, and would happily vote to forbid both altogether.
—Ruakh_TALK 17:10, 5 August 2008 (UTC)

I've now created Wiktionary:Signatures (and new shortcut WT:SIG), based mainly on this page, and labelled it as draft policy. Please comment as you see fit. I will start a vote to make it official in a few days, probably Thursday night UK time, but this is obviously dependent on what discussion it generates, etc. If someone else wants to write the vote proposal then please do, as I'm not brilliant at such things. Thryduulf 13:40, 10 August 2008 (UTC)

Deleting old anonymous talk pages

I seem to remember that it is our practice to delete old talk pages from IP addresses after a decent interval. My understanding is that this is because many IP addresses are shared by multiple users, and old messages and warnings would be confusing for them. Have we ever decided what such a "decent interval" should be, and has anyone ever been tasked with carrying out such deletions? SemperBlotto 21:22, 31 July 2008 (UTC)

Could it be done by bot? --nwspel ^tork _kontribz 21:24, 31 July 2008 (UTC)

Yes, but deletions can only be done by sysops. So the bot would have to have sysop privileges - something that none of our bots have, and I'm not at all sure that such a thing would be approved. SemperBlotto 21:33, 31 July 2008 (UTC)

I have no problem with a sysop bot, we already have one run by Ullmann (deleting redirects). I think that deleting the pages is a nice thing to do, but not overly important. If someone with lots of free time wants to write a bot we should run it - say 6 months as old, and it could also get rid of any remaining permanently blocked userpages (which we are also supposed to delete?), but otherwise it sounds like the perfect aggression relieving task for when one is unable to be productive. Conrad.Irwin 21:41, 31 July 2008 (UTC)

Is there a list of these pages to be deleted? If the active sysops here take a chunk at a time, we can probably knock out a good number in short order. Unless the parameters are tightly controlled, I'm not sure I'd like to see a bot doing the job. bd2412 T 22:58, 31 July 2008 (UTC)

But if done manually, it would require a constant check of every talk page to see how old it is, and thus to decide whether its time to delete or not. A bot could check that by itself. --nwspel ^tork _kontribz 10:49, 1 August 2008 (UTC)

See User_talk:Conrad.Irwin/oldanonpages for all anonymous talk pages last edited before 2008. There are a lot to go through by hand, but it might be worth doing a random spot check to see if anything has got through the net before we release a bot onto it (If we do so). Conrad.Irwin 11:05, 1 August 2008 (UTC)

Can you also provide a handy listing of the number of contributions by that anon and most recent contribution date? Sorry to be a pain, but those are the things I need to see. bd2412 T 01:38, 5 August 2008 (UTC)

Perhaps we can get a flood flag. I've raised the issue in the GP.—msh210℠ 20:57, 20 August 2008 (UTC)

Referencing Wiktionary in Wikisaurus

Thanks to the good people in Signatures (above) I've figured out how to reference specific text on other pages. This would come in really useful for wikisaurus and for the whole project to keep definitions in line with wiktionary and to ease the burden of the amount of data that has to be crowded on one wiktionary page.

What I'd like to propose is that I have a mouse-over, tool-tip reference back to the original definition for that specific synonym. I wouldn't have to worry about validity of the word because it will have been validated in the original wiktionary page. What do you think? (I would be happy to set the anchors as I need them) Amina (sack36) 23:00, 31 July 2008 (UTC)