Wiktionary:Grease pit/Form of

form of

{{plural of}} and {{alternative spelling of}} aren't giving me the default bolding, although {{form of}} with the same code is. What's going on? ∂ανίΠα 19:14, 17 June 2006 (UTC)

Because "form of" has hard-coded bolding, which it shouldn't. — Vildricianus 19:17, 17 June 2006 (UTC)

Ah, thanks for fixing it. So bold isn't the default though? ∂ανίΠα 19:33, 17 June 2006 (UTC)

No. Should it? Paul raised the issue earlier today as well. Perhaps there should be some BP discussion on it (again!!), but don't count on me for starting it :-). — Vildricianus 19:42, 17 June 2006 (UTC)

What is the default supposed to be? Looks like it is plain text: "Plural of foot". Now, my objection to this is twofold:

1. The grammatical term should be wikified, IMO, as we often do for "comparative". Of course, almost everyone knows what a plural is, but many people don't know what a comparative is. Although someone could guess what a comparative is by looking at the definition, such as "thicker: comparative of thick", it is possible that someone with very little knowledge of English grammar might guess wrongly: it's just (note, just) possible that someone reading "flatter: comparative of flat" might guess that this means the verb "to flatter" and that "better: comparative of good" means someone who bets, despite the part of speech headers. Unlikely, but possible. So I think that "comparative", etc, should be wikified in these templates to aid the reader.
2. I'm not too bothered whether or not we do that, but I think it would be helpful to some readers. However, one thing that I do feel strongly that we must do, especially if we decide to wikify the grammatical term, is to show that the term in the cross-referred is being referred to and not being used.

Here's an example of the second issue (a slightly contrived one, but it illustrates my point very well):

"older: comparative of old"

As there is no typographical distinction between "comparative" and "old", this reads in the same way as "comparative of old", as it would appear without the wikification (for example, in a print dictionary). Now, written in that way, that means "a former comparative; a comparative used in olden times", which is not at all the intended meaning.

The adjective being compared must therefore be typeset in some way that indicates it means we are referring to it rather than using it. This can be done with quotation marks or, better still, because we have it available, with typography. This could be emboldening (compare Chambers), italicisation or small capitals (compare the OED and some US dictionaries). For myself, I think emboldening is ideal as this suggests a headword, since headwords appear in bold in Wiktionary.

So my personal choice would be for bold text to be the default, as I think this works best. It is however my personal choice, so I'm not going to force it on others.

So the formats can include "comparative of old", "comparative of old", with variations in capitalisation, etc, but what I feel we must not allow as the default or even as an option is unadorned text, for the reasons I give above. No print dictionary would give "comparative of old", "plural of foot", etc, and neither should we. — Paul G 07:01, 18 June 2006 (UTC)

No, a print dictionary would certainly use ‘Comparative of old.’ That's because wikification is a typographic distinction, and the fewer used, the better. However, you're right that if we wikify the grammatical term, then the referrant needs an extra level of distinction (I agree with you that bold is the best choice). But I personally think it's cleaner and neater without wikifying the ‘comparative’ bit. Widsith 06:42, 19 June 2006 (UTC)

If I am not mistaken, the practice of using links as a form of "highlighting" words has been discouraged in HTML since its early days and is furthermore specifically discouraged on Wikis. — Hippietrail 02:56, 20 June 2006 (UTC)

Sounds like sound policy to me, considering how easily things are linked (e.g. future wikification) or unlinked (e.g. in printing). Davilla 04:10, 20 June 2006 (UTC)

Compare [1]. With the new possibilities, there's no need to repeat that, fortunately. Bold is a reasonable idea, which I've stated before. Not allowing unadorned text doesn't sound very wiki, and is technically not possible (but who cares anyway what people prefer through their css files?). — Vildricianus 14:12, 18 June 2006 (UTC)

Do we need to have another vote for default bolding of sub-class mention? The primary objections were for non-Latin scripts. Is there any way to determine the script and only bold the Latin ones? Davilla 04:10, 20 June 2006 (UTC)

That last one sounds like a bad idea to me. But if it's the only possible solution, fine then. For that part, either italics or bold is bad in non-Latin. Underlined doesn't work either of course. Strike-through perhaps? :-) — Vildricianus 10:57, 20 June 2006 (UTC)

Why is there a need to do anything with non-Latin scripts? They stand out enough as it is on their own. Davilla 16:12, 20 June 2006 (UTC)

What? That I don't get at all. — Vildricianus 16:18, 20 June 2006 (UTC)

The use/mention distinction that Paul raised above. Introducing inconsistency merely because something is not in the set [a-z,A-Z] is a bad idea. When "Printer friendly link" is clicked, we do need to still distinguish between use and mention. Wikification of the term is ignored (entirely) on the "Printer friendly link." I'd prefer double quotes (ASCII, not Microsoft quotes) since we have syntactic problems with every other choice, except perhaps bold. I think the quotes convey the distinction much better than bolding them. --Connel MacKenzie T C 16:28, 20 June 2006 (UTC)

Who said anything about [a-z,A-Z]? There are many more Latin-based characters than that! The point that had been raised during the plural-of bot discussions was that Chinese, Japanese, Russian, Greek, Arabic, Hebrew, etc. did not consistently conform to being bolded or italicized, but that it didn't matter because the use-mention distinction was pretty clear in these cases. Now if you want to put quotations around Chinese characters because you're not sure where the distinction between use and mention might occur, then you can go ahead and find a CSS solution for that. My only point is to narrow the argument down to those scripts for which all of the proposed solutions could actually apply. Davilla 08:14, 21 June 2006 (UTC)

Sorry to say, but this sounds really like an ugly reasoning. The fact that something is in a non-Latin script doesn't mean we shouldn't distinguish use/mention anymore. I don't bother too much about the whole u/m thing, but one thing I'd like is to keep things consistent across languages. — Vildricianus 12:43, 21 June 2006 (UTC)

I disagree completely. The fact that a mentioned word in an English sentence is in a non-Latin script does mean we don't need to mark it explicitly as mentioned. The following examples illustrate:

• Mentioning a Latin-script word:
  • Blend of spice and ham.
  • Short form of foxterrier.
• Using a Latin-script word:
  • Blend of spice and ham.
  • Short form of foxterrier.
• Mentioning a non-Latin-script word:
  • Blend of 辣 and 火腿.
  • Short form of терьер.

Note that there is no need to italicize (or otherwise to make distinct) non-Latin-script words. Doing so actually obscures the mention because italics have undesirable affects on non-Latin fonts:

• Undesirable effects of italicizing non-Latin fonts:
  • Blend of 辣 and 火腿.
  • Short form of терьер.

The Chinese words are practically illegible and the Russian word looks completely different from what was intended. Use-mention distinction in English Wiktionary need only apply to Latin fonts and, if we use italics (the standard wiki way of mentioning words) we should do so only for Latin script words. Rod (A. Smith) 16:08, 21 June 2006 (UTC)

How about this:

• Accusative of HTML-код. Latin or Cyrillic?
• Genitive of paca. Latin or Cyrillic?

— Vildricianus 17:06, 21 June 2006 (UTC)

Good points. I hadn't even considered mentioning text of mixed scripts or the fact that Latin script text can appear to be Cyrillic, so I retract my "I disagree completely" statement. :-) Those problems combined with the illegibility of italic hanzi/kanji make italicization problematic. Should we always use quote marks? Rod (A. Smith) 19:25, 21 June 2006 (UTC)

That seems (to me) to be the only viable option. --Connel MacKenzie T C 19:32, 21 June 2006 (UTC)

No, italics can't go for said reasons. Bold is slightly better but still bad. Underlining is a non-starter. Small caps? Mmm, same problems. Well I can't think of anything else than quotation marks then, indeed. I certainly don't like them but it seems there's no other option. — Vildricianus 20:27, 21 June 2006 (UTC)

Present participle of

Connel brings up a good point below regarding {{present participle of}}. Perhaps it's time to define our long-range scheme for English and non-English "form of" templates. [The following is copied from User talk:Rodasmith:]

Is {{present participle of}} as intended? Not all p.p. are gerunds (although they can be, not all are.) Is there a separate template for normal p.p. without the gerund thing? --Connel MacKenzie T C 18:25, 20 June 2006 (UTC)

Hmm. I guess I didn't thoroughly think that template through. After creating the "form of" templates, I realized that there actually needs to be a distinction between the "form of" templates for English and for other languages, at least in the case of verbs. The problem you cite is but one of the issues. The other is that with English verbs, we want "to" to appear in the output, but obviously not with verbs of foreign languages. I'd be quite happy to see them changed, but I haven't thought of the best convention for the template names. Any thoughts? A suggestion to resolve the "pres part/gerund" issue is to have a "ing form of" (or some other short-named) template for "present participle and gerund of to..." since that's by far the most common in English. Then "present participle of" could be used for the exceptions where the verb has no gerund. That doesn't address the English ("to") vs. non-English issue, though. I welcome any improvements. Rod (A. Smith) 18:33, 20 June 2006 (UTC)

1. "to" should indeed remain part of the template, as we may want to remove and re-add it from time to time :-)
2. Templates for other languages are necessary, but don't need to affect the English ones. We'll need to use ISO codes then, e.g. {{fr-present participle of}} or whatever.
3. I'd recommend keeping all names obvious, as they are now (so nothing like {{pp}} or the like).
4. I'd also recommend to leave out the "verbal noun" thing or anything else, since these forms aren't likely to appear under a =Verb= heading. Or are they?

Cheers! (Feel free to copy this discussion over on the GP, it'll need more ideas there and I think there's already a thread on it). — Vildricianus 18:39, 20 June 2006 (UTC)

Extra note: ideally, we'd have an exhaustive list of "form of"-templates and corresponding definitions, which we should stick to, in order to remain consistent across all entries. — Vildricianus 18:53, 20 June 2006 (UTC)

This is the ugliest wording and it has to come to this basic point to get rid of it. In case any of you forgot, English verbs are defined from the base form, not the infinitive. I actually taught from a storybook last year in which all of the verbs were listed in the glossary under "to", can you imagine!? So if you wanted to look up "reach", you'd have to know it was a verb under the 't' section because you wouldn't find it under 'r'. There were other problems with the book by the way. And there are some problems with Wiktionary in this respect. Under reach I find "to reach", "reaches", "reaching", and "reached". Nowhere is the plural and first/second-person singular "reach" listed.

Back to the templates. The solution to this problem is very simple. Just get rid of "to" and the templates can be shared across languages. {{present participle of|word}} should be used whenever the page name is a present participle of word, period. Davilla 08:02, 21 June 2006 (UTC)

Is anyone here familiar with the "&uselang=" thing? I'm pretty sure that \Mike used to use it for MediaWiki:Nogomatch/sv by setting the default language in his user preferences? (BTW, that should move to MediaWiki:Noexactmatch/sv right?)

Would the better long-term plan be to have {{en-noun/sv}}, {{en-noun/zh}} etc.? Whenever the language-specific sub-page doesn't exist, the MW software is supposed to just give the English version of the page, right?

I won't suggest {{en-noun/en-uk}} for adding spurious "u"s, just yet. --Connel MacKenzie T C 19:43, 21 June 2006 (UTC)

I don't get it. Please clarify where you would expect to apply "en-noun/sv" and "en-noun/zh". Rod (A. Smith) 20:49, 21 June 2006 (UTC)