Wikipedia talk:Manual of Style/Archive 106

This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 100

←

Archive 104

→

Plurals

the paragraph on plurals recognizes the difference between British and North American use, but doesn't indicate which is used by Wikipedia. For example, the article on the band Genesis has both ('Genesis are' and 'Genesis has'). Is there one use required by Wikipedia? or does it depend on what the article is about?--Richardson mcphillips (talk) 19:10, 30 November 2008 (UTC)

AP Stylebook recommends that a singular collective noun should take a plural verb only for bands and teams. British English often takes plural verbs with collective nouns. I'd go with "are" for AmEng and BrEng, but I don't know how other countries handle this. - Dan Dank55 (send/receive) 21:57, 30 November 2008 (UTC)

P.S. I'd appreciate feedback on the special case of words which are being used as synonyms for "a lot of"; "a bunch of kids are going to the movies" sounds right to me, but "a bunch of flowers is..." also sounds right. - Dan Dank55 (send/receive) 21:57, 30 November 2008 (UTC)

See WP:ENGVAR. Septentrionalis PMAnderson 22:42, 30 November 2008 (UTC)

thanks all. pmanderson, I conclude that the paragraphs your link sends me to imply that an article on an English (UK) rock band should use the English (UK) style. --Richardson mcphillips (talk) 03:42, 1 December 2008 (UTC)

Yes, it probably should (and very likely does). Septentrionalis PMAnderson 17:35, 1 December 2008 (UTC)

The Collins COBUILD English Grammar (HarperCollins, 1990, reprinted 2000) says, at pp. 16–17, "When you use a collective noun, you can use either a singular verb or a plural verb after it. You choose a singular verb if you think of the group as a singular unit, and a plural verb if you think of the group as a number of individuals. ... The names of many organizations are collective nouns, and can be used with a singular or plural verb." The following examples were given:

"Our little group is complete again."
"A second group are those parents who feel that they were too harsh."

"The BBC is sending him to Tuscany for the summer."
"The BBC are planning to use the new satellite next month."

"The government has said it would wish to do this only if there was no alternative."
"The government have made up their minds that they're going to win."

Hope this is useful. — Cheers, JackLee ^–talk– 10:07, 3 December 2008 (UTC)

Glossary needed with articles with non-English word usage?

See Talk:Vithoba#Glossary. User:Alastair Haines] has suggested use of a glossary at the beginning (though i would prefer one at the end) or use of sidebars to express foreign term usage. I have never seen this on any article, but I like the idea. This issue has effects technical topics, which use have a heavy use of jargon. Currently we have used "unique genre of devotional lyric, the abhanga" a short summary like this one or brackets "Vitthala is composed of the words vittha (ignorance) and la (one who accepts), thus meaning 'one who accepts innocent people who are devoid of knowledge'". Is it enough? --Redtigerxyz ^Talk 14:05, 3 December 2008 (UTC)

I would not be in favour of a glossary anywhere in an article, as I think it would look clumsy. The current techniques for explaining technical terms mentioned in Redtigerxyz's message above are fine. If there is a Wikipedia article on a certain technical term (e.g., abhanga), that can be linked to. If not, it would be better to create a suitable entry in an appropriate Wiktionary and link to that. — Cheers, JackLee ^–talk– 14:41, 3 December 2008 (UTC)

See Portal:Contents/List_of_glossaries for examples of Wikipedia's glossaries, and Wikipedia:WikiProject Glossaries for more info. There are all sorts of ways to include short glossaries within articles (in the "notes" section, or in their own article-section, or as a "legend" at the end of a table, etc), as well as the many full glossary articles. -- Quiddity (talk) 00:14, 4 December 2008 (UTC)

Full glossary articles seem fine to me. Have you any specific examples of short glossaries used in "Notes" sections, etc? I'd like to see how they look. — Cheers, JackLee ^–talk– 03:57, 4 December 2008 (UTC)

As I look at this discussion, the primary message i get is - Add notes, links, if necessary. Brackets or short explanations in the text do work. Correct me, if I am wrong. --Redtigerxyz ^Talk 05:36, 4 December 2008 (UTC)

Talk:Vithoba#Glossary has examples. --Redtigerxyz ^Talk 05:37, 4 December 2008 (UTC)

I'm for notes and links where relevant – it's better than leaving people to jump from article to linked article and back for instance though the link can be followed anytime. Something like a glossary at the top might scream: Watch out – something over your head comes this way and isn't encyclopedic besides. my two bits Julia Rossi (talk) 06:24, 4 December 2008 (UTC)

Footnotes can be used "to add material that explains a point in greater detail". I can't find any examples immediately, but they're everywhere, and often used as "glossary" locations, usually quoting a definition from a specific source. See Jabberwocky#Glossary and Neuromancer#Glossary for a large in-article glossary (again, there are more, but that's all I can find immediately). -- Quiddity (talk) 22:51, 5 December 2008 (UTC)

I've had a look at "Talk:Vithoba#Glossary". I would not be in favour of terms being defined in little floating boxes or floating tables scattered throughout the text – that would be distracting and would give them undue prominence. Notes and links within the text as mentioned above are better. A section setting out a glossary at the end of the article may be all right, though if there are a lot of terms that require defining it may be better to put them in a separate article as a list. — Cheers, JackLee ^–talk– 10:46, 6 December 2008 (UTC)

Consistency within articles

WP:MOS#Consistency within articles:

Each article should consistently use the same conventions of spelling and grammar. For example, center and centre are not to be used in the same article. The exceptions are: . . . titles (the original spelling is used, for example United States Department of Defense and Australian Defence Force)

Sounds reasonable, except what is a “title”? The examples imply that title is being applied more broadly than in its conventional meanings: the title of a work or the honorific of a personage. The article tank uses British English, but we are guessing that M8 Armored Gun System should retain the U.S. spelling because it is a proper name (a “title”?).

So can we change the guideline to say proper names instead of titles? Would this account for all cases? —Michael Z. 2008-12-07 16:56 z

I don't think the term proper name is usually applied to the title of a work. It might be technically correct, but I don't think most readers of the MOS would think of titles of works when they see proper name. --Gerry Ashton (talk) 17:11, 7 December 2008 (UTC)

Isn't proper name a precise synonym for proper noun, except maybe not restricted to the discussion of grammar? Fahrenheit 451, Mona Lisa, and One Flew Over the Cuckoo's Nest are proper names, in contrast to common names. —Michael Z. 2008-12-07 17:35 z

Separated out book titles, as being both proper names and quotations. Our examples are intended to be proper names, although there probably is a book called United States Department of Defense. Septentrionalis PMAnderson 16:29, 8 December 2008 (UTC)

November Update

The monthly WP:Update is done for November changes to content policy and general style guidelines. WP:CITE and WP:WAF were added to the pages to be updated next month. - Dan Dank55 (send/receive) 19:44, 9 December 2008 (UTC)

Encyclopedic style

Is there a more general, fundamental style guide that addresses the actual writing style, as opposed to the little details? Imagine an article written like this:

Cardiologist Dr Hyan Mighty, MD, PhD, Chair for Life of the Extremely Important Department of the World-Renowned Medical College (WRMC), in his 2007 Journal of Medicine paper, writes that proper scientific studies prove that half of women in America die from heart disease.<ref>Might, H. ''What women die from.'' 2007. JM.</ref>

As this is not a statistic that is significantly disputed, the attribution is unimportant -- and possibly even inappropriate, since the fact of attribution suggests that this is a view held by only one person. But I keep seeing articles that throw in these details. I think the typical goal is POV pushing, but sometimes it's real ignorance of what constitutes an encyclopedic summary. Good editors rapidly grok what we're after, but others don't. Is there somewhere an essay about "encyclopedicity"? WhatamIdoing (talk) 20:10, 22 November 2008 (UTC)

We shouldn't need a rule for everything, but I agree that in general if a fact is undisputed then it should simply be presented by itself, with attribution in a reference. Only if there is reason to believe that opinions differ amongst reliable sources should the proponents be mentioned in the article body, so as to avoid weasel words. Chris Cunningham (not at work) - talk 11:06, 23 November 2008 (UTC)

In most articles such a sentence would be trimmed to: "half of women in America die from heart disease.<ref>Might, H. ''What women die from.'' 2007. JM.</ref>". We need the cite, what we don't need in most cases, as WhatamIdoing points out and Chris supports, is the in body naming of the source unless the statement is contentious. Science articles tend to use Wikipedia:Parenthetical referencing, though that would appear like this: "half of women in America die from heart disease (Might 2007).<ref>Might, H. ''What women die from.'' 2007. JM.</ref>". More information may be found at Wikipedia:Citing sources. SilkTork *^YES! 13:50, 23 November 2008 (UTC)

So it sounds like we don't have any essays on this subject. I'd really rather not write one myself, but it would be handy. I'm thinking that it should cover some basic issues like:

Just state generally accepted facts (according to mainstream views) without a lot of folderol about the specific source. That's what your ref is for. Don't blather on about a source's credentials. If they're that important, then write a bio about them.
Write dispassionately. We're not trying to convince anyone of anything. Wikipedia is WP:NOT advertising/promotion/public awareness/anything else.
Write concisely. Bits are free, but the reader's time and attention is not. Present the most important information. Supporting details, like who said what to whom, or how many people completed the survey, are normally too trivial to include.

If anyone has other ideas, I'd be happy to hear them. WhatamIdoing (talk) 04:05, 27 November 2008 (UTC)

In certain circumstances it is acceptable to quote where the information from, and doing so helps the reader understand the situation better and scrutinize it, and makes them think without requiring their own impetus (if the average reader cared enough to check the source). I think it should be handled on a case-by-case basis, rather than just ending any discussions on the matter by having part of the MoS to refer to. Not stating the details of an investigation is a common technique amongst amateur academics seeking to make a POV - quote a stat, and bury the details in a footnote (so the problem exists in doing this aswell)...meanwhile, Wikipedia has a reputation for being poorly sourced and it would do the site a great favour if it were to quote sources in the text, in the case of disputed information. The ideas for an essay don't say that this guideline is limited to widely held beliefs only.

I think that "Write dispassionatly" is a bit vague: what is passionate or not is difficult to define (I think WP:WEASELWORDS and WP:NPOV are sufficient and we should try not to micromanage prose). Yohan euan o4 (talk) 20:33, 5 December 2008 (UTC)

Thanks for your comments. In general, I agree with your comments. But what I'm looking for is a description for the newbie of what you probably need to fix when someone slaps {{Unencyclopedic tone}} on an article. The instance that I typically encounter is someone that is trying to puff up a (boring, widely accepted) fact with utterly unimportant details. See, for example, this exchange, in which an editor asserts that if (his commanding officer's?) name isn't advertised in Wikipedia as having authorized a particular study, then safety research will never be authorized in the future because the person authorizing the expense won't get famous for it (I paraphrase liberally). Or see many early versions of Wilderness acquired diarrhea, which describe the details of many, many medical studies, including the kind of study, the location, the people that ran it, the number of people participating, the number of days they were hiking, and so forth -- when, in fact, whole sections should have been reduced to single sentences like "Washing hands is associated with a reduced incidence of diarrhea among backpackers.^[1][2][3]" However, it could be anything. There are many ways to be "unencyclopedic" without screwing up grammar and punctuation (which is largely the point of this document).

So: Assume a basically clueless, trying-to-be-helpful newbie. Assume that the problem is not with grammar and punctuation issues, but with the way of selecting and presenting information. Assume that it's not a person that easily mimics styles and honestly doesn't understand why the example I give at the top of this section is not "encyclopedic". Assume that the person is confused and frustrated by the unexplained "bad tone" tag. What resources would you point this person to? WhatamIdoing (talk) 05:47, 8 December 2008 (UTC)

Okay. I certainly see the use in it. I wouldn't want to see an essay being misused to purge Wikipedia of all source details, in the main text -- especially in articles that are framed around a debate, or a contentious theory. I don't think there's much precedent as such information is not usually under the remit of encyclopedias. I also think it's useful in the cases of new information, that could be still up for debate, or could be disproved/isn't widely accepted -- some editors make the mistake of thinking that because something is recent, it's more accurate. In these cases, it would be best to give the details of new developments. Yohan euan o4 (talk) 20:14, 10 December 2008 (UTC)

National variants of brandnames

I'm not a hundred percent sure this is the right place to ask this, so if I'm wrong, please bear with me. The other day I came across an article about an Australian topic (chocolate crackles) which included talking about a breakfast cereal which in Australia is called Kellogg's Rice Bubbles. In many (all?) other nations it's called Kellogg's Rice Krispies. But the article gave the heavier weight to the non-Australian name: "Rice Krispies (known as "Rice Bubbles" in Australia)". This didn't seem right to me as I know that articles about Australia are supposed to use Australian English, articles about the USA are to use American English and so on. I just assumed the same would apply to brand names as well. However someone reverted my edit citing, in part, that "global brand name is fine". Now it is probable that the other reasons that this person reverted my edit were more significant to them (and besides I know they're probably right about the other - stupid CITE rule ;-) ), and they may not have worried if it were only the Bubbles/Krispies issue. But I'm curious to know if there is a policy in regard to this issue. ☮ AussieDingo1983 (talk•my edits) 06:49, 4 December 2008 (UTC)

I don't know the policy, but expanded the article to get around the problem. (I hope) Julia Rossi (talk) 22:57, 4 December 2008 (UTC)

I agree that the original wording was wrong, but question whether Rice Krispies is merely a brand rather than a generic term in the US? (Surely Kellogs don't have 100% market share in that product). In New Zealand, the generic term is generally rice bubbles, but some people use the (market leader) Sanitarium trademark name "ricies" generically. Other brands are using names like rice pops and rice poppas. I'm not sure what Kellogs call them (it tends to be the dearest brand here, so I never buy them), but I've never heard of Krispies here - other than as a longstanding trademark name for coconut-flavoured thin biscuits. Generally I'd say we should avoid all brand names and use the most common generic name relevant to the country.

OK, I've just checked NZ's food bible, the Edmonds Cookbook, and they use the generic term "puffed rice breakfast cereal" dramatic (talk) 23:39, 4 December 2008 (UTC)

They're called Rice Krispies in the US, Canada, and the UK. And yes, they do have a huge market share. The nearest generic puffed rice product is unpopular possibly because it is soft, not crisp. The analogous desert is Rice Krispie squares, usually made with marshmallow and sometimes with chocolate.LeadSongDog (talk) 05:39, 5 December 2008 (UTC)

This is not the right place for this discussion – I would suggest the article's talk page. But you may find the principles in "Wikipedia:Disambiguation#Is there a primary topic?" helpful, even though the guideline relates to article titles rather than content. — Cheers, JackLee ^–talk– 10:49, 6 December 2008 (UTC)

*Anderson, I don't accept that because hard-copy style guides dither on an issue, we should too. A specific point: most readers have their prefs set to justify their text. When it wraps around figures and infoboxes, the column-width is relatively small. Does this affect the appearance and readability of the jumbo-spaced dots in the ellipsis symbol? That was always in the back of my mind, as well as (1) my subjective, personal preference for the closer placement of successive periods, and (2) the ease with which anyone can type them in without having to find out about the symbol. Tony (talk) 11:21, 10 December 2008 (UTC)

Um, wrong place Tony? Anderson hasn't edited this section. Woody (talk) 11:24, 10 December 2008 (UTC)

Oops, thx Woody—I've been looking for it but the phone rang. Tony (talk) 11:46, 10 December 2008 (UTC)

Em dashes, en dashes and ellipsis

Since these characters have been added pre-made to the symbol box at the bottom of edit page shouldn't we encourage people to use them rather than the code? It prevents clutter of the edit window with unneeded code. Also, if we prefer to use three spaced dots rather than the actual ellipsis character why has that been added tot he edit box?--Patton 123 17:02, 4 December 2008 (UTC)

Patton, the MOS guideline on ellipses needs revisiting. It is not perfectly consistent, and it was never subjected to full scrutiny here. But one thing is clear: three spaced dots are not recommended in the guideline. Three unspaced dots are recommended. As for the pre-made ellipsis, we say this:

Pre-composed ellipsis character (…); generated with the … character entity, or by insertion from the set below the edit window. This is harder to input and edit, and too small in some fonts. Not recommended.

Myself, I endorse that. It has been a pretty stable part of a once-contentious guideline.

Inclusions and exclusions under the edit box are not determined by editors at MOS. It is unclear who decides such things, and where and by what consultative process. I would make radical changes to it. In particular, I would remove many of those seldom-used symbols (and add the root sign √, which is obviously more useful than ₪, ৳, ₮, or ₩); I would re-order the modified Latin characters so that things could actually be located; and I would add a suite of basic music symbols, including ♭, ♮, and ♯.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 06:33, 6 December 2008 (UTC)

I believe Noetica is saying not to use ". . ." for an ellipsis, and most people understand this, but I've seen people misread that as "don't put spaces on either side of the ellipsis". AP Stylebook likes spaces on both sides of the ellipsis whenever it's used. - Dan Dank55 (send/receive) 15:24, 6 December 2008 (UTC)

That reading is just possible, Dan; but it is clearly not the natural way to interpret what Patton wrote. Never mind; it doesn't matter. As for people misreading the guideline itself, I would be surprised if they did not grasp the clear recommendation that three unspaced dots be used to form the ellipsis. This could only happen by their inattention, or by verbosity and inconsistency later in the guideline. After all, at the start of the guideline we don't talk about spaces before and after, but simply define and show three ways of making the ellipsis mark itself: three unspaced periods (...); pre-composed ellipsis character (…); and three spaced periods (. . .).

Here's the rub:

Many guides are unclear about what is part of the ellipsis itself, and what merely accompanies it. Accompaniments may be spaces of various sorts, or one additional dot. This is very poor! CMOS is guilty of both muddles, in one of its most thoroughly confused guidelines (11.51 ff.).
No printed guide covers online text adequately, as we must.
No printed guide covers collaborative editing adequately, as we must.
Properly, the spacing around an ellipsis is not a matter of deploying standard characters, such as in wordprocessing or other normal entry or editing of text. It is a sophisticated matter of typography. MOS cannot address such matters well, though it improperly strives to for nested quotes – and that needs fixing. Nor does any major guide for editors or writers attempt such a thing with awareness and diligence.
Most guides are inconsistent, incomplete, and arbitrary in their treatment of ellipses.

All of these considerations have inclined me not to revisit this topic in MOS. There are major issues to be sorted out before we can approach it rationally; and there is not enough sustained insight or energy here to work on those issues. Nor are there appropriate protocols. This applies in some degree to a great deal more in MOS, if not to all of MOS.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 22:48, 6 December 2008 (UTC)

What this page lacks is respect for Wikipedia and its editors. If we confined this section to a discussion of the advantages and disadvantages of the various forms of ellipsis, that would be far more useful - and probably more convincing - than the present text. I'm glad to see the section has improved in that respect.

One of Patton's questions assumes we prefer three internally spaced dots to three dots with no spaces [that is, the three precomposed dots 23:36, 8 December 2008 (UTC)] . I think, from the history, that not recommended is a more favorable view than Strongly deprecated. But if these boldface commands aren't being understood, why include them? Septentrionalis PMAnderson 15:42, 8 December 2008 (UTC)

SMcCandlish is responsible for the wording Strongly deprecated, applied to three individually spaced dots. I sympathise, though I would not have put it the same way. That option is the worst for us, even though CMOS allows it and nothing else (except the conceptual confusion of a fourth dot).

I agree that discussion should generally be on content. I suspect I have examined more sources on the ellipsis than anyone else here, and followed the arguments more closely. But I do raise issues of process in this case, because they are unavoidable when we approach the ellipsis – at the boundaries of typography, ordinary management of text, and markup. The topic is mishandled in most forums and in most style guides, because of inattention to superordinate issues. I would revisit it when those issues are dealt with.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 23:25, 8 December 2008 (UTC)

So list the superordinate issues, unless you mean the numbered issues above. I don't see why we need do more than take out the bold-face comments, and cite sources for the value judgments we do make, but I'll consider anything. On the numbered issues, which mostly seem beyond MOS's brief:

We are discussing typography, with a fairly simple layout language; we must treat two descriptions which result in the same final appearance are equivalent.
Aside from browser issues, which we usually ignore, what difference does being online make to an ellipsis? We're not going to link them.
We could use a general guide to collaborative editing; but how does it affect ellipses either?
Neither can our edit space address complex typography adequately; that is a condition of our existence. All we can do is write MOS for the present set of symbols, and decide when we should appeal to the developers.
Most guides are inconsistent and arbitrary on ellipses. True; a standard we should not follow; but all we can do about it is to accept them as they are - as reasons for one method or another, but not decisive reasons. Septentrionalis PMAnderson 23:48, 8 December 2008 (UTC)

Anderson, if you have not yet grasped them I will yet again list the superordinate issues (that meaning overarching, logically prior issues, opposite of "subordinate", as I thought you would know). But first I want matters resolved in the section below concerning respect and common purpose. As you know, if those matters are not resolved I will leave here soon. They are the first of the superordinate issues, in fact. Meanwhile, among my recent housekeeping changes are a few in the Ellipses guidelines. To answer your five points above:

It is the transfer of theory and practice apt for the printed page to Wikipedia that is difficult. Hic labor est. To my knowledge, no one has yet addressed these issues squarely for our sort of environment, let alone found solutions. We have to be pioneers. That's the Wikipedia way.
The differences have to do with collaborative editing, and with breaks at the spaces that may be associated with ellipses (compounded by that collaborative editing and by unpredictable end-use, with various browsers configured in various ways). And so on.
See 2.
I agree with you; and generally I stand against mere high-tech kludges to solve our problems. We need elegant solutions, or we're better keeping things simple. The hard space, bless its cotton socks, is something we desperately need to sort out (see recent posts at WT:MOSNUM); but as Montaigne said and McLuhan quoted, "the thing of it is, we must live with the living." We must live with the limited and clunky resources Wikipedia so far has available, and with the reluctance of our peers to recognise good sense when they trip over it.
Quite so. The punctuation guidelines we had especially (some of which I initiated) reflected common practice in the seriously imperfect established guides, adapted as well as we can manage to Wikipedia's online environment. Now, with the slight housekeeping I have done for ellipses, we manage this even better. I have studied treatment of the ellipsis in no fewer than thirty sources, and I am confident that what we have now is ecumenical and workable. None of the MOS guidelines that I have anything to do with is arbitrary, I can assure you. Some of us do read beyond Fowler's, you know (though I also have all major editions of Fowler's, in my armamentarium).

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 10:46, 10 December 2008 (UTC)

Anderson, I don't accept that because hard-copy style guides dither on an issue, we should too. A specific point: most readers have their prefs set to justify their text. When it wraps around figures and infoboxes, the column-width is relatively small. Does this affect the appearance and readability of the jumbo-spaced dots in the ellipsis symbol? That was always in the back of my mind, as well as (1) my subjective, personal preference for the closer placement of successive periods, and (2) the ease with which anyone can type them in without having to find out about the symbol. Tony (talk) 11:48, 10 December 2008 (UTC)
- In ahort, Tony wishes to redesign the English language, and make every article use his undithering Newspeak. We are not entitled to do that; we cannot succeed in doing that. Septentrionalis PMAnderson 16:08, 10 December 2008 (UTC)

In short, Anderson, Tony is not doing that and nor am I. Both Tony and I work from a base of long-settled practice in editing for publication, and we respect precedent and industry-standard sources. We also respect the work of others who do the same. Have you ever once acknowledged this point? Have you ever noticed that, beyond Fowler (early in the 20th century) and OED (wrongly quoted, misunderstood, but easily corrected by my checking), you yourself do no such thing? See immediately above; and see passim. And stop repeating such slanders.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 20:20, 10 December 2008 (UTC)

I have no evidence of any such point. I have seen you assert you reviewed the sources; Tony doesn't even do that much. I have hardly ever seen either of you actually cite any. Dank55's cites CMOS, which is at least a source, far more often. Septentrionalis PMAnderson 23:51, 10 December 2008 (UTC)

Every once in a while I see . . . which is even worse. While the dots in "…" appear much closer together than with "..." in the font of the edit box, but appear 1 or 2 px. when viewing the saved page, at least on my system. — CharlotteWebb 21:14, 10 December 2008 (UTC)

Commas

Incorrect: Oranges are an acid fruit, bananas are classified as alkaline.

Might I humbly suggest that this is actually correct usage, as the comma stands in the place of a missing word:

Oranges are an acid fruit but bananas are classified as alkaline.
Jubilee♫clipman 09:29, 10 December 2008 (UTC)

Certainly it is good to see your comment, J. In fact, a few style guides do allow for a comma in these cases, but only for rapidly shifting conversational use, or in formulaic sayings:

She couldn't do it, I wouldn't do it, we didn't do it!
Life is short, art is long.

This use of the comma is generally informal, and unencyclopedic.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 10:11, 10 December 2008 (UTC)

This is, in short, much like the use of the exclamation mark inside sentences, discussed above: generally of non-encyclopedic tone, but not to be corrected in quotes and dialogue. We could also acknowledge that very short sentences can do without the semicolon: He shot the guard, his brother drove the van. Septentrionalis PMAnderson 16:13, 10 December 2008 (UTC)

Anderson: That example is ungrammatical. What is the relationship between the two ideas? Additive? If it were to be "... and his ..", just why the "and" suffices can be judged only on basis of the surrounding context. Tony (talk) 16:22, 10 December 2008 (UTC)

No wonder this page has no authority; it's filled with statements with no basis in fact. The example is something WP might actually want to say, supported by idiom. Tony would suppress a legitimate choice of cadence, which may be appropriate to quick action. Style guides should not worsen style; this one does. Septentrionalis PMAnderson 16:33, 10 December 2008 (UTC)

Jubilee: Is there an intended meaning in the different treatment of what appear to be apposite statements? Acids are, but bananas are classified. Huh? Tony (talk) 16:25, 10 December 2008 (UTC)

Has Tony noticed that Jubilee is quoting the example presently in our text, both with comma and with semicolon? Go ahead and fix it; that won't change the discussion in the least. Septentrionalis PMAnderson 16:33, 10 December 2008 (UTC)

Thanks all for the feedback! As I think about this issue, I realise there is a very good reason to avoid this particular usage of the comma: clarity.

He shot the guard, his brother drove the van. This sentence could mean:

a) He shot the guard but his brother drove the van.

b) He shot the guard before his brother drove the van.

c) He shot the guard after his brother drove the van.

d) He shot the guard although his brother drove the van.

e) He shot the guard and his brother drove the van.

f) er... etc!

The style of an encyclopaedia clearly has to avoid ambiguity, unless it is intended.

Methinks I'll concede defeat on this one! Unless anyone has any better examples? ;)

Jubilee♫clipman 16:57, 10 December 2008 (UTC)

PS Tony1: bananas are classified? I never knew that! lol!

Yes, we should only use this form when it is unambiguous in context, but that may be narrowing down the exception to something that can be handled by WP:IAR . Septentrionalis PMAnderson 17:00, 10 December 2008 (UTC)

Yes, we have no [alkaline fruit]. Wikipedia appears to be innocent of this salient dietary distinction, except that MOS once more rides to the rescue.

Anderson makes this point: "This is, in short, much like the use of the exclamation mark inside sentences, discussed above: generally of non-encyclopedic tone, but not to be corrected in quotes and dialogue. We could also acknowledge that very short sentences can do without the semicolon: He shot the guard, his brother drove the van." Well, we could. As I have said above, "a few style guides do allow for a comma in these cases" (with strong restrictions). But again the superordinate issue of detail arises. If detail is to proliferate every time someone finds some hair-fine Hegelian ramification to add, we will elicit justified complaints. Why allow the occasional question mark within a sentence (or even exclamation mark, in transcription of spoken English), and not these commas? That is exactly the sort of thing we can discuss! But such work is ruinously inefficient and vulnerable, until we agree on a common purpose here, and on common protocols. That's of the highest priority; that's superordinate.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 20:08, 10 December 2008 (UTC)

Detail has already proliferated; furthermore, it is erroneous detail, phrased as or taken to be imperative. That's what wrong. Septentrionalis PMAnderson 23:29, 10 December 2008 (UTC)

A confident reply, Anderson. A pity it is not supported by facts or argument. Erroneous detail? Ha! However, flexible as always, I'll once more respond to your proposal and incorporate something appropriate in MOS.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 02:35, 11 December 2008 (UTC)

Forcing Lead image

I thought we were clear that there was no reason, and no consensus for size forcing the lead image purely because it is the lead image. Examples of where the lead image have been forced can be traced to the other criteria listed. I will remove this statement: "Lead image, which should usually be no larger than 300px." I feel we need to have some rationale for size forcing the lead image before we can include that statement. SilkTork *^YES! 16:49, 21 November 2008 (UTC)

I dunno, that statement's been in the guideline for easily long enough to become established (I mean, to the extent that we need consensus to take it away, not consensus to leave it in). Maybe we could look at recent practice on FAs, to see what people have been doing with lead images recently. It seems natural to me, from the point of view of the appearance of the article, to start of with a sizeable image as a main illustration of the article subject.--Kotniski (talk) 17:07, 21 November 2008 (UTC)

(After looking at FAs) Well, the last few have had infoboxes rather than lead images, so not much evidence to go on there. But it strikes me that since infobox images are generally larger than standard size thumbs, it makes sense to allow the lead image (which is basically an infobox image without an infobox) to be at least as large as the infobox ones.--Kotniski (talk) 17:15, 21 November 2008 (UTC)

(thinking aloud here) And MOS:INFOBOX gives 300px as a typical width for an infobox, so maybe that's where the 300px for the lead image comes from. Seems illogical to insist on making the image significantly smaller just because it hasn't got an infobox to wrap it up in.--Kotniski (talk) 17:19, 21 November 2008 (UTC)

It also depends on the size at which the lead image is useful. The one at Paleontology needs to be big, but summarises many aspects of the subject. The one at Cambrian substrate revolution summarises half the article, and again needs to be big - and in this case it's hard to think of what an alternative lead image would look like. As Kotniski points out, many lead images are in layout structures such as infoboxes, and it would look silly if the images were much narrower than the boxes. --Philcha (talk) 17:38, 21 November 2008 (UTC)

That wasn't quite what I meant; I was assuming that. I meant that if we allow ≈300px images on many pages because they appear in infoboxes, then it would be inconsistent (from a stylistic point of view) to discourage ≈300px lead images on other pages that don't currently have infoboxes.--Kotniski (talk) 17:46, 21 November 2008 (UTC)

And the examples you mention are perhaps not the best ones to consider, because they would qualify for the larger size under other criteria (the need to show detail). We have to consider images that have no other reason to be larger than standard thumbs except that they are lead images. Not having any better example off hand, let's consider this article (it's an old revision, since changed to an infobox). Here I made the lead image 300px (in accordance with this guideline), for no other reason than that it was a lead image. Thoughts please (or better examples if you can). Is the picture OK? A bit too big? A lot too big? --Kotniski (talk) 17:55, 21 November 2008 (UTC)

I certainly don't think "we were clear" on this, no. There may have been a temporary majority on the odd thread here about it, but this certainly isn't clearly supported. Chris Cunningham (not at work) - talk 17:50, 21 November 2008 (UTC)

It's an interesting question. Speaking with the bias of a copyeditor, my favorite content policy page is the one that says Wikipedia is an encyclopedia. Look at any encyclopedia, online or print, and you'll see that, when they have an image that helps define the topic, which is the function of images in Wikipedia lead sections, the image is not a thumbnail and doesn't require a click for viewing. When we do things the way everyone else does them, it facilitates the free flow of information into and out of Wikipedia, and enhances our reputation. I believe this is the thinking that produced the long-standing guideline (which SilkTork has just removed) that lead images can have their width specified, as long as it's kept to no more than 300px. I'd be in favor of restoring that guideline. I understand the argument that many people use smaller screens to view Wikipedia, and some people have slower connections and don't want to see large images unless they specifically ask for it by clicking on it. Over the last year at WT:MOS, the most contentious arguments have always been the ones that require the devs to do something they haven't done yet. This problem needs a dev solution, including at least another option or two in user preferences, and perhaps some help for non-logged-in readers as well. We can't make up for the lack of dev action here, and we cause a lot of trouble when we try. - Dan Dank55 (send/receive) 18:32, 21 November 2008 (UTC)

If there is a rational need for an image (any image) to be forced then we account for that in the existing criteria (or add extra criteria as needed). Unless we give a reason why a lead image by itself should be forced then by default the statement "Lead image, which should usually be no larger than 300px." allows all lead images to be forced. If a lead image is in an Infobox it automatically gets adjusted to the size of the Infobox, and so there is no need for us to make any comment on this page other than what is already there "Start an article with a right-aligned lead image or InfoBox."

It is not a convincing argument that because the "lead image" statement has been around since April 2007 that it has consensus. It was inserted into the guideline in dubious conditions. It was backed up by a link to the debate as though the debate had led to a consensus that the statement should appear - but when reading through the discussion, the conclusion was the opposite - it was decided to remove the lead image comment. Since then there has been repeated questions about it, as linked to above. I am here myself because there is some dispute about the issue. What might be an advantage is to clarify the situation because there has been no consensus to size force the lead image purely because it is the lead image, and the wording which allows size forcing of the lead issue has been questioned and challenged.

The question is: Is there a reason (other than those already given on the guideline page) why a lead image should be size forced? If there is a reason why an image should be forced then that reason is pertinent for ALL images within an article. Unforced images in the lead position appear to be fine - and there are many of them. SilkTork *^YES! 20:13, 21 November 2008 (UTC)

I voted for you in your ongoing and wildly successful RfA; I very much respect your work. But you apparently didn't think much of my arguments, and I think this might have something to do with the lack of sufficient intersection between the communities at WP:AfD and WT:MOS. The focus at AfD is learning all the ways that the world has gotten it wrong, and how Wikipedia is different: how to use better judgment than other information sources do and how to do a better job of finding consensus. I think if you surveyed people at AfD on why they think Wikipedia has become the top online content site in the world by hits, they'd focus on the kind of arguments you hear at AfD. But over here at WT:MOS, we focus more on similarities: which phrases and orthography and page design elements are easy for a wide range of readers to digest because they've seen them before. My argument that lead images should be big enough to see clearly because that's how all encyclopedias present their lead images is probably a persuasive argument over here; it might not be the kind of thinking that AfD people are used to, but if so, you guys should come hang out over here more often :) What we do is important, too. - Dan Dank55 (send/receive) 21:43, 21 November 2008 (UTC)

It seems most editors favour keeping a mention of forced lead picture sizes in, for which reasons have been given. This was after all Silk Tork's original proposal a few screens up, though I am ok with "to 300px" per Ty, or as a minimum. Johnbod (talk) 01:18, 22 November 2008 (UTC)

To Dank. Sorry, I hadn't fully taken on board your comments when I wrote the above. I was responding mainly to the responses above yours, and I had to finish up quickly as I was going out. I don't feel I am in any particular camp - I contribute to guidelines as much as I contribute to AfD (possibly more) - nor do I feel there is a conflict between these camps. If there is any perceived conflict that is possibly down to misunderstandings. And misunderstandings certainly happen when people take text from a policy and then change the wording on the guideline. Where policy statements are used in guidelines then it seems to me to make sense to stick with the same wording OR change both policy and guideline wordings so they match. When we have differing (and possibly conflicting) statements the whole point of having guidelines gets lost - and people can pick and choose the statements that most favour what they want to do. You don't like forcing image size, then quote Image Policy. You want to force the image size, then quote MoS - Image. No wonder people keep raising this same issue! It appears to me that the Image Policy is fairly clear on the point of size forcing: we don't do it unless there is a reason, and points to MoS for those reasons. Currently we have the following reasons:

Images with aspect ratios that are extreme or that otherwise distort or obscure the image subject(s)
Detailed maps, diagrams or charts
Images in which a small region is relevant, but cropping to that region would reduce the coherence of the image
Images containing a lot of detail, or where the detail is important to the article

Those seem decent enough reasons, and if a lead image fits one of those reasons then it can be forced, and it would be appropriate to do so. Added to that, is the provision that an InfoBox can be placed in the lead section, and that would allow an image to be forced within the InfoBox. So, that appears to cover every instance where a lead image might appropriately be forced.

Your argument is that an image in the lead section should be "big enough to see clearly", and that is an excellent suggestion. I would add to that, that ALL images should be big enough to see clearly. I think that's a given. And, I hope, that the provision for making images big enough to see clearly is given in the criteria we already have, and which I quoted above. If people feel that the current wording is not clear enough to guide people to make the decision to force an image so it can be clearly seen, then we need to address that. But that would be a general criteria, not one specific to the lead image. If you take a look at an article close to my heart - Beer - you'll see two lead images. Neither image is forced, and the top one shows the subject very clearly. Take a look at Wine. Is that image not large enough? And that is not forced. I took a look at some Featured Pictures such as Innocence and Catrina, for examples of unforced Featured images used as the lead image, but that was difficult because InfoBoxes prevail, and also because Featured Pictures are not always used as the lead image (which surprised me). But I hope that with the few examples I have shown, that it is clear that the lead image does not by default need forcing. And that if the lead image does need forcing to be seen clearly, then I hope our general forcing criteria would be sufficient. Am I making sense? SilkTork *^YES! 08:56, 22 November 2008 (UTC)

When you say "currently" above, you meant "currently, since I just changed it from the long-standing guideline". The best I can tell from the above discussion, consensus hasn't changed from what it's always been. I follow that you're trying to remedy a difference you perceive between image policy and WP:MOS. I don't know what was going on over at the image policy page; I keep up with content guidelines and style guidelines, and there are a lot of people who know a lot more about image issues than I do. Regardless of what some guy stuck in on that page, I'm pretty sure that the consensus of editors who show up at GAN and FAC, and certainly the consensus of article reviewers, is to prefer images that are larger than thumbnail in lead sections. I've reverted your change. - Dan Dank55 (send/receive) 21:35, 22 November 2008 (UTC)

This "no larger than 300px" appears to be a reversal of how it had been for a very long time. Without actually checking back, I'm certain that the size was "no smaller than 300px", on the basis that 300 was the largest size that a user could set in preferences and that the wording here made sure that the forced size was no smaller than that user pref size. --AliceJMarkham (talk) 07:46, 23 November 2008 (UTC)

I find all this historical analysis rather unhelpful. We recently had a stable guideline for several months; on a highly visible page like this, people would have been objecting right from the start if it didn't have consensus. I believe it did say something like "no smaller than 300px" at one point, based on the logic Alice cites, but surely we can see that that logic is flawed (it's a repeat of the date autoformatting fallacy - why should an insignificant aesthetic effect on 0.001% of users cause us to adopt a less desirable display style for the other 99.999%?) Then I think it was changed to "no larger than..", but people objected to that too, so we settled on "often...about 300px". If anyone seriously objects to that then they can propose a new wording, but please ensure that there is consensus for any change so we don't get into another pointless edit war.--Kotniski (talk) 11:26, 23 November 2008 (UTC)

Let us look for some acceptable wording for size forcing the Lead Image. Personally I'm not interested if the lead image is forced or not. But I am concerned that the issue is contentious (as based on the diffs I gave above), that forcing the lead image because it is the lead image is not a valid reason and rubs against the wider consensus of policy, and that the lead image clause was inserted in this guideline erroneously. Suggested wording:

"The Lead Image may be size forced only when it meets the size forcing criteria"
"The Lead Image may be size forced (beyond the size forcing criteria) when the Lead Image needs to have impact."
"The Lead Image may be size forced (beyond the size forcing criteria) when in the editor's judgement the lead image needs greater impact."
"The Lead Image may be size forced (beyond the size forcing criteria) in all cases, provided the image is kept within the 180 - 550 pixel range. This is because the Lead Image sums up the topic, and needs to have greater impact than other images in the article."

I would favour either 1 or 4. SilkTork *^YES! 13:31, 23 November 2008 (UTC)

Option 1, personally. The default should always be used unless there is a specific reason for a particular image to be resized. Just being near the top of the article doesn't make an image need to be resized. Also, there are practical problems with large lede images and tables of contents on small screens; these are less serious when the lede image is smaller. — Carl (CBM · talk) 13:42, 23 November 2008 (UTC)

I blame myself and other style and layout people much more than I blame anyone else for the fact that we don't have huge piles of professional copyeditors participating at WT:MOS who would gladly jump on this question. We're not doing a good enough job of being friendly and welcoming on the one hand and holding conversations to a high and professional standard on the other. So, to be clear, I'm not disappointed in SilkTork and Carl, in fact I highly respect both of you, I'm disappointed in myself. I agree with you that specifying image sizes is a problem on small screens, but that's a problem the devs need to fix. Having said that, the option you two are suggesting is truly, abysmally awful, for the following reasons:

As I said above, some communities at Wikipedia are right that what what they do is critically important but wrong about the importance of listening to publishing professionals on some issues, especially copyeditors. There isn't any great disagreement among people who produce online and printed reference materials that attractive lead images are important, as long as the images are relevant and helpful and there's no difficult tradeoff (such as breaking someone's back when they try to pick up your encyclopedia). 140-pixel (the default for upright thumbnails) and 180-pixel images may be functional, but they're rarely attractive. This wouldn't even be a discussion if professionals were weighing in and if their views were taken seriously.
Any MBA will tell you that it's a classic mistake to assume that you know what magic is responsible for your company's success without asking your co-workers and your customers. Don't radically change the look of your product just because some guys in the back room thought it would look "cooler" that way, or some engineers told you that you could save a little money (or in this case, a little page layout landscape) with a different design. The style guidelines haven't required thumbnails before, and most contributors have chosen lead images more in the range of 300px than 140px, and they've probably made those choices because they and their readers thought it looked better. Don't assume they're all wrong.
SilkTork says above that it's not relevant that a recommendation for around 300px has been in WP:MOS for a long time; by implication, he's saying that the many conversations over the years on the subject at WT:MOS, style pages and article review pages that led to the recommendation are irrelevant. I disagree; I'd rather listen to a wide range of people, and I especially listen to people who also like to listen to a wide range of people. Those who don't read history are doomed to repeat it. - Dan Dank55 (send/receive) 18:39, 23 November 2008 (UTC)

The strength of Wikipedia comes from the diverse talents, skills, knowledge and experience that people bring to the project. Copyeditors are part of that. All views are respected and listened to - that's the Wiki way. And that's what we are doing here. I'd like just to address Dan's third point about the 300px recommendation being in WP:MOS for a long time. I'm thinking that at this point it might be best if we look at current arguments rather than hark back to what has been said in MoS previously. However, if we are going to do that then I'd like to point out that Policy takes precedence over guidelines and WP:Image use policy has been consistent since at least 2004 on the view that by default images should not be forced, though accepts that there are exceptions. Exceptions are discussed at Wikipedia:Mos#Images. Until 20 April 2007, there was no exception for lead images. That change came about through a misreading of this discussion. The bringing in of the statement that it is acceptable to size force "a lead image that captures the essence of the article." was done so against the advice and consensus of the discussion, and specific image policy. It didn't at that time specify any limits to or recommendations to size. The first mention of 300px came with this edit from this discussion. It's a little unclear the intention, as the edit gives a maximum of 300px at the same time as a minimum of 300px. After a period of uncertainty this clearer wording was offered in July this year. A short while later we return to the contradiction of having a maximum 300px and a minimum 300px. The contradiction has been observed and fixing the size at 300px is offered. A bit of flexibility is introduced. These are selected difs, there are plenty more which reflect the varying views on the issue. My point is that there has not been a stable period of harmony as regards the size forcing issue - and in particular the 300px size. My suggestion is that, anyway, it's not helpful to us to look at what has gone on in the past 18 months, but that we would be concentrating on discussing the best way forward. SilkTork *^YES! 23:52, 23 November 2008 (UTC)

I am bothered by the fact that this discussion, about an issue which affects ALL editors, is receiving input from only 3 people (with a single comment from 5 others). The reason for the lack of participation I can only put down to being unaware of this debate. Even so, in a month's time the 10 000 or more WP editors will find a decision based on this debate tagged as having been arrived at by consensus. WP is in desperate need of an effective way of notifying people about important discussions that will affect their editing lives. Does this worry anyone else? Rotational (talk) 05:19, 24 November 2008 (UTC)

2c: size forcing should be avoided. People may have very good reasons for their chosen user thumb size preferences; e.g. small images because of low bandwidth; or large images because of eyesight problems. It is discourteous to dishonour these preferences without a good reason. The vast, vast majority of size forcing lacks any such reason. This includes lead image size forcing: it is claimed that size forcing the lead image is sometimes necessary to achieve more impact, but I am yet to see a concrete example of that. I very much like the idea of the MoS discouraging size forcing. Hesperian 05:46, 24 November 2008 (UTC)

I have tagged this as a centralised discussion following Rotational's comments above. SilkTork *^YES! 08:24, 24 November 2008 (UTC)

I think it's very wrong the the image use policy has a section on displayed image size. Image size is a style issue that should be treated in a flexible way. Policies are meant for more fundamental issues such as legal and ethical issues, basic principles, technical limitations, etc. These are type of issues that occupy most of the WP:Image use policy page. Having a section on display size there sticks like a sore thumb; I imagine someone added it to try to provide a helpful summary and link to the style guidelines, not to try to provide a "size policy" to trump the MoS. --Itub (talk) 10:44, 24 November 2008 (UTC)

Can you see the irony there? Allowing people to set their own defaults is termed "inflexible", while over-ruling those defaults with a fixed image width is termed "flexible". Hesperian 22:21, 24 November 2008 (UTC)

No irony. Not all images need to be the same size, and the best person to decide is the editor. Now, if the user preferences had an option for relative scaling of the images, that would be something useful. --Itub (talk) 13:23, 25 November 2008 (UTC)

... which is why I submitted an enhancement request over three months ago.[1] Meanwhile, we're stuck with what we have, which is the requirement that we choose whose preferences we are going to honour: those of the user or those of the editor. Both choices can equally validly be termed "flexible" and "inflexible" depending on your point of view. Hesperian 06:04, 26 November 2008 (UTC)

I'm not sure that centralising all discussion of image sizes under the heading "Forcing Lead image" is a good idea, as I think lead images have specific issues that do no apply to others.

I've stated at Forced v unforced image sizes (again) why I think editors sometimes need to have control over the sizes of images, so for now I'll stick to lead images.

I'm aware of a number of issues about lead images:

Many articles have Wikiproject-specific infoboxes, which provide a standard summarisation that is helpful to readers. Infoboxes vary in width from about 180px to 260px. A 180px or 200px image may look silly in a 260px infobox. It would be helpful to ask Wikiprojects why they chose a particular infobox width. It would be silly to set a rule about lead image widths without considering infoboxes and very silly to set a rule about infobox widths without consulting the Wikiprojects that use infoboxes.
Ignoring infoboxes, some lead images are just eye-candy and some attempt to give information about the subject. In the latter case they may need to be greater than the default thumb size. I know of one lead image, in a GA, that illustrates several aspects of a subject. Because that image is complex, it needs to be shown at 400px. Before even suggesting it I checked that it worked both on widescreen monitors (aspect ratio 16:10 or 16:9) and "traditional" (4:3). I offered the GA reviewer a more modest alternative image that conveyed less information, and he was happy with the large one. Note the GA reviewer for that article is a very experienced reviewer and editor of biology-related articles.
We would also need to consider what I'd call "pseudo-images" i.e. illustrations produced by methods other than Image tags. I'm familiar with 2, cladograms, see for example those at Arthropod#Evolutionary_family_tree, and chess diagrams, see for example those at Ruy Lopez. Right now I can't think if a case where a cladogram should be the lead image, but Ruy Lopez quite rightly has a chess diagram as its lead image. I know the one at Ruy Lopez is in an infobox, but it should be the same size (about 250px) even if there were no infobox, just so that readers can see clearly what the defining position of this chess opening is.

The argument about users with visual difficulties needing larger images is weak because such users can enlarge both images and text in any decent modern browser, even Internet Explorer 7 (I've tested with IE 7, Firefox, K-meleon and Opera).

The argument about low-bandwidth users is totally spurious, because transfer time depends on image file size, not on displayed size. Wikimedia displays images by means of a standard (X)HTML IMG tag, using the WIDTH and HEIGHT attributes to control the displayed size. Anyone can see this if they do a "view source" to inspect the generated (X)HTML.

I'm not convinced by arguments based on the difficulty of viewing large images via small-screen clients such as WAP phones. Small-screen clients make reading any article like viewing a landscape through a keyhole. For example try wapedia.mobi/en/Precambrian_rabbit with your browser window reduced to mobile phone size or even PDA size (the Tungsten E2 is 4.5 in high by 3 in wide) - it's no fun, and that article is fairly short and has no images.

Finally, I agree with Rotational's (05:19, 24 November 2008) comment "I am bothered by the fact that this discussion, about an issue which affects ALL editors, is receiving input from only 3 people (with a single comment from 5 others)". I've advertised this discussion at Wikipedia:Village_pump_(policy)#WP:MOS_-_discussion_of_image_sizes.2C_especially_lead_image_sizes, and any other means of advertising this discussion would be welcome. --Philcha (talk) 11:29, 24 November 2008 (UTC)

Sorry Philcha but your "file size versus display size" argument is a load of bollocks. Wikipedia uses server-side thumbing and always has. About a week ago, server-side thumbing was turned off for png images, which may be the source of your confusion. Hesperian 22:21, 24 November 2008 (UTC)

I'm happier now that this discussion is getting wider publicity. I can't remember off the top of my head another time when I've said that someone's approach was terrible, and I groan to hear myself say that (it sounds so much like the kind of dismissive, unhelpful comment that I will generally complain about), but there's a significant danger of coming across as autocratic here and losing valuable contributors in the process. Given the choice between a lead section image in the 250- to 300-pixel range (and as SilkTork points out in the link from 2004, 250px used to be fine according to our image policy page) and a 140-pixel lead image (which is now the default upright thumbnail size), a wide range of editors have chosen the larger size for a wide range of reasons, and they didn't have to go through any tribunal to justify that image size. Experienced article reviewers know this, and people experienced with online and print encyclopedias, books, magazines and newspapers know that it looks better without having to think twice about it. What image size looks right is a subject that has been discussed constantly in review processes. To ignore all of these conversations and preferences because someone wasn't doing their job at WP:Image use policy and let it get out of sync with community expectations would make me very sad. - Dan Dank55 (send/receive) 16:50, 24 November 2008 (UTC)

I've just encountered an example of gratuitous size forcing on Beer purely because of the Lead Image wording. At least one editor is going round making ad hoc changes based on an assumption that MOS:Images said it was OK. This is clearly wrong. We are still in discussion here. There has been dispute about this size forcing of the lead image since the wording appeared in the guideline (as indicated by my many diffs). Until there can be an agreement on the wording of the Lead Image, it would prevent edit wars and disputes if the contentious wording under dispute remains out of the guideline. In the interim if having no statement is problematic to people, we could have "Size forcing of the Lead Image is currently in dispute." SilkTork *^YES! 11:38, 25 November 2008 (UTC)

Don't blame people for making changes based on what the MoS says - that's what it's here for. And please don't start making changes based on what you personally consider to be gratuitious and wrong - other people have different opinions, and you must gain consensus to change an established guideline. A guideline doesn't cease to be one just because it's under discussion and a few people disagree with it. I'll mark it with an {{underdiscussion-inline}} tag.--Kotniski (talk) 11:45, 25 November 2008 (UTC)

That said, I certainly agree with you about the beer example. The wording perhaps needs to be changed to make it absolutely plain that lead images don't have to be 300 px if inappropriate. --Kotniski (talk) 11:50, 25 November 2008 (UTC)

SilkTork, I have to agree that the lead image of Beer (before you adjusted it to the default size) was absolutely ridiculous an dyou were right to make it take the default size. It's a perfect example of the type of image I've described as "eye-candy". I also acknowledge that there are other advantages of a standard image size, such as ease of reading - even where I specify image sizes, if there's a series of images close together I usually make them the same width so that starts and ends of lines of text have the same X co-ordinates.

OTOH there are other cases where fixing the size is desirable. North Sea contains a couple of examples. The lead image there would be indistinct at a smaller size, and the infobox can't be much narrower because it contains a large amount of information. If you look at section "History of the coastlines" you'll see a couple of techniques I've used with images: the one on the left is cropped and zoomed to focus on the relevant part of a large image; and the two images side by side illustrate stages in geological history that would otherwise require a much longer prose explanation. I've already mentioned Paleontology, where a very experienced GA reviewer was happy with the choice and size of the lead image. I've only mentioned cases from from my own experience, but I'm sure there are dozens of other situations where editors have needed hands-on control of image sizes. Hence I think legislating about image sizes is a bad idea and the issue should be left to editors and GA / FA reviewers. --Philcha (talk) 12:12, 25 November 2008 (UTC)

North Sea is a good example of an article that looks ridiculous to me because my default thumb size is 300px, but the infobox mandates a smaller size, making the lead image smaller than other images in the article. Your "the lead image there would be indistinct at a smaller size" reveals your myopia with respect to default sizes other than your own. Just because forcing the size of the lead image makes the article look better for you, doesn't mean you're making it look better for everyone! Hesperian 13:21, 25 November 2008 (UTC)

That's interesting. What are the physical size and resolution of your monitor? My impression is that the layout of a lot of article swoudl be at least fractured if all images were 300px wide, and taller pro rata. Do you get a lot of whitespace in articles with a lot of images? Or articles with images producing a chequered flag effect because they can't all right-float in the same place? Or images next to some text other than what they're supposed to illustrate?

BTW re "myopia", my eyesight is pretty poor. --Philcha (talk) 14:30, 25 November 2008 (UTC)

My image thumb default size is also set to 300px and it works just fine when all images in an article have been left at thumb size. North Sea, though, has a hodge-podge of sizes with a lot of thumbs thrown in, and the net effect is not pretty. I experimented with North Sea, tweaked all the image sizes to thumb and the layout improved immensely. Perhaps, just as a test, you should set your preferences to the same and see the effect - you might even become a convert!! Rotational (talk) 19:05, 25 November 2008 (UTC)

No offence to anyone, but it really isn't important what you see on your screens with non-standard preferences set. Our readers get the default, and it's them we're writing the encyclopedia for. If it actually broke something for preference-setters I might be a bit concerned, but if it's just aesthetics then they can be safely ignored (unless it makes absolutely no difference to ordinay readers, but in this case of couse it does).--Kotniski (talk) 19:38, 25 November 2008 (UTC)

Exactly my point - if the readers' default is set to 300px instead of the ridiculous 180px, and if all images in the article were set to thumb, then readers might actually see something like a sensible layout. Rotational (talk) 05:47, 26 November 2008 (UTC)

I think Kotniski's point is that the majority of readers are not registered, so they can't set preferences.--Philcha (talk) 10:28, 27 November 2008 (UTC)

It's the unregistered readers I'm talking about - they can't set their preferences, but WP can by upping the default size of thumbs to 300px Rotational (talk) 13:48, 28 November 2008 (UTC)

Almost none of our readers are logged in experienced Wikipedians who have modified their preferences. We need to cater to the general reader - not to ourselves. Setting almost every image to a width of 180px is detrimental for the reader. There's nothing wrong with specifying the width that seems most appropriate in each case - we don't need to stick the general reader with a standardized width of 180px just so that a few veteran Wikipedians can tinker with their display settings. There isn't a choice between a) "forced size" and b) "flexible size" - there's a choice between a) 180px and b) editors' judgment on how big a given image should be in a given context. I'll take the editors' judgment every time. Haukur (talk) 14:26, 27 November 2008 (UTC)

I agree with Haukurth. The aesthetics of an article are as important as the text. We shouldn't be interfering with the judgment of the editors. In any event, guidelines and policy should describe best practice, and the best editors don't (that I've ever seen) pepper their articles with thumbnails, especially not the lead. SlimVirgin ^talk|edits 22:46, 27 November 2008 (UTC)

Unfortunately almost no editors can be aware of the relative frequency of the range of screens and settings used by readers, and naturally tend to chose sizes they think appropriate for their own set-up and tastes. A default of unforced images has been policy for a good while, but like most people here, I think lead images should be one of the explicit exceptions, and the default size should be increased. Johnbod (talk) 02:54, 28 November 2008 (UTC)

The typical editor has a typical set-up, that's pretty much a tautology. Those who don't have a typical set-up are usually aware of that fact. And, this being a wiki, if an inappropriate size is chosen then another editor can come along and make things better. I'd much rather have sizes chosen by editors in each case than a roundly inappropriate 180px across the board. Sure, choosing the best size, taking everything into consideration, can sometimes be difficult. But that's no reason to decide that we shouldn't even try. If not setting a size would result in images automatically adjusting to the set-ups and tastes of each reader, that would be excellent. I would be very much in favor of that. But that's not an option we have - not setting a size just means setting the size to 180px for almost all of our readers almost all of the time. Haukur (talk) 16:47, 28 November 2008 (UTC)

Indeed, and that's what we want - almost all the time, the default size should be used. The problem with "editorial judgment" is that most editors have no criteria at all by which to judge image sizes, apart from what happens to look good on their particular monitor with their particular browser settings. So in most cases, we really should simply not try. I think that increasing the default to 240 px would be reasonable, but having every editor make a more-or-less random decision about which size they happen to prefer isn't reasonable. — Carl (CBM · talk) 03:30, 30 November 2008 (UTC)

It just isn't that dependent on monitors and browsers. The editor primarily needs to answer questions like: can I comfortably make out the relevant part of the image at this size? If not, it probably needs to be bigger. Of course not everyone agrees on the best size in every case but that's just the way we bake the cake here on the wiki. I would probably support a bigger default but that doesn't look like it's on the horizon. Until something changes we need to work with the system we have. Haukur (talk) 20:29, 1 December 2008 (UTC)

This section needs a break; replying to Carl at the bottom of the next subsection. - Dan Dank55 (send/receive) 13:48, 30 November 2008 (UTC)

Scaling factor instead of fixed size?

Noting that there is now the ability to use a scaling factor such as by adding the parameter upright=2 to an image, should we perhaps be moving to recommend this method of making the lead image (and indeed any image) larger? This would scale to a multiple of the default for most readers, or of the logged in editor's preference. As long as the factor is kept reasoable, this seems like a much better compromise than arguing incessantly about whether or not to use absolute fixed sizes. --AliceJMarkham (talk) 15:08, 28 November 2008 (UTC)

If the scaling factor isn't buggy, I think you've got a good argument that it's desirable; bigger screens would want bigger images, and smaller screens would want smaller. Anyone see a problem? - Dan Dank55 (send/receive) 22:51, 28 November 2008 (UTC)

The top image is "thumb", the middle is forced at 180px, the bottom is forced at 300px. Top and bottom images appear the same on my screen because of my settings. Aren't scale factors just putting another name to the same problem. Rotational (talk) 13:54, 29 November 2008 (UTC)

No, a scaling factor is a multiplier times the size it would be if no size was specified. Hence, if like me you have your pref set to 300px, an image scaled for a factor of 2 would display at 600px wide for you. However, for someone with the default size of 180, that same image would display at 360px. I think that a scaling factor of 2 would probably be a bit too big in most cases, and that a size in range of 1.5 to 1.8 would probably work best in most circumstances. --AliceJMarkham (talk) 09:06, 30 November 2008 (UTC)

I've now added a 4th image that is scaled using upright=1.5. The size of this image will vary in proportion to the configuration that you're using. Comments? --AliceJMarkham (talk) 09:17, 30 November 2008 (UTC)

I support all of what Carl/CBM said just above, and as usual with Carl, I can't think of any way to improve on what he just said, either by subtracting or adding. I also support recommending scaling factors in the lead section and in most (but not all) cases rather than fixed sizes. What the scaling factor should be for the lead section depends on what our default size is; I agree that 180/140 is too small, and it's harming the appeal of the encyclopedia every day it remains, because editors are making layout decisions based on that look that won't be valid when it's increased. - Dan Dank55 (send/receive) 13:48, 30 November 2008 (UTC)

This still feeds on the fallacy that size preferences are widespread and important. Even if I pretend for a moment that they are I think most people who've set their preferences to 300px haven't done so because they have limited vision but because the 180px size is just too small for the typical eyes and the typical image. Anyway, scaling factors would still be a step up from the one-size-fits-all 180px system and if this more convoluted way of achieving what amounts to the same thing is somehow more acceptable then I suppose that's all right by me. Haukur (talk) 20:49, 1 December 2008 (UTC)

I'm concerned that scaling factors might work as well as most compromise solutions, i.e. not very well. In the short term editors will use it to make images the sizes they want them to be - and I've already mentioned enough cases where I think that's necessary. But if some unknown person changes the default size, articles get broken - lead images that needed to be larger than standard size suddenly become humungous, images further down pile up and create a chequerboard effect, etc. --Philcha (talk) 21:27, 1 December 2008 (UTC)

Excellent point, Philcha. If we do decide we want to recommend scaling rather than fixed pixel width, say for image in lead sections, in WP:MOS, then we'd better have a really good argument that we know what the default thumb size will be from now on. - Dan Dank55 (send/receive) 02:46, 2 December 2008 (UTC)

That's a reason not to use the proportional scaling, then. We don't want to set up a system where it would become impractical to ever change the default size. Especially because monitors are likely to continue getting better over the next few years. — Carl (CBM · talk) 02:53, 2 December 2008 (UTC)

Damn. Okay, we'll need something special from the devs for people with smaller screens, then; we don't want to be forcing 300px images on them if they don't want them, and they should probably have some better alternative than simply turning off the display of all images. - Dan Dank55 (send/receive) 17:29, 3 December 2008 (UTC)

A wicked problem

It's no wonder that we continue to go in circles with this issue, which fits the formal definition of a wicked problem — that is, one that can never be fully solved.

For instance, if Wikipedia were a conventional website, the problem might well be solved by appealing to principles of web design, where concern for usability and user preferences are paramount. But in this case the users are also editors, hence producers of content whose concerns and preferences also matter, and will be asserted in any case, no matter how we feel about it. Editors' concerns and preferences can clash with those of other editors, and those of non-editing readers, in unpredictable and contradictory ways. Therefore, by "solving" the problem from one perspective, we often create new problems. Changes beyond our control (e.g. to Wikimedia code, browser technology, user opinion) also introduce new problems.

This implies that the best we can do is to take actions that make things relatively better, or that seem to cause more benefit than trouble. Of course we won't agree on what those actions are. — ℜob C. alias ᴀʟᴀʀoʙ 19:13, 12 December 2008 (UTC)

Small ranges of relatively large numbers

I see this sort of thing a lot, stuff like:

pg. 1011–2 to actually mean "pages 1011–1012", and presumably not "page 1011, paragraph 2" or somesuch.
Until 1910–12 which could mean "until some uncertain end between 1910 and 1912" or "until December of 1910".
elevations of 1200–700 m. could mean "from 1200 m. up to 1700 m." or "from 1200 m. down to 700 m." depending on whether the sentence flows up-hill or down-hill, i.e. whether we are talking about mountain-climbing or stream hydrology.

Really, wouldn't it be better to use complete numbers in all cases? — CharlotteWebb 12:39, 30 November 2008 (UTC)

Agreed that all 3 of your examples have undesirable ambiguity. - Dan Dank55 (send/receive) 13:51, 30 November 2008 (UTC)

Single-digit closing year-ranges are against the recommendation at MOSNUM. Try "pp. 1011–12", or if you must, the full "pp. 1011–1012". "Until 1910–12" might be OK if three-year ranges are at issue (three-year agreements from 1903–06 until 1910–12"), but otherwise, best to stay away from the jarring of spelt-out lead word then a symbol for "to"; yes, render the whole thing fully. MOS's advice on en dashes says to avoid constructions such as "of 1200–1700 m" (use "to"); a reverse range in running prose is probably confusing unless the reader is warmed up to it; render in full. Tony (talk) 14:14, 30 November 2008 (UTC)

otherwise, best to stay away from the jarring of spelt-out lead word then a symbol for "to" ← I've read this again and I don't know what you're saying.

I did look at MOSNUM and abbreviating the second year to two digits is okay if they are in the same century. I would prefer that we avoid altogether except in tables where there is a compelling reason to save space, or at least avoid them where there they can be potentially confused with the YYYY-MM of ISO format.

However it doesn't say anything about page numbers in ref templates, or other units of measurement. I suppose I will continue changing these to complete numbers wherever I see them. — CharlotteWebb 16:19, 30 November 2008 (UTC)

Indeed, I didn't express myself very well. "... from 1998–2009" is better as "... from 1998 to 2009"; one preposition spelled out and one rendered as a symbol can be a little jarring. Tony (talk) 08:52, 6 December 2008 (UTC)

Here is a real-life example of two-digit numbers being used after a year to variously imply a month in some cases or an end-year in other cases, in the same section [2]! Wouldn't it be nice if we had some way to semantically distinguish date elements? Err, wait a second… — CharlotteWebb 19:26, 30 November 2008 (UTC)

Really, now; do we have to invent a whole new English language? Yes, I suppose we do; that 1910–12 (please note the useless endash) is English usage, and clear enough in context (or how was it emended) seem to be irrelevant. Septentrionalis PMAnderson 21:12, 30 November 2008 (UTC)

If you'd read the above example you'd see where "2006-07" was used to mean 2007–2008, but 2007-10 was used on an adjacent line presumably to mean "October 2007" (as 2010 is in the future—if not for this, I would have sincerely misinterpreted it). Please try not to belittle a very real ambiguity. — CharlotteWebb 21:18, 30 November 2008 (UTC)

And if it was clear, that is not a problem; if it were not clear, it should have been tagged. MOS cannot remove ambiguity from English, which is a natural language. Bear has more than a dozen meanings, from the animals to the force of a spring to a kind of tent. This does not mean we should replace it, everywhere it means the animal, by Ursus sp. to be unambiguous; so here. If 2006-07 is clear in context, leave it alone as idiom; where it's ambiguous, fix it, as you would any other piece of ambiguous language. Septentrionalis PMAnderson 22:34, 30 November 2008 (UTC)

I got a question on my talk page about this matter. Note that I objected to "until 1910–12", not "1910–12". AP Stylebook and NYTM are silent. TCMOS, 6.83, doesn't mention the word "until" specifically, but they do say not to use a dash if the range is preceded by "from" or "between", and from the examples they give, I don't see them allowing any other prepositions if the dash is used in the range, either. For my part, I can't figure out what "until 1910–12" means ... until some unspecified or unknown time in that range? until December 1910? until the beginning of some "season" that lasted from 1910 to 1912? If someone asks me, that's my answer, but I'll be happy to ask around if you like; maybe the phrase has some consistent meaning that I've missed. - Dan Dank55 (send/receive) 02:17, 1 December 2008 (UTC)

As far as dates go, this has been endlessly discussed in the past, with firm concensus for allowing both 1910-12 and 1910-1912, but not 1910-2 nor 1913-4. Johnbod (talk) 04:34, 1 December 2008 (UTC)
- Yet for some reason I see articles with all three of those on a regular basis, which suggests that one form is no less confusing, on average, than the other two. The issue is not limited to years, nor should it be. I gave examples of how any of these shorthand number ranges could be ambiguous. — CharlotteWebb 19:39, 1 December 2008 (UTC)
Dank, I suppose "In 1953–57" is in a class of its own [3]? — CharlotteWebb 21:55, 1 December 2008 (UTC)

If someone wanted me to copyedit their article, I'd change it to "From 1953 to 1957", because that's what I'm used to seeing, and because AP Stylebook doesn't mention hyphens or dashes in this context. - Dan Dank55 (send/receive) 02:36, 2 December 2008 (UTC)

For the record, Charlotte Webb has misread the text of which she complains, which read 2007-10/2007 followed by 11/2007. 10/2007 means October 2007, but we could do worse than to emphasize that October 2007 is preferable to avoid the ambiguity. Septentrionalis PMAnderson 19:09, 12 December 2008 (UTC)

ALT text

A recent edit to WP:MOS suggests that every image should have ALT text, and gives a link to WP:ALT, which gives this simple example of alt text to explain an image to the visually impaired: "This animation shows the gradual transformation from the straight-chain form of glucose to its ring form. The straight chain form consists of four C H O H groups linked in a row, capped at the ends by an aldehyde group C O H and a methanol group C H 2 O H. To form the ring, the aldehyde group combines with the O H group of the next-to-last carbon at the other end, just before the methanol group." Any thoughts on this recommendation? - Dan Dank55 (send/receive) 20:43, 7 December 2008 (UTC)

Yes, I would appreciate feedback on this new addition to the MoS, which has been discussed at WP:ACCESS. That particular example is perhaps too complex to be a good teaching example, but I was pressed for time and couldn't find a better. I wanted a challenging example, such as an animation, and something from a Featured Article, to illustrate the need for ALT text. I chose the first example that came to my mind, and wrote the ALT text myself; I believe that anyone who knows the subject, such as Tim Vickers, will agree that it's accurate.

Wikipedians can be rightly proud of their FA's, but I hope that they'll also want to share them with the blind. Once we're agreed on that goal and the need for ALT text to fulfil it, several pages will need updating, as described on WP:ACCESS. Please feel free to pitch in! Thanks, Proteins (talk) 22:51, 7 December 2008 (UTC)

Please keep in mind that Wikipedians and the blind are not exclusive clubs. —Michael Z. 2008-12-07 23:37 z

I do, indeed! :) I spoke that way only for brevity. I'm indebted to Graham and I'm conscious of his many contributions to Wikipedia; and I also know that he is but one of larger community of visually-impaired Wikipedians. Proteins (talk) 04:17, 8 December 2008 (UTC)

I think this is a good idea. The aim of Wikipedia is to make information available to everyone. The example given is probably one of the more complex images, and yet it's still possible to come up with a good description in less than a thousand words. I used to do "recordings for the blind" and we routinely described all images, from artworks to charts. I can't see any downside to this, unless they become another issue for content disputes. With some topics almost any edit is disputed so that's not a sufficient reason to avoid this. ·:· Will Beback ·:· 23:45, 7 December 2008 (UTC)

Well, I'll cut to the chase, there are a number of problems. For each of these problems, even if you believe we'll always make the right decision at WP:FAC, what are the odds that every anon editor will always make the right decision?

Since we just boldly proposed this addition to the MoS, I of course don't expect that it should be accepted merely on our say-so. Still, I can't imagine that you disagree with the goal of making Wikipedia more accessible to the blind. Proteins (talk) 04:17, 8 December 2008 (UTC)

Per WP:V, one of our content policies, any editor can challenge any text, and if you can't provide a citation, which normally means an inline citation, in a reasonable period of time, then the material must be removed. Okay, I'm challenging your alt text. How will you provide an inline citation for it?

As you point out below, and people at WP:ACCESS have pointed out repeatedly, alt text does not supplant the caption or the article text. It is intended for those who cannot see the image, whereas the caption and the main text are intended for everyone. I'm no expert at writing alt text, so I don't doubt that this one can be vastly improved. Perhaps it's unfair to judge the entire class from my fumbling first attempts?

That said, the ALT text needs only to describe the image, not to interpret it. Therefore, sighted users should need only their eyes to judge whether the alt text agrees the image. I think you'll agree that the central four groups in the animation are indeed C H O H groups and that the two ends are C O H and C H 2 O H. If your eyes can follow the animation (it's rather quick), you'll also see that the C O H end connects with the group just before the C H 2 O H end, forming a ring from a straight chain. Perhaps someone may doubt that the C O H group is indeed an aldehyde group, but that's basic chemical nomenclature and easily checked. I suppose we could eliminate that or provide a reference, written out to be intelligible to screen readers. Proteins (talk) 04:17, 8 December 2008 (UTC)

How will we ensure that alt text complies with all our content policies, especially WP:OR since they're describing the image in their own words, when most editors and most readers won't notice that it's there? The whole "everyone can edit" thing comes apart when most editors won't see the text to edit it, and most of the ones who do see it will have trouble with wikimarkup they've never seen before.

As noted above, the alt text only describes the image, and should not allow much room for OR. I also hope to make the alt text visible to sighted Wikipedians as a hovering tooltip over images (see below). Proteins (talk) 04:17, 8 December 2008 (UTC)

Adding extra text to explain the salient points of every image is hard work. It involves judgment calls, and potential embarrassment, and editors might feel that it's work that they're not used to doing and that they never volunteered to do. It will increase general resistance to reading and following the style guidelines.

I see your point, but I'm not persuaded yet. As I'm sure you're aware, there are many stylistic guidelines imposed at FAC that are unfamiliar to most casual writers, even those who publish regularly in scholarly journals. Surely, some volunteers will rise to meet the need, just as they do to check sourcing or image licenses or em-dash spacing. Proteins (talk) 04:17, 8 December 2008 (UTC)

Alt text is going to be a vandal magnet. Only people who check the edit window, or right-click on the image and select "properties", or use a screen reader, will even know it's there. Long alt text is going to make the job of finding the hidden vandalism all that much harder.

Yes, I'm conscious of that. I've been busy lately, but I've been planning on asking simetrical to alter their code so that the hover tooltip shows the alt text instead of repeating the caption. Then sighted people will be able to read the alt text easily and compare it with its image. Proteins (talk) 04:17, 8 December 2008 (UTC)

A lot of ideas for hatnotes (at the top of articles) that could make articles more readable have been argued and rejected (various kinds of disclaimers, descriptions of the subject matter or reading level of the text, "prerequisite" articles to be read before this article, etc). The strongest argument, the one it usually comes down to, is that the extra information might be nice, but it only adds marginally to the article. The argument goes: the extra time spent arguing on how to "classify" an article could almost certainly be more productively spent writing another article. ALT text could cause the same problem, times 10 for an article with 10 images. Is that a happy elephant or an exploited elephant in the image? Is this image trying to show us what a typical college campus looks like, or is it highlighting the differences, or is it more about the students than the campus? Blah blah blah. Lots more to argue about.

I see the danger, and I truly dread starting fresh arguments that inhibit productivity on Wikipedia. However, the alt text need not interpret, only describe, which might forestall some controversies. I also feel strongly that we should not sacrifice the comprehension of blind readers because of the modest work involved in writing a few alt texts. Even the longest articles have rarely more than 10 images; if ten captions can be written to consensus, why not ten alt texts? I hope that alt text will become an easy and routine courtesy, once we work out best practices and people get into the habit of writing it. Proteins (talk) 04:17, 8 December 2008 (UTC)

"This is an image of a beautiful girl enjoying a bath. This is a sunset. These are happy graduating students." [Subtext: what a pity you can't see any of them. We just thought you'd want to know.] Obviously, you and I are going to be sensitive enough to try to stick to essential facts. Is every anonymous editor going to be so careful?

Forgive me, but none of us can assume responsibility for all offensive material on Wikipedia. There will always be inconsiderate people, but I expect that most people will recognize immaturities or accidental slights as such and move on. Proteins (talk) 04:17, 8 December 2008 (UTC)

Note that all FAs are required to follow all style guidelines, which means, if we add this to the style guidelines, consistency would require that we insert alt text for every image of every past and future FA. Who wants to volunteer?

I will volunteer, now that you ask so kindly. ;) So will Aquatique, as discussed already at WP:ACCESS. So will, perchance, the original FA authors, if we ask them nicely enough and explain the motivation. Experience suggests that Wikipedia does not lack for volunteers willing to help in a noble undertaking, quixotic though it be. You of course are also cordially invited to join in the fun. :) St. Crispin's Day and all that, Proteins (talk) 04:17, 8 December 2008 (UTC)

The impulse to try to make Wikipedia the same experience for people with screen readers as it is for readers is laudable, but is it really going to be the same experience?

This touches upon a philosophical question that I'm ill-equipped to address. ;) Speaking broadly, I feel that we should do our best, even if we cannot reach perfection. Proteins (talk) 04:17, 8 December 2008 (UTC)

For all these reasons, it's possible that alt text may be a short-lived experiment. If I'm wrong and we're going to have to live with it for a while, I'd suggest the following guideline: alt text should contain the minimum number of words necessary to make the caption and the page text make sense if you can't see the image. It should never repeat information either present in or implied by the captions or page text. - Dan Dank55 (send/receive) 00:30, 8 December 2008 (UTC)

As mentioned above and discussed already on WP:ACCESS, we agree completely with your proposed guideline, although we hadn't specified "minimum". If you would like to shorten my alt text for the animation, please feel free! :) I'm sure that we can learn best practices from each another's experience. Proteins (talk) 04:17, 8 December 2008 (UTC)

I am concerned that this will become "Oppose promotion, images have no alt text", when (as WP:ALT points out) some images should have no alt text, because the image is described in the article. Again, what is the useful alt text here? "Man in coat, velvet vest, and stickpin, black and white"? Septentrionalis PMAnderson 00:21, 9 December 2008 (UTC)

I don't think we've mentioned in this thread that you can see ALT text if you right-click on the image and select "properties" (in Firefox), and I substituted that above.

Okay, the recommendation in favor of ALT text is still in WP:MOS. Not sure what to do about this; it seems like a bad idea to me for the reasons above. Proteins is giving the counterarguments a good try. I'm not convinced, but even if I got 3 or 4 guys on my side, that wouldn't come close to constituting "consensus" on a change that has all the potential ramifications I mentioned above. I'll ask over at WP:Image use policy and WP:VPP for more eyeballs on this thread. - Dan Dank55 (send/receive) 16:15, 10 December 2008 (UTC)

P.S. I see "images need not have alt text; a good test is whether anything significant would be lost for sighted readers if the image were blanked." This doesn't fix the problem; people are likely to make the case that if the image didn't add anything, it wouldn't be there, so alt text is required. In fact, this gives us one more thing to argue over for each image: whether alt text is required. - Dan Dank55 (send/receive) 16:37, 10 December 2008 (UTC)

I am surprised that I didn't see W3C guidelines in Wikipedia guidelines. The comments here are not unique to Wikipedia. Accessibility is a transfer of pain from reader to author. Note that alt text does not just benefit the partially sighted, many editors read pages in text mode. There will always be a few images whose meaning is difficult to describe, but it doesn't mean that we shouldn't strive to supply alt text for the vast majority of images with a meaning that is easy to describe. Lightmouse (talk) 16:47, 10 December 2008 (UTC)

- Not a few: most pictures of particular people will be like Viscount Amberley above: the picture does add something to the article for sighted people (the appearance of the subject, insofar as it is not describable in words), but the name (and background, if any) will be in the caption or the text. Septentrionalis PMAnderson 16:58, 10 December 2008 (UTC)

It wouldn't "break the wiki" to adopt some kind of recommendation for Alt text, but I agree with Sept on this one, the problem with what I read from the supporters is that they're underestimating the amount of time that's going to get devoted to this. It's not just a few images. If we recommend Alt text, there will probably be an Alt text noticeboard, and people will become specialists in it, and we'll have FACs on hold waiting for someone to verify that the Alt text is not too much, not too little, not POV and not OR. - Dan Dank55 (send/receive) 17:37, 10 December 2008 (UTC)

I feel ill-qualified to champion this cause, being sighted myself and having ever written alt texts for only one article (Acid dissociation constant). But you shouldn't feel that we're rushing this or pushing it on people. Everyone at WP:ACCESS is busy with other things, and I sincerely believe we won't hector people over alt text. For example, the guidelines for alt text are still half-written, since everyone's taking it slowly. We'll work on adding alt text to the FA's article-by-article and if others choose to join us, then great! As noted above, I feel that alt text fits Wikipedia's mission and it's pretty light work for the reward of clarifying the article for people without access to the images. I see the possibility of controversy as you do, but honestly, if people are set on being rancorous, there are much more efficient ways of doing it. ;) If FA authors don't want to add alt text themselves — as Petergans did not for Acid dissociation constant — I feel sure that someone will take on the task, as I did. And if they don't immediately, oh well, we'll get to it someday. Here's hoping cool heads will prevail in such teapot tempests, Proteins (talk) 18:32, 10 December 2008 (UTC)

I'm sure you won't hector people; I've never seen WP:ACCESS do that. The problem with MOS, however, is that somebody will treat every sentence that can be taken as a command as a non-negotiable condition at FA or a justification for edit-warring across WP. Would you consider, until you are prepared to set up a alt-tagging WikiProject, phrasing this as "please consider" or "where possible"? Septentrionalis PMAnderson 19:28, 10 December 2008 (UTC)

Sure, how about "Editors are encouraged to alt text to all images for which it seems reasonable." I almost wrote "heartily encouraged" but that seems slightly POV. ;) Proteins (talk) 19:47, 10 December 2008 (UTC)

As one would expect for a new and complex topic, this thread covers several sub-topics that should be separated. Here goes, with my own suggestions for discussion ...

Should MOS require ALT text?

Not until all the sub-topics are resolved. --Philcha (talk)

I believe we may have reached consensus on that yesterday? The present MoS guidelines do not require alt text. No one wants to start arguments over alt text and no one wishes to prevent deserving articles from being recognized as Featured Articles. However, all editors are heartily encouraged to add alt text to images, so that the article is more accessible to those who can't see the images. Surely that's consistent with Wikipedia's mission? I agree that there might also be a clever technical solution that makes it easier for editors.

I encourage more people to offer their thoughts about the alt-text guideline, especially after they've tried their hand at writing alt text. As I wrote above, such experimentation, carried out over weeks and months, will help us to identify best practices and the most sensible guidelines. To help people imagine what the article would be like without access to the images, I changed my script so that it blanks images with no alt text. I or other people at WP:ACCESS will also be happy to offer advice on how to install screen readers so that you can experience first-hand what the alt text (or its absence) sounds like. Proteins (talk) 18:02, 11 December 2008 (UTC)

Are there situations where ALT text is unnecessary?

In some cases substantial ALT text is unnecessary, e.g. in the pic of Bertrand Russell's father the caption does it all. IIRC if there is no ALT text at least some screen-readers recite the image's URL, which is a waste of time, and in (X)HTML ALT="" is recommended . I suggest this should be the default, i.e. provided by the Image "tag" and not entered by editors. --Philcha (talk) 15:53, 11 December 2008 (UTC)

As you say, the default is now alt="", as noted in the guideline and on WP:ALT. Until October, the default for images (not equations) was alt=the caption, which led to the caption being read twice by screen readers. However, in images where the thumb or frame parameter is not set, the final parameter becomes the alt text. You can see examples of this on the Main page; the alt text of the captionless images is shown in the hovering tooltip.

Speaking for myself, I don't think that the caption describes the Viscount's image. I might say instead, "Antique black-and-white photograph of a young man with a thin moustache, sloping shoulders, and dark hair swept back over low-set ears. The young man is wearing a spread-collar shirt with a stick-pinned ascot tie under a high-buttoned, velvet waistcoat and a peaked-lapel jacket." Even though Septentrionalis and I disagree on the alt text, I don't think people are going to argue endlessly about alt text, which will be seen/heard by so few readers. There will be more important things to attend to. Proteins (talk) 18:19, 11 December 2008 (UTC)

How detailed should ALT text be?

It depends on the way the image is used in the article. Some images are just eye-candy, and the caption is enough. Others are quite technical diagrams or otherwise make a point, which should be summarised in the ALT text. --Philcha (talk) 15:53, 11 December 2008 (UTC)

Even in complex images I suggest ALT text should be as simple as possible and certainly no more detailed than the text the image illustrates. E.g. for the animation Image:4-Stroke-Engine.gif I'd suggest "Animation: the four-stroke combustion cycle". In this example the article Four-stroke engine gives a fair explanation, although it omits the valve timimg and the fact that ignition is triggered by a spark in petrol engines and compressive heating in diesel. In general the main text should fill any gap between the ALT text and what the image needs to show. --Philcha (talk) 15:53, 11 December 2008 (UTC)

Others may have a different opinion, but I think "Animation: the four-stroke combustion cycle" is inadequate as alt text for Image:4-Stroke-Engine.gif. Although it says what the image depicts, it doesn't describe the image. I agree that the alt text needn't be a paragraphs-long description worthy of a 19th-century novel; perhaps there's something between that extreme and nothing? Proteins (talk) 18:26, 11 December 2008 (UTC)

Verifiability of ALT text

There are two perfectly adequate places for citations: the caption and the Image description page. I often use both. --Philcha (talk) 15:53, 11 December 2008 (UTC)

A vandal magnet?

Changes to ALT text should show up in watchlists and diffs. --Philcha (talk) 15:53, 11 December 2008 (UTC)

They will show up in diffs for the page; the alt text is part of the image description, which is just more text to the diff generator. Septentrionalis PMAnderson 17:27, 11 December 2008 (UTC)

A new burden for editors

Simple - the criterion for ALT text should be that it's an intelligible but concise summary, and MOS should not apply to the content of ALT text. The content should be enough to make it clear why the image is considered worth including, without necessarily using anything like these words. --Philcha (talk) 15:53, 11 December 2008 (UTC)

Whether it's tactful

Even raising this issue is political correctness providing its own reductio ad absurdum. If the point of an image is that's accepted as beautiful, the ALT text should say the subject is beautiful. --Philcha (talk) 15:53, 11 December 2008 (UTC)

It's "political correctness" when we try to be more inclusive by making assumptions what the blind want rather than asking them. WP:ACCESS has been largely successful exactly because we've been respectful to the opinions of people who rely on screen readers. (Agreed with the above point that this isn't just for the blind, it's also for people who need much larger fonts, and also for people who are using screen readers to listen to Wikipedia because they're commuting or jogging or whatever.) What I'm saying is: don't assume that the information that you would miss if you couldn't see the image (if you're sighted) is exactly what the blind would want to know; ask at WP:ACCESS. - Dan Dank55 (send/receive) 17:24, 12 December 2008 (UTC)

Some people with perfectly good vision view pages in text-only mode. This is useful when data download is constrained/expensive or screen space is limited. Think of viewing Wikipedia on a small portable device when paying expensive data roaming charges in a foreign country - many people turn images off and only turn it on and redownload the page for worthwhile images - good alt text will give a clue to that. Lightmouse (talk) 00:53, 13 December 2008 (UTC)

Capitals

I have attempted to move a number of articles which incorrectly use capitals for legal designations. I have provided the primary statutory sources which created the designations, but a group of editors is opposing the move, citing sources which (incorrectly) use capitals as more relevant than the actual legislation which created the designations (without capitals). Can anyone help? See:

*(Special Protection Area, Area of Conservation, Area of Outstanding Natural Beauty, Site of Special Scientific Interest, and Special Area of Conservation?) Mooretwin (talk) 09:51, 12 December 2008 (UTC) Forget it. Mooretwin (talk) 16:13, 12 December 2008 (UTC)

As discussed on those talk pages, the terms used are capitalised in official use to show that an area or site has been designated by an official body under the UK legislation as worthy of statutory protection, because of its landscape value, scientific importance, etc. The articles concerned relate to those areas which have been so designated, not to other areas in a wider sense. Mooretwin is simply in error in stating that capitalisation in these cases is "~~wrong~~ incorrect". An example of the official usage - that is, the usage now employed by the organisation responsible for administering one such area - is here, which unequivocally states: "The correct title of ‘Shropshire Hills Area of Outstanding Natural Beauty’ should be used. The words have capital letters as it is an official designation." Ghmyrtle (talk) 10:06, 12 December 2008 (UTC)

See WP:MOSTRADE; we are not bound by official styles. If, as has been claimed in the discussion, UK legislation never capitalizes, its usage does not determine between uc and lc. (Presumably for lack of ambiguity, just as legal forms did not use to employ punctuation.) Septentrionalis PMAnderson 15:00, 12 December 2008 (UTC)

The original message to this section seems to violate WP:CANVASS as it was posted after a number of RMs were proposed by Mooretwin. DDStretch (talk) 15:19, 12 December 2008 (UTC)

You win. Congratulations. Mooretwin (talk) 16:14, 12 December 2008 (UTC)

Boldface in film articles' cast lists?

Hello, I was wondering if boldface was permissible for the names of actors and their roles in a film article's "Cast" section. I'm a coordinator at WikiProject Films, and from what I can tell, we have used boldface for cast lists where there are multiple lines for a bulleted entry (to make them stand out). Lately, I've questioned this approach since it did not seem in compliance with MOS:BOLD. Another coordinator referenced an older discussion here during which this revision of David E. Kelley seemed to permit this. However, some time since then, it was related by Proof as an example of a definitional list. Since definitional lists seem more for disambiguation pages and their setups, I was not sure if they could be applied to cast lists. Can it be clarified whether or not boldface would be acceptable for cast lists? We have generally pursued the rule of thumb in only applying boldface when there are multiple lines. When there is a simple list (just the actor and the role), we avoid boldface since, well, it's a whole lot of unnecessary bold. The approach is imperfect, though, since some bulleted entries may have multiple lines, while others will not. Here is an example... Tropic Thunder with boldface and without boldface. Also, a tangential question -- is it even applicable in prose where there are paragraphs tied to a character's name? See The Dark Knight (film)#Cast and characters as an example. I am just trying to understand if MOS:BOLD's scope is explicitly limited or if there is some permitting boldface in circumstances like with cast lists. —Erik (talk • contrib) 00:24, 11 December 2008 (UTC)

If there is consensus to do so, then it's permissible. WP:POL prefers that guidelines arise from actual good practice. There may be some argument that this is not good practice, but I find the boldface much easier to read. We should acknowledge and document this here. Septentrionalis PMAnderson 02:38, 11 December 2008 (UTC)

Yes, I was hoping to see what kind of consensus could exist. I ask in part because our guidelines have mentioned boldface use for some time, and I've been working on updating the guidelines, including a draft for the "Cast" section. I just wanted to clarify what kind of formatting would be acceptable for presenting the actors and roles under the MOS. —Erik (talk • contrib) 15:59, 11 December 2008 (UTC)

If every name is linked, as there, the boldface is less necessary. Tony (talk) 10:08, 13 December 2008 (UTC)

But as in the example of Tropic Thunder, the characters don't have articles to link (and shouldn't); they should be boldened for comprehensibility - especially when the actors' names are marked by bluelinking. Septentrionalis PMAnderson 20:54, 13 December 2008 (UTC)

Double digit numbers

I've been real curious about the numbers for awhile. Although most people like to use actual numbers after the first few (ex. nine, then 15 and 18), I actually prefer to use the spelled out versions all the way up to twenty (ex. nine, fifteen, eighteen, and even twenty). Of course I know that it should be 3.5 million, 121 lbs, etc. But what about when you have a number that's flat even but it's pretty big? Like how would I go about formatting forty or sixty? Should it be spelled out or should it be written 40 or 60? Everything over a hundred, I just write it out as the actual number (ex. 200, 400). I'm sorry if this was a confusing question or has been asked before. ★Dasani★ 02:29, 13 December 2008 (UTC)

You can do it either way. See Wikipedia:MOSNUM#Numbers_as_figures_or_words. (As a rule of thumb, when all other things are equal, I use words when they are one or two syllables, and digits otherwise. But the most important thing is being consistent in each context: 5 cats and 32 dogs or five cats and thirty-two dogs, not five cats and 32 dogs.) -- Army1987 – Deeds, not words. 02:38, 13 December 2008 (UTC)

Leadership. Nice. It’s rather amazing (since I am deep in the atomic-level details of this highlighting technique), but the first time I read your example, I didn’t even notice the formatting technique used to differentiate the example text; I noticed only the point of your post and the thoughts being conveyed and didn’t get sidetracked by the distractions of a method used to convey those thoughts. I’m quite convinced this is an excellent way to accomplish this. Greg L (talk) 02:49, 13 December 2008 (UTC)
- I hope we will not use the white background; on a yellow page, it makes the text much harder on my eyes. Septentrionalis PMAnderson 04:14, 13 December 2008 (UTC)

Leadership, Greg? I suspect that is as dangerous a notion around here as authority, or prescription, or guidance. But this is not the thread for explaining such things. Please contain such talk in Respect for MOS and its editors, above. Still not resolved.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 04:50, 13 December 2008 (UTC)

You lost me, Noetica. I suspect your above post is somewhere in the facetious/irony/jest/frustration flight envelope, but I’m not sure. Greg L (talk) 05:42, 13 December 2008 (UTC)

Ah no, Greg. More towards the serious. Some editors here are spooked by ideas like leadership, authority, prescription (as opposed to flaccid description). They really are! The ensuing lack of clarity and common purpose impedes the work of MOS, as it would impede the work of any guide to style. See the long thread I refer you to.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 06:09, 13 December 2008 (UTC)

That’s their problem. If I believe something to be the truth, I say it. In the dictionary under Responds to social pressure:, it says Don’t even bother considering Greg L. Army1987, above, used an example of the new proposal to highlight some text. Doing so was a fine way to show off the new highlighting method. It was a paradigm technique for Wikipedians to lead by example to affect change for the good. Greg L (talk) 06:30, 13 December 2008 (UTC)

To PMAnderson and Army1987: A) do we have to prescribe a background color? And B) If we want to use this on MOS and MOSNUM, do we really have to worry about non-white backgrounds? Answering my own question to “B,” I suppose if one wanted to use maroon-highlighted text in a {{quotation}} template or within a green-div box, specifying a background color would be as ugly-ass as Bruno Magli shoes on O.J. So, Army1987: can we de-specify (leave transparent) the backround in your <span> technique? Greg L (talk) 05:49, 13 December 2008 (UTC)

Greg, these points, including this reponse, should be moved into the thread above with the RfC, right? Background colour needs to be considered. In my opinion, yes: keep it transparent. People have different local setups affecting background and the like; and in any case three attribute changes is going a bit far. WP already uses two attributes changes for links: colour and underlining. I like underlining, as I explained in the discussion at WT:MOSNUM; but since links use it already, I guess it should be set aside as an option for our purposes.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 06:09, 13 December 2008 (UTC)

Go ahead and move this color stuff then. Greg L (talk) 06:30, 13 December 2008 (UTC)

I left this thread intact, but added a note with relevant points in the RfC thread above.–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 07:36, 13 December 2008 (UTC)

Editors: Remarks about example text are valuable in the thread above, which involves an RfC. Please contribute such analysis there.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 07:36, 13 December 2008 (UTC)

Hyphens after -ly adverbs (rationalised section)

I've moved the content of the two misleadingly titled sections just above down here, so that the topic is treated in one place. Below the existing material, I've posted a more detailed explanation of why the long-standing text should not be anodysed as Anderson has done twice in the past day.

In regard to the question of using a hyphen after an -ly adverb, User:Pmanderson has just informed me that "the question of English usage ... differs between the national varieties of English." The same editor has also changed the wording of Wikipedia:Manual of Style#Hyphens, section 3, point 4. I looked for more information about the matter on this talk page, but found none. If those differences exist, what specifically are they?
-- Wavelength (talk) 05:18, 11 November 2008 (UTC)

Reverted; thanks for alerting us, Wavelength. The reason for the rule is that an -ly adverb clearly flags that it will quality a subsequent verb (usually immediately after). Tony (talk) 14:27, 11 November 2008 (UTC)

This is why I quote "specially-designed sound cards", where the adverb is not qualifying a verb, but an adjective (here a participle, but it could be any adjective). Septentrionalis PMAnderson 22:38, 13 November 2008 (UTC)

A participle is not an adjective. Ilkali (talk) 00:42, 14 November 2008 (UTC)

'Tis where I come from, but is there anything more here than a purely verbal argument? Septentrionalis PMAnderson 16:07, 14 November 2008 (UTC)

The comment I replied to was predicated on classification of participles as adjectives, which is contrary to mainstream linguistic theory. If the comment in question had a point (and I can't see one), it is defeated by that fact. Ilkali (talk) 17:13, 14 November 2008 (UTC)

Citation please. Septentrionalis PMAnderson 23:34, 14 November 2008 (UTC)

The old wording encouraged, indeed mandated, the dehyphenation of compound adjectives, such as specially-designed sound cards. These are being "corrected" by a single-purpose account, who gives little scope for the use of the hyphen here in avoiding ambiguity, and none to the difference in this respect between American and British English.

He is also finding his instances by searching on ly-. This is a recipe for bad writing; although I have to admit it could be worse; he could be using a bot.

I think nothing more is needed here than a toning down; and have done so. Septentrionalis PMAnderson 23:48, 11 November 2008 (UTC)

Please see User:David Kernow/Internet sources re hyphens and adverbs,

which includes American, British, and Canadian sources.

(By the way, mine is not a single-purpose account, but that is irrelevant anyway.)

If "specially-designed" is a compound adjective,

then "the-quick-brown-fox" is a compound noun,

and "jumps-over" is a compound verb,

and "the-lazy-dog" is another compound noun.

If "specially designed sound cards" is ambiguous,

then a disambiguation here would be helpful,

because I can see only one possible interpretation.

-- Wavelength (talk) 22:49, 12 November 2008 (UTC)

Response by Tony1: I've reverted Anderson's second attempt to dilute into meaninglessness a useful guideline in MOS that I believe has improved the readability of WP's article text. I've received a substantial email on the matter from User:Noetica, who is on an extended Wikibreak, but noticed Anderson's edits. Noetica writes:

"I subscribe to the Chicago Manual of Style online (recommended). Here is the Chicago ruling:

5.93 There are exceptions for hyphenating phrasal adjectives: (1) If the phrasal adjective follows a verb, it is usually unhyphenated—for example, compare a well-trained athlete with an athlete who is well trained. (2) When a proper name begins a phrasal adjective, the name is not hyphenated {the Monty Python school of comedy}. (3) A two-word phrasal adjective that begins with an adverb ending in ly is not hyphenated {a sharply worded reprimand} (but a not-so-sharply-worded reprimand).

Absolutely definitive, confidently prescriptive, and in accord with our long-standing text. Other guides issue much the same edict."

Anderson, please stop trying to impose your anti-centralist, do-as-you-please agenda on our style guides by anodyning its guidance bit by bit.Tony (talk) 08:10, 12 November 2008 (UTC)

In general, CMOS is a hasty piece of shorthand for writers and editors in a hurry. We are not; we can afford to be prescriptive only when English usage is without exception. (And in fact, we omit most such points, because we don't need to specify them; when English usage is genuinely without exception, it's not controversial here also.) Septentrionalis PMAnderson 17:25, 12 November 2008 (UTC)

I deeply regret and oppose the attitude that guidelines must be prescriptive; they can simply guide, as the other clauses in the same section do. Septentrionalis PMAnderson 18:28, 12 November 2008 (UTC)

- I'm beginning to think Tony and PMA are using different definitions of "prescriptive". Perhaps they'd care to elaborate their intended meanings for the term in the above, because to my reading the CMOS quote above is more descriptive than prescriptive. LeadSongDog (talk) 20:06, 12 November 2008 (UTC)

Also, hyphenation is here, as elsewhere, more common in British than American. CMOS is a guide to American English, and does not have WP:ENGVAR. I will try another draft. Septentrionalis PMAnderson 22:35, 13 November 2008 (UTC)

[Please do not insert comments within this post, but respond after it.-N]

PMAnderson, you flagged the guideline concerning ly-adverb modifiers as disputed. The way you did it disrupted the markup and the numeration. If you re-apply a dispute tag, check the effects of doing so. Otherwise you gratuitously blemish a page consulted by well over a thousand people a day. While you do such checking, take the time to read the section you are tinkering with. The section on hyphens includes the following proviso:

Hyphenation involves many subtleties that cannot be covered here; the rules and examples presented above illustrate the broad principles that inform current usage.

I have not joined in discussion here for ages. Nor have I edited the page for many months, though I have been one of its most prolific editors, if the statistics are to be believed. The reason? I thought progress was impossible beyond a certain level, because editors like you, PMA, stood doggedly against MOS as a genuine corpus of guidelines. You just don't seem to be able to separate the issues. Every style guide must guide. Every style guide adopts a prescriptive tone. That is its virtue, and its reason for being. That is what people value in it, and why we get our thousand hits a day (even after allowing for the activity of MOS editors themselves). We disappoint those who come here for guidance, if we offer only bland nullities. Don't tell them this, for example: "The hyphen may be omitted after an -ly adverb"! That will license text like this: "we had a really-good time". Tell them what all major guides tell them. That's what the current text does perfectly well:

A hyphen is not used after an -ly adverb (wholly owned subsidiary) unless part of a larger compound (a slowly-but-surely strategy).

And don't purvey such unsupported conceits as this, to a readership of one thousand a day:

A hyphen is generally not used after an -ly adverb (wholly owned subsidiary) unless part of a larger compound (a slowly-but-surely strategy); sometimes, especially in British English, a hyphen may be used to clarify exactly what the adverb modifies.

Almost worthless. Can you even cite any authority for that assertion about British English? I have surveyed many style guides, and not one says anything like that. I don't approve of everything CMOS says, but the guides are pretty well unanimous concerning ly-adverbs and hyphens. Find one that contradicts our current guideline. In fact, ours is more complete than most, since it includes the most important kind of exception (a slowly-but-surely strategy). There are other exceptions sanctioned implicitly by some authorities; I know about them, through my extensive investigations. But you have not identified them, and you have not brought to this discussion any respected guide that so much as hints at them. Can you do that? I doubt it! And even if you could, such a source would have to be weighed against the united voice of the American CMOS, the British New Hart's Rules, and the other guides that I could cite.

If this matter did not affect the convenience of one thousand readers a day, I would have simply stayed away in graceful retirement, PMAnderson. Ever tried graceful retirement? Think it over. I thoroughly recommend it, for the new perspective and tranquillity it affords on old issues one has cared too closely about. But I cannot stand by, leaving Tony to struggle alone against such unreflective and obtuse depredations.

MOS cannot function at its best if it is not allowed to guide. It must not be a draconian imposition on editors, and it is not; but to guide, it must answer the questions that people bring to it. The thousand a day are looking for a concrete prescription, not the vapid deliverances of Desiderata.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 00:21, 14 November 2008 (UTC)

Yes; it is part of Fowler's discussion of hyphens. (He does not prescribe for American.) The old wording, sans exceptions, is being used to justify bot-like removal of hyphens in this construct, even when they are grammatical and do add to clarity, despite the catchall sentence. When moderate wording indicating that hyphens can occur in this construction, although they normally do not, is agreed upon, the tag certainly can and should go. Septentrionalis PMAnderson 00:28, 14 November 2008 (UTC)

I do not understand the claim that the tag disrupts the numbering. On my system, the only things actually numbered in WP:HYPHEN are the three main subsections, and they are numbered 1, 2, and 3. Septentrionalis PMAnderson 00:31, 14 November 2008 (UTC)
I have stated the example that annoyed me, and it was not "really-good time". I am very tired of having red herrings dragged across MOS to defend it every time it says something stupid. Septentrionalis PMAnderson 00:36, 14 November 2008 (UTC)

PMA, the tag is designed to be placed at the top of a section, as its wording suggests. With some browsers (like recent versions of Mozilla Firefox) it interferes with the display of subsequent text if it is used irregularly. As for a really-good time, it is no red herring. You continue to avoid the issue. The wording you had proposed does not rule out such errors; the wording you attempted to replace did rule them out. That is why our MOS, along with every other major style guide, gives such a guideline. As for Fowler's, note first that it is not a style guide. Then I must ask: Which edition are you talking about? What does it say, exactly? Why should the fact that it typically does not address American English warrant our saying, to well over one thousand readers a day, anything at all about differences across the Atlantic concerning these compound ly-adverb premodifiers?

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 01:15, 14 November 2008 (UTC)

Yes, the old language ruled out much that was undesirable, it also ruled out much that was desirable; that's what was wrong with it. See Type I and Type II errors. Septentrionalis PMAnderson 06:33, 14 November 2008 (UTC)

I personally find adverbly-adjective constructions acceptable, but I recognise the value of conventions and adamantly object to Pmanderson's constant attempts to erode the MoS. In this case, his proposed alternative is no better than having no text at all. I am in favor of the original wording. Ilkali (talk) 00:41, 14 November 2008 (UTC)

Then you agree on the point at issue. Why should we insist on something which is not English? Septentrionalis PMAnderson 01:12, 14 November 2008 (UTC)

If I may presume to give the obvious answer on Ilkali's behalf, he prefers this uniformity (which is a perfectly serviceable approximation for all purposes) to insisting on his own preferences. He works toward the greater good. (Many of us do that, PMA. Take note.) In this case, the greater good is a style guide that actually guides. I don't know that Ilkali accepts a really-good time and the like; but he seems to want to give hundreds of thousands of Wikipedians each year access to a ruling that will keep us from becoming a laughing stock.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 01:22, 14 November 2008 (UTC)

I don't, of course, accept a really-good time. What I should've said is that I generally find it acceptable to hyphenate -ly adverbs with past participles, but there are probably some exceptions. Likewise, there may be instances of hyphenating with simple adjectives that I find acceptable, despite really-good being aberrant. Without some way to concisely articulate my intuitions, and without any indication that others share those intuitions, I'm happy to accept a simple rule that sometimes excludes constructions I would accept, but reliably excludes ones we'd all reject. Ilkali (talk) 03:44, 14 November 2008 (UTC)

[Further amendment: I am obviously in verbal mode today, not numerical. I have again amended all of my references to the number of people consulting WP:MOS. It is in fact around 2,000 per day. After allowing for hits from MOS-editors themselves, the conservative figure is still well over a thousand users per day seeking a ruling on style.–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 05:47, 14 November 2008 (UTC)]

That still offers a wide scope for interpretation. Indeed, I would hope that many looked to see what this page said, and had better sense than to treat what they found as a ruling; and of those who sought "rulings", perhaps half-a-dozen looked at any given sentence (perhaps one; we cannot tell how many passed through on their way to subpages). Your language shows what is wrong with MOS: we are not a race of kings. Septentrionalis PMAnderson 06:26, 14 November 2008 (UTC)

Please don't pretend we agree, PMan. As Noetica suggests, the difference between us is that I am willing to discard my personal sensibilities in support of something universally beneficial, such as the clarity and unequivocality of the text you want to replace. "Not English"? Absolute drivel. Is specially designed sound card "not English"? Ilkali (talk) 03:44, 14 November 2008 (UTC)

Very well, we do not agree. You speak as a disciple of Carlyle and Lord Shang: Nothing is useful, except a direct unconditional order. Actually describing the English language is of secondary importance, if any. Septentrionalis PMAnderson 06:22, 14 November 2008 (UTC)

At this point, I have tried four different wordings in a search for compromise. Is there any reason, other than this "we must speak prescriptively" argument, why we should not say, as the other clauses do, that hyphens are not normally used after -ly? Septentrionalis PMAnderson 16:07, 14 November 2008 (UTC)

Reasons have already been given. Do you accept that, for any hyphenated form excluded by the original guideline, the unhyphenated equivalent is legitimate English? Do you accept that your version of the guideline permits illegitimate English? Ilkali (talk) 17:10, 14 November 2008 (UTC)

Diff for the reasons, please. As for these questions: no, I do not accept that "specially designed sound cards" is good English; if sound were a true adjective, rather than an attributive, it would be worse still. Nor do I accept that "normally" permits illegitimate English, any more than the normally in the following clause: A hyphen is normally used when the adverb well precedes a participle used attributively (a well-meaning gesture; but normally a very well managed firm, since well itself is modified); and even predicatively, if well is necessary to, or alters, the sense of the adjective rather than simply intensifying it (the gesture was well-meaning, the child was well-behaved, but the floor was well polished). Septentrionalis PMAnderson 23:34, 14 November 2008 (UTC)

"no, I do not accept that "specially designed sound cards" is good English". Well, I don't know what to tell you other than that the purpose of the MoS isn't to license all of your personal preferences. Ilkali (talk) 03:14, 15 November 2008 (UTC)

Style guides are by nature prescriptive. If you wish to abolish the MoS, kindly approach the matter directly and overtly rather than trying to disassemble the document one piece at a time. Ilkali (talk) 17:10, 14 November 2008 (UTC)

I do not wish to abolish MOS; that is my next to last choice. But you are mistaken: style guides are not by nature prescriptive; they are descriptive, and advisory: They describe what English does, and recommend between the available choices, giving what reasons there may be for each. As an example, I quote the text for the second section of WP:HYPHEN.

There is a clear trend, not yet complete, to join both elements in all varieties of English (subsection, nonlinear), particularly in North America. British English tends to hyphenate when the letters brought into contact are the same (non-negotiable, sub-basement) or are vowels (pre-industrial), or where a word is uncommon (co-proposed, re-target) or may be misread (sub-era, not subera). North American English reflects the same factors, but tends strongly to close up without a hyphen when possible. Consult a good dictionary, and see WP:ENGVAR.

That offers facts and advice, which varies by national dialect. It makes no command, except the inarguable one to consult a dictionary, but it guides the reader all the same. If there were compilations of punctuation comparable to dictionaries, we would recommend that they be consulted for the rest of this paragraph. Septentrionalis PMAnderson 22:48, 14 November 2008 (UTC)

According to What Is Grammar - Descriptive and Prescriptive Grammar,

By their nature, all popular style and usage guides are prescriptive, though to varying degrees: some are fairly tolerant of deviations from standard English; others can be downright cranky. The most irascible critics are sometimes called "the Grammar Police."

-- Wavelength (talk) 23:43, 14 November 2008 (UTC)

By searching on the web, one may find almost any set of keywords; it helps, however, to read the text found and see what it actually says. This is addressing, as may be plainly seen, the opposite issue: whether a style guide should tolerate "deviations from standard English". That is not what is in question here: Ilkali and Wavelength are demanding that MOS forbid a usage that is standard English, and which is sometimes clearer and less ambiguous than the alternative. Septentrionalis PMAnderson 02:52, 15 November 2008 (UTC)

"style guides are not by nature prescriptive; they are descriptive, and advisory". Welcome to contradiction theatre, folks. Ilkali (talk) 03:14, 15 November 2008 (UTC)

You can find information about web pages linking to http://en.wikipedia.org/wiki/Wikipedia:Manual_of_Style,

at Site Explorer - Search Results.

-- Wavelength (talk) 07:23, 14 November 2008 (UTC)

Most of which appear to be WP pages, either subpages of MOS or pages with such tags as {{duplication}} or {{cleanup}}. Septentrionalis PMAnderson 15:46, 14 November 2008 (UTC)

Please see Site Explorer - Search Results "Except from this subdomain"

-- Wavelength (talk) 16:50, 14 November 2008 (UTC)

Does Wavelength look at search results before posting them? That list begins with a wikimedia user page, continues with Wedding, then on to es:Wikipedia:Manual de estilo. This at least has found some results which are not WP and its sister projects, mostly one Stephen Downes' blog, but is there a point here? (Aside from the question of whether {{rewrite}} should redirect here; the objections at Wedding are not MOS concerns.) Septentrionalis PMAnderson 03:10, 15 November 2008 (UTC)

Above the list of links, you can see the number of links. At the bottom right corner of the page, you can click to see subsequent pages of the list of links. The distinction here between Wikipedia pages and non-Wikipedia pages is of minor importance. Even links to the project page from other Wikipedia pages are an indication of the importance of the project page. The main point here is to show the importance of the page by the number of pages linking to it, and thereby to supplement what Noetica said about the number of times the project page has been viewed. -- Wavelength (talk) 07:11, 15 November 2008 (UTC)

This has gone on far too long. If Anderson were able to present any evidence for the claim that he is now reduced to (inclusion of the word normally), we might have settled things earlier. He has referred us only to Fowler's. Which version, I asked earlier? No response! Fowler's does not, despite Anderson's assertion, deliver anything relevant to ly-adverbs that would support his case. Since Anderson cannot do or refuses to do the necessary work, I will. Garner's Modern American Usage (Bryan A Garner, OUP, USA; 2nd edition 2003, 928 pages) is a meticulous and authoritatively prescriptive work, at ease with its role and enormously well received by a wide readership. Garner devotes seven dense columns of text to phrasal adjectives. His essential message? Use hyphens. He begins with a general rule:

A. General Rule. When a phrase functions as an adjective preceding the noun it modifies [...] the phrase should ordinarily be hyphenated. [...] Most professional writers know this; most nonprofessionals don't. (p. 604)

This is followed by the solid reasons for such a practice, and long lists of examples. Then this (with my underlining):

Some guides might suggest that you should make a case-by-case decision, based on whether a misreading is likely. You're better off with a flat rule (with a few exceptions noted below) because almost all sentences with unhyphenated phrasal adjectives will be misread by someone. (p. 605)

Then follow illustrations of what can happen when hyphens are omitted. And last, the exceptions referred to. Only one of these is relevant to the guideline we are discussing:

B. Exception for -ly Adverbs. When a phrasal adjective begins with an adverb ending in -ly, the convention is to drop the hyphen [...example...]. But if the -ly adverb is part of a longer phrase, then the hyphen is mandatory (the not-so-hotly-contested race).

That's it. Nothing more to be said that's relevant to our guideline. No qualification with normally, no sometimes, no British do this but we do that. After establishing the baseline assumption that hyphens are generally to be used in phrasal adjectives, and listing several dozen examples to reinforce that assumption, Garner finds only one class of phrasal adjectives with -ly that calls for a hyphen; and that is exactly the class that we mention. CMOS, New Hart's, and all other major style guides agree. If they mention any exception at all, they mention that same exception. Now, I am aware of further considerations, and certain very particular sorts of further exceptions that no style guide I have seen addresses. But I don't think they amount to much, and I have no more time to devote to this exercise. Anderson should now accept that the weight of almost all who write in this field is against him, and that he has no body of support here. I have removed the attenuated qualification with normally that he inserted, pre-empting discussion here. And I have removed the dispute tag, for such an ill-founded and one-sided assault on our plain, useful, and perfectly justified wording. Let him waste our time no longer, without bringing detailed argument and evidence to the table. This has been a serious imposition on my time, and on the patience and time of many others here.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 00:51, 16 November 2008 (UTC)

Dispute tags

I have once again removed a dispute tag that was improperly applied in the middle of the section. See my edit summary. The edit appears anonymous because of a local system glitch. (I am not at my usual location.)–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 22:11, 14 November 2008 (UTC)

OK, there is no question that there is a dispute in the section. If the dispute tag cannot be next to the disputed statement, it should be at the beginning of the section. — Arthur Rubin (talk) 15:04, 15 November 2008 (UTC)

Please no, it makes people think that the whole section is in dispute. Better to leave the tag out completely - it serves no purpose except to half-direct people to what independent observers will conclude, sadly, is a very lame discussion.--Kotniski (talk) 17:22, 15 November 2008 (UTC)

Nonsense. There is a legitimate dispute. If there's no way to tag just the disputed statement as disputed, the tag is necessary for the article to be a legitimate guideline. The alternative, I suppose, is to use the disputed guideline tag at the top; which we'd both like even less. — Arthur Rubin (talk) 18:12, 15 November 2008 (UTC)

We don't insert dispute tags every time someone disagrees with part of the article. Is there some reason the tag is appropriate for this issue but not any of the others being discussed on this page? Ilkali (talk) 18:53, 15 November 2008 (UTC)

Please see Wikipedia:Template messages/Disputes.

-- Wavelength (talk) 19:26, 15 November 2008 (UTC)

Or more relevantly perhaps, Wikipedia:Template messages/Wikipedia namespace.--Kotniski (talk) 19:42, 15 November 2008 (UTC)

Of course there is no reason to flag this particular section as disputed any more than the others. These disputes over dispute tags are becoming tiresome and disruptive - let's remember that guideline pages are not private battlegrounds, but are intended to be read by editors, who don't know what to make of a section that has a red question mark at the top (particularly misleading when the dispute is only over one insignificant detail of a large section; and the tag is added without even using the parameter that takes people to the relevant talk page discussion). Rather than continue battling the issue out here, I've tried to formulate some guidance for use of such tags, or rather to clarify and expand what was already written at WP:Policies and guidelines. Please comment on that talk page on my recent changes there. Of course I'm not saying anything I wrote there is applicable (anyone's perfectly entitled to slap a disputed tag on that), but I think it would currently imply that the correct solution in this situation is no tag or, if people really think it's useful, an {{underdiscussion}} tag, with the talk page discussion section noted.--Kotniski (talk) 19:53, 15 November 2008 (UTC)

I am sympathetic to both sides here, and the general question needs to be discussed elsewhere; so thanks to Kotniski. In the present case, the tag disrupts markup if placed exactly where it is relevant, but if placed at the beginning of the whole section its scope is misrepresented. Luckily, the dispute has now diminished to exclusion or inclusion of the single word normally. No user of WP:MOS will be misled or inconvenienced by the tag's absence; but some might be by its presence. Accordingly, I am removing the tag. Let anyone reinstating it give a justification that counters the reason I have just articulated.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 00:08, 16 November 2008 (UTC)

This discussion would not have begun if the absence of normally, or some such qualifying phrase, were not itself inconveniencing users of the encyclopedia. An irresponsible editor is taking its absence as an excuse to remove useful hyphens. Pleading the disputed clause as a justification for this form of vandalism should be more difficult if this tag remains. An in-line tag, like {{dubious}}, would be ideal, but I'm not sure one exists. Septentrionalis PMAnderson 15:56, 16 November 2008 (UTC)

"An irresponsible editor is taking its absence as an excuse to remove useful hyphens". At the moment this can't be taken as anything more than a single person's opinion. If nobody else agrees, there is no consensus either to make the change you suggest or to mark the section as disputed. Ilkali (talk) 16:16, 16 November 2008 (UTC)

This is a repetition of part of a post which I made at 22:49, 12 November 2008 (UTC).

If "specially designed sound cards" is ambiguous,

then a disambiguation here would be helpful,

because I can see only one possible interpretation.

-- Wavelength (talk) 17:14, 16 November 2008 (UTC)

Is that supposed to be a poem? But on the question of the tag and the section's wording, can we all eat our tea if we include "normally" (like it makes any difference) and omit the (clearly misleading) section tag?--Kotniski (talk) 20:57, 16 November 2008 (UTC)

I did not intend it to be a poem. I put each clause on a separate line, to make the sentence easier to read. I am still waiting for PMAnderson to disambiguate the expression, and to explain how a hyphen after "specially" makes a difference.

-- Wavelength (talk) 21:35, 16 November 2008 (UTC)

Paragraph 4 of Electric guitar#Sound and effects contains the expression "specially-designed sound cards" [sic]. I removed the hyphen at 23:33, 10 November 2008, but it was restored at 23:38, 11 November 2008. If I remove it now, do I have the support of the Manual of Style (with the inclusion of the word "normally" in the relevant guideline)?

-- Wavelength (talk) 04:22, 17 November 2008 (UTC)

Unless it is, as I would hold it to be, one of the abnormal hyphens which does clarify: in this case, by indicating that "specially designed" acts as a unit in modifying "sound cards". Septentrionalis PMAnderson 06:10, 17 November 2008 (UTC)

I have to agree with Wavelength - at least, until you can tell me what other interpretation you can put on that phrase that we've so far missed.--Kotniski (talk) 07:02, 17 November 2008 (UTC)

Manderson, all style guides are bound to be both descriptive and prescriptive. Now, tell me, as a child were you thrashed regularly by your father for disobedience? I'm trying to work out exactly why you have such a bee in your bonnet about rules. Tony (talk) 00:48, 17 November 2008 (UTC)
- No, actually. Were you? Septentrionalis PMAnderson 03:51, 17 November 2008 (UTC)

Since PManderson has just reintroduced the dispute tags, I feel I have to ask: PMan, in what circumstances would you drop the issue and cease edit-warring over the tags and the wording? Clearly it's not enough that you've faced near-unanimous opposition. What more do you need? Ilkali (talk) 18:18, 20 November 2008 (UTC)

An acknowledgment that "is not hyphenated" is false; I will settle for "is not normally hyphenated" and would consider stronger language (I can't think of any, or I would have suggested it); simply enough to discourage the automatic "correction" of wholly-owned, especially in articles in British English, where it is plainly usage, and may be predominant usage. We are not here to invent a new language, but to guide editors through the existing one. Septentrionalis PMAnderson 18:28, 20 November 2008 (UTC)

I asked what it would take for you to drop the issue. Am I to take from your answer that you will continue edit-warring until you get your way? Ilkali (talk) 18:34, 20 November 2008 (UTC)

Evidence that the English language is not as I understand it. Consensus does not justify the assertion of undocumented falsehood. Septentrionalis PMAnderson 18:42, 20 November 2008 (UTC)

So you will ignore the opinions of all other editors and continue edit-warring until you are convinced that you are wrong? Ilkali (talk) 18:54, 20 November 2008 (UTC)

I think Andersen et al. have done enough to show that this rule is not an absolute in good English writing. Let's remove the dispute tags that deface the whole section over this one trivial point, include "normally" as Andersen suggests, and (if anyone considers it important enough) try and find some more pertinent examples that illustrate the use and non-use of hyphens in this situation.--Kotniski (talk) 19:02, 20 November 2008 (UTC)

You're entitled to think that (and it's why I said "near-unanimous"), but it doesn't answer my question.

Again, PManderson: Are you resolved to edit-war over this until you get your way, ignoring the opinions of any editors who disagree with you? Ilkali (talk) 13:13, 21 November 2008 (UTC)

I give limited weight to opinions which insist upon language which is demonstrably contrary to reliable sources. Septentrionalis PMAnderson 13:49, 21 November 2008 (UTC)

Explaining hyphens after -ly

Since a literate editor like Kotniski is puzzled by what seems clear to me about "the specially-designed sound cards", let me compare with "really-good time".

"really good time" and "good time" are both meaningful, and mean much the same thing with different force; really modifies good only and directly.
But "the designed sound cards" does not mean at all the same thing as "the specially-designed sound cards"; indeed, "designed" is a piece of mush which an editor should remove or recast. Grammaticé, "specially-designed" as a whole modifies cards - this is marked by the hyphen; omitting the hyphen will tend to trip the reader by making xem parse the clumsy phrase "the designed sound cards" first. Septentrionalis PMAnderson 17:56, 17 November 2008 (UTC)

No, -ly means already that the qualified word is coming, usually immediately. A hyphen is redundant, and indeed obstructs. You need another thrashing, Anderson. Tony (talk) 09:12, 18 November 2008 (UTC)

I have attempted to draw a parse tree for the noun phrase "specially designed sound cards".
The periods have no meaning in the diagram; I put them there to help with the display.
The word "designed" looks like a verb in the past tense, but here it is a participial adjective.
The word "sound" looks like a noun, but here it is a noun adjunct.
For comparison, I have also shown "very good style guides", which has a similar parse tree.

                      Noun phrase
             ______________|_____________
             |                          |
    Adjectival phrase               Noun phrase
     |            |                 |         |
  Adverb      Adjective         Adjective    Noun
     |            |                 |         |
 specially     designed           sound     cards
   very          good             style     guides

-- Wavelength (talk) 20:08, 17 November 2008 (UTC)

Yes, of course, designed is a (passive) participle, and sound is in adjunction, but a sentence has more than cladistics: very and good form a different sort of adjectival phrase than specially and designed; as removing the adverbs will show. Septentrionalis PMAnderson 20:31, 17 November 2008 (UTC)

I do not know what you mean by "cladistics"; most references give a biological meaning.

Please see cladistics - OneLook Dictionary Search.

I am still analyzing what you said about removing the adverbs. I have revised the parse tree.

-- Wavelength (talk) 21:26, 17 November 2008 (UTC)

(to Sept, primarily) I think we see the difference, but still I don't see the need for the hyphen to assist parsing in one case but not the other - the parse tree shows that the two sentences need to be parsed in the same way. Indeed the example given in the MOS - "wholly owned subsidiary" - which you seemed to leave unaltered when you edited the section - is one of the kind which you now seem to be claiming can use a hyphen.--Kotniski (talk) 17:22, 18 November 2008 (UTC)

"Wholly-owned" probably should be hyphenated, insofar as it is British English; I missed it. Reading with an American accent, I suspect; thanks. Septentrionalis PMAnderson 03:53, 20 November 2008 (UTC)

When I am reading "PCs with specially designed sound cards" (as in paragraph 4 of Electric guitar#Sound and effects), I am reading in a forward direction, from left to right. After I read the adverb "specially" I anticipate an adjective to follow. It could be a passive participial adjective such as "designed". It could be an active participial adjective such as "functioning". It could be a non-participial adjective such as "strong". In this example, I read "specially designed". Then I read the word "sound". If there is any temporary "tripping" by the string "specially designed sound" (if I analyze "designed" as modifying "sound", with "sound" as the head of "specially designed sound"), it is soon corrected when I read the next word ("cards"). Then I immediately understand "specially designed sound cards" as a correctly parsed noun phrase.

-- Wavelength (talk) 00:38, 19 November 2008 (UTC)

The taxonomy above is nested in the type of traditional grammar that teaches no one to write well (although is possibly useful in the early stages of learning Eng. as a second language). For example, epithets and classifiers are both tagged as adjectives, whereas their grammatical roles are quite different. See Halliday. Tony (talk) 00:14, 18 November 2008 (UTC)

Thank you for the reference to Halliday. To me, this is new and interesting. I have skimmed the following articles: Traditional grammar, Michael Halliday, Systemic functional grammar, Epithet, Classifier (linguistics), and Taxonomy.

I found this statement at Traditional grammar#The role of "traditional grammar":

The main benefit of "traditional grammar" is that it gives learners a basic understanding of the building blocks of language, which can help in improving their writing skills.

-- Wavelength (talk) 04:03, 18 November 2008 (UTC)

I see that Cole has been running around changing the "discussion" tags at the three style guides relevant to the dates issue back to "disputed". In addition, he has been spattering the talk pages of numerous users with a threatening message about their work in improving the date formatting of the project. I ask that he calm down and look clearly at the issues rather than pursue this frenzied campaign against hard-working editors. Tony (talk) 03:24, 18 November 2008 (UTC)

Response to Wavelength on the taxonomy: I agree that for second-language learners, trad. grammar word classes are probably essential. But they are sadly inadequate—and often misleading—when it comes to the underlying complexities of the grammar, and hardly ever improve written and oral skills. As a matter of interest, in Halliday's SFG (systemic functional grammar), "good" is an interpersonal epithet (I'm convincing you of my attitude towards the style guide), and "sound" is a classifier (in this sense, you can't say that they're very sound cards—it's a cut-and-dried category of card, e.g., not a playing card). "Style" is ambiguous, since it might be said to have merged into "guides", and is indeed merged by some writers into a compound word. If not merged, it's a classifier. "Stylish" would be a plain epithet.

You may wish to have a look at nominal group, which has direct relevance to this taxonomic exercise. The concept and application of nominal groups are probably more important in English than any other language: ours is a very highly nominalised language, which is one reason it's so good for scientific and technical writing, where Things are so important. Another taste of SFG is the short article on thematic equative, another important feature of English grammar (trad. grammar calls it a cleft something or other, but is rather wobbly—superficial, actually—WRT its grammatical function). A Swedish client has objected to my claim that English may be the only language to have such a grammatical feature; I'm not convinced that he's right, but the question is open. Tony (talk) 04:36, 19 November 2008 (UTC)

It gladdens my heart to see this surge of elevated linguistic analysis at our normally pedestrian talkpage. But let us not lose sight of a salient fact: nothing has been said to support any weakening of this guideline:

* A hyphen is not used after an -ly adverb (wholly owned subsidiary) unless part of a larger compound (a slowly-but-surely strategy).

Even if something above does count as evidence supporting change, there is still no consensus to diverge from the unanimity of major style guides. PMAnderson has twice attempted to smuggle the word normal into the guideline, with these edits: [4] and [5]. Each time he concealed the change with an edit summary that masked the change, and misled editors who were patiently discussing the matter with him. Clearly this editor does not really like the idea of discussion and consensus-building. Add to this his reluctance to produce, when challenged, evidence that he claims exists (from Fowler's, which I have checked in every one of its major editions).

MOS should not surrender to this relentless death of a thousand pin-pricks. Anderson would do better to study MOS and learn what it actually says – and perhaps get up to date with developments since the decline and fall of his mentor Edward Gibbon (see Anderson's self-inflating reference, ~~above~~ below [where many a true word was spoken in ironic jest]–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 13:19, 20 November 2008 (UTC)). For example, he might learn to avoid mistakes and misrepresentations like this one.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 00:25, 20 November 2008 (UTC)

Only one style guide has been cited; unanimity usually requires at least two minds to unite.
The OED and all three of its quotations hyphenate wholly-owned subsidiary (wholly adv. def 3). That should trump an inference from CMOS 's brevity. Septentrionalis PMAnderson 03:53, 20 November 2008 (UTC)

In the interest of a balanced debate, we might hope that you will eventually get something right, Anderson. For a start, two serious and respected authorities have been cited against you: Garner's and CMOS. And I can cite more. Would you like me to do that? Fine: I'll give two further authorities for every one that you can adduce – starting with your first, please.

Meanwhile, OED does not rule on these matters of hyphenation. In fact it scarcely remarks on them at all, and does not, in its catholicity, care one way or the other (as we, with our altogether different brief, must). It is sheer luck that its examples for wholly-owned are hyphenated. Other occurrences in its citations have wholly owned unhyphenated and without comment (see entries for "network, n." and "refinance, v."). To see how little OED cares about hyphenation, look at this from its entry for "made, ppl. a."

[III. 8.] b. With adv. (or sometimes adj.) giving the sense ‘made in a certain manner, having a certain quality or kind of make’, as in badly-, neatly-, well-made; often with reference to the ‘make’ or ‘build’ of the body (= -built), as in loosely-, powerfully-, stoutly-, strong(ly)-made. Most of these combs. are treated under their first element, or in their alphabetical place as Main words.

Very well; but when we look for these words under their first elements, or as main words, we find only one of the combinations mentioned cited with a hyphen! (Stoutly-made, cited at "stoutly, adv." and "standard, n. (a.)": both from the same source dated 1833.) In fact OED has many combinations with ly-; but that is entirely predictable, given its universal and historical scope.

I am again reverting your unilateral tinkering, Anderson. You are fumbling badly with this, and we shouldn't have to bear the consequences. You wanted discussion? You have had it. No one has supported any of your desperate attempts to sabotage a perfectly sound guideline. I wish we could move on; this is wasting so much precious time.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 04:53, 20 November 2008 (UTC)

Lynne Truss writes that the OED used to say anyone who took hyphens seriously would go mad, and that Churchill's rule was to avoid them at all costs. PMA, they're intended only to aid understanding. Truss gives the example of the pickled-herring merchant, someone who sells pickled herrings, rather than the pickled herring merchant, someone who sells herrings and is drunk.

Where they're not aiding understanding, they're not needed. When you're dealing with an adverb, "ly" is the hyphen, as it were, in that it signals that it's about to qualify a verb. There's no need to signal that twice. SlimVirgin ^talk|edits 05:09, 20 November 2008 (UTC)

Indeed. Anderson, there's only one thing worse than omitting a sorely needed hyphen: inserting a redundant one. This is what you want to promote, apparently. Noetica is right: move on, please. Tony (talk) 05:35, 20 November 2008 (UTC)

"Wholly-owned subsidiary" - so written - is a legal term in the UK, defined in S 1159 of the Companies Act 2006 [6]. I find the onjections to making this an ENGVAR matter mysterious. Johnbod (talk) 09:29, 20 November 2008 (UTC)

Do you indeed, Johnbod? Then cite the non-American authority on style that agrees with this hyphenation, as a systematic feature of practice. It is not in New Hart's Rules, Complete Plain Words (3rd edition), Cambridge Guide to English Usage, Ridout and Witting's The Facts of English, or Australia's AGPS Style Manual, all of which are squarely in accord with our current guideline. Nor is your exception given in Oxford Dictionary for Writers and Editors, which is an alphabetic list of special cases. Yes, there are some spellings in some British usage that disagree with our guideline, and these are pointed to here and there in some books. The world is wide and various. For example, British dictionaries typically spell newlywed optionally with a hyphen. Are we to list all such curiosities? Are we to clutter our guideline by accommodating exceptions ignored by other major guidelines – American, British, and Australian? We could. In fact, I have prepared a draft that makes sense of all this, and explains certain exceptions. But I don't propose it, because our present guideline is industry-standard and uncomplicated. That is the special virtue of a good guideline!

Once more I should draw attention to a line in our section on hyphens, since no one else in this discussion seems to have taken note of it:

Hyphenation involves many subtleties that cannot be covered here; the rules and examples presented above illustrate the broad principles that inform current usage.

That certainly is adequate to cover the odd sort of case that you and Anderson raise. I understand your point, and your contribution to these long deliberations. Anderson's case is different: he attempts to foist successive variations on us, under cover of misleading edit summaries, and will not relinquish his stand no matter what weight of opinion or evidence is laid out plainly before him.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 13:12, 20 November 2008 (UTC)

If "wholly-owned subsidiary" is enshrined in UK law, though, we should at least find a different example to replace it in the MoS.--Kotniski (talk) 13:50, 20 November 2008 (UTC)

Anderson, it's not a joke or game, and the MoS is not your private plaything; everyone sees your antics as tiresome and disruptive. It's time to stop now, and to start making constructive contributions. Your silly contrarian tactics undermine any semblance of authority you might have had. Try to learn from highly skilled writers like Noetica and Kotniski, please. Tony (talk) 15:32, 20 November 2008 (UTC)

MOS should not be a joke or a game; but it is both. As long as handfuls of language cranks set up rules they are quoting out of context, have misunderstood, or have invented, and broadcast them all over Wikipedia, contrary to English usage or some major variant of it, it will continue to be a joke - or, rather, it would be one if it did not do so much harm both to the civility and to the text of Wikipedia. Septentrionalis PMAnderson 18:16, 20 November 2008 (UTC)

As for this matter, which is almost as arbitrary, and fully as silly as "Mebibyte", I present below a quote from the Second Edition of the OED. A hyphen is not used after an -ly adverb (wholly owned subsidiary) has the disadvantage of being false; it is at best an unacknowledged and unfounded exception to ENGVAR, which has genuine consensus, unlike this invention.

Wholly

....

3' Comb.: wholly-owned a., applied to a company all of whose shares are owned by another company.

1964 Financial Times 11 Feb. 12/1 The directors..have decided to give the holders of Ordinary shares the opportunity of acquiring an interest in the wholly-owned subsidiary. 1972 Accountant 21 Sept. 360/1 The UK company is a subsidiaryalthough not wholly-owned. 1976 Scotsman 20 Nov. 3/2 The plan is recommended by the boards of all the companies, who will become wholly-owned subsidiaries of the new Malaysian group.

Obsrerve all three quotations; last I heard the FT had a style guide. Septentrionalis PMAnderson 18:16, 20 November 2008 (UTC)

As for SV's citations of anyone who took hyphens seriously would go mad and they're intended only to aid understanding, I agree with both; we would have a better section if all the subordinate bullet points under WP:HYPHEN 3 were replaced by that language. The only good I expect from MOS is to get out of the way of an editor who hyphenates for clarity - or does not hyphenate for clarity. Septentrionalis PMAnderson 18:38, 20 November 2008 (UTC)

Anderson, I have already dealt with the OED case that you quote at length, and shown that elsewhere OED cites examples with wholly owned (no apostrophe). You have utterly ignored my painstaking analysis. Let me make it clear once again: OED rarely concerns itself with matters of hyphenation. In the case of made, to give the example once again, it presents a number of compounds with hyphens: loosely-, powerfully-, stoutly-, strongly-. But this is only for expository convenience in lexicography. The dictionary is founded on citations; but in fact, only one of these very compounds is cited in examples: stoutly-made, twice in different articles, from the same 1833 [sic] source. In fact, strongly made occurs in nine OED entries, but each time without a hyphen. So much for how OED presents phrases with hyphens, outside of its citation texts. It even uses strongly made in its own definitions, for example at the entry "strong-box".

OED goes about its monumental work of recording the rich chaotic variety of our language, but style guides do not have the same brief. CMOS, New Hart's, and the others I refer to above (and more!) offer straight guidance to help editors and writers confronting hard choices. The guides I cite are unanimous on the point we are discussing. And in fact, with our wording we help more than some: not all guides show compounds like a slowly-but-surely strategy.

The fact that a phrase turns up in OED is not compelling, for style guides. Nor does its fixed technical use in UK statute law. It is easy to see why, in special circumstances, special treatment of these phrases will occur. I remind editors a third time that we have long included this proviso:

Hyphenation involves many subtleties that cannot be covered here; the rules and examples presented above illustrate the broad principles that inform current usage.

Given this proviso, and given our brief to guide rather than minutely govern choices in writing at Wikipedia, the wording we have is perfectly adequate. If we qualify this guideline with its own normally, where a general qualification is already in place, we weaken the guidance and make it harder for editors to be sure how robust this particular provision might be. In fact, the present formulation is very robust: close to unanimously endorsed by style guides, regardless of practice in the drafting of UK statutes and regardless of the quite different imperatives under which OED operates.

Now, as I have revealed, I have a draft that is more complex and less like the standard that is accepted by almost all guides. I am not absolutely against a more complex guideline! But so far, Anderson has given only feeble reasons for pusillanimous changes, some of which he attempts to introduce by stealth under misleading edit summaries. That we should resist. His attacks have gone on for too long. If he used evidence, if he offered cogent arguments (and responded to cogent arguments!), the case would be different.

I might accept normally if it were not the result of such a flawed process of deliberation here. I would accept a different example than wholly owned subsidiary; but I note that to avoid this example is to avoid guiding exactly where an editor will want guidance. It is hard cases that bring editors to MOS for help, and hard cases that we must therefore address. We can do that briefly (which we do perfectly well already, with the general qualification that I have now exhibited here three times); or we can do it at length. But we should not do it weakly, and we should not do it without genuinely guiding practice.

Anderson, you reveal this about yourself:

The only good I expect from MOS is to get out of the way of an editor who hyphenates for clarity - or does not hyphenate for clarity.

But such an editor does not write in isolation; in this community an editor's work is subject to review and improvement, with rational discourse back and forth. To help that editing, that review and improvement, and that discourse, we offer guidance. Please stop your campaign aimed at ensuring that MOS is only ever "out of the way"!

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 21:57, 20 November 2008 (UTC)

Oh, MOS could do a great deal better than stay out of the way; but it is clear that it will not be. In the meantime, a statement which is factually incorrect is not atoned for by a qualifier in a different paragraph; we all know that MOS is usually consulted and quoted out of context. I am prepared to adopt any solution which satisfies these criteria, but I am fresh out of suggestions. Since I agree that a tag at the head of the section is imprecise for a problem with a single clause, I will be restoring the {{dubious}}. Kotniski, do you have any further suggestions? Septentrionalis PMAnderson 23:54, 20 November 2008 (UTC)

I do not contend that wholly-owned is the only British usage, but it is plainly a legitimate one; doubly so if it is the term of art in the law. Septentrionalis PMAnderson 23:54, 20 November 2008 (UTC)

Don't come here mouthing half-truths about factual correctness, Anderson. Not with your record! Concerning the Financial Times, you write "last I heard the FT had a style guide". O really? If they do have one, it is no "authority" for hyphenation in the phrases we are considering. Some recent snips from FT:

a licence to apply to open wholly owned units in the country [10 September 2008]
two employees of his wholly-owned EasyGroup [16 November 2008]
Sir Stelios's wholly-owned Cayman Islands registered EasyGroup [whew! 15 November 2008]
a purely private, wholly unregulated mode of transportation [14 November 2008]
StudioCanal's Optimum Releasing, a wholly owned subsidiary of Canal Plus [1 November 2008]
created by wholly-owned stores to support wholesale sales [8 November 2008]
Sales of Volvo, wholly-owned by Ford, plummeted [3 November 2008]
the size of his wholly owned online delivery business [8 November 2008]

I think we can do better than follow their example!

And don't appeal to terms of art in law (in just one jurisdiction, as far as we know) to determine our general recommendations here. It is because they are terms of art that they are special cases, and should not affect standard consistent usage for Wikipedia. We cannot note all special uses; we give the best general guidance we can. Some of us do, I mean.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 00:18, 21 November 2008 (UTC)

Yes, I agree; we could do worse than to follow their example: permit both. Of the seven examples of wholly-owned, four are hyphenated (one of those may be unwise); wholly unregulated is like really good, and should not be hyphenated. Septentrionalis PMAnderson 01:01, 21 November 2008 (UTC)

In light of the evidence presented, is there any reason why people keep removing "normally" and re-inserting what has been shown to be a patently bad example (the wholly owned thing)? Or tagging the whole section when only one small point is considered dubious?--Kotniski (talk) 07:01, 21 November 2008 (UTC)

Because this is a style guide, and that's the chosen style. Has anyone produced a style guide that advocates for a hyphen after ly? SlimVirgin ^talk|edits 07:09, 21 November 2008 (UTC)

Kotniski, I cannot speak for others. For myself, there are two answers to your question:

The various suggested changes weaken a strong and useful guideline that is in accord with all major style guides – British, American, and Australian, at least. The guideline already falls in the scope of a general qualification that I have quoted three times above. Further qualification is unnecessary and counterproductive.
As I have written above: "Now, as I have revealed, I have a draft that is more complex and less like the standard that is accepted by almost all guides. I am not absolutely against a more complex guideline! But so far, Anderson has given only feeble reasons for pusillanimous changes, some of which he attempts to introduce by stealth under misleading edit summaries. That we should resist. His attacks have gone on for too long. If he used evidence, if he offered cogent arguments (and responded to cogent arguments!), the case would be different. / I might accept normally if it were not the result of such a flawed process of deliberation here. I would accept a different example than wholly owned subsidiary; but I note that to avoid this example is to avoid guiding exactly where an editor will want guidance." I should add to that: wholly owned subsidiary is an excellent example for us to retain, since it shows how our chosen style (indeed, that of all major style guides) resolves even that unstable and uncertain case!

So I do not mean to be intransigent. I am taking a stand, and so are some others, against an editor whose pernicious activity has been steadily against everything we (nearly all of us) are working to achieve here. He is, for whatever unfathomable reason, zealously against the natural program of every style guide: to provide genuine guidance; to prescribe solutions to problems of style, whether or not they are ultimately adopted by those who seek that guidance. We sometimes capitulate to Anderson, because life is too short. But this time I do not capitulate, and I invite others not to also. At the very least he should not amend the guideline while there is still discussion of it here. But he doesn't even do us that courtesy.

Kotniski, much more detail is in the dense forest of verbiage above. I have answered every one of Anderson's points that has any relevance; he has addressed precious few of mine, and seems determined to have his own way against the majority here no matter what evidence or what argument is brought to bear. If the majority think we should give in to pressure from such an isolated and ill-motivated editor, I shall happily return to my retirement from this talkpage and from editing WP:MOS. Yes, I withdrew entirely, many months ago, because of the impossibility of dealing with Anderson. I came back reluctantly because I saw that an unfair burden fell on Tony, who labours to keep Anderson from white-anting Wikipedia's Manual of Style. Unlike some, I have no psychological need to stay here; and I will not, if MOS once more appears to me to be a lost cause.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 07:38, 21 November 2008 (UTC)

I understand your sentiments and hope that you will stay with us and make more of your valuable contributions, but on this specific point, if the law itself says "wholly-owned", this must at the very least be the wrong example to use, and it probably also serves as evidence that the MoS rule can be watered down at least to the extent of adding "normally" (or possibly by trying to formulate the exception more precisely, maybe with a restriction to British English, if this kind of hyphen grates with those from the US and Oz).--Kotniski (talk) 10:44, 21 November 2008 (UTC)

Kotniski, never mind the sentiments: feel the evidence. And weigh the arguments. As for wholly-owned being enshrined by what you with wide eyes call "the law itself", so is wholly owned. From the British Shipbuilders Act 1983, UK:

any activities which were carried on, immediately before the date of transfer, by a company which, by virtue of this Act, becomes a wholly owned subsidiary of British Shipbuilders
none of whom need be a wholly owned subsidiary of British Shipbuilders

And more of the same within that act. Then there is this from the Transport Act 2000, UK:

The expressions “subsidiary” and “wholly owned subsidiary” have the meanings given by section 736 of the [1985 c. 6.] Companies Act 1985 or Article 4 of the [S.I. 1986/1032 (N.I. 6).] Companies (Northern Ireland) Order 1986.

Seems pretty definitive and consistent, right? Or the Railway Heritage Act 1996, UK:

any wholly owned subsidiary of the Board

"The law itself", even within the single jurisdiction of the UK legislature, is inconsistent. Must we therefore be? No. We needn't strive at every point to pursue a half-baked pluralism. No style guide does that. It sets a style, and prescribes remedies for those seeking specific help with style issues. We let them down if we explicitly permit every recorded variant, or cravenly step around giving a decision in problematic cases. Where would it end? They can still use their judgement; no one is saying they shouldn't or can't.

Burchfield's revision of Fowler's ran into a great deal of censure for failing in exactly the way Anderson advocates. The reprinted first edition gets 4.5 stars out of a possible 5 stars at Amazon; Burchfield's edition gets 3 stars, and the reviews tell you why. Then compare Garner's Modern American Usage that I have referred to: absolutely clear and prescriptive, and because of that it gets a near-unanimous 5 stars from its 25 reviewers (one of whom gives it only 4 stars). You can be sure that not one of those reviewers felt under any compulsion to abide by Garner's expert rulings. But they all appreciated his guidance.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 11:38, 21 November 2008 (UTC)

O, but I simply must add these two, both from the very same page of the Crossrail Act 2008, UK:

from a body corporate which is a wholly-owned subsidiary of CLRL
from a body corporate which is a wholly owned subsidiary of the Greater London Authority, the London Development Agency or Transport for London

Anderson might take this as precedent for allowing inconsistency even within one article; I take it as decisive evidence that a clear ruling is needed.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 12:12, 21 November 2008 (UTC)

I see where you're coming from, but the Companies Act actually defines wholly-owned subsidiaries (and uses that form consistently). It would be perverse to use a different hyphenation if writing about that Act, for example. This may be the type of exception that is so rare as not to require mention, but we could at least find an equally valid and illustrative example that isn't itself likely to be the subject of such an exception. (Maybe even the "specially designed" one that started this whole thing off, though that might look a bit pointy.)--Kotniski (talk) 13:23, 21 November 2008 (UTC)

Along with never minding the sentiments, Kotniski, never mind "where I'm coming from". Mind, rather, the hard arguments and evidence I bring to bear. I am tired of repeating points that readers miss. Brevity is most valuable and compelling; but neither brevity nor long-windedness nor even the oratory of angels will have any effect, if people fail to pay attention. I have already argued for our not being swayed by practice in UK legislation, and explained how that inconsistent practice should have us retain the controversial example, not weakly sidestep it, abnegating our responsibility.

See my response to Wavelength, in the section below.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 22:50, 21 November 2008 (UTC)

I would prefer to tag the clause only, but every effort to do so has been reverted (we'll see if Kotniski's sticks); as far as I am concerned, a tag is preferable to edit-warring.

I don't see why the addition of another normally to a section which already has one is being decried as chaos come again. Why is "we normally do X" not clear? If I were genuinely uncertain between X and Y, I'd take it as advice to do X. Septentrionalis PMAnderson 13:58, 21 November 2008 (UTC)

Kotniski's tag is a brilliant innovation, and much to be preferred. Why should it be removed, so long as this discussion drags on?

No, Anderson: normally would not amount to "chaos come again". But it is one more tiny worm eating away at the robustness and utility of our Manual of Style. And you are the Wurmmeister.

If this must continue, let it be in the section below, for the convenience of us all.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 22:50, 21 November 2008 (UTC)

Alternative examples

Here are 12 alternative examples for use in the manual. Maybe they include at least one which everyone can accept. These include four active participial adjectives, four passive participial adjectives, and four non-participial adjectives. They include eight plural nouns and four singular nouns. They can even suggest other examples which may be better.

vaguely familiar voice
gracefully blooming roses
intricately embroidered curtains
softly falling snow

brightly shining sun
quickly galloping horses
carefully placed napkins
majestically tall sequoias

surprisingly good results
eagerly awaited event
fondly remembered faces
nearly ripe apples

-- Wavelength (talk) 20:05, 21 November 2008 (UTC)

All of that avoids the core issue, Wavelength: an issue that is illustrated by the particular guideline we are looking at, but one that pervades all of our work at MOS. That is why I, for one, am holding on tight.

The core issue? It is easy to find alternative examples and workarounds. It is fatally easy to evade the task of settling the hard questions that will bring editors to MOS in the first place. Sure, we could go the way of Burchfield, and give no useful ruling. Look what reception that gets from users (see above)! Or instead, we can say this, in effect: "Here is a useful and robust guideline. You are wondering whether to write wholly owned subsidiary or wholly-owned subsidiary? Ah yes: both are used, sometimes inconsistently by the same source. But we, along with every other major style guide, recommend the form without the hyphen." And we have our reasons for this recommendation, what's more.

I am reminded of uncertainties with terms addressed in Diatonic and chromatic. In initiating and developing that article I surveyed a vast number of sources, as the fifty footnotes might suggest. I saw very many "authorities" that crept around the central problem with the term diatonic, which has two logically incompatible meanings. Such craven pussy-footing is not a good look, and it is worthless in a style guide. To guide, we must confront hard cases. UK legislation is inconsistent, but that is no good precedent for us.

Keep the example we have now. If we do not, either we fail to guide or we must make a considerably more complex guideline. I am not absolutely against a more complex guideline; but I think others might be.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 22:34, 21 November 2008 (UTC)

The core issue is that Noetica insists on making MOS assert falsehood. "Wholly-owned subsidiaries" is usage, as xer own examples attest; in some contexts it is mandatory usage. Having an example which denies that makes MOS look illiterate and bullying; I should prefer not to.

The grammatical point here is the difference between two slightly different constructions: "majestically tall sequoias" can become "tall sequoias" with a slight change in force. But "X Company sold off its owned subsidiary Y Technologies" is scarcely intelligible - and that's why it is usual to hyphenate wholly-owned. The examples Wavelength cites are of the first, normal, type, and that's why we should use one of them. Septentrionalis PMAnderson 15:56, 22 November 2008 (UTC)

The legal document cited above is about commercial law, and not about regulating the English language. It is not a style guide. An error slipped into the document in many places, and no one corrected it. Someone should contact the people responsible for the document, and tell them about the mistake. Maybe there will be a corrected version published in the future, or, at least, an addendum of errata.

If I were a British lawyer composing text about the act, I would not hyphenate the expression "wholly owned subsidiary", in spite of what the act says. If I were a British legal secretary composing text for a British lawyer who wanted the expression hyphenated, I would tell the lawyer that the authors of the document made a mistake, and I would try to persuade the lawyer to follow a reliable style guide.

Whether "designed sound cards" or "owned subsidiary" is meaningful is irrelevant to a reader who reads in a forward direction, from left to right. (Presumably, the meaningfulness is questioned because all sound cards are designed and all subsidiaries are owned, and so the unmodified adjectives add no information.) Faulty parsing appears to be the crux of PMAnderson's unusual point of view.

I understand better now why Noetica wants to retain the long-standing example "wholly owned subsidiary". In the interests of brevity, a person consulting the Manual sees that, even in a case like this, where some people might have seen in the legal document a reason to hyphenate, the omission of the hyphen is recommended. Therefore, in easier cases, such as "specially designed sound cards", that is likewise the course to follow.

Unfortunately, my good intentions in providing alternative examples may have made things more difficult for Noetica and for me. (I started this discussion because my dehyphenation to "specially designed sound cards" in paragraph 4 of Electric guitar#Sound and effects was challenged.)

-- Wavelength (talk) 22:00, 22 November 2008 (UTC)

Wavelength, I cannot begin to tell you how pleased I am that someone, at least, can be persuaded by sound argument. Thank you for paying attention, and for graciously acknowledging a change of heart, as we all should from time to time. That is enormously to your credit.

Anderson, you claim that I "insist on making MOS assert falsehood". But you misrepresent me. Two points:

I am discussing the issues; you insist on making changes to a long-established guideline that is tagged for discussion. I have just reverted your last attempt (which was garbled, by the way). Who is grasping for control, here?
With a guideline in the form of a plain indicative utterance, a style guide cannot genuinely assert a falsehood. Not in any interesting or relevant sense. The guideline is a stipulation. When God says thou shalt not kill and the recipient of the instruction does kill, has God uttered a falsehood? Only on a twisted understanding of how language operates. We have already covered this: the surface structure of a guideline can be a plain statement (thou shalt not kill), or it can be a modal formulation (thou mayst not kill), or it can be an imperative (do not kill). But none of this affects its manifest force as a guideline (or a recommendation, or an injunction, depending on the case and the context).

You claim that my examples attest to wholly-owned subsidiaries. Sure, that is used: here and there, inconsistently. It is one sort of problem case; loosely speaking, it is a crux for anyone setting out to guide style. For that very reason we should confront it! As I have pointed out (and you have not answered the point), it is exactly in such crucial instances that editors will consult MOS. You would have us supply information that most people already have. Few will have trouble with majestically tall sequoias; many will with wholly owned subsidiary. Yes, even UK legislation is all over the place with that one! If we stick boldly with a difficult case like wholly owned subsidiary, we cover all contingencies: we advise grandma without telling her how to suck eggs, and a fortiori we advise those who do need advice in the basics.

A clear and confident MOS will not look bullying, any more than Garner's book (mentioned earlier) is seen that way. Rather, like Garner's book it will have earned respect, and actually be used: it sets a consensual style, and delivers it boldly and clearly. The newest Fowler's fails to guide; you would have us do the same.

I agree that there is a difference between most purely adjectival and many participial cases. I have analysed this thoroughly myself, and could say more. But all major style guides give the guideline that we currently give; and it is mysterious why we should want to complicate things, and be over-delicate where they are robust and clear.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 22:23, 22 November 2008 (UTC)

But changing the example wouldn't need to do either of these things. There must be other equally crux-ish examples that don't carry the objection that the term is defined (not just mentioned) in law using a different hyphenation style. I'm sure most people would consider it incorrect to write "the UK Companies Act defines a wholly owned subsidiary as..." if we know it uses a hyphen there - factual correctness must trump stylistic correctness. --Kotniski (talk) 11:49, 23 November 2008 (UTC)

If I were writing about the Act, I would omit the hyphen. If I were quoting the Act, I would use the hyphen, but I would put "[sic]" after the expression (see sic). I would do so, even if it meant doing both in the same sentence. See paragraph 2 of my previous message.

-- Wavelength (talk) 18:02, 23 November 2008 (UTC)

OK, I think we're just gonig to disagree about this. Doing what you suggest would, I suspect, be viewed as annoying or incomprehensible pedantry by most readers (but then the majority aren't always right...)--Kotniski (talk) 07:21, 24 November 2008 (UTC)

English is a language which allows flexibility. Both The Guardian[7] and The Daily Telegraph[8] arbitrarily use "wholly owned" and "wholly-owned". Our style guideline cannot be prescripive, if general practice is not. Therefore editors can use either construction as they prefer. Ty 12:32, 23 November 2008 (UTC)

Prescriptiveness has already been discussed. Please read #Hyphens after -ly adverbs (rationalised section) from the beginning, or search for "prescri" on this page.

-- Wavelength (talk) 18:55, 23 November 2008 (UTC)

Wavelength:

Your care with citation of the act would go well beyond the care that the UK legislators take, as we have seen. Analysis of occurrences of wholly[-] owned subsidiary using the advanced search facility at www.publications.parliament.uk is revealing. Like Google, the facility does not detect hyphens. The first 50 hits on a search for wholly owned subsidiary show, in the heads that appear on the screen:

33 hyphenated instances
22 unhyphenated instances

Not an impressive majority, for Hansard and all other parliamentary uses. When we restrict the search to actual legislation (bills or acts), we find in the first 20 hits:

14 hyphenated instances
16 unhyphenated instances

Clearly the definition in the Companies Act is not about English usage, as you have pointed out. Nor is it taken as guiding style at the official site of the Parliament that framed it. Nor do the Hansard transcribers appear to give a hoot one way or the other. Should we care what the UK Companies Act says, then? Of course not! Nor should any other style guide. And – unanimously – they don't.

Ty:

You write that "Our style guideline cannot be prescriptive, if general practice is not." But that is singularly unhelpful. "General practice" is that which stands in need of prescription (or does not, if your view is different). To say that general practice either prescribes or fails to prescribe is to make a basic category mistake. Are CMOS, New Hart's, and the other major style guides in awe of "general practice"? Of course not. They prescribe standards, just because "general practice" is chaotic and inconsistent. And people subscribe to them in droves because they want such guidance. Please follow Wavelength's suggestion, and read what has already been said about this.

Kotniski:

It is not clear whom you are addressing when you write "Doing what you suggest...". Please try to keep a long and complex discussion orderly. See, for example, how I am structuring this post. Beyond that, the issue here is not something on which we can simply agree to disagree. Through this seemingly trivial example (hyphenating wholly owned subsidiary) we approach core dilemmas for MOS: Err on the side of detail or the side of simplicity? Err by prescribing too narrowly or by barely prescribing at all? And more.

Anderson:

Again I have reverted your unilateral meddling with an established guideline that is tagged for discussion here. (I hope others will assist in reverting such uncollegial edits.) You write: " 'X Company sold off its owned subsidiary Y Technologies' is scarcely intelligible". But that is not so. The details vary according to the jurisdiction in which a company is set up, but strictly a subsidiary is not always owned. Some subsidiaries are not even minority-owned subsidiaries. What makes a company a subsidiary of its parent company is that the parent company has control of the subsidiary. This is normally underwritten by at least majority ownership, but it need not be. Google searches are dangerous, as we all acknowledge; but a search on "the owned subsidiary" (the quote marks are essential) gets 1,350 hits. I could say much more about the motivations for hyphens in such premodifiers, and more tellingly than in the case of wholly owned subsidiary. But I am inclined not to. As I have said, major style guides are effectively unanimous on this point; and they agree with our own long-standing and simple guideline. You have not made a case for our doing differently; and you have not made a case for us holding back from delivering genuine guidance in difficult cases.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 23:00, 24 November 2008 (UTC)

There would appear to be consensus above that wholly owned subsidiary is not the best example, from Johnbod, Kotniski, and Wavelength, for various reasons. Any of Wavelangth's dozen examples above would not be hyphenated, even in the most formal British English; they are, therefore improvements over wholly owned subsidiary. Perhaps Noetica can explain why xe keeps reverting to the weakest example we can find, which is, even by xer own research, hyphenated as often as not? Septentrionalis PMAnderson 22:17, 24 November 2008 (UTC)

Anderson, once more you distort the tenor of the discussion. You refuse to focus on the lines of argument and the enormous amount of evidence that I can easily bring against your flat assertions. Do not persist in making changes to a guideline flagged for discussion. Do not continue to pretend that you have unambivalent support for such changes, let alone consensus. Far from wholly owned subsidiary being the weakest example we can find, it is much bolder and gives much clearer guidance than any of the listed alternatives – alternatives that Wavelength now regrets offering, if I read correctly. Please: let us read correctly! And let us not attempt to mislead other editors. If they read through what has gone before, the deception is easily uncovered.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 23:12, 24 November 2008 (UTC)

Anderson, what on earth makes you think I want, or anyone else here wants, "a compromise". I do not compromise when it comes to redundancy in language, and nor should you. Ilaki, I don't understand why you find the hyphen acceptable when it is redundant. Tell me now that it's not redundant, and we can get down to the nub of it. Hyphens are useful—indeed indespensable—to good writing, but only where they're useful. To add a redundant hyphen is to distract the reader, and to make the reading harder, not easier. I have no intention of accepting this watering down of a useful and long-standing guidance to our editors that has helped to make WP's prose better for our readers. Tony (talk) 00:43, 25 November 2008 (UTC)

WP:Consensus. Those who are unwilling to compromise should find some other hobby; those who are uncompromising in order to impose a novel variant of the English language should find a Newspeak Wiki. Wholly-owned subsidiary is plainly British usage; we should not attempt to reform that - it is also (less common) usage elsewhere for the sound reason that it indicates a difference in construction from majestically tall sequoia. Septentrionalis PMAnderson 00:53, 25 November 2008 (UTC)

Perhaps one who is "opposed to all efforts to impose uniformity" has mistaken Wikipedia for a contrarians' playground. -Ac44ck (talk) 01:12, 25 November 2008 (UTC)

Anderson:

We are not attempting to reform the language, but to guide style in Wikipedia articles. That is the whole purpose of MOS.
The "variant" you seek to ridicule has support from all major style guides. You have not succeeded in bringing to the table even a minor style guide that supports your own preferred variant.
Compromise is often desirable. Capitulating to an extreme position (refusal to allow guidelines to guide; refusal to work collegially; refusal to represent others' positions accurately; refusal to join in nuanced analysis of the facts) is not a good kind of compromise.
Those who seek to undermine all that MOS is attempting to achieve – they are the ones in need of "another hobby".
Wholly-owned subsidiary is definitely not settled as British usage, as the carefully presented evidence above amply shows. It is not accepted as such in any major style guide. It is becomely more and more apparent, Anderson, that argument and evidence count for nothing, for those implacably advocating their own long-held prejudices. What is the point of anyone arguing rationally, and presenting clear evidence, if argument counts for nothing, and is simply passed over in favour of repeated assertion?

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 01:16, 25 November 2008 (UTC)

This discussion keeps popping up in my watch list. My interest is in another topic on this talk page. The rehashing here is annoying. I experienced the same repetition of prior statements in my dealings with User:Pmanderson. It is apparently known behavior. There comes a time to cut people loose. -Ac44ck (talk) 01:37, 25 November 2008 (UTC)

"Ilaki, I don't understand why you find the hyphen acceptable when it is redundant". Am I "Ilaki" for the purposes of this conversation? All I said is that, in my own personal version of English, hyphenation is acceptable in most of those instances (including wholly-owned subsidiary). I do not think this should affect the design of our guideline, and I agree with you and Noetica on what the wording should be and what example should be given. Ilkali (talk) 01:30, 25 November 2008 (UTC)

PMAnderson, I request that you refrain from citing me as a supporter when you change the Manual. Although I did suggest 12 alternative examples, I am now advocating for retention of the long-standing example ("wholly owned subsidiary"). This may baffle you, but the reason for my previously suggesting the alternative examples is also the same reason for my now advocating for retention of the long-standing example, namely, its difficult nature. The example is potentially difficult for many users of the Manual (but not too difficult for editors of the Manual who prescribe according to reliable style guides), and therefore it can be used as a benchmark for deciding less difficult cases. Whereas previously I was offering examples for easy acceptance by editors of the Manual (although with less helpfulness for users of the Manual), I am now advocating for greater helpfulness for users of the Manual (although with more difficult acceptance by editors of the Manual). I am also hoping for an example that would support "specially designed sound cards" as clearly as possible, without actually using that as the/an example. Would you rather live in a straw house, a wood house, or a brick house? Constructing any of them requires work.

Noetica, please do not give up on this case. Since it began on November 11, no one has proven to be more qualified than you in discussing it.

--Wavelength (talk) 02:49, 25 November 2008 (UTC)

So why not simply use that as an example? It's just as robust as the wholly owned one, and (as far as we know so far) isn't subject to any objections of the type raised with respect to w.o.s. We do the MoS no favours by deliberately choosing an example to which people are able to make quite valid objections, thus weakening the authority of the MoS.--Kotniski (talk) 11:09, 25 November 2008 (UTC)

Kotniski, the long-standing example ("wholly owned subsidiary") is a better and a more robust example, as I just explained in my previous message at 02:49, 25 November 2008 (UTC); also, in paragraph 4 of my message at 22:00, 22 November 2008 (UTC). How did you miss it? Please read again my messages, carefully. Also, please read again Noetica's explanations of the same point, carefully.

-- Wavelength (talk) 18:48, 25 November 2008 (UTC)

Yes, I've read it all, but I still see no reason why it's any better or more robust than any other of a thousand similar examples. Instead it just invites people to discredit the whole rule by pointing out that the law defines this particular term otherwise. Still, as long as they are not doing so (or attempting to write about such companies), I'll let it go.--Kotniski (talk) 19:32, 25 November 2008 (UTC)

Another kind of exception

Wikipedia:Manual of Style#Hyphens, subsection 3, point 2, refers to the use of a hyphen to disambiguate, as in little-used car, where little is an adverb, and not an adjective, as it usually is. It happens that, of the English words that can be both adjectives and adverbs, some end in -ly.

half-hourly, hourly, daily, nightly, weekly, fortnightly, semi-monthly, monthly, quarterly, yearly

early, kindly, likely, only

I am still pondering what would be a good example of a phrase requiring disambiguation. This kind of exception can be added to subsection 3, point 4.

-- Wavelength (talk) 22:27, 28 November 2008 (UTC)

This point about adjectives in -ly is made in some style guides, Wavelength. But these are not strictly exceptions to any statement that is about adverbs. Kindly can be an adverb or an adjective. It is an adverb in Kindly given gifts are better than grudgingly given ones; it is an adjective in a kindly-seeming teacher, like good in a good-looking girl.

But I don't think we should strive to educate users on every such point of grammar!

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 19:48, 29 November 2008 (UTC)

Here are some illustrative sentences.

Jolly Restaurant is famous for its yearly barbecued venison.

Jolly Restaurant is famous for its hourly-barbecued venison.

This box contains red and yellow apples.

This box contains red-and-yellow apples.

This box contains only yellow apples.

This box contains only-yellow apples.

This box contains the only yellow apple.

This box contains the only-yellow apple.

-- Wavelength (talk) 19:32, 2 December 2008 (UTC)

This international expedition has one zoologist from each country.

The likely American zoologist is John Smith.

This international expedition has only one zoologist.

The likely-American zoologist is John Smith.

-- Wavelength (talk) 22:44, 3 December 2008 (UTC)

Comparing characteristics of this extant basal angiosperm, more derived flowering plants, and the fossil flowering plants may give us some idea of the characteristics of early flowering plants and how they have evolved, or changed through time. (Amborella, paragraph 1)

Viola beckwithii, also known as Great Basin Violet or Sagebrush Pansy, is an early-flowering plant of sagebrush country in the Western United States.

-- Wavelength (talk) 16:08, 5 December 2008 (UTC)

This harbor is a natural bay (thus providing protection from easterly or westerly storms) and sheltered on the north by Grand Island. (Munising Rear Range Light, paragraph 1)

For instance, the Continental Divide in North America divides the mainly easterly-draining Atlantic Ocean and Arctic Ocean basins from the largely westerly-flowing Pacific Ocean basin. (Stream#Drainage basins) (In this sentence, the dominant criterion may be the presence of the two additional adverbs, "mainly" and "largely".)

-- Wavelength (talk) 23:32, 5 December 2008 (UTC)

All very interesting, Wavelength. In the last few days (and for the next few) I have only restricted internet access; but I have been thinking about such examples myself, especially some with only and early. These two show up as adjectives or adverbs, and they are exceptional in that they are not analysable as adjective (in current English) plus -ly (in the manner of kindly). These two, and any of the others that are like them in that way, are genuine exceptions; but no major guide to style mentions them as exceptions. Your early-flowering plant has adverbial early; and it is good, I say. I am sympathetic also concerning the mainly easterly-draining ~~Atlantic Ocean~~...basins for the same reason: we want it to be immediately clear that early and easterly function as adverbs modifying the participial forms flowering and draining in compound premodifiers for plant and ~~Atlantic Ocean~~basin, and not as adjectives directly modifying those nouns. Sure, the meanings might often exclude such a possibility, after brief reflection; but the immediacy is important. Anything that delays or jeopardises the reader's understanding is regrettable.

I think your suggestion has some force: "In this sentence, the dominant criterion may be the presence of the two additional adverbs, 'mainly' and 'largely'." But it is not decisive. Note that in your thus providing protection from easterly or westerly storms, easterly and westerly are unambiguously adjectival.

And so on. But these question always come back: How much detail on borderline or complex cases should we give? How much training in grammatical analysis should we engage in? Most style guides minimise such detail and such training. The present brief guideline already does more than some major style guides do on this point, and none does more. I would need to see a strong argument for departing from that standard, and for weakening a clear statement that covers the great majority of cases. Let us remind ourselves again: already there is a general caveat that not all has been said, but that these are the important principles.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 07:47, 6 December 2008 (UTC)

Thank you for your reply. The noun phrase "Atlantic Ocean" becomes a noun adjunct phrase, together with "Arctic Ocean" modifying the noun "basins", and the compound adjective
"easterly-draining" also modifies the same word "basins".

If we leave it up to editors to use "reasonableness" in cases which have not been explicitly decided, then disputes may arise when editors have different points of view about what is "reasonable". As far as I know, no two people always think exactly alike. We could say the following: "A hyphen is not used after an -ly adverb (wholly owned subsidiary) unless part of a larger compound (a slowly-but-surely strategy) or part of an expression where disambiguation is important (early-flowering plants)". How does that weaken the current statement?

Whatever your response to this message may be, I probably will not pursue this subsection much further. Participating on discussion pages is more time-consuming than simple editing, partly because of the importance of being balanced with assertiveness in communication, without being aggressive or passive.

[At this point, I wish to suggest that Wikipedia have one or more pages of instruction in assertiveness, and one or more quiz pages on assertiveness. They can be organized to progress from elementary concepts to more advanced concepts. Even the editor most adept in assertiveness (whoever that may be; I am not that person) can benefit from such instruction, because of refreshing forgotten points, and because of having to deal with difficult people. Designing these pages can be a task for one or more editors with professional experience in assertiveness training. I may or may not make this suggestion at another time and in another place. Is there a place where new pages in the Wikipedia namespace can be requested?]

-- Wavelength (talk) 00:55, 14 December 2008 (UTC)

Wavelength, thank you for your efforts in finding an acceptable solution. While I have stood implacably against mere laissez-faire hand-waving, I am not against rational changes that do not weaken MOS. Since you and a few others favour specific focused changes to the premodifer guideline, and since Anderson clearly wants to allow for exceptions (though his suggestions have been regrettably vague and erosive), I'll now incorporate something consensual arising out of this long discussion. It will have to be a little technical – beyond the detail of any other style guide. But if that's what editors want, that's what I can competently offer them, based on your work and others'.

As for the rest, I agree that we must be assertive. Sometimes valuable contributions are not pushed enough, or not even offered, through a lack of this quality. Surely Wikipedians are encouraged to be assertive, here and there (WP:BOLD springs to mind). Equally, training is important for editors (and here Tony's fine tutorial program springs to mind). But in my opinion, combining these desiderata so we offer training in assertiveness is straining what Wikipedia ought to be aiming to achieve.

Whatever your own perceptions of your role here, Wavelength, I for one think highly of your work. It seems to be getting more and more assured and insightful. Please continue!

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 04:29, 14 December 2008 (UTC)

Thank you, Noetica, for incorporating a mention of these cases, and thank you for your estimation of my work. -- Wavelength (talk) 06:37, 14 December 2008 (UTC)

[Unindent] Fine, Wavelength. Sorry I didn't also respond to your correction about what modifies what (basins, etc.). Of course you are right; I was blinkered, and got those details wrong while I homed in on what I thought was central. We don't disagree at all about those details, in fact. I have patched things above so that no one would be distracted by my error.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 10:42, 14 December 2008 (UTC)

Respect for MOS and its editors

I draw editors' attention to a contribution at this talk page:

*Support but oppose change to emdash. Like much of the Manual of Style, WP:DASH is the product of language cranks, who see WP as a means to reform the English language to their liking. Ignore in case of nonsense. Septentrionalis PMAnderson 19:16, 7 December 2008 (UTC)

Anderson's remarks were prompted by another editor's misunderstanding of hyphens and dashes, and ignorance of the guidelines in WP:DASH. I posted a response to that editor, pointing out what our guideline actually recommends. Anderson, on the other hand, was either ignorant of those recommendations, or blatantly ignored them and took the opportunity to denigrate our work and to diminish the standing of MOS in the eyes of the Wikipedia community.

As one of the editors who has worked most to refine our guidelines on punctuation, especially those on dashes and hyphens, I strongly object to this behaviour. It goes beyond Anderson's typical style of misrepresentation (at the Language Reference Desk, for example). It is an unusually focused sniping attack, from afar. We have a tradition of vigorous discussion within this talkpage, but Anderson's comment is made on a distant talkpage for an article few of us are likely to see, where he probably thought we would not notice it.

Here are my suggestions:

That Anderson apologise to us here, and wherever he has mounted such assaults on our efforts at MOS.
That Anderson withdraw from involvement here. He plainly does not support the idea of a Manual of Style that provides actual consensual guidelines, painstakingly researched from respected sources and adapted to Wikipedia's needs by discussion: argument, weighing of evidence, and reasonable compromise between the opposing requirements of a realistic pluralism and the consistency that is essential in any large encyclopaedia.
That editors working here explicitly acknowledge our central aim: to guide style (not to dictate it, not to force it, not arbitrarily to constrain it).
That editors debate MOS's role within Wikipedia in appropriate other forums (since that is ultra vires, for us here), but without misrepresention or slander.
That editors work collaboratively and constructively here, or get out.

We are not "language cranks" simply because we make guidelines as all style guides must. We achieve a better result than many guides, in fact, because of Wikipedia's unique collegiality.

Some months ago I withdrew from editing here (though statistics still show that I am the second most prolific editor of MOS). I did so without saying why. In fact, it was mainly because of Anderson's relentless negativity, which I thought made progress impossible beyond a certain plateau. I came back in order to address Anderson's continued insidious weakening of a guideline, in an area where I was singularly well equipped. Now once again I intend to withdraw, after this current matter is dealt with: tranquilly, and without any sort of personal difficulty. It's simply a matter of not wasting time.

Since there is reason to believe that Anderson will continue unrepentant and uninsightful, I suspect that I will only return to editing here if Anderson does in fact withdraw.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 01:38, 8 December 2008 (UTC)

Don't say that, Noetica—it's exactly what Anderson wants, since he appears to have time to burn; his is a "win by waiting them out" approach. I endorse your suggestions. Tony (talk) 01:50, 8 December 2008 (UTC)

My judgment is based on this talkpage, and its even more arbitrary and provincial sister pages. While I regret that Noetica has not read what I wrote with any care (I did not say all editors, which would indeed have included Noetica and myself), I stand by it. Septentrionalis PMAnderson 15:35, 8 December 2008 (UTC)

A transparent attempt at evasion, Anderson. I have read what you wrote with care, and I have made a considered and full response. But you have not read what I have written.

For what it is worth, I am obviously among the targets of your sniping – indeed, I am at the cross-hairs more than anyone else. I acknowledge responsibility for WP:DASH, the very section you contemn as "the product of language cranks". Having recently been bettered in open debate concerning a related guideline, you have lashed out from a distant knoll.

The five suggestions I make above still stand. You have only succeeded in demonstrating their urgency.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 22:57, 8 December 2008 (UTC)

I didn't realize that Noetica claimed ownership of WP:DASH; I thought I recalled xem complaining of its inadequacies. We can discuss what needs to be fixed later, when xe has calmed down. I do regret stomping on a corn.

I do not propose to withdraw. This is not a private club; all Wikipedians are welcome here, despite the hostile response which feedback from the unwashed masses who actually write articles usually gets.
I don't see why it should be ultra vires for a talk page to discuss the nature of its main page; that's what talk pages are for.
I will believe that the central aim of this page is to guide style when it is phrased so as to do so; indeed, I came here precisely to secure a page which would guide style; the present text is precisely an effort "to dictate it, to force it, and to constrain it arbitrarily", according to the whims and prejudices of the editors.
Likewise, I agree that this page should be "actual consensual guidelines, painstakingly researched from respected sources and adapted to Wikipedia's needs by discussion: argument, weighing of evidence, and reasonable compromise between the opposing requirements of a realistic pluralism and the consistency that is essential in any large encyclopaedia." When has this talk page done any of these things? Not recently.

Septentrionalis PMAnderson 23:09, 8 December 2008 (UTC)

I too endorse Noetica's suggestions, for all of Tony1's reasons. After I added some shortcuts to what I thought was a reasonably stable MOS front page (to quote Jim Carrey, "I was way off!"), I noticed the whole hyphen debate/edit war thing. He kept trying to change a working sentence unilaterally, in a very uncivil manner, and now it is flagged "under discussion". Now he is making personal attacks against the style guide's writers and trying to, as Tony1 very accurately said above, "dilute the authority of the MoS" by only wanting it to list a bunch of different styles of English, and (also mentioned above) he has attracted prior complaints. In my eyes, he's free to go and free to block. As for Noetica, please stay and continue your work on this important guide; do not feed the troll with another exit. --an odd name 05:34, 9 December 2008 (UTC)

"to guide style (not to dictate it, not to force it, not arbitrarily to constrain it)": eloquently put. I reckon it belongs in the WP:MOS lead. Hesperian 23:14, 8 December 2008 (UTC)

Well, fancy meeting you here. I second this suggestion; I would include Noetica's editing program as well. Septentrionalis PMAnderson 23:26, 8 December 2008 (UTC)

Not to stir a well-stirred pot, but the section heading is a tad... um...; anyhow, when it came up on my watchlist, I had to investigate. Hesperian 23:34, 8 December 2008 (UTC)

There are two circumstances under which I would withdraw.

If this set of pages gave up its pretenses to an "authority" it has not earned, and does not deserve; it ceased to be as justification for bots, scripts, and other forms of autopilot vandalism; and the blanket references to it were removed from WP:WIAFA and WP:WIAGA. Then editors would be free to consult this page, and would believe as much as they chose. This would mean that the self-appointed language reformers would have to actually resort to persuasion instead of bullying, but every benefit has its cost.
If this did indeed become an effort to guide, to explain to editors the advantages and disadvantages of the various possibilities in English, to treat other Wikipedians like adults, and gave sources for what it claimed, my work here would be done. But I don't expect that before WP:DEADLINE; surprise me.

Septentrionalis PMAnderson 23:26, 8 December 2008 (UTC)

Concerning the accusation: "WP:DASH is the product of language cranks"—Anderson, please make me a "Language crank lives here" template to position at the top of my talk page. Tony (talk) 02:42, 13 December 2008 (UTC)

OK, so first of all I have to say that I don't have any real problem with what I've seen about the dashes. Most of it looks pretty good; not that I've read it in detail. As a general rule the MoS people seem to come up with stylistic recommendations that are by-and-large reasonable most of the time, and I have to give them that. A lot of Sept's objections to them do strike me as kind of contrarian for its own sake.

However I have to commend him for fighting the good fight against the arrogance and self-importance of the MoS regulars, which is really seriously out of bounds. Tony keeps talking about "diluting the authority of the MoS". The MoS has no "authority"; that it does is a self-delusion, one that doesn't get challenged much on this page by anyone but Sept, for the obvious reason that anti-centralists are for the most part not interested in participating here, because they don't value the MoS project as much as the centralists do. That fact does not provide any support whatsoever for the centralist view!

I generally can't be arsed to worry about it. I'm thankful there's someone who can. --Trovatore (talk) 09:19, 9 December 2008 (UTC)

Indeed. Having some guidelines on optional things like how to format dashes is useful, but giving a central page "authority" about such details against the wishes of content contributors can be counterproductive. Important things like sourcing and neutrality of articles are not uniform across Wikipedia; achieving a consistency in the formatting of dashes and ellipses is a far less important question in building a quality encyclopedia. Demands for "respect our authority" are very much against the usual spirit of Wikipedia. Please try to work constructively with others, do make up some rules and expect everybody to follow based on your own perception of a nonexistent authority. Kusma (talk) 10:25, 9 December 2008 (UTC)

As the (quote) "ignorant" editor (no offence taken, don't worry) at the talk page who set this ball rolling, I have to say that, although MoS is indeed very useful guidance, I thankfully don't take it half as seriously as you lot do. Ghmyrtle (talk) 00:03, 11 December 2008 (UTC)

Well said, Trovatore, Ghmyrtle, and PMAnderson! The MOS is merely a recommendation or an exhortation. It is not policy that can be enforced on an article against the consensus of the editors for that article. Using Bots, AWB, and scripts to force compliance with the MOS in hundreds of thousands of articles is completely out-of-bounds so long as MOS remains anything less than mandatory. Tennis expert (talk) 05:13, 15 December 2008 (UTC)

"instructions for blank alt text" edit revert by AnOddName

AnOddName reverted this edit, calling it opinionated and disruptive. Since the text linked to this section of WP:ALT, which says:

In Wikipedia, providing completely empty alternative text requires the use of a <nowiki> element, like this:

[[Image:source-of-image|<nowiki></nowiki>]].

and had linking text which said: instructions for blank alt text are here, what is the opinion, let alone the disruption? (I thank Xer Oddity for the chance to improve the wording since).

If I ever have to install alt texts, I'll want that instruction; it's not obvious. Septentrionalis PMAnderson 18:32, 9 December 2008 (UTC)

For the <nowiki> stuff: you're right, that part is non-obvious. After your past edits, I've begun to lose faith in you, but I won't revert that instruction again. --an odd name 18:46, 9 December 2008 (UTC)

"Please try to work constructively with others, do make up some rules and expect everybody to follow based on your own perception of a nonexistent authority." Can someone untangle that for me? Tony (talk) 16:28, 10 December 2008 (UTC)

A trivial emendation yields "Please try to work constructively with others, don't make up some rules and expect everybody to follow based on your own perception of a nonexistent authority." The slip occurs when Kusma shifted from paraphrasing the position he disagreed with to stating his own. Any competent editor should be able to do that much. Septentrionalis PMAnderson 17:34, 11 December 2008 (UTC)

A minor clarification about the new software introduced in October 2008: The <nowiki></nowiki> trick above is needed to blank the alt text only when no thumb or frame parameter is specified. In the latter cases, which I believe hold for most images in articles, the alt text is blank by default. In other words, if nothing is specified for the alt text in the Image tag, then the alt text is automatically empty. Proteins (talk) 18:10, 10 December 2008 (UTC)