According to the [U.S. Navy Style Guide] under "ship names," "Do not use 'the' in front of a ship's name: 'USS San Jose,' not 'the USS San Jose.' " Darkfrog24 (talk) 13:44, 25 July 2009 (UTC)[reply]

Another source, The Guardian (UK) Style Guide, uses the phrase "mariners shun the definite article" when talking about ships. An observation I personally might add is that the confusion may stem from the concept of ships as structures, as with "the Empire State Building" or "the Brooklyn Bridge". This use of the definite article is different because the structure name is in the form adjective-noun, building or bridge being the noun. Ships, to me, are not unlike thoroughbreds in that the name is a proper name like "Secretariat" or "Man o' War" (!) although the ship name has often had "the" added by writers in the past because of the confusion with buildings. Therefor, I would say that the above summation by Ekem is wrong on the third point, it should be "the article is generally not used" with the PDF I'm linking above providing a further explanation (under the "Military tendency" heading). Sswonk (talk) 14:56, 25 July 2009 (UTC)[reply]

I think a global rule on the use of the definite article would be a mistake. For instance, I would never say "I got seasick a once while riding James." when I meant "I got seasick once while riding the James." FWIW, in day-to-day operations, the only time I don't use "the" before my ship's name is on the radio, for example i.e. "Suez Pilots, this is Fredericksburg," but "The Fredericksburg was a deathtrap." and "I'm so glad we scrapped the Fredericksburg." Examples on both sides go back at least as far as the Argo. Cheers. Haus^Talk 04:12, 28 July 2009 (UTC)[reply]

There are strong examples for the use of the article and it will be hard to ignore Melville, Conrad, O'Brian or the New York Times (i.e. [1]). This is a nice observation of the dual use, though, and I concur (personal observation) the article is omitted on the ship radio, yet when talking about a boating experience the article tends to be used. Checking on cruise ships, companies refer to their ships without article (at least the two lines I looked at), but people would say "we went on the xxx" when talking about their cruise ship. So, is there a distinction between a "mariner/operational use" and a "literary/general use" - I am struggling with these terms but throw them out here to help us move forward - ? Ekem (talk) 13:33, 29 July 2009 (UTC)[reply]

Certainly it seems that when the ship is personified or being portrayed as an active agent (as in, for example, when it is the ship herself that is figuratively making a radio call), omission of the article is much more common than in other situations. Powers ^T 14:50, 29 July 2009 (UTC)[reply]

All very good points. One of my motivations to make the suggestion above, "the article is generally not used", was that "the" isn't used in a lot of the articles (this can be dangerous i.e. Article (grammar) vs. Article (publishing)) in WP:SHIPS that are FAs, see USS Constitution and SS Kroonland. Another good example might be Vasa (ship) where "the" is used sparingly (once) when talking about the ship but multiple times when talking about books, a museum, in the map and so on. So if a standard isn't spelled out properly I would fear having to chase down edits to these and other articles where the well meaning editor decides all of the mentions of the ship need to have a definite article. Sswonk (talk) 15:09, 29 July 2009 (UTC)[reply]

Are you saying here, it does not matter what may be the correct answer as we moved already into the "no article" version?Ekem (talk) 01:46, 31 July 2009 (UTC)[reply]

No, that wasn't me saying it doesn't matter. I wrote about FAs under the scope of the WP:SHIPS project and also the very good and careful eyes of many of its members. What I meant to convey there was that these articles have passed our highest form of vetting and stand as good examples. There are also the examples of the two published style guides cited and the comments I and other editors provided above. The problem is popular usage, which sometimes represents a different standard than what is correct. This happens with subjunctive (also see) mood verbs: "I wish I were going to the park" is correct for formal writing but "I wish I was going to the park" though incorrect is often found in popular usage. Judging by Google hits, "was" is more common by a wide margin: searching the exact string "I wish I were" produces 3.6+ million hits while "I wish I was" produces 16.1+ million hits. The same might be true here. So I would absolutely not say "the article is generally used"; in spite of the usage in novels (not formal writing, possibly in character etc.) and the New York Times (anecdotal examples), the other examples show a strong preference for the proper name style among mariners and people here who write about ships. Presenting both versions with explanation in this style guide might be possible, but ultimately pointless, as I don't think we are going to be able to say either form is truly correct or even "generally used". Sswonk (talk) 05:03, 2 August 2009 (UTC)[reply]

I respect people who are writing here immensely, but -just on their own- they are not reference points for style. I would think that reference points are literature, popular use, style books, and major newspapers (this includes the New York Times - the Times also writes about "the USS Ronald Reagan"^[1] - and the Times' use of the article has been consistent over time and (print)space, not "anecdotal"). It appears to me that the omission of the article is seen primarily in writings by mariners and in some recent mariner literature. Ekem (talk) 01:54, 3 August 2009 (UTC)[reply]

To clarify, I used the word "anecdotal" to describe the use of random example articles to show usage in the Times as opposed to use of an entry in the Times official style book, if such an entry has existed at any point in time. Of course I did the same thing with the example FAs, it is just a distinction of the source of the information that I was making. Sswonk (talk) 13:13, 3 August 2009 (UTC)[reply]

Here is another try to summarize: Literature, major newspapers (not only the NYT but others including the above mentioned Guardian), Style book(s) (only one mentioned above), and general use apply the definitive article before the name of a ship (including "the USS (xxx)"), while mariners may use or omit (including the US Navy) the definitive article (above are examples for both). A pronoun within the ship prefix (ie HMS) makes, of course, the definite article grammatically superfluous. In Wikipedia editors have been applying both usages, even within articles, apparently based on the preference of the editor. Ekem (talk) 00:13, 13 August 2009 (UTC)[reply]

That seems accurate to me. The guideline Wikipedia:Naming_conventions (ships) has allowed both cases for years. Here's another old conversation on the same topic. Cheers. Haus^Talk 00:17, 14 August 2009 (UTC)[reply]

(Linking to this from numerous places)

There's some debate at Talk:List of film duologies regarding whether the nascent term "duology" should be used on Wikipedia. It refers to a two-part work of fiction; it's not listed in any established dictionary, but is widely used on the web. The issue seems to be one of brevity versus clarity. Here are a few of the places the term is in use:

• Duology is a redirect to Sequel
• wikt:duology
• List of film duologies
• The Hand of Thrawn duology, Category:The Hand of Thrawn duology
• Cygnus X-1 duology
• Template:The Waxwork Duology
• Template:The Terror Toons Duology

There was some prior discussion over this last August at Wikipedia_talk:Naming_conventions_(films)#Problems_with_film_series_titling_guideline. The consensus at that time appeared to be that film series should use "(film series)", and by analogy I presume that "(series)" or "(book series)" is appropriate for literature. That leaves open the question of the use of the term in the bodies of articles, and the use of it in the names of list articles like List of film duologies (should it be List of two-part film series?). What's the final verdict on "duology"? Dcoetzee 05:33, 28 July 2009 (UTC)[reply]

Well I hope the term "duology" is correct, as I've been using it professionally to refer to a pair of literary works on a common theme by the same author! I dislike the use of the term "series" to refer to only two works – in publishing, a "series" is four works or more. Etymologically, it should be "dilogy", but "dilogy" means a word or phrase which has two meanings: "dilogy" is not a dilogy, although it would be kind cute if it were… Physchim62 (talk) 08:31, 28 July 2009 (UTC)[reply]

Yes, the word duology is ugly; but so are most new words. And yes, it is not well-formed from Greek; but then, neither is television (first part Greek, second part Latin). And yes, dilogy is already taken (as well as "an ambiguity", as Physchim points out, it can mean "2. Repetition of a word or phrase, in the same context. In recent Dicts" (OED). Looks like it fills a niche that diptych does not fit into neatly. We can't recommend against it without also recommending something better in its place.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 09:28, 28 July 2009 (UTC)[reply]

Agree with Noetica ... except that I have no objection to Greek/Latin combos! Tony (talk) 11:25, 28 July 2009 (UTC)[reply]

Normally, I am fine with "duology" and "trilogy", but my problem is with the {{Film series}} template that goes beyond these: tetralogies, pentalogies, hexalogies, heptalogies, octologies, ennealogies, decalogies, and polylogies! I mean, come on. "Trilogy" is the most well-known of the bunch, and "duology" is known to a lesser extent, but the rest of them seem like an overly complicated way to be consistent with trilogy and duology. I think we should pursue a different naming convention -- something like "List of film series with two entries". —Erik (talk • contrib) 12:00, 28 July 2009 (UTC)[reply]

Perfect illustration of how the template mentality can go awry. Tony (talk) 14:29, 28 July 2009 (UTC)[reply]

Agreed. It seems that even the film world stops using number-specific terminology once they hit four films. As for "duology," I'm of a divided mind on the word itself, but "List of film series with two entries" seems quite unobjectionable. However, it also sounds as though not all film series that happen to have two entries are true duologies, just as not all film series that have three entries are true trilogies. Therefore, our real criterion should be what exactly it is that we're trying to describe. Darkfrog24 (talk) 17:04, 28 July 2009 (UTC)[reply]

When I skimmed the lists of film series with x entries, they seemed well-updated. If a film series gets one more film, the list of films is moved to the successive "list of film series" article. My real concern is for the article titles. "Duology" and "trilogy" can be somewhat misleading because it sounds like an intended setup, where it's not always the case. For example, American Psycho had a DTV sequel that really had very little to do with the first film (and actually completely missed the purpose of it). Using "List of film series with x entries" makes the issue more ambiguous since unlike the tetralogy/pentalogy/hexalogy names, it does not sound like a preconceived setup. —Erik (talk • contrib) 17:39, 28 July 2009 (UTC)[reply]

Leaving aside the quality of the articles, Wikipedia has a policy where neologisms should be avoided. As "duology" cannot be established in any dictionary, this alone should be enough for it not to be used, but as additionally there is already controversy as to whether "dilogy" should be used in its place, then there is no question something else should be used. Serious consideration should also be given as to whether pentalogy / hexalogy / etc should be used as these words do not appear in dictionaries. As a couple of other posters point out however, with the articles in question (as well-updated and meticluously put together as they might be) it isn't an issue, as they are not lists of true "trilogies", etc as they include works that do not follow a overall cohesive narrative arc, etc, so to my mind these should all be moved. One could argue that some of the entries on these lists do not even form true series, but that's probably an argument best left for another day.

I'm unsure what to suggest as an alternative for the word "duology", and they do get more clumsy as the numbers get higher. If in publishing three is a "trilogy" and four or more is a "series" (noting that four is not a "tetralogy"), this would therefore seem to be the appropriate convention to follow for films with more than three in a series. As, in the case of the film lists, this would be a simple movement even for the "trilogy" list as as we have established above, a lot of these do not form coherent trilogies. "List of film series with x entries" would seem to be the appropriate form as others mention above, even for two. Others can argue as to what counts as part of the series and what doesn't - I believe Star Wars is in under "ennealogies", which for me is two trilogies, or a series of 6 films.

I think that we just don't have a word for a series of two works (although i don't have a problem with "a series of two" myself). I don't think that it is up to us to invent one, and neither should we use "duology" until "they" put it in the dictionary. Robsinden (talk) 12:21, 5 August 2009 (UTC)[reply]

To satisfy my own curiosity, it would appear "Star Wars" has moved from "ennealogies" to "decalogies" and now back to "hexalogies". I wonder about the necessity of these lists! Robsinden (talk) 12:32, 5 August 2009 (UTC)[reply]

Please see Talk:Film_series#Requested_move - Robsinden (talk) 12:13, 10 August 2009 (UTC)[reply]

Why is it that most FACs I review now have microscopic images that are utterly useless unless you double-click on them? Is there scurrilous guidance here that is making nominators ruin this already-weak part of WP (criticised in an external report written up in the most recent Signpost, I think). It seems weird. Have our readers got Internet connections that are so snail-paced we have to go backwards to the 20th century? And, BTW, who wants to see tiny-wrap skyscraper-tall captions with one-to-three words in most lines? They look ridiculous. See here, for example. Tony (talk) 12:17, 3 August 2009 (UTC)[reply]

As an aside, I agree with Tony that the example was ridiculously small, and have fixed it in the Donnchadh, Earl of Carrick article. On the left is original the example again, so that you can see what it was like before it was fixed. Eubulides (talk) 23:13, 3 August 2009 (UTC)[reply]

Good heavens, is it possible for you to write a criticism without sounding apoplectic? Anyway, the article you linked looks fine to me. Maybe you can adjust your thumbnail size in preferences? Powers ^T 12:38, 3 August 2009 (UTC)[reply]

Tony, there seems to be something in the MoS that makes editors think only thumbnails are allowed, except for the lead image. Certain editors go around removing image sizes, citing the MoS. With a couple of articles I've written, I've suggested on the talk page that we could try to get them to FA standard, and one of the first responses has been, "You'd need to get rid of the fixed image sizes." The problem with using thumbnails is that the sizing is inconsistent. Most of the time they took tiny, and sometimes enormous; see right for an example of the latter. SlimVirgin ^{talk|contribs} 12:48, 3 August 2009 (UTC)[reply]

Re "something in the MoS that makes editors think only thumbnails are allowed, except for the lead image" - it's the "one size fits all" approach of some MOS zealots, and their forgetting that MOS is a guideline, i.e. WP:IAR applies. The idea that "only thumbnails are allowed" has clear weaknesses, e.g.:

• It's no use to readers, most of whom are unregistered so can't set prefs.
• It depends on the pic. E.g. a simple pic such as most flags can be shown pretty small, but many maps and diagrams need to be larger.
• It also depends on the use. For example File:Atrax robustus.jpg can be shown if it's just a pic of the Sydney funnel-web spider or of spiders in general, but needs to be enlarged if it's used to illustrate the modification of the chelicerae into venom-injecting fangs, which is spiders' signature feature.

Such points should be left to the judgement of editors and reviewers. --Philcha (talk) 13:06, 3 August 2009 (UTC)[reply]

We've had this discussion several times on this page that I recall, and every time there is no consensus to force people to use thumbnails, yet somehow it has crept into the MoS: "Generally, use the thumbnail option ("thumb"), which is available in the image markup. ... As a rule, images should not be set to another size (that is, one that overrides the default)." SlimVirgin ^{talk|contribs} 14:02, 3 August 2009 (UTC)[reply]

Something I've suggested on the WT:FAC page that I think could be considered: We ask for thumbnails for images when they are to be flush alongside text. As we are still (?) designing towards a 800px resolution monitor, these should never be beyond 300px to avoid blocking off too much text. Yes, we should avoid setting a size and let "thumb" do its job, but that shouldn't be a hard-set rule. But there are points where images need to realistically be larger than 300 px horizontally to do their job on the printed version of the page. In this case, I propose that if the image needs to be displayed (and for print) at a size larger than 300px (and all NFCC considerations are met for that, if non-free), that we allow for "inline" images that do not run flush on text (see, for example, the Los Angeles panoramic view). This would need to use some special formatting to make sure there's text clear breaks on both sides if the image is >300 but less than 800 px, but technically not an issue. --MASEM (t) 14:10, 3 August 2009 (UTC)[reply]

Hi Masem, two questions: first, why do we ask for thumbnails? There has never been any consensus for that, so far as I know. And secondly, what do you mean by "flush on text"? I don't really understand the point you're making. SlimVirgin ^{talk|contribs} 14:20, 3 August 2009 (UTC)[reply]

I would support a change to the MoS allowing editors to use the image size best suited to the article. Darkfrog24 (talk) 15:05, 3 August 2009 (UTC)[reply]

I personally don't know the initial reasons why we ask for thumbs, but I would say that now they serve two purposes: they help to normalize the sizes of images in articles when the various images are all ranges of sizes (from what could be freakin' huge 2000x2000 free images to, say, reduced non-frees of standard TV/computer size dimensions (eg 320x240); and using thumbs from 180 to 300 px help to keep the page layout across WP pretty consistent without having to worry about browser/OS/font size issues (eg, we are trying to avoid pixel-perfect placement).

"Flush" means, as the example at the top of this section, the text runs as close as possible (with some margins) next to the image. This is compared to, say, if I were to put an image on a line by itself with no text running flush along either side of it. --MASEM (t) 15:21, 3 August 2009 (UTC)[reply]

BTW, the image I cited at the top of this section has a caption that is ... wait for it ... 14 lines deep, most of them three or four words across. Here is the current text that appears relevant to this issue:

This is what I suggest, although I'm an amateur at image control, so please weigh in with your suggestions and we can create another draft below if necessary. In particular, I've no idea about the 300 versus 400 pixel issue:

Tony (talk) 16:34, 3 August 2009 (UTC)[reply]

Nice. The line about editorial judgment is exactly what we need. However, I think it would be best to distinguish more clearly between description and instruction. This stuff is confusing for beginners and I don't think we want to confuse information on what images are with instructions on what we want the users to do. For example, "Most images are 300 pixels wide" is confusing because I've seen images of all shapes and sizes.

I think any solution that puts the "thumb" option as the secondary option under "allow the editors to decide" is going to be a problem. Knowing those that love to use images, if you give them up to 300px, they will use 300px, even if its not necessary for an image. You've also just nuked the reason a user may choose a different image thumb preference.

An optional approach may be to consider changing the default thumb size from 180 to around 220-250px (a setting in the WP wiki config, I believe), with the noted allowances for going beyond this if necessary. This would fix most of the "walls of caption text" that seem to be at issue, and retain the user preferences aspect. If anything, we then need to emphasize that it is appropriate to break out of the "thumb" size when it makes sense, being a lot more lenient and IAR-ish about that fact. The only limiting factor is why we chose 180px in the first place - if it's for NFCC issues or the like, we may have to stick to that. (That subtly comes into play here, it should be noted). If the 180px was due to file size and typical transmission issues, that's likely no longer a valid reason of concern with the proliferation of high-speed today. --MASEM (t) 17:13, 3 August 2009 (UTC)[reply]

• How about starting: "On Wikipedia, most images are narrower than 300 pixels, with a maximum for most articles of 400 pixels wide."? Tony (talk) 17:40, 3 August 2009 (UTC)[reply]
  • I think we've got two aspects going on here. We want to try to funnel images to <300px (what that is, well, that's an issue). The next question is that when an image has to go over 300px for good reason, how do we deal with it? And that's where the 400px is - it's not really a maximum image size, but a size where we now span over half the page at 800px, and thus needs special treatment beyond that. If it is larger than 400px then it should be placed without flush text per the {{wide image}} template; if it's less than 400px, then it should be placed aligned in text. --MASEM (t) 17:47, 3 August 2009 (UTC)[reply]

Three things about Tony's proposal:

• He's right about editorial judgement being the final arbiter. Whether it's deciding what info the pic needs to convey or how to solve layout problems (which are harder on widescreen monitors), the editor can often produce a better solution than a set of rules can.
• I agree with Masem that the "thumb" option should be presented as the preferred option in most circumstances, simply because thumbs adapt more easily to change - for example perhaps we'll all have portrait monitors in 5 years.
• We need to avoid the word "should" because it's ambiguous - some treat it as "is preferable", others as "must", and we already have enough arguments over interpretation of "should" and "should not".

So here's my attempt for y'all to shoot at:

--Philcha (talk) 17:53, 3 August 2009 (UTC)[reply]

I'm not in favor of saying the thumb is better, or the default, because it encourages editors to go around removing image sizes, which is the problem with the current wording. I prefer Tony's version, where we stress editorial judgment. SlimVirgin ^{talk|contribs} 18:02, 3 August 2009 (UTC)[reply]

This makes two separate issues: the legitimate complaint against the MOS's prescription of using "thumb" which presently defaults to 180px, mostly due to 180px being too small, and the behavioral problem of those that forget that the MOS, while policy, needs to be treated in an IARish manner, and that going around removing "thumb" without consideration of editorial decisions is not good editing practice. The latter is something outside the bounds of MOS beyond using the best language possible to convey this, but continued "abuse" of the MOS in this manner should be dealt with in appropriate dispute resolution.

So if you take the behavior out of it, we're still at the case where, at least to me, we want to still encourage and funnel users to use "thumb", with strong consideration of upping the default thumb size to 220-250px, but emphasizing that they may consider other sizes if they deem it necessary. --MASEM (t) 18:21, 3 August 2009 (UTC)[reply]

You can't take the behavior out of it, and it's not reasonable to ask us to engage in dispute resolution every time someone tries to enforce the MoS. The wording needs to be changed, so as not to encourage people to impose thumbnails on articles. There's anyway no good reason to use thumbnails. I have no problem setting parameters—no smaller than X, no larger than Y—but otherwise editors have to be allowed to use their judgment. SlimVirgin ^{talk|contribs} 18:26, 3 August 2009 (UTC)[reply]

If people are interpreting, as it is presently worded "Generally, use the thumbnail option ("thumb")...", as that we require thumb sizes, then those people are overstepping their editing bounds. This is a human-decision-based clause, it cannot be done purely mechanically. --MASEM (t) 18:32, 3 August 2009 (UTC)[reply]

• How about changing the default thumb size as suggested by MASEM, and changing the guideline to keep the basic encouragement to use the thumbnail option, but adding encouragement to use proportional image scaling where appropriate. Suggested wording:

That way we get away from the problems which seem to be associated with |px=400 or other fixed sizes. The ideal or maximum ratio for |upright= should be given, alternatively clear advice could be given on using {{tall image}} which seems to duplicate the |upright= parameter. . dave souza, talk 18:09, 3 August 2009 (UTC)[reply]

Dave, what is wrong with letting editors set image sizes (in words a completely non-technical person will understand, if possible)? SlimVirgin ^{talk|contribs} 18:28, 3 August 2009 (UTC)[reply]

I wrote the above before seeing your comment, but left it as "Basic practice is to use the thumbnail option" simply says what the basic way of doing it is, without saying you must do it. Don't mind if you've got better wording, perhaps "The simplest approach is...." My aim is to set out clearly the best ways to change image sizes, and state what the limitations are, if any. The proportionate method using |upright= sounds as though it may fix the problem of |size=400px, but technical advice is needed. Part of the problem is that the same thumbnail view or enlarged view looks much larger in Camino than it does in Safari, don't know about other browsers. The aim is a simple way for editors to play about with image sizes and reach workable and reasonable layouts. . dave souza, talk 19:20, 3 August 2009 (UTC)[reply]

Yes, exactly. Everything depends on the image, the shape of it, what it shows, what it's being used for, where in the article it's going, whether it'll be near other images, templates etc. We can't have one-size-fits-all. Look at the image I posted at the top of this section, for example. That's a thumbnail, but clearly too large. I think we need to remove anything that suggests thumbnails are required or preferred. SlimVirgin ^{talk|contribs} 19:30, 3 August 2009 (UTC)[reply]

This section got pretty long, so I created a new section #Image size rewording below for further comments. Eubulides (talk) 23:13, 3 August 2009 (UTC)[reply]

As it happens, over the past couple of weeks I've rewritten WP:Picture tutorial, particularly its image-size section. I read through the above discussion and a lot of it makes sense, but there are some problems:

• The current and proposed text sometimes suggests 300px, sometimes 400px. It should be consistent. The 300px limit comes from old desktop and laptop displays that are 800px wide. These displays are obsolete and are no longer of practical importance to Wikipedia. Currently, the narrowest displays in widespread use on desktops and laptops are 1024 pixels wide. Of course there are mobile devices with displays narrower than even 800px, but they use different techniques to show Wikipedia images, which won't be affected much by changing suggested max width from 300px to 400px.
• Dave souza's draft has some good suggestions, particularly about proportional image scaling, but it also has some problems, e.g., it doesn't address non-thumbnails and it uses some syntax like "px=300" that doesn't work.
• It's OK to say that thumbnails are in common use (as they are); this doesn't mean they are required or preferred.

The following draft attempts to address the above issues. A sandbox diff shows the difference between this text and what's currently installed.

(This draft is now obsolete; please see the #MistyRose proposal below. Eubulides (talk) 08:06, 4 August 2009 (UTC))

• Most pictures should be between 100 and 400 pixels wide. Common practice is to use the thumbnail option ("thumb"), which is available in the image markup and normally floats the image to the right. This results in a default width of 180 pixels, although logged-in users can set a different default in their user preferences.
• A picture may benefit from a size other than the default. Adding the parameter "|upright=2.2" (or "|frameless|upright=2.2" for non-thumb images) scales an image to about 400 pixels wide by default; making the number larger or smaller changes the image size accordingly. A floating image should generally be no more than 500 pixels tall and 400 pixels wide, so that it can be comfortably displayed next to text on the smallest displays in common use; a nonfloating image can be somewhat wider if it stands alone. The {{Wide image}} and {{Tall image}} templates can display images that would otherwise be impossibly wide or tall. Examples where adjusting the size may be appropriate include:

• Images with aspect ratios that are extreme or that otherwise distort or obscure the image
• Images containing important detail. For example, a map, diagram, or chart may contain important text that would be unreadable at the default size.
• Images where detail is relatively unimportant. For example, a national flag may be easily recognizable even at a small size.
• Images in which a small region is relevant, but cropping to that region would reduce the coherence of the image
• Lead images, which should usually be no wider than "upright=1.67" ("300px")}}

Eubulides (talk) 23:13, 3 August 2009 (UTC)[reply]

I see no problem with the above (and yes, agree the 300px is probably out of date.) 400px is probably the best limit to stick with as to consider the issues with NFCC (as if you take most NFC, being from screenshots, a 400px width image will get you an image file that is just around 0.1 megapixels, which has been suggested as the "ideal" low resolution image, though by far not a hard limit).

I still say we talk about increasing the default "thumb" size to 250px based on the same logic above. --MASEM (t) 23:18, 3 August 2009 (UTC)[reply]

The default thumb size could well be increased too. The above proposal is designed to make sense no matter what the default is, though of course its specific numbers would need to be changed to match whatever the default is. Eubulides (talk) 03:06, 4 August 2009 (UTC)[reply]

If this is aimed at newcomers as well as the technically proficient, it might be a good idea to avoid the word "parameter", which startled me as a non-techie, even though its meaning later becomes apparent. There are several other phrases that mean nothing to newcomers, such as "floating", and others which suggest complexity such as the templates. The style just looks too technical and daunting, I'm sorry to say.

I think, however, that it might be a good guideline if converted into the language of an ordinary reader who just wants to post a picture (which is quite complicated enough already by the multiple-stage process of uploading not to Wikipedia but to the Wikimedia Commons, certifying the rights, and then transcluding it — if that's the correct term).

As for image size, my Compaq FS7600 monitor (on Windows Vista Home Basic) is I think 800 x 600 pixels, but I deliberately used a 400-pixel-wide image of a New York Times cartoon so that the important words within the image would be legible on small screens like mine. (Originally it had been 550 pixels wide, but that caused problems for other users. See Talk:New York City mayoral election, 1917#Formatting and the article itself. See also, for comparison, Talk:New York City mayoral election, 2009.) Although I'd agree that it's bad style to have too much of it, I see nothing wrong in occasionally squeezing wrap-around text so long as the text is suitable for such squeezing (in the 1917 mayoral example, it was a set of bullet points) and you're conscious of what you're doing. —— Shakescene (talk) 06:00, 4 August 2009 (UTC)[reply]

• Thanks for the comments. I attempted to reword the draft to make it less technically daunting, and the result is in the #MistyRose proposal below. Unfortunately some of the stuff can't easily be omitted (e.g., "upright=2.2"), but the words you mentioned ("parameter" and "floating") can be removed, and wikilinks can be added for parameter names like "upright".
• The Compaq FS7600's recommended resolution is 1024×768, as per its user's guide (PDF). Like most CRTs, one can run it in 800×600 mode and it will display larger text that way; this sort of thing used to be more common but I think it's relatively rare nowadays. But perhaps this point is moot if 400px looks OK on your screen.

Eubulides (talk) 08:06, 4 August 2009 (UTC)[reply]

You're right; once I remembered the easy way to look up my resolution, it was indeed 1024 x 768 pixels. The FS7600 is smaller than my previous 19" screen, so the higher resolution gets swallowed up by the tiny display (sometimes commas are hard to distinguish from periods/full stops; and umlauts, tildes and circumflexes from each other.) While I don't have the time or patience to do it right now (being exhausted by trying to reconstruct who was Chairman and who President of the Reconstruction Finance Corporation, and when), perhaps someone would like to see what in fact does happen to New York City mayoral election, 1917 when seen in an 800 x 600 resolution. —— Shakescene (talk) 08:21, 5 August 2009 (UTC)[reply]

All of the above proposals are articulate on why it is good to set image sizes, and make no mention of why it is good to use the default. Some people have tiny screens or terrible bandwidth. These people can improve their browsing experience by setting their preference for very small thumbs. But only if we honour those preferences. When you decide that an image looks better at 350px than your default 300px setting, you're also telling people with a 120px preference that they are going to take this image at 350px whether they like it or not.

There are so many different screen sizes, font sizes, browser layout engines, aesthetic preferences, etcetera, that you are fooling yourself if you think you can construct a layout that looks nice to everyone. In reality, all you are doing is making the layout look nice to you, on your screen. I wish I had a dollar for every edit where someone shrunk an image (for me) with edit summary "bigger image".

My preference is for honouring thumbing. But, failing that, at least don't give us a policy statement that provides a persuasive rationale for only one side of the debate.

Hesperian 23:33, 3 August 2009 (UTC)[reply]

Bingo. There is a reason why thumbnails were preferred, and it's so that users can see smaller images if they want to see smaller images. Or larger ones if the have the space; my default is set to 250px, for example. Powers ^T 02:02, 4 August 2009 (UTC)[reply]

The #Beige proposal (above) addresses both of the previous comments, because it suggests using the "upright=Factor" option instead of absolute pixel sizes. The upright option causes the image to scale according to your preferences. For example, if an article's editor wants a larger image and uses "upright=1.5" to get it, you the reader will see a 375px image if your preference is 250px, and no reader will get an image smaller than they would have otherwise got. Eubulides (talk) 03:06, 4 August 2009 (UTC)[reply]

Just to be clear, 99% of the people who read Wikipedia are not logged in, and have no option to change their viewing preferences. Since the default thumb setup (which is what unlogged readers see) is clearly unsatisfactory to the majority of viewers here, it can be extrapolated that it is likely a problem for unlogged readers too. I can assure you that when I was using a public computer to view some stuff (and was not logged in), the images were so small they were nearly pointless. Thanks to those who are working to let us have pictures that are big enough to see. Risker (talk) 04:12, 4 August 2009 (UTC)[reply]

Risker puts it very well. In this respect, it's like our discovery last year that date-autoformatting didn't work for 99.99% of readers, and prevented us from noticing the glitches that lie behind the display, which had always been there for our readers to see. As a general rule, I strongly advocate that the editors who are responsible for crafting the appearance and formatting of our articles should see what the consumers see—they are the only readers who really count. I'd go so far as to say that WPians should disable their thumbnail-size preference, so they can make proper judgements as to detail and text wrapping for each image in an article.

Dave Souza's version(s). (1) It seems to have lost some of the good bits of mine, in particular the need for editors to balance the amount and size of detail in a pic with the need to avoid the uncomfortably narrow wrapping of text. I think we need to emphasise editorial judgement; we need to encourage editors to treat image size (and placement) seriously as an integral part of a high-quality product. (2) In the green box, Dave's bullets refer to both up- and down-sizing. In the beige they refer to down-sizing (?). It needs to be explicit so normal editors can easily understand it. (3) Except for flags/icons etc, I find 180px to be almost never worth bothering with. Dave's text still seems to encourage or condone the wide use of 180px. If you want a good example of how numerous superb historical pics we are able to use are basically ruined from an in-article reading perspective, have a look at the current FAC "Operation Charnwood". The vividness and life of the large infobox image are completely lost further down. You'll first need to disable your thumbnail size prefs, if you haven't already. This is nano-pic land; whereas this is after I've gone through to force a bigger size on the pics, all still thumbnails. The "Handley Page Halifax" pic under "Preliminary attacks" no longer looks like a close-up of a mineralogical sample: now it's an airplane above the chaos of war. I'd have it even larger—the accompanying text seems to demand it—but I'm being cautious.

Norse–Gaelic example still dysfunctional. I still have a problem with the map I exemplified at the opening of this debate. It is now displayed just below Slim's plastic bottle example of why we need to move on from the one-size-fits-all. Eubulides enlarged the map to 1.8 thingies, but I found that still too small to work out up from down, not even considering the nanoprint. I've raised it in the article to 2.0, but still, the print is undecipherable; you are forced to divert to the original image to learn what any word actually is, and this despite the fact that the caption refers one by one to words on the map. It is dysfunctional in these ways. So ....

Font size. Until now, we have included no advice about the size of lettering on maps and diagrams. It's a pet issue for me, since my clients often have to be encouraged to use larger font-sizes on their diagrams and pictures to avoid irritating the assessors. I don't know why people aren't fussy about this when there's plenty of space in which larger font-size can lie, such as on that map. (PS In fact, I find it very hard to decipher the wording in the original, large version. The dark-green colour doesn't help one bit, but even the yellow-background font needs significant boosting. I note also that some editors have complained about the inclusion of image titles in the image itself, as here; I have sympathy for this complaint.)

Brevity in captions. Whenever I look at a caption, I seem to be able to find ways of removing words with no loss of meaning. We have never encouraged editors to be vigilant about language bloat in their captions—particuarly the usual redundancies. I don't advocate stubby, truncated language in captions, but I do think it's even more important to avoid the vertical rise effect where it's easy to do so.

Therefore, I wonder whether there is support for include brief, "soft" advice to editors on font size within images and the need to avoid redundant wording in captions. This could be included at the same time as we get the advice on image size right. Again, the emphasis should be on encouraging editors to skill up and use their creative judgement for our readers' sakes, by outlining the editorial issues at stake. Tony (talk) 04:30, 4 August 2009 (UTC)[reply]

We do need to set up some guidelines for fonts in maps and charts, and in fact the encouragment to generate free versions of these in SVG when possible to avoid the destruction of readibility when images are shown at low size. But that's a much larger issue and doesn't seem as pressing to make sure we guide those actively seeking to limit images to "thumb" size to avoid doing that.

That said, there is still an issue of why 180px was picked in the first place that my Google-fu is not easily figuring out from talk page archives. It may be associated with NFCC policy, in which case, then, there's likely good reason to stick with it as the default thumb size. But if it's "just because" then damn the torpedos and lets get the default thumb up to 250px and emphasis IAR in allowing larger images. --MASEM (t) 04:55, 4 August 2009 (UTC)[reply]

The NFCC concerns are not related to default thumbnail widths; please see #NFCC and default width below. Eubulides (talk) 18:17, 6 August 2009 (UTC)[reply]

Just in case it's getting lost in the details, the original problem that we're trying to correct is that some users (mis)interpret the current phrasing of this section of the MoS to mean that thumbnails are required rather than recommended. It is my understanding that people have been going in and changing other-sized images to thumbnails solely because they believe it is required by the MoS. The conversation has transitioned toward discussing default image sizes, but we must be careful to remember to correct this problem as well.

It is also my understanding that a phrase that precedes a colon must be an independent clause, even if it is introducing a list. Ergo, our final text should read "includes the following" rather than stopping at "includes."

As for adding soft advice, I don't think it's a good idea, not even if we can all agree on whether short or long captions are best. We've already seen people can misinterpret "recommended" to mean "required." I feel that overly verbose captions are the sort of thing that we should allow users to fix as they go. Darkfrog24 (talk) 05:02, 4 August 2009 (UTC)[reply]

People here do seem to agree that editors shouldn't be going around changing image sizes to thumbnails over objections. I've therefore removed the parts of the MoS advice that were causing that, particularly the sentence, "As a rule, images should not be set to another size ..." [2] SlimVirgin ^{talk|contribs} 05:51, 4 August 2009 (UTC)[reply]

Nice, but we could use a "the following" after "are not limited to." Darkfrog24 (talk) 05:58, 4 August 2009 (UTC)[reply]

Re SlimVirgin's "People here do seem to agree that editors shouldn't be going around changing image sizes to thumbnails over objections" (05:51, 4 August 2009), I'd add a clause: "except where the objection is only that use of thumbs is standard practice" or equivalent - in other words the objectors must also consider the effect on the (unregistered) reader of the size of the specific image in the specific place in the specific article. --Philcha (talk) 06:33, 4 August 2009 (UTC)[reply]

Several good points were made in the previous comments. The specific wording proposals that I gleaned from them, relative to the #Beige proposal above, are

• SlimVirgin's wording change, already installed.
• Shakescene's suggestion to use less jargon.
• Darkfrog24's suggestion to put "the following" before ":".

I incorporated these suggestions into the #MistyRose proposal below. Eubulides (talk) 08:06, 4 August 2009 (UTC)[reply]

Re "this is after I've gone through to force a bigger size on the pics, all still thumbnails" above: A problem with changes like those, which inserted exact widths like "200px" into Operation Charnwood, is that this makes the images smaller for readers who've expressed their preference for larger thumbnails by changing their default widths to 300px. It's better to use "upright=1.1" instead, as this does not shrink the image for anybody. I changed the article to use "upright=Factor" rather than "Widthpx"; this won't affect the appearance for IP users or for users who have not changed their default image widths, and will behave better for users who have changed their defaults. Eubulides (talk) 08:30, 4 August 2009 (UTC)[reply]

I made my edit earlier not realizing the page was protected. I've requested unprotection. If it isn't granted, I'll revert my edit. SlimVirgin ^{talk|contribs} 09:50, 4 August 2009 (UTC)[reply]

Thanks for making that change, SV. I don't think anyone here disagrees with it, so I hope it's not reverted. I'm off to tell WT:FAC the good news. Tony (talk) 11:03, 4 August 2009 (UTC)[reply]

This is an updated version of the #Beige proposal for improving the guideline for image sizes; it takes into account the comments noted above.

• On Wikipedia, most pictures should be displayed so they are between 100 and 400 pixels wide. The thumbnail option ("thumb") results in a default width of 180 pixels, although logged-in users can set a different default in their user preferences.
• A picture may benefit from a size other than the default. The "upright=2.2" option (or "|frameless|upright=2.2" for plain pictures) resizes an image to about 400 pixels wide by default; making the number larger or smaller changes the image size accordingly. An image should generally be no more than 500 pixels tall and 400 pixels wide, so it can be comfortably displayed next to the text on the smallest displays in common use; an image can be somewhat wider if it uses the "center" or "none" options to stand alone. The {{Wide image}} and {{Tall image}} templates can display images that would otherwise be unreasonably wide or tall. Examples where adjusting the size may be appropriate include, but are not limited to, the following:

• Images with aspect ratios that are extreme or that otherwise distort or obscure the image.
• Images in which detail is relatively unimportant (for example, a national flag may be easily recognizable even at a small size).
• Images containing important detail (for example, a map, diagram, or chart may contain important text that would be unreadable at the default size).
• Images in which a small region is relevant, but cropping to that region would reduce the coherence of the image.
• Lead images, which should usually be no wider than "upright=1.7" ("300px")

The difference between the currently-installed version and this version is available. Further comments are welcome. Eubulides (talk) 08:06, 4 August 2009 (UTC) updated 20:09, 4 August 2009 (UTC) updated again 05:59, 5 August 2009 (UTC)[reply]

Is there any benefit to recommending editors write "upright=x," as opposed to "xpx"? —Preceding unsigned comment added by SlimVirgin (talk • contribs) 08:45, 4 August 2009 (UTC)[reply]

Yes, as explained above; an explicit size like "250px" undesirably makes a thumbnail smaller for a reader whose width preference is 300 pixels. In contrast, "upright=1.4" makes the thumbnail larger for all readers. Eubulides (talk) 09:21, 4 August 2009 (UTC)[reply]

Very few readers have set their width preferences. We need to cater to the average reader, and if an editor wants an image to be 250px, there's no reason not to allow him or her to do that, surely. SlimVirgin ^{talk|contribs} 09:24, 4 August 2009 (UTC)[reply]

Yes, there's no reason not to allow 250px settings; in some cases they make perfect sense, for example when sizing one image to fit against another small-resolution image. The proposed wording does not disallow or state a preference against settings like that. Eubulides (talk) 20:09, 4 August 2009 (UTC)[reply]

We should not be encouraging editors to try to produce pixel-perfect layouts. There's too many target platforms that WP serves that the work of a user Editor-Using-IE8-on-1200x1068-Monitor-with-10pt-Ariel to make the layout perfect by assigning pixel sizes to image to get it to look right on their monitor will undoubtedly screw up on any other platform.

Hmm. Anyone know if there's a way to have images show up as a percentage of the box width that it is in? --MASEM (t) 15:33, 4 August 2009 (UTC)[reply]

That would be a very nice thing to have. As far as I know it is not possible with the current MediaWiki software. If there's some way to do it, I will document it in WP:PIC. Eubulides (talk) 20:09, 4 August 2009 (UTC)[reply]

Agree. Please let's not get to prescriptive here. --Joopercoopers (talk) 15:08, 4 August 2009 (UTC)[reply]

• "Most pictures should be between 100 and 400 pixels wide" -- I hope this doesn't mean image uploaders will only upload images between 100 and 400 pixels wide now, citing the MOS "policy". Perhaps, "Most pictures, when included in articles, should be resized to between 100 and 400 pixels wide"? Matthewedwards :  Chat  17:13, 4 August 2009 (UTC)[reply]
• Agreed. Let the actual stored image size be a concern of NFC, let the display size be the concern here. --MASEM (t) 17:18, 4 August 2009 (UTC)[reply]
• Thirded. That's a very good way to phrase it. Darkfrog24 (talk) 17:19, 4 August 2009 (UTC)[reply]

• Thanks for this suggestion. I adjusted the proposal to say "Most pictures in articles should be displayed with widths between 100 and 400 pixels.", which I think captures the suggestion more succinctly. Eubulides (talk) 20:09, 4 August 2009 (UTC)[reply]

It's a bit ungainly. Try this: "On Wikipedia, most pictures should be displayed so that they are between 100 and 400 pixels wide. [Explanation about monitor size.]" Darkfrog24 (talk) 00:34, 5 August 2009 (UTC)[reply]

Thanks, that's better, and I reworded it that way. Eubulides (talk) 05:59, 5 August 2009 (UTC)[reply]

Eubulides and the others who have contributed, Misty Rose looks very well-worded—nice work indeed. I took the liberty of tweaking in a few places (no substantive change in meaning). A few queries:

• "so it can be comfortably displayed next to the text on the smallest displays in common us"—will that get us into trouble if Blackberry/mobile displays are construed as this? Perhaps not, but those more familiar with such issues might consider the option of simply finishing the statement on "text".
• Can we remove "somewhat", since it doesn't seem to add anything?
• "The {{Wide image}} and {{Tall image}} templates can display images that would otherwise be unreasonably wide or tall." Can we remove "can" (or don't those templates always have good effect)? And just from my dumb perspective (which is a good target to address), the alternative to using those templates would be what? Would it not be clearer to write: "... templates moderate the dimensions of images that would otherwise be unreasonaly wide or tall."?
• As I mentioned above, the bullets mix the reasons to up-size and the reasons to down-size; it would be a little easier if the reasons to down-size were all listed together (say, first), and then the reasons to up-size. I've swapped the location of one bullet point to this end, but I'm wondering whether the last one (lead, no more than 300px) is meant to be up- or down- or either? If either of up-size only, it could stay at the end. If down-size only, it could go up the list with its siblings.

There's just one issue rootling around in my mind: several editors have expressed a clear dissatisfaction with the thumbnail default of 180px, which is how the vast majority of images that have been thumbnailed because of the MoS wording will appear to our readers. Is there any reason we shouldn't lobby for this to be increased to, say, 200px? If there were consensus, this could be done as a second phase, after misty rose is implemented. Tony (talk) 07:55, 5 August 2009 (UTC)[reply]

Indeed we should; this is often raised at WP:IMAGE & elsewhere, but is not pursued. Johnbod (talk) 15:50, 6 August 2009 (UTC)[reply]

• One minor wording issue I have is that none of the drafts above make it clear that even with forced sizes you still need to include the "thumb" parameter or you lose the caption & formatting. I certainly support editorial choice in the matter, allowing as far as possible for different screen sizes etc. Johnbod (talk) 15:50, 6 August 2009 (UTC)[reply]

• Could you propose brief wording to fix this? Eubulides (talk) 20:06, 9 August 2009 (UTC)[reply]

There seems to be consensus for this change, so I installed it, with Tony1's suggestions to remove "can" and "somewhat". Further improvements are of course welcome. Eubulides (talk) 20:06, 9 August 2009 (UTC)[reply]

Heck, I'd lobby for 250px instead of 200px. The only issue which I cannot google-fu well enough to determine is if there is any NFCC concerns for non-free images by going above 180px as the default. Those that complain about that being too big on their layout are likely registered users and can change it. --MASEM (t) 15:54, 6 August 2009 (UTC)[reply]

• Blackberries and mobile displays are not an issue. They ignore our width advice entirely.
• The NFCC argument seems to be based on a misunderstanding of how default widths apply. If an image's native width is (say) 200px due to NFCC concerns, then any default width above 200px is ignored; the image is never displayed above its native resolution merely due to default widths. So if we change the default width to (say) 250px, there cannot possibly be any NFCC concerns with the resulting article that are not already present in the native image (which is readily available). Now, there may indeed be concerns about blowing up a 200px image to (say) 600px, using an explicit width; but a default-width image should always be OK.
• Tony, removing the "somewhat" and the "can" is fine. However, the Wide and Tall image templates do not "moderate dimensions"; they display the images in a little window that's scrollbarred. I'm not sure it's worth putting details about that here. If you can redo the wording to make it more logical, please do so.

Eubulides (talk) 18:17, 6 August 2009 (UTC)[reply]

MASEM, sure, I'd go for 250px too. Does anyone object? WP is being criticised for being weak on images. Tony (talk) 09:37, 7 August 2009 (UTC)[reply]

I'd say 250 as well. Cheers, Ian Rose (talk) 14:48, 15 August 2009 (UTC)[reply]

the major changes regarding possible correct forms of possessives of nouns ending in a single s don't seem like an improvement to me: what's now presented as a third option is not clarified well, and i have serious doubts about the assertion that it is "the most commonly used" (which i have now removed). can we please revert to the previous version of this section and discuss any proposed changes? thanks Sssoul (talk) 05:39, 6 August 2009 (UTC)[reply]

It may not be obvious to you why changes were urgent in that section, Sssoul. But all can be explained, and will be as required. I had sought to made minimal and obvious changes, but there is a great of confusion about possessives in English, and our guideline was not free of them. Nor are most WP editors, nor are most MOS editors, nor are most published style guides. It's all quite tricky.

Of course if all the changes I made to that section are controversial, or likely not to reflect consensus, then they should all be reverted until all have been discussed. I acquiesce quite happily in your reversion of the recommendation that I added; I am confident enough that it will be found acceptable after analysis and discussion.

The section as it stood was incomplete and aberrant. It is simply not true that there are two main bodies of practice, each taking an extreme view. In fact, while that was said, other points were made that were inconsistent with that dichotomous view. I am surprised, Sssoul, that you think the third option that I added is not explicated well; and I am surprised that you have serious doubts that it is "the most commonly used". I can assure you, from a long study of style guides major and minor, and of practice from many publishers, that most adopt some version of that compromise position.

Anyway, let's do some analysis. (As a preliminary, it might best if editors wanting to participate read the English-language sections of the article Apostrophe, which surveys the field in great detail.)

Here is the section as it stood, before I amended it:

Possessives section before Noetica's editing

For thorough treatment of the English possessive see Apostrophe. See also Apostrophes, above. It's is the short form of it is or it has (as in it's a nice day, it's been a nice day); the possessive its (as in the dog chased its tail) has no apostrophe, nor do hers, ours, yours, theirs, and whose. The possessive of a singular noun is formed by adding 's (as in my daughter's achievement, the boss's wife, Glass's books, Illinois's legislature). For a singular noun ending in one s, there are two widely accepted forms:[2] Add 's: James's house, Euripides's plays, Moses's early life, Brahms's music. This style is more common for modern names and common nouns. Add just an apostrophe: James' house, Euripides' plays, Moses' early life, Brahms' music. This style is more common for biblical and classical names (Socrates' wife; Moses' ascent of Sinai; Jesus' last words). Either of those forms may be acceptable in Wikipedia articles, as long as consistency is maintained within a given article. According to some it depends on perceived pronunciation, such as whether one would say James' house or James's house, but at least be consistent within an article. The possessive of a plural noun ending in s is formed by adding just the apostrophe (my daughters' husbands). The possessive of a plural noun not ending in s is formed by adding 's (women's careers, children's toys). Official names (of companies, organizations, or places) should be made to conform to a specific style.

And the section after my editing (before Sssoul's removal of the recommendation that I added, and with left parenthesis now placed properly):

Possessives section after Noetica's editing

For thorough treatment of the English possessive see Apostrophe. See also Apostrophes, above. What follows are the most basic principles. It's is the short form of it is or it has (as in it's a nice day, it's been a nice day); the possessive its (as in the dog chased its tail) has no apostrophe, nor do hers, ours, yours, theirs, and whose. The possessive of a singular noun is formed by adding 's (as in my daughter's achievement, the boss's wife, Glass's books, Illinois's legislature). For a singular noun ending in just one s (a pronounced s, not as in Illinois or Descartes), there are three practices:[3] Add 's: James's house, Euripides's plays, Moses's early life, Brahms's music, Dickens's novels, Doris's opinion. Add just an apostrophe: James' house, Euripides' plays, Moses' early life, Brahms' music, Dickens' novels, Doris' opinion. Add 's or just an apostrophe according to the assumed pronunciation: Socrates' wife; Moses' ascent of Sinai; Jesus' last words (and similarly for most classical and biblical forms); but Doris's opinion, and James' house or James's house depending on the assumed pronunciation. Any of these three practices may be acceptable in Wikipedia, with consistency in an article. But the third option is recommended as by far the most commonly used. The possessive of a plural noun ending in s is formed by adding just the apostrophe (my daughters' husbands). The possessive of a plural noun not ending in s is formed by adding 's (women's careers, children's toys). Official names (of companies, organizations, or places) should not be altered to conform to a specific style. For example, St Thomas's Hospital, being the official designation, should not be rendered as St Thomas' Hospital, even for consistency.

Here's a good place to start in considering these two versions. Suppose an editor comes to MOS wondering how to manage the four possessives in this sentence:

Sentence A. These are Doris'[s] copies of Morris'[s] books on Socrates'[s] and Descartes'[s] philosophies.

(Rewording would just be an evasion, and is to be thought unavailable.)

Question 1: What forms should MOS recommend for the whole of Sentence A?

Question 2: Why?

Question 3: How well and how clearly does the unmodified guideline settle things for the editor? (Explain in detail.)

Question 4: How well and how clearly does the guideline with modifications settle things for the editor? (Explain in detail.)

I suggest that editors offer their answers, below. For myself, I cannot give good answers regarding the earlier version; but I can for the modified version.

Please post after this contribution.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 06:17, 6 August 2009 (UTC)[reply]

... why is rewording "just an evasion and to be thought unavailable"?? i don't see any value in using implausible examples. (for the record, i don't consider "Illinois's legislature" a felicitous example either, since one would normally talk about the Illinois [state] legislature .)

i find the explanation of Option 3 far less clear than the previous concise (and true) note that some people (and some styleguides) consider it dependent on pronunciation. but a] since other people (and styleguides) do not view it as dependent on pronunciation, what are the grounds for recommending that as the "most commonly used" option? and b] how does one determine whose pronunciation "counts"? (by the way, in the old version as well, examples of pronunciation should use IPA - simply spelling them doesn't clarify anything.)

it's also thoroughly unclear how your proposed Option 3 relates to the principle of consistency within an article - until further notice, it seems like you're recommending that consistency be abandoned in favour of someone's ideas of pronunciation. which is not a helpful guideline. Sssoul (talk) 08:25, 6 August 2009 (UTC)[reply]

Sssoul, I am sorry that you do not grasp the value of the "implausible example" I propose as a starting point. Of course I could have proposed something more "lifelike", but it would have been much longer and far less efficient. Concerning implausible examples generally, note that they are purified "laboratory specimens" that help us to approach problems rationally and scientifically. Apart from that, all sorts of oddities really do turn up, and often you can't simply reword. Sometimes you need to represent a speech in written form, and you must choose the punctuation without changing the wording.

I designed the example with great care, to elucidate the differences in the two versions of the guideline: along with the questions that I then put forward. It is unfortunate that you did not engage with this exercise. Have you read the sections of Apostrophe that are relevant to our guideline?

For the rest, it is irrelevant that you deem Illinois's legislature infelicitous as an example. First, it is present in both versions anyway; second, it or something similar to it could easily turn up (Arkansas specifically legislated for the form Arkansas's, because the question did come up all the time); third, it represents a whole class of English words ending in silent s that turn up in our articles – distinct from cases like Descartes'[s]. You prefer the way pronunciation is dealt with in the unmodified version? Well, if you worked through the task that I recommend above, you might see why it is seriously inadequate. Both versions call for consistency; but with the unmodified version, consistency is undefined, since the recommendations are scattered illogically, with contradictions and avoidance of the very cases that those consulting MOS will want guidance with. Far from being infelicitous, they are the very cases that test the robustness of our guidelines. We need to know what to say about Illinois'[s], Doris'[s], Morris'[s], Socrates'[s]; and we need to have guidelines that show what our recommendations are. We cannot (and should not) prescribe every detail: no style guide attempts that for all cases that might turn up; but we do need to be comprehensive and decisive in laying out principles. Finally, IPA pronunciations are a red herring in this context. We simply appeal to what editors take to be the best pronunciation for the possessive of James. If they say James's, they can confidently write that; if they say James', then they can write that. If their understanding of the pronunciation happens to be questioned, the matter can be resolved by discussion. That sort of simplicity and clarity is the best we can hope for!

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 11:06, 6 August 2009 (UTC)[reply]

(outdent) if simplicity and clarity are the priorities, the previous version was simpler and clearer: there are two correct/accepted forms - and also differing views about whether or not pronunciation has anything to do with it - which means that in Wikipedia articles, there's no reason to change from one correct form to the other, except to maintain consistency within an article. changing that to a recommendation that the choice of form should be based on one's preferred pronunciation is a very major change, which calls for discussion and consensus, so i hope other editors will state their views. Sssoul (talk) 11:57, 6 August 2009 (UTC)[reply]

I think I like Noetica's amended text. However, may I make a plea that bold face not be scattered about? If anything, I'd make the pre-colon opening of each of the three bullets bold, but nothing else. Tony (talk) 12:12, 6 August 2009 (UTC)[reply]

Sure, Tony. I tried to salvage as much as I could of the earlier version, so I retained that highlighting. Of course formatting can be improved – and other details too.

With respect Sssoul, all that has been demonstrated is that you have not understood either version of the guideline: neither the way it was, nor my modified version. This may suggest a need for further clarification. That I concede!

If the previous version was "simpler and clearer" as you assert, it remains illogical, indeterminate, and therefore an inadequate guideline. In particular, it recommends consistency but fails to show what would count as consistent! Is an article consistent if it has both Socrates's and James'? That version does not give an answer; the version I put forward does.

You ignore my evidence and my arguments, and my suggestion that you read the highly relevant material at Apostrophe, and my specific analytical points above, and the challenge that I construct above (to demonstrate the differences in the two versions). I have laboured long to address all that you say, and I can help you no more. Indeed, let's see what others have to say. (I cannot do any more here right now, since I have work to do in the real world.)

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 12:23, 6 August 2009 (UTC)[reply]

peace, okay, Noetica? the fact that i'm trying to keep my replies brief doesn't mean i'm ignoring your points. my main point is simply these changes need to be discussed - by more than just the two of us - and whatever is decided on needs to be phrased clearly. thanks for conceding that! Sssoul (talk) 13:08, 6 August 2009 (UTC)[reply]

I consider the original version clearer. We should attempt to explain the choice between 's and the single apostrophe within the confines of their respective bullets, rather than presenting the information as a third "practice". Do not construe this preference as an endorsement, though – the original does need work, and does not read well. It's the basic structure that is superior. Be aware that this is merely a cursory opinion given whilst skimming through the MoS and its talk page. —Anonymous Dissident^Talk 12:39, 6 August 2009 (UTC)[reply]

Very briefly: AD, that's part of the problem. On a superficial reading the earlier version appears to be OK. But analysis shows that it fails to give determinate guidance, exactly where guidance is needed. And it contradicts itself. If people would read the points I make above, and take the test that I propose, they would see how.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 12:49, 6 August 2009 (UTC)[reply]

I see that the old version fails in effectively instructing the reader in when to use the apostrophe-s and when to use the apostrophe; this concession is (intendedly) implicit in my first comment. My point is that your layout creates a problem in removing a problem – that is, it introduces a bullet structure that is not intuitive (option to use 's, option to use single apostrophe, explanation of the choice) for the sake of clarifying the disparity. Tests aside, I think we should work on giving the explanations within a two-bullet layout, because it simply makes better sense; two options: two bullets. To give due credence to your point of view, though, I will be sure to return in a few hours to analyse your comments more thoroughly and to take the test of which you speak. —Anonymous Dissident^Talk 13:16, 6 August 2009 (UTC)[reply]

There seem to be two issues with Noetica's changes. 1. Noetica altered the wording. 2. Noetica added a third option for possessives where only two had been before. The second is more important. Is this third option used in the English language? Sure. The question is whether or not it should be used here. The purpose of the MoS is not to describe what people do elsewhere; it is to tell Wikipedia editors what to do here. I consider this to be the sort of change that is substantive enough to merit discussion, if not ahead of time, then now.

I would also add that, with regard to intra-article consistency, it is my understanding, it is the principle that must be consistent, not the appearance of the punctuated material. Otherwise, the American system would be the only acceptable way to deal with quotation marks on Wikipedia (though it does make things look neater and more professional, heh heh). Therefore idea that the third offered option is not consistent should not, alone, be used to discredit it. Darkfrog24 (talk) 12:46, 6 August 2009 (UTC)[reply]

Again very briefly, Darkfrog: yes, I did transparently alter what is demonstrably a poor guideline. Practice with possessives, here and everywhere, is complex and subject to exceptions. I happily accept reversion (for now) of the recommendation that I put in, but a close reading of the versions and the discussion above will show that what it replaces is very poor. Especially, I stress yet again, there is no point calling for consistency if we don't say what consistency would amount to!

Please read those parts of Apostrophe; please take the challenge I propose above. I suggest a definite procedure. It might be useful if some editors would give that carefully designed procedure a try.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 12:57, 6 August 2009 (UTC)[reply]

... by "that carefully designed procedure", do you mean the sentence about Doris, Morris, et al? my principles for possessives of nouns ending in s cope with that one just fine - but that doesn't mean it would be appropriate for me - or the MoS - to impose my principles on Wikipedia articles in which another correct/acceptable standard is used consistently. (as for what "consistency" means: yes, there is a school of thought that would support "James's portrait of Euripedes' wife", but within one article we should not have both "James's portrait" and "James' portrait".) Sssoul (talk) 13:30, 6 August 2009 (UTC)[reply]

[Outdent] Patiently and with all due respect, Sssoul: It will be worth discussing this with you if you answer my clearly articulated points, just as I respond to yours. My "carefully designed procedure"? Yes: I posed four numbered questions in my first contribution, above. Please give answers, instead of fragmentary assertions addressing nothing anyone else has said.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 23:44, 7 August 2009 (UTC)[reply]

Anonymous Dissident (relevant to others' comments also), you write:

Tests aside, I think we should work on giving the explanations within a two-bullet layout, because it simply makes better sense; two options: two bullets.

My points in response:

• First, the test I set up above is systematic and revealing. Please do return to it, as you said you would.
• Second, your brief analysis does not fit the structure of the problem. No one disputes that there are two endings used to form the possessives of nouns: -'s or simply -'. But there are not just two extreme practices to distinguish. Hardly anyone uses just one of the two endings for all nouns ending in a single -s; and only a couple of minor or old-fashioned guides come close to proposing that. Current guides advocate an intermediate way, and so should we. Clearly! That is implicit in the old and confused version of our guideline; it is explicit in the version I put forward.
• Third, I stress yet again that the field is vexed and complicated. We cannot evade the issues in a flurry of hand-waving, with genuflections toward consistency and motherhood. Sssoul now says, above, that consistency would be not using both "James's portrait" and "James' portrait" in the same article. Well of course! That much is obvious. Now how about "Gibbs's portrait" and "Biggs' portrait"? "Descartes's philosophy" and "Dumas' novels"? "Socrates' discourse" and "Rameses's tomb"? "Socrates's discourse", "James' dissertation", and "Carlos's debt"? To settle these consistently – to give guidance – we need an articulated principle. The old version did not offer one. Most style guides do offer that, and the guideline I put forward follows their broad consensus.

Try it, using the test I propose.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 00:16, 8 August 2009 (UTC)[reply]

I've given this matter some thought over the past few days. It seems to me that punctuating possessives according to pronunciation, even variable pronunciation (James'[s] can be pronounced "James" or "Jame zz.") is a correct system and we do not have a clear and substantial reason to tell Wikipedia editors not to use it. However, I also feel that we should not recommend this system above the other two as either more common outside of Wikipedia or more desirable within Wikipedia, especially not without a source. It is not clear to me that it is either more effective or more common. More than that, I am of the opinion that the MoS is not the place to make descriptive statements like that. We should offer a link to an article where the matter can be discussed in detail in its proper place. Darkfrog24 (talk) 00:34, 8 August 2009 (UTC)[reply]

I was just over there taking another look around and I noticed something: We explain singular nouns ending in just one s and we explain plural nouns, but this will leave the reader wondering about singular nouns like "princess" and "governess." Darkfrog24 (talk) 00:37, 8 August 2009 (UTC)[reply]

it may be that a few bits of the previous version could use rewording for clarity, grace, etc. but the previous version deliberately described two correct/acceptable options without prescribing either of them. if you look into archived discussions you'll see that this 's-vs-just-apostrophe question has been a hugely divisive issue for Wikipedia editors. it's not okay to switch to a prescriptive stance (prescribing either of those options, or a "third" version that mixes the two) without first establishing a clear consensus on the issue.

and yes, Noetica, it is indeed obvious what "consistency within an article" means, and there's no need to cloud it up. thank you for conceding that.

Darkfrog, nouns ending in a double s take apostrophe-s form of possessive. an example (the boss's wife or something like that) is included among the non-controversial forms (unless its been eliminated in recent revisions). Sssoul (talk) 05:41, 8 August 2009 (UTC)[reply]

Darkfrog:

• Sssoul is right about this: both versions say what to do with nouns ending in -ss.
• We do have a reason for recommending the system that respects pronunciation, for singular nouns ending in just one s. We have a solid reason to set aside uniform use of -'s, and also uniform use of -'. Even without consulting my collection of published style guides, I can assure you that these three current and respected works support the third option that I have put forward:

1. Gregg Reference Manual, 10th edition 2004. (This celebrated and meticulously detailed style manual, 688 pages long, gets 5/5 stars from 47 of its 50 reviewers at Amazon.)
2. Perfect Punctuation, new edition 2007. (A comprehensive and widely distributed guide from an established writer on English style.)
3. How to Punctuate: Penguin Writer's Guide, 2006. (Penguin's current 304-page flagship treatise on the topic.)

Several more can readily be found.

• You write: "It is not clear to me that it is either more effective or more common." Well, is it clearer now? Please furnish your sources to support such alternatives as these: "Gus' ideas"; "the bus' helpful driver"; "Xerxes's expedition".
• You write: "More than that, I am of the opinion that the MoS is not the place to make descriptive statements like that." So, what do you want? No description? No prescription? I am mystified.

Sssoul:

• You write: "... the previous version deliberately described two correct/acceptable options". May we see your sources to support your claim for those options, as I have done for the third option (and can do again)?
• You write: "it's not okay to switch to a prescriptive stance (prescribing either of those options, or a 'third' version that mixes the two) without first establishing a clear consensus". You assert that it is not OK to "switch to a prescriptive stance", but my version does not prescribe: it recommends, while listing alternative ways also. What's more, it recommends with firm backing from the broad consensus of major published guides. Nevertheless, I immediately accepted your reversion, and am quite happy to discuss the matter. But ever since you raised the matter by starting this thread, you have been evasive. Why raise it at all, if you won't take carefully articulated responses on board?
• You fail to understand my point about consistency, yet again. You write: "yes, Noetica, it is indeed obvious what 'consistency within an article' means". I have shown that it is not obvious. Of course everyone knows not to put both "James's portrait" and "James' portrait" in the same article. You completely ignore the issues of consistency beyond that primitive level.

Quite frankly, Sssoul, someone has to give this advice: if you don't know things, don't say things; if you have no evidence, don't make substantive claims; if you can't answer fair questions honestly (or at all), don't raise topics in discussion here; if you don't know the sources (even our own article Apostrophe), don't pretend to be able to make MOS guidelines until you remedy that situation. You wouldn't do it at Nuclear fission, would you? So why do it here?

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 06:51, 8 August 2009 (UTC)[reply]

(outdent) ... peace, Noetica, okay? you've already made it clear what your strongly-held view is; it's still the case that changes like the one you're proposing need to be based on consensus and collaboration. so let's proceed with that, all right? without the dissing, preferably - it doesn't become you. thanks.
as you've noticed, i'm not interested in engaging in a debate with you over which option is "best". i'm asking whether or not interested Wikipedia editors (note the plural!) agree that it's time for this part of the MoS to "recommend" a particular approach, instead of just describing acceptable options. Sssoul (talk) 10:50, 8 August 2009 (UTC)[reply]

I will clarify what I said earlier: There seem to be three practices for plural nouns that end in s: 1. Add an '. 2. Add an 's. 3. Go by pronunciation. I am saying that "Go by pronunciation" is a correct style. I am saying that the MoS should not forbid writers to do what is correct and proper unless it has a very clear, solid, real and good reason to do so. I am saying that we have no such reason. In other words, I do not feel that the MoS should recommend a particular approach instead of just naming acceptable options. However, nothing shows me that this third option is any better than the other two—and yes, Noetica, I have read your posts quite thoroughly—so neither should the MoS put this option above the other two. Darkfrog24 (talk) 11:23, 8 August 2009 (UTC)[reply]

Sssoul:

I'm interested in rational discussion, not this dissing that you speak of. But you show complete disrespect for fellow editors and the principles of Wikipedia. You start up a thread for discussion, presumably towards understanding and consensus, then you refuse to engage. How are we to proceed, if not by putting questions for others to answer, and presenting hard evidence and solid arguments for colleagues to address? SEE ABOVE! It is now clearer to me. Unfortunately, you are not able to focus on the questions and evidence that I lay out for discussion: with regard to precedent, clarity, ranking and selection of alternative recommendations, or anything else. If you refuse to budge no matter what the evidence shows, then there is nothing anyone can do. You have your opinion, and that's final. We assume rationality here; but if it is missing, we hit a dead end. Don't blame me: I've done all I can. I will not labour to present further evidence, since there is obviously no point.

Darkfrog:

I still can't tell what you want us to do here. You don't think we're ready to make any prescription concerning these possessives, but you don't want descriptions either (see earlier: "I am of the opinion that the MoS is not the place to make descriptive statements like that"). Have I got it right? So what do you want to see concerning possessives in MOS? I have given solid evidence that major guides favour a mixed approach. I have asked for, and still await, your evidence of guides supporting the two extreme approaches that our guideline lists. It really does seem that no imaginable evidence would make the slightest difference to your stance – whatever that stance is, which remains elusive. You write: "nothing shows me that this third option is any better than the other two". Well, if the bizarre examples I give above do not shift you a millimetre, and if you completely ignore the verdict of weighty authorities (and you cannot present countervailing precedents), then I see no reason to continue this discussion. As I said above to Sssoul, why bother to show any more evidence, or present the further arguments that I have available? No one takes the slightest notice. Your claim to be paying attention is belied by everything else that you say.

The level of discussion here has seriously degenerated. I hope we will see editors returning who are capable of supple, rational, discourse toward consensus. I don't mind being proven wrong; but I do object to wasting my time with editors for whom precedent, argument, evidence, and rationality are mere trifles that we can dispense with on a whim.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 13:00, 8 August 2009 (UTC)[reply]

Obviously, editors here feel strongly about many aspects of the style guide. However, I want to make a strong plea for people to cool it. I believe it's quite within our ability to interact more positively than this section suggests is occurring at the moment. As an incentive, are you all aware that ArbCom is scheduled to examine the stability of the style guides next month? I suggest that we voice our differences in a way that suppresses overt irritation with each other, and heightens a sense of collaboration and friendliness. We all have skills to offer. Tony (talk) 13:07, 8 August 2009 (UTC)[reply]

I agree. First, changing MoS makes lots of existing articles that were conformant now not conformant. Second, we recognise people speak differently and write differently and that all varieties of English are equally good.

Now, to the possessives in particular. Fowler on this for example says write -s's if that makes sense and just s' if that makes sense. That is bound to depend on your pronunciation. Personally I think it is best that the MoS keeps quiet on the matter or says, at often does, just be consistent within the article. As a regular sub I often have a dificult to choice to make and I go one way and someone corrects it under MoS. That is fine, but if it is the toss of a coiin it is just down to one editor's preference over another's. Proper spelling and grammar is very important I think, but if it is simply a toss-up well what is better, improving the content or fighting over apostrophes? SimonTrew (talk) 13:39, 8 August 2009 (UTC)[reply]

Now, I haven't been following this section, but I'm VERY concerned that my colleagues are at loggerheads about the "Possessives" section ... and you are all experts in your areas. I'm keen to resolve this swiftly and amicably, guys, since we need to work together to improve the MoS—it won't happen overnight.

Here's the section as Noetica edited it (from the edit history). I've removed the MoS-breaching scattered bold face, added "either", and changed "but" to "however"; these changes make no difference to the substantive meaning, but I can't help myself. The bits in contention are underlined.

A quick flick upwards suggests that Sssoul and Darkfrog are uneasy about:

• (a) the insertion of the third bullet point, starting "Add 's or just an apostrophe ..."; and
• (b) the sentence below, italicised in the box: "But the third option is recommended as by far the most commonly used." (This one has been removed since Noetica's edit.)

Could we start afresh, respecting each other's expertise, by getting to the nub of why this third option, slightly favoured, might or might not be unwise to include in the style guide? I myself would have been happy with just the third bullet alone, and I wonder why having all three bullets might not be construed as a compromise already, even with the slight favouring of the third bullet in the text at the bottom? (I think I remember it from as far back as Fowler.) But I should shut up and let others talk.

I invite concise statements by Darkfrog and Sssoul first, please, and from anyone who hasn't yet contributed who would like to. Can Noetica hold off until at least DF and S have written here? Tony (talk) 14:38, 8 August 2009 (UTC)[reply]

That pretty much sums it up, Tony. All three of these are correct systems, and I do not feel that Wikipedia should forbid a correct system without a significant, non-imaginary reason to do so, which we do not have. However, I do not feel that Wikipedia should recommend the third system above the other two. I feel that this would cause problems with people believing that the third system is required and give us the same sort of edit wars that we've seen over thumbnail sizes.

In fine: Keep the bullet point specifically allowing the by-pronunciation system. Lose the line saying that the by-pronunciation system is recommended. Darkfrog24 (talk) 15:23, 8 August 2009 (UTC)[reply]

I haven't commented here yet, but I prefer the third bullet above as our advice. The first two bullets do not provide advice for tricky cases, which is why people will come looking. I will note that none of this is what I was taught: Single apostrophe if the word is one syllable (James's book) and apostrophe-s when two or more (Delores' radio). Does anyone know what reference teaches that approach? I must be for simplicity's sake. --Andy Walsh (talk) 15:25, 8 August 2009 (UTC)[reply]

I found something similar in Fowler's Dictionary of Modern English Usage. See next section. Hans Adler 21:40, 8 August 2009 (UTC)[reply]

thanks, Tony. i don't object to the idea of the third option, as long as there's (civil) discussion of it and evidence of consensus for adding it. i do object to the idea of recommending any option over the others without evidence that there's broad support for that kind of change, because there's a history of editors getting pretty het up over attempts to prescribe one style of possessive.

if there is consensus for adding a third option, the addition needs to be phrased clearly, and it needs to address questions like "whose pronunciation counts?" IPA (or something a lot like IPA) should be used to clarify what pronunciation James' book and James's book are supposed to represent – spelling doesn't convey pronunciation, and in this proposed third option, something needs to convey the pronunciation of the examples given.

whether or not the addition is made, the section could use improvements. the examples used should be plausible within the context of Wikipedia (William Morris'[s] designs would be more apt than Doris'[s] opinion); and it's worth pointing out that when in doubt (or in conflict) rephrasing may be an excellent way to go (eg the trial of Socrates, the Illinois [state] legislature, the bus driver). Sssoul (talk) 16:23, 8 August 2009 (UTC)[reply]

Sssoul, it seems to me that either you missed Noetica's point about useful examples or you chose to ignore it. Which is the case? If the point is invalid, why is it invalid? Earlier, Noetica said, at 11:06, 6 August 2009 (UTC):

Concerning implausible examples generally, note that they are purified "laboratory specimens" that help us to approach problems rationally and scientifically. Apart from that, all sorts of oddities really do turn up, and often you can't simply reword. Sometimes you need to represent a speech in written form, and you must choose the punctuation without changing the wording.

You have just said that "the section could use improvements. the examples used should be plausible within the context of Wikipedia". [sic]

Suppose that you are in a bookstore, looking for a travel guide to assist you during your trekking in the Australian outback. You tell the bookseller that you specifically want guidance for the possibility of encountering a crocodile in the wild. The bookseller tells you that such an encounter is extremely rare, and claims that the likelihood of crocodile attacks against humans has dropped dramatically in recent years. Despite your insistence on guidance for such an encounter, the bookseller persists in urging you to purchase a travel guide which obviously lacks any mention of crocodiles. What would you think of that bookseller? What would you think of that bookstore?

-- Wavelength (talk) 15:00, 9 August 2009 (UTC)[reply]

yes, i read what Noetica wrote about the implausible example s/he proposed as a "test case"; what i wrote refers to examples used in the MoS itself. what i mean is: this is the Wikipedia MoS, so (i feel) Wikipedia-like examples are more useful to people consulting it. right, rephrasing isn't always possible - but that doesn't mean "the bus'[s] friendly driver" is a good example of anything except poor usage. "Illinois's largest employer" would be a clearer-cut example than "Illinois's legislature", since the latter is normally referred to as "the Illinois [state] legislature". and "William Morris'[s] designs" is a type of construction more likely to occur in a Wikipedia article than "Doris'[s] opinion", so why not use it? it illustrates exactly the same point.

please note that i'm replying to you to be courteous, not because i'm trying to insist that my view of this must prevail. i'm not interested in debating; i'm interested in collaborating. and in any case the choice of examples is not a major issue compared with establishing whether or not there's consensus for including the third option and (if so) getting it formulated clearly, including practical principles for deciding whose pronunciation "counts" when there are disagreements. Sssoul (talk) 15:38, 9 August 2009 (UTC)[reply]

Sssoul, comprehensiveness is more useful than plausibility in style guides. You might buy a travel guidebook that advises you what to do if you encounter an echidna, a kangaroo, a kiwi, a kookaburra, a platypus, or a wombat, but says nothing about crocodiles. You would probably consider that guidebook to be seriously deficient if you encountered a crocodile.

You might be considering the purchase of an insect repellent or insecticide. Do you want protection against only the most common insects, or do you want protection also against the most dangerous ones? If your trek takes you into territory inhabited by insects which transmit dengue fever, would you want to omit being protected against the risk, just because the risk is not common, not plausible, and not likely to occur?

Buyers and sellers of insurance tend to be realistic about assessing the relative likelihood of different events, the relative seriousness of different events, and the relative usefulness of different insurance coverages. An encounter with a kangaroo is probably more common than an encounter with a crocodile, which you might consider when you buy travel insurance. However, which insurance coverage is more useful? You would probably say that the insurance coverage for an encounter with a crocodile is more useful (and therefore more expensive), especially if it also includes coverage for encounters with kangaroos.

Noetica's example is comprehensive enough for some unusual cases, but does not hinder guidance for more common cases. -- Wavelength (talk) 20:06, 9 August 2009 (UTC)[reply]

[I revised my message. -- Wavelength (talk) 20:43, 10 August 2009 (UTC)][reply]

I found the following possessive forms in current versions of Wikipedia articles, linked here with permanent links. A search for a character string ending with an apostrophe finds both forms if they exist, whereas a search for the same character string followed by a space finds only the shorter form if it exists.

Doris' and Doris's: 0 instance and 1 instance respectively

Doris' and Doris's: 1 instance and 0 instance respectively

Doris' and Doris's: 1 instance and 0 instance respectively

Doris' and Doris's: 0 instance and 3 instances respectively

Doris' and Doris's: 0 instance and 1 instance respectively

Doris' and Doris's: 1 instance and 1 instance respectively

Doris' and Doris's: 3 instances and 0 instance respectively

Doris' and Doris's: 0 instance and 1 instance respectively

Morris' and Morris's: 2 instances and 27 instances respectively

Morris' and Morris's: 3 instances and 6 instances respectively

Morris' and Morris's: 21 instances and 8 instances respectively

Morris' and Morris's: 1 instance and 0 instance respectively

Morris' and Morris's: 2 instances and 4 instances respectively

Morris' and Morris's: 0 instance and 1 instance respectively

Morris' and Morris's: 2 instances and 3 instances respectively

Morris' and Morris's: 0 instance and 1 instance respectively

Socrates' and Socrates's: 34 instances and 1 instance respectively

Socrates' and Socrates's: 12 instances and 6 instances respectively

Socrates' and Socrates's: 13 instances and 0 instance respectively

Socrates' and Socrates's: 2 instances and 0 instance respectively

Socrates' and Socrates's: 1 instance and 0 instance respectively

Socrates' and Socrates's: 1 instance and 0 instance respectively

Socrates' and Socrates's: 7 instances and 0 instance respectively

Socrates' and Socrates's: 2 instances and 0 instance respectively

Descartes' and Descartes's: 26 instances and 1 instance respectively

Descartes' and Descartes's: 11 instances and 0 instance respectively

Descartes' and Descartes's: 4 instances and 0 instance respectively

Descartes' and Descartes's: 3 instances and 0 instance respectively

Descartes' and Descartes's: 1 instance and 0 instance respectively

Descartes' and Descartes's: 2 instances and 0 instance respectively

Descartes' and Descartes's: 5 instances and 1 instance respectively

Descartes' and Descartes's: 4 instances and 1 instance respectively

-- Wavelength (talk) 21:26, 10 August 2009 (UTC)[reply]

I have started preparing answers to Noetica's four questions, but I have not finished them, and I might change my answers. At this stage, they express my personal preferences, although I am ready to make allowances for preferences of other editors. I spell the possessive of Doris as Doris' and I pronounce it the same as Doris.

I have here some information which is causing me to reconsider re-evaluate my own preferences. The following are given names: Andrea (English female), Andreas (German male), Carlo (Italian male), Carlos (Spanish male), George (English male), Georges (French male). Therefore, the following statements are ambiguous.

• These are both Andreas' books.
• These are both Carlos' books.
• These are both Georges' books.

The English language has many surnames formed by the simple addition of s to a male given name, for example, Andrews, Edwards, Peters, and Williams. Therefore, the following statements are ambiguous.

• These are both Andrews' books.
• These are both Edwards' books.
• These are both Peters' books.
• These are both Williams' books.

Someone who owns a duck called Emus (rhyming with deem us) might say: Emus' wings are not flightworthy. In written form, that is an ambiguous statement.

I am prepared to accept a single, simple standard for Wikipedia, even if it happens to be different from my own preference. (Incidentally, when I have a choice, I do not follow any supposed supposedly "national" variety of English, and I doubt that many people do so completely.) If I travel to a country where the traffic moves on the other side of the street, I am prepared to adapt to the difference. Following the middle of the road in that situation would not be appropriate.

(Also, please see my comments at Talk:Apostrophe.)

-- Wavelength (talk) 20:56, 8 August 2009 (UTC)[reply]

[I revised my message. -- Wavelength (talk) 20:43, 10 August 2009 (UTC)][reply]

• I am very happy with Noetica's version, but it's worth noting that there are indeed a few other conventions out there in the wild. The AP Stylebook rule is to use an apostrophe without s in the following situations only: (1) Nouns ending in s with the following exception: Common nouns that are singular in form and meaning and which have a different plural form get 's unless they are followed by a word that starts with s. (2) In the two special expressions "for appearance' sake" and "for conscience' sake".

I don't consider this a particularly good rule. Their formulation is also confusing and unclear/inconsistent in some corner cases such as "the headlice'(s) symptoms". I guess the reason for the complicated rules is that they tried to codify a possible variant of pronunciation practices. Hans Adler 15:28, 8 August 2009 (UTC)[reply]

• The Economist Style Guide has a more practical rule: Use an apostrophe without s in the following situations only: (1) Nouns ending in s if they are plural in form and function. (2) Proper names ending in s if they are plural in form. It also gives specific guidance on how to treat "people", although that follows from their general principles. (Whereas according to a literal reading of the AP Stylebook it must be "the American people' power".) Hans Adler 15:47, 8 August 2009 (UTC)[reply]

the exact statement of The Economist's principle is here: "Use the ending s' on plurals that end in s — Danes', bosses', Joneses' — including plural names that take a singular verb, eg, Reuters', Barclays', Stewarts & Lloyds', Salomon Brothers'."

• Fowler's Dictionary of Modern English Usage (1965 edition) says general practice is moving from the second bullet (apostrophe-only for words in s) to the third (depends on pronunciation). It follows from this that in poetic or reverential contexts the apostrophe-only version always applies, and that in ordinary speech the 's version applies for all one-syllable words and possibly for all others as well. A special rule governs "for X's sake": If X ends in an s sound and has at least two syllables, this becomes "for X' sake" or even "for X sake". Hans Adler 16:08, 8 August 2009 (UTC)[reply]
• Strunk & White (1918 and 2003 editions) says, citing the United States Government Printing Office and Oxford University Press as authorities, that possessives of nouns which end in s always get 's, the only exceptions being Jesus and ancient proper names ending in es or is. Hans Adler 16:22, 8 August 2009 (UTC)[reply]
• a short form of OUP's recommendation is: Use apostrophe-s to form the possessive of singular nouns except when the final syllable (of a noun ending in a single s) is pronounced like the verb is: Moses' mother and Lloyd Bridges' career, but Isis's husband and William Burroughs's reputation. Sssoul (talk) 14:06, 11 August 2009 (UTC)[reply]

The discussion seems to have come to an end. There appears to be more consensus for Noetica's proposal than for any other version, here and—I think I'm right in saying this—as most often recommended in other style guides. I propose that the most straightforward solution is that the sentence slightly favouring the third bullet be reinstated, along with the trivial copy-edits I made in the yellow box above. Tony (talk) 05:22, 11 August 2009 (UTC)[reply]

smile: the Wikipedia MoS very often adopts principles that are not "what most other style guides" recommend; and the "consensus" is among ... what, five or six editors? but anyway!

Tony and others, what practical guidance should the Wikipedia MoS include for determining whose pronunciation "counts" when applying Option 3?

and are you suggesting that the "slight favouring" of the third option should override the usual Wikipedia MoS point about not changing from one acceptable style to another without compelling reasons for the change? Sssoul (talk) 06:04, 11 August 2009 (UTC)[reply]

in the spirit of collaboration, this is as far as i've been able to get in formulating something clear. i hope someone can address the parts i've put in bold, because they are really puzzling me:

The possessive of a singular noun is formed by adding 's (as in my daughter's achievement, the boss's wife, Glass's books, Illinois's largest employer). For a singular noun ending in just one s (a pronounced s, not a silent s as in Illinois or Descartes), three practices are acceptable in Wikipedia articles:^{[footnote - see below]}

• Add 's: James's house, Euripides's plays, Moses's early life, Brahms's music, Dickens's novels, Morris's designs.
• Add just an apostrophe: James' house, Euripides' plays, Moses' early life, Brahms' music, Dickens' novels, Morris' designs.
• Add 's or just an apostrophe according to the pronunciation used by ... well, by whom? For many people this may entail:

°adding just an apostrophe to most classical and biblical forms: Socrates' wife; Moses' ascent of Sinai; Jesus' last words;

°adding 's to most nonclassical/nonbiblical names: Morris's designs, Dickens's novels, Burroughs's birthdate, Thomas's career;

°using either James' house or James's house, depending on the particular editor's pronunciation. When editors' pronunciation differs, the choice should be based on ... well, what?

Any of these three practices can be used on Wikipedia, but should be applied consistently within an article. As usual when Wikipedia accepts more than one style, ArbCom's stipulation about stability applies: Editors should not change an article from one style to another without a substantial reason unrelated to mere choice of style.

the footnote: For details on the different forms and the rationale for their use, see Apostrophe. Evidence that this issue is largely unsettled among professional style guides and among Wikipedians can be found in the archives at WT:Manual_of_Style/Archive_92#Possessives of proper names ending in "s", WT:Manual_of_Style/Archive_100#Singular possessives ending in "s", WT:Manual_of_Style/Archive_102#Possessives, WT:Manual_of_Style/Archive_104#Possessive, and WT:Manual_of_Style/Archive_105#Possessives of common nouns in s.

hope that helps get this proposed change formulated so that it provides Wikipedia editors with clear, practical guidance. Sssoul (talk) 08:22, 11 August 2009 (UTC)[reply]

Sssoul, this is not one of my areas, I have to admit. When there is more than one pronunciation that might suggest a different usage WRT bullet 3, I'd say it could be negotiated by the editors at the time; what do you think? Are there any examples? "Mars' desire" as Marz or Marzuz desire? I'm not sure.

Isn't the principle about not changing from one to the other announced at the top of the MoS? It seems unnecessary to repeat it here—I'd have thought it would be good enough to refer to it WRT its more prominent location.

To put it in summary, I don't care much which (I myself, in my own RL editing, go for the "always" ad 's no matter what—it's just simpler), but I appreciate the need to keep the peace and to maintain within-article consistency. Tony (talk) 12:01, 11 August 2009 (UTC)[reply]

Tony, there are miles of cases where pronunciation varies/might vary, and (smile: believe it or not!) editors can get pretty heated over things like this (check out the archived discussions linked in that footnote). the MoS needs to provide clear guidance.

i pronounce the possessive of both Jesus and Moses as three syllables, but i know there are people who pronounce one or both of them as two syllables. someone earlier in this discussion stated that they pronounce the possessive of Doris as two syllables, while many people pronounce it (and similar possessives) as three; some editors might not feel sure how they'd pronounce Dos Passos'[s] parents; and so on. so whose pronunciation should count, under Option 3?

also, since i have no way to guess how another editor pronounces the possessive of Lloyd Bridges'[s] surname, when i see it spelled with just an apostrophe in a wikipedia article, does that mean i should follow Option 2 or Option 3 in that article?

that kind of thing needs to be set forth clearly in the MoS if this is supposed to be a practical guideline that lets editors know what to do in "tricky situations". the other two options do provide clear guidance: use 's in all cases, or use just an apostrophe in all cases, without worrying whether it's a "tricky case" or what other style guides recommend. Sssoul (talk) 13:45, 11 August 2009 (UTC) ps: once again: i am not against the idea of including an Option 3, but it needs to be expressed clearly enough to actually help editors who disagree, and/or who aren't sure how many syllables there are in Mars'[s] or Burroughs'[s].[reply]

One good way to "slightly favor" one style over another two would be to simply list it as the first bullet point of the three. Darkfrog24 (talk) 12:42, 11 August 2009 (UTC)[reply]

You're right, DF. Tony (talk) 13:00, 11 August 2009 (UTC)[reply]

(outdent) like Tony, i personally am satisfied using apostrophe-s all down the line, but if a pronunciation-based option is wanted, would the OUP option i just posted under "what other styleguides say" (above) be acceptable to people? it has the advantage of being clear-cut, which is a pretty major advantage on Wikipedia: Use apostrophe-s to form the possessive of singular nouns except when the final syllable (of a noun ending in a single s) is pronounced like the verb is: Moses' mother and Lloyd Bridges' career, but Mars's sister, Isis's husband, Burroughs's reputation, Euripedes's plays and Delores's nickname. Sssoul (talk) 14:18, 11 August 2009 (UTC)[reply]

... are there any ideas about how to decide whose pronunciation counts when editors disagree about how to apply Option 3? to be of any practical use as a guideline, it really needs to address that. Sssoul (talk) 07:17, 13 August 2009 (UTC)[reply]

The rule as you described it seems almost fine to me, except it also refers to pronunciation, albeit much less critically. But your example Isis's husband irritates me like hell because I can't guess what pronunciation you are using that takes you there. Do you really mean it as an example, not as an exception? (An exception would make some limited sense because the final s is a Greek addition and pedants might decide not to pronounce it.) Or perhaps you just pronounce the verb is in a way that I am not familiar with, so that it sounds more like in Delores than in Isis??? The other problem is that your rule seems to require Jesus's, and Jesus' being the one constant case throughout most of the different rule sets this simply doesn't seem right or practicable. Hans Adler 08:01, 13 August 2009 (UTC)[reply]

sorry Hans Adler - i've only just now noticed your question. the final syllable of Isis (like the final syllable of Delores) is not pronounced like the verb is. the verb ends with a /z/ sound; both Isis and Delores end with an /s/ sound. and yes, the OUP rule recommends Jesus's (which matches my pronunciation anyway, so it's fine with me). Sssoul (talk) 14:08, 22 August 2009 (UTC)[reply]

(I'm kind of surprised that it's not more biased toward apostrophe-s if at least Noetica, Tony, and Sssoul [and I] all prefer it). I know I've arrived late, but really wish it were firmer toward that in most cases. I don't know which it recommends to the unfamiliar person for these names. Although its pronunciation should be common sense, I'm disturbed that on René Descartes people think it's "Day-kart influence" instead of "Day-kart's influence". I doubt the MOS would change much, but here's my general preference:

• The possessive of all singular nouns is formed by adding 's. Since the rule for plurals is to add only an apostrophe, the only possible exceptions are those that are pronounced like a plural: James (Jame), Mars (Mar), Socrates (Socratee), not anything ending in ss (boss), silent s (Illinois, Descartes), or wouldn't make sense pronounced without it (Doris, Morris).

But since I think the consensus above is that Descartes's pronunciation requires it, could someone take that to René Descartes? Reywas92^Talk 23:14, 13 August 2009 (UTC)[reply]

Sorry, I did not notice that is says "(a pronounced s, not as in Illinois or Descartes)". However, it doesn't directly say that those do need an apostrophe-s. That article still needs to be corrected though. Reywas92^Talk 23:10, 15 August 2009 (UTC)[reply]

All,

Could we use full stops after post nominals such as knighthoods etc. (eg. Horatio Nelson K.B. instead of Horatio Nelso KB). This would help indicate that there is more than one word in the post nominal (eg. O.B.E. (three words) instead of OBE (one word)). However, with titles from universities the common version could still be used (dPhil and not d.Phil. etc.).

Please discuss and share your opinions.

With compliments.

DAFMM (talk), 6th August 2009.

What would be the advantage of so indicating? Barnabypage (talk) 18:05, 6 August 2009 (UTC)[reply]

It's pretty standard these days to omit the full stops; see, for example, the British monarchy official offical website. I don't think the MOS should legislate full stops here, when even the Queen omits them nowadays. Eubulides (talk) 18:17, 6 August 2009 (UTC)[reply]

There is a trend away from these unnecessary dots. Capitalisation is enough to show that it's not a word ... that and simple facts like that there is not such word as obe. JIMp _talk·cont 18:18, 6 August 2009 (UTC)[reply]

This would be a step back to the 1970s and before. Throughout the English-speaking world there has been a strong trend to drop the fly-specks from abbreviations (at least those that end with the word-final character), acronyms and initialisms. As a result, a long-standing redundancy has been removed from usage, and text looks more attractive. We still have one mountaint to climb: you dot es dot. While we have to accept that some US writers still insist on this, by the end of the next decade it will be a thing of the past, as in almost every other US acronym (USA). Tony (talk) 09:32, 7 August 2009 (UTC)[reply]

I'm in favour of fewer dots in general. As Tony says, the text looks better without them; and without such superfluities essential punctuation stands out more. Functionality, clarity, elegance.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 22:38, 7 August 2009 (UTC)[reply]

I find that we should be very careful about predicting what will happen with language. They're tricky things and don't always do what they're told. The MoS should deal with what is, not with what we think might be in ten years. That being said, the trend away from dots in acronyms is stronger in the U.K. than in the U.S. and, since this issue has strong national ties to the U.K. over the U.S., it is reasonable to follow the British preference. Darkfrog24 (talk) 15:33, 8 August 2009 (UTC)[reply]

Nobody has mentioned what the MoS currently says, & I can't be bothered to look. The full stops should certainly not be mandated, per those above, but they should not be forbidden, especially given the ENGVAR aspect. Johnbod (talk) 15:12, 9 August 2009 (UTC)[reply]

Even the English don't normally use full stops in general initialisms these days (GPO vs. G.P.O., etc), but some old hands still seem to want have postnominals taking up as much space as possible, so any comma or full stop that can possibly be squeezed in is seen to be fair game. It's old hat, really. -- JackofOz (talk) 21:58, 13 August 2009 (UTC)[reply]

John had fish and chips, Giuseppe had spaghetti with garlic, oil and chilli peppers, and Akira had some sushi.

These are different things. On the one hand, there are commas between the three clauses of the sentence, and the three clauses constitute the entire sentence. On the other hand, there are commas between the individual things that Giuseppe had. I think the rule applies to the latter more than the former. --Auntof6 (talk) 21:25, 12 August 2009 (UTC)[reply]

See Semicolon#English usage, point 3, subpoints 1, 2, and 3. -- Wavelength (talk) 21:53, 12 August 2009 (UTC)[reply]

That editor at Quark was me. I fixed some small problems with references, something in alt-text, some grammar, and these serial commas. Since it is a featured article, Quark is required to conform to MOS guidelines, including this provision:

Both styles [serial commas by default, or no serial commas by default] are acceptable in Wikipedia, with consistency in an article; but in a case where including or omitting the comma clarifies the meaning of the sentence, that solution should be adopted.

The effect is clear: an article can have serial commas consistently, or it can have no serial commas, with exceptions only as required to "clarify the meaning". Since Quark had many serial commas that were not strictly needed to clarify the meaning (though they were stylistically better), I simply made the article consistent. No changes of mine diminished clarity. As I pointed out in copious edit summaries, the wording in instances of inclusion and omission was closely parallel. Not a good look.

So much for the editing in question. On the more general matter of the guideline, I say this: it is too long and detailed. I would like it trimmed to something much tighter. We don't need all that theoretical to-and-fro, since the linked article Serial comma does that admirably.

I declare my interest: Argument, evidence, precedent, authority of the best style guides, and long experience convince me that the best policy is serial comma by default – with the only exceptions being some formular or proverbial utterances, some similar omissions for marking prosody, and very rare complex cases that probably call for rewording anyway (if that option is available). [This third exception added later.–N]

As things stand, I am prepared to support the two-style guideline as we now have it, just as I support the choice between consistent unspaced em dashes and consistent spaced en dashes (which I prefer) as sentence-level punctuation. I will argue strenuously against any departure from either of these two-style compromises.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 01:08, 13 August 2009 (UTC)[reply]

I do not feel that the MoS should rely on links to other articles to clarify its viewpoint on this matter for two reasons: 1) Writers come here looking for answers. There's enough link-jumping going on in the expanded MoS to make a writer too weary of it to be expected to do it again. 2) These other articles often describe grammar and punctuation principles that are not endorsed by the MoS. In this case, the MoS gives two good examples of problems that can be expected to come up on a semiregular basis. The explanation is short but not incomplete and it makes no false or overly hypothetical statements. Darkfrog24 (talk) 11:49, 13 August 2009 (UTC)[reply]

I understand that your edits conformed to the MoS, and that they don't decrease clarity; I was just questioning the point of requiring "consistency in an article". That's why I posted here and not at Talk:Quark. That's not a purely stylistic point such as the spaced en dash and the unspaced em dash, which are always 100% equivalent and thus there is never any point in mixing them in the same article; ditto for the spelling of colour; the serial comma instead affects pronunciation. I can understand in the case of parallel constructions (I had missed them in Quark); but requiring consistency for all constructions is, IMO, pointless. Consider a list of very short items, especially not at the end of a sentence: "... are designated x, y and z, and remain unchanged ...". Now consider a list of sentence-length items, especially one ending a sentence: "... it allowed for a better description of the weak interaction (the mechanism that allows quarks to decay), equalized the number of known quarks with the number of known leptons, and implied a mass formula that correctly reproduced the masses of the known mesons." If you read them aloud, you'll see that the latter will have a more "natural" rhythm if you make a pause before "and implied", whereas the former will if you make no such pause before "and z". A foolish consistency is the hobgoblin of little minds. --   A. di M. 13:13, 13 August 2009 (UTC)[reply]

A di M:

First, a foolish consistency, perhaps; but not a wise consistency. Clichés will get us nowhere.

Second, as you can see from the article Serial comma, there are not only good arguments for consistency, but also for consistent default use of the serial comma. See especially the quote from one of the most eminent sources, The Oxford Style Manual:

For a century it has been part of OUP style to retain or impose this last serial (or series) comma consistently, [...] but it is commonly used by many other publishers both here and abroad, and forms a routine part of style in US and Canadian English. [...] Given that the final comma is sometimes necessary to prevent ambiguity, it is logical to impose it uniformly, so as to obviate the need to pause and gauge each enumeration on the likelihood of its being misunderstood – especially since that likelihood is often more obvious to the reader than the writer. (pp. 121–122)

I highlight with bold and underlining some points that are especially relevant to our articles, which are written and edited by dedicated amateurs, many of whom are unaware of such traps.

Third, the requirement for consistency in matters of style is fundamental and general, and has a solid basis. If you want to address that requirement, please raise it generally.

Fourth, yes: it may be worthwhile examining some edits in the case of Quark, to illustrate the point about consistency as applied to serial commas with concrete examples that you have raised. In none of these do I detect prosodic points that carry any weight in favour of inconsistency, and in fact I disagree with your prosodic analysis. As I have discussed a couple of years ago with an editor at the talkpage of Serial comma, we notoriously do not know how we read things out aloud. We don't know where we pause, often. In any case, nothing could be prosodically clearer than saying this, for three items: A [pause] B [pause] and C, and that is how we typically do it, with a very quick and quiet and. (Believe it or not; try it: dispassionately and with an open mind.) So even if prosody and rhythm did have primacy, the prosody we naturally use is matched by the punctuation with serial comma anyway: A, B, and C.

The text had this already, with two serial commas:

a red, a green, and a blue arrow canceling out to white, representing a baryon; a yellow ("antiblue"), a magenta, and a cyan ("antired") arrow canceling out to white, representing an antibaryon.

And this without:

There are three types of color charge, arbitrarily labeled blue, green and red.

And this without:

Each of them is complemented by an anticolor—antiblue, antigreen and antired.

And this without:

Just as the laws of physics are independent of which directions in space are designated x, y and z, and remain unchanged ...

I added serial commas to those last three. The absence of them is worse than distractingly inconsistent: the expectation in this highly intricate and technical article is that punctuation will follow sense more than sound, and that it will do so in a principled way. It is not poetry, it is difficult scientific prose. Now, consider the last case I cite above. The reader scans for a group of items, and has been primed to expect a serial comma by earlier uses. So momentarily there is a tendency to parse x, y and z, and ... as two items (x, then y and z) followed by a serial comma and an and, with the expectation of some third item to follow. This sort of thwarted expectation incrementally slows reading and hinders comprehension. Why introduce even the possibility of such inconvenience, when a simple, consistent, and rational policy will prevent it?

Darkfrog:

I still don't quite understand how you want MOS to work, any more than I did when I raised the question in an earlier section. We have excellent articles elaborating on some of the topics that MOS must deal with; we can usefully refer to them. Earlier you seemed to like that! As for the rest, I take it that you approve of the serial comma guideline as it stands, is that right? Well, I'd like to look at trimming it to make it more focused on guiding and less on detailed justifying. But I would like to do that later. There are other issues that have not yet been settled on this talkpage first.

So I recommend not doing much more on this now. It's a perennial issue, usually very poorly handled, and we have other business.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 15:11, 13 August 2009 (UTC)[reply]

I believe that the current policy on the serial comma, which I interpret as recommending consistency within each article whenever reasonably practical, is right and proper and does not need modification at this time. I believe that the two examples of cases in which the serial comma can either create or prevent confusion are also good and should remain. This is because they directly address the way in which we expect writers to use the serial comma on Wikipedia, regardless of how serial commas were and are used in other works. They do not state beliefs; they give instructions: "When you see one of these, use the serial comma. When you see one of these, don't." Darkfrog24 (talk) 17:32, 13 August 2009 (UTC)[reply]

Thanks Darkfrog. I understand your meaning better now. Myself, I'd still like to see the guideline trimmed. But that issue can wait.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 22:41, 13 August 2009 (UTC)[reply]

I wouldn't use a longer pause before and in "A, B and C" than the normal foot break I would also use in "B and C"; YMMV. --   A. di M. 18:11, 14 August 2009 (UTC)[reply]

I have made a proposal at Wikipedia talk:Colours#Proposal: No colors in body text to prohibit the use of colored links, colored text and colored backgrounds in the body of the article. Comments welcome. Jpatokal (talk) 08:46, 13 August 2009 (UTC)[reply]

A couple of years back, someone made mass changes to a page I was working on, converting most references to the people involved to use the people's last names after the initial mention. It was quoted that this was done by MOS. I've used this style since. Now, on another page I'm working on, I'm getting revered to first name references. I went looking for the MOS guideline to back up my version, but have been unable to find such. So before I raise the level of discussion on a MOS point that maybe does not exist, I figured I would ask if anyone else can point me to such a guideline somewhere in the MOS over the use of personal/fist vs family/last names in an article that refers to people multiple times. (The article is current question is the one on the current season of America's Got Talent, and the names at issue are the names of the judges, who get referred to over and over throughout the article.) - TexasAndroid (talk) 13:19, 13 August 2009 (UTC)[reply]

You want WP: SURNAME, Texas Android. Be advised, in some naming traditions, particularly the one they use in Iceland, it is more proper to use a person's given name than his or her last name. Darkfrog24 (talk) 17:24, 13 August 2009 (UTC)[reply]

Understood. We're talking about American and British names here, so I do not think that'll be a problem. Thanks. - TexasAndroid (talk) 21:40, 13 August 2009 (UTC)[reply]

Some people are usually referred by given name even when they do have surnames, e.g. Dante, Prince, Galileo, etc. Dunno about the US, but in Italy this is the case for many reality show participants. How are those people most commonly referred to by the media? --   A. di M. 18:20, 14 August 2009 (UTC)[reply]

WP: SURNAME specifically mentions this, actually. Dante, Prince and Madonna are covered under the same rule. However, even in the cases of celebrities who are only known mostly by one name (rather than nearly exclusively as with Madonna) the MoS instructs editors to use the surname in subsequent mentions: Beyonce Knowles and then Knowles, rather than Beyonce Knowles and then Beyonce. This seems to be where the contestants would fit in; they're known mostly by one name. It is my opinion that even when the entertainment news is informal, Wikipedia should maintain an encyclopedic tone. In this case, that means surnames for TV show contestants. Darkfrog24 (talk) 21:47, 14 August 2009 (UTC)[reply]

TexasAndroid said it was the judges, not the contestants. The judges are definitely "Hasselhoff, Osborne, and Morgan" not "David, Sharon, and Piers". I'm stunned anyone could argue that. Powers ^T 12:52, 15 August 2009 (UTC)[reply]

I suggest adding something to the section "Follow the sources" to reduce the endless debates and huge amounts of time constantly wasted in usage discussions on hundreds of WP articles. Discussions on content usually improve the quality of WP articles, but most discussions on usage do not improve article quality or make WP less amateurish. The main cause of the current chaos (including polls decided on the basis of the personal preferences of people who happen to notice a poll or are alerted by others) is that Wikipedia does not yet prevent OR in deciding usage.

Most editors are at best interested in or even experts in an article's topic, but they are almost never experts on English usage in general. Even those that are aware that usage may be different in general reference works or other English-speaking countries from what they're used to tend to bitterly defend either their personal preferences or at best those of one or more experts in a field. A good example is the endless discussions on the capitalisation of compact disc, whose result is in addition simply amateurish and would not be possible in any major publishing house. As shown here, Wikipedia is apparently the only major reference source that currently uses the uppercase spelling "Compact Disc" and apparently all major UK and US reference works use the lowercase spelling "compact disc" (or "compact disk").

The arguments against lowercase spelling presented in that discussion (and arguments in thousands of other discussions) show that most WP editors don't understand how modern dictionaries and other reference works are compiled. Many seriously believe they can do a better job at deciding what is current usage than the constantly updated online versions of dictionaries. Most WP editors for example don't understand that it is simply OR trying to decide if a trademark has become genericised. I therefore suggest the following:

Many points of usage, such as the treatment of proper names, can be decided by seeing what other writers do. Unless there is some clear reason, it is generally a good idea to follow the usage of reliable secondary sources in English on the subject. If the sources for the article can be shown to be unrepresentative of current English usage as a whole, follow current English usage instead—and consult more sources.

It is however very important to realize that it is very unlikely that you are an expert on English usage even if you have collected many reliable sources or are perhaps even an expert on a topic. The same reasoning that prohibits use of primary sources and original research in determining the contents of a Wikipedia article implies that editors should not try to decide what language usage a Wikipedia article follows if usage in secondary sources varies or is different from that in major dictionaries. Even when secondary sources specifically address a usage issue, they are almost always simply the opinions of someone who is an expert in the field but not in English usage.

Often experts in a field believe they should and can dictate or even change how terms from their field are used in general English. This kind of prescriptive approach to usage was common in dictionaries in the past, but nowadays all major English dictionaries try to be descriptive. They use extensive professional experience in collecting and processing extensive collections of citations from reliable sources, so they should be the main sources for usage questions in Wikipedia articles.

• Instead of trying to determine usage on the basis of secondary sources when these do not agree with each other or with major dictionaries, which would be original research, it is better to follow the usage that dictionaries and other reference works say is the most common.

• If a majority of major dictionaries and reference works agree on a certain usage, this should be used in Wikipedia even if some experts (on a topic, not usage) disagree, or if, for example, the preferred usage of a copyright owner is different. In cases where most British dictionaries record a different usage than American dictionaries do, the general MOS principles on national varieties of English apply.

• Exceptions to this rule include new technical terms (and slang) that are not (or not yet) recorded in reference works or, for example, names of countries that have changed since the reference works were published.--Espoo (talk) 14:14, 13 August 2009 (UTC)[reply]

"The same reasoning that prohibits use of primary sources and original research in determining the contents of a Wikipedia article... is not true. Primary sources are allowed in determining contents, although must be used with caution. --Jc3s5h (talk) 16:56, 13 August 2009 (UTC)[reply]

Perhaps instead of telling people not to bother trying to figure out proper usage, we should instead encourage each Wikipedia editor to focus on his or her area of expertise: content for some, usage for others, content and usage for a few. However, I doubt that this or anything else will stop people from arguing about usage when they feel that it is the right thing to do. If you feel that the editors in question are wasting their time, then the only thing for it is to step back and let them waste it. Sometimes it results in better articles. Darkfrog24 (talk) 17:37, 13 August 2009 (UTC)[reply]

DF writes: "We should instead encourage each Wikipedia editor to focus on his or her area of expertise: content for some, usage for others, content and usage for a few." I couldn't agree more. I fervently wish that editors who are not experts in style – and who know little about precedents, style guides, and such basics – would concentrate more of their effort away from this page; or at least that they would see how their competence is limited. Until they learn more. (I say all this generally, not as a comment on the present discussion.)

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 23:03, 13 August 2009 (UTC)[reply]

The problem is not confined to writing style. It is the same in most subject areas, where you see well meaning, clueless people who think they understand something arguing with university professors and other experts. The problem is endemic to, and inherent in, wikis where the contributors' respective qualifications are unknown to one another. (I do not believe that well meaning people requires a hyphen.) —Finell (Talk) 19:04, 15 August 2009 (UTC)[reply]

And, to a lesser extent, that's going to be a problem on an "encyclopedia that anyone can edit." It's the tradeoff that we make. I think this is something that we've just got to accept. With regard to Espoo's specific proposal, "Follow the sources," I don't think that it's our answer. As TexasAndroid has brought to our attention in the matter of first names vs. last names, many valid sources, in that case in the entertainment media, deliberately (and often rightly) choose a tone less formal than the encyclopedic one preferred on Wikipedia. So the way I see it, following the exact expressions used in the sources may not be best even when no one is doing anything explicitly incorrect. Darkfrog24 (talk) 19:46, 15 August 2009 (UTC)[reply]

"The same reasoning that prohibits use of primary sources and original research in determining the contents of a Wikipedia article implies that editors should not try to decide what language usage a Wikipedia article follows if usage in secondary sources varies or is different from that in major dictionaries." Nonsense. In the unlikely case that I found ten major abstract algebra textbook giving the same identical definition for field (mathematics), and two major dictionaries giving a slightly different definition (e.g. omitting the requirement for commutativity, essentially defining field as a synonym of division ring), I would use the definition of the textbooks, not that of the dictionaries, and refuse to use the word field to refer to a non-commutative division ring. "English usage" depends on what you are speaking about, and the people who most often speak about something are the ones who generally know better how you should speak about them. --   A. di M. 20:50, 15 August 2009 (UTC)[reply]

(As for "well-meaning", that's a compound adjective, not the adverb "well" modifying the adjective "meaning", which would make no sense. So I'd use a hyphen in it, and the dictionary currently closest to my hand agrees with me. --   A. di M. 20:54, 15 August 2009 (UTC))[reply]

The "hyphen" article in Fowler's 2d ed. (my fave), if I understand it correclty, counsels that in an adverb adjective noun construction, a hyphen is only necessary where its presence or absence would change the meaning, as where the adverb might be mistaken for another adjective. For example (his), "a little used car" versus "a little-used car". According to the "well and well-" article, hyphenating "His well-known courage" is common but unnecessary. —Finell (Talk) 23:20, 15 August 2009 (UTC)[reply]

Well now that it's come up, the Bedford Handbook places hyphens in compound adjectives when they precede the noun, giving as one example, "well-known candidate." Fowler's is British, isn't it? This might be a U.K./U.S. thing. Darkfrog24 (talk) 00:38, 16 August 2009 (UTC)[reply]

¶ For those contortionists here who can contrive no better way of spending Saturday night than tying themselves in Gordian knots,

1. If you want to become completely confused, try reading the article on "hyphen" in the first edition (1926) of H.W. Fowler's Modern English Usage (dedicated to the memory of his brother F.G. Fowler, who didn't make it with him through the war that began 95 years ago this month). Among many other things, Fowler goes to great lengths to avoid hyphens and the situations we try to resolve here with en-dashes, em-dashes and spacing around them (e.g. the Campbell-Bannerman — Asquith — Lloyd-George — Bonar Law succession of 1905-1923).
2. I don't think that "well" in "well-meaning" is either an adverb or an adjective. It's an object or predicate of some kind governed by "meaning". "Well" is an adverb in the Well-Tempered Clavier, because it governs "Tempered" (and not "Clavier"). But other uses are far murkier to my feeble eyes, such as well-intended, well-meant and well-meaning. Consider what happens when you replace "well" with "harm" as in "he means you no harm" or "he intended no harm". "Harm" is a noun serving as the direct object of the verb. I almost always hyphenate in such a case, if only because English verbs usually come after the subject and before any object. (Thus, "I glimpsed a gut-wrenching scene through my view-finder."). Now replace "harm" with "ill" as in "he means you ill." Are "well" and "ill" nouns (and thus objects) here? Do "well-meant" and "well-meaning" have the same syntactical structure and the same (...well...) meaning? —— Shakescene (talk) 04:32, 16 August 2009 (UTC)[reply]

I must say that was well-argued. You've convinced me. ;-) Hesperian 04:53, 16 August 2009 (UTC)[reply]

(I couldn't resist that rogue hyphen, but I'm not trolling you. I really do think you've made a pretty compelling point there.) Hesperian 04:53, 16 August 2009 (UTC)[reply]

"Let every nation know, whether it wishes us well or ill, that we shall pay any price, bear any burden, meet any hardship, support any friend, oppose any foe..." — John F. Kennedy's inauguration speech, 20 January 1961. Well-wishers are always welcome, especially the well-educated and the well-spoken. ;-) —— Shakescene (talk) 10:02, 16 August 2009 (UTC)[reply]

I see someone has added a comment in the project about this.

The formal names of the titles of the British Prime Minister are First Lord of the Treasury and Minister for the Civil Service. The first is what is written on the brass plate on the door of number 10. So there is no formal title of Prime Minister, at all. SimonTrew (talk) 21:48, 13 August 2009 (UTC)[reply]

No, those may be the only formal portfolios the PM holds, but the title of his/her office is still "Prime Minister of the United Kingdom". -- JackofOz (talk) 21:52, 13 August 2009 (UTC)[reply]

This problematic "formality" is not relevant anyway. It is a matter of what to do with the usual terms for these offices and the people who occupy them. I have amended the guideline, trimming it of glosses that are obvious and removing the modifier formal.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 22:49, 13 August 2009 (UTC)[reply]

It seems to me that this MOS could do a better job of differentiating between the use of hyphens for conjunction and en dashes for disjunction. Hyphens are properly used to group things, as in Mason-Dixon Line, surveyed by Messrs. Mason and Dixon. The en dash is used to separate things, as in Pennsylvania–Maryland border. As currently written, the examples might lead an editor to think that whenever and could be used, the en dash is the proper form of punctuation. This is not so. In the example "Mason-Dixon Line," the proper nouns function as a compound adjective, and therefore fall under the "compound adjectives and adverbs" usage of hyphens. But since no similar examples are given, an editor might instead believe that this example more closely matches the "substitute for some uses of and" usage for en dashes. ThreeOfCups (talk) 04:05, 21 August 2009 (UTC)[reply]

It's ridiculous to expect editors to worry these increasingly hair-splitting distinctions - especially as they are not visible in our standard CSS, so there's no benefit to unregistered readers. --Philcha (talk) 05:34, 21 August 2009 (UTC)[reply]

Thanks for your minority opinion, Philcha. But all respectable style guides address such matters. See below:

In fact, Three, Mason[]Dixon Line is spelt with either a hyphen or an en dash. It is very similar in form to Michelson–Morley experiment, which is an example MOS uses for en dash as opposed to hyphen. Mason–Dixon Line is correct according to MOS principles; the matter seems not to have been discussed for the article Mason-Dixon Line, which currently has a redirect from Mason–Dixon Line. Such cross-linking conforms to MOS also, except that MOS would have it the other way around.

There is variation for all such combinations, in various dictionaries and style guides. Garner's Modern American Usage favours the en dash in the present case, as do the Oxford Style Manual and SOED. Chicago is against such use of the en dash, and favours the hyphen: Mason-Dixon Line. Perhaps the matter is settled clearly enough here at MOS in favour of the en dash. Apart from the example Michelson–Morley experiment, we have this:

As a substitute for some uses of and, to, or versus for marking a relationship involving independent elements in certain compound expressions (Canada–US border, blood–brain barrier, time–altitude graph, 4–3 win in the opening game, male–female ratio, 3–2 majority verdict, Lincoln–Douglas debate, diode–transistor logic; but a hyphen is used in Sino-Japanese trade, in which Sino-, being a prefix, lacks lexical independence.)

But to make things even clearer I'll now add an internal link. Thanks for raising the point!

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 05:46, 21 August 2009 (UTC)[reply]

I updated the Hyphens section to state that en dashes should be used instead of hyphens to join "independent elements" (rather than "equal elements"). This is consistent with the wording in the En Dashes section. It's also consistent with the guiding statement that hyphens are used for conjunction and en dashes for disjunction.

I also removed Mason-Dixon Line as an example because I believe the jury is still out on that. I checked my copy of Garner's Modern American Usage (1998 ed.), and saw nothing to indicate that an en dash is appropriate in Mason-Dixon Line. The punctuation entry indicates that an en dash is used for joint authors, but that's a specific case, and I don't think it's appropriate to extrapolate from that. (The latest edition of Modern American Usage is just out, so when I get my new copy, I'll check to see if this section has been updated.)

I'm not trying to be disruptive, controversial, or controlling. Strictly from a usability standpoint, preferring en dashes to hyphens (in those cases where either one will do) leads to a lot of needless editing. Most people don't know what an en dash is, much less how to type one. Where sources disagree on usage, why not opt for simplicity? The hyphen is familiar and easy to type on a standard keyboard. In an encyclopedia that anyone can edit, that seems the most efficient choice. ThreeOfCups (talk) 02:53, 22 August 2009 (UTC)[reply]

Three, I respond to this below, rather than fragment the dialogue. Let's continue down there.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 04:09, 22 August 2009 (UTC)[reply]

I think this was discussed not too long ago, and may arguably belong in MOSNUM, but I think the connector in article titles, headers and sub-headers should never be anything but a hyphen, because no one is going to search for "...X&Ndash;Y..." (I capitalized the "n" in "&ndash" to avoid conversion into a dash), and very few keyboards have any connector besides the hyphen. The example I adjusted today was restoring a subhead I'd written as World Series#World Series record by team or franchise, 1903-2008. "1903–2008" ("1903&Ndash;2008") may look better in the body text, and may or may not be better aesthetically, stylistically or grammatically, but as a practical matter, we shouldn't make it essentially impossible for an ordinary reader to search for a subhead or article from the search box. We already (and to my mind utterly and shamefully unnecessarily) make many searches painfully difficult (and often unsuccessful) even for experienced old hands, by having case-sensitive software that makes you guess blindly at the capitalization. Google would never have succeeded with such rules. —— Shakescene (talk) 07:07, 21 August 2009 (UTC)[reply]

There was discussion of category-naming, Shakescene. But as for page names, the guideline remains unchanged in WP:MOS:

En dashes in page names
When naming an article, a hyphen is not used as a substitute for an en dash that properly belongs in the title, for example in Eye–hand span. However, editors should provide a redirect page to such an article, using a hyphen in place of the en dash (e.g., Eye-hand span), to allow the name to be typed easily when searching Wikipedia. See also Wikipedia:Naming conventions (precision). The names of a page and its associated talk page should match exactly.

For Wikipedia's search facility, finding the hyphen-variant redirect page will do the trick, yes? I agree that the software needs reform, and I think also that many articles using the hyphen form need to be moved (not complicated, really); and many more redirect pages need to be made. Bot-makers, take one step forward please!

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 07:48, 21 August 2009 (UTC)[reply]

But you can't do that for every heading and subheading; there are far too many of them, and they change far too often. —— Shakescene (talk) 07:52, 21 August 2009 (UTC)[reply]

That's a separate problem, with distinct searching problems. And then, browser search facilities are implicated as well. There is not much that can be done, since we are already committed to en dashes as distinct from hyphens, for solid stylistic reasons. (En dashes are semantically, not merely typographically, distinct from hyphens; the matter is ontologically subtle.) This may be one area in which the various technologies need to catch up, rather than us capitulating to their inadequacy.

I suggest that you examine the archives to see what's been said about this before, Shakescene. Then if necessary, open a new section to discuss it further. It would belong here, not at WT:MOSNUM; and please: in any discussion of such matters let's all invariably specify precisely which sort of search facilities we're talking about:

1. Web search facilities (like Google)
2. Wikipedia's search facility
3. Browsers' search facilities

Discussions have usually been muddled in the past by people confusing these three types.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 08:27, 21 August 2009 (UTC)[reply]

I agree that the MoS should provide instruction on subtle distinction like this one, even if readers wouldn't immediately notice. The writers who come to the MoS are the ones looking for such instruction. However, I do not feel that the MoS should ban any correct and modern usage of hyphens or similar dividers. "Mason-Dixon/Mason–Dixon" seems to be one of those times when either would do.

But just because there are hundreds or thousands of subject headers that would need to be changed doesn't mean that we should not attempt it. In any such case, we should decide whether the change is desirable on its merits, regardless of how long it would take to accomplish. Then we would change (or not change) the MoS, putting the word out to editors. Then we would accept that it would take a number of years before most Wikipedia articles would be corrected. Darkfrog24 (talk) 12:31, 21 August 2009 (UTC)[reply]

The whole point of having a manual of style is to give writers and editors guidance on which style to use. When people are moving pages based on their own personal interpretation or opinion of whether a hyphen or en dash is better, the manual of style has failed. Strictly speaking, the manual of style does provide guidance that hyphens are used to join things (conjunction) and en dashes are used to separate them (disjunction). But unfortunately, insufficient examples are given to explain what conjunction and disjunction mean. En dashes are best used when the meaning is "versus," and hyphens when the meaning is "and." In certain specific cases, en dashes may be used when the meaning is "and," but these are the exceptions. The MOS needs to make this clear. And = conjunction, not disjunction. ThreeOfCups (talk) 01:43, 22 August 2009 (UTC)[reply]

Three, I have reverted addition of new material that needs discussion because it adds something new of substance:

When the elements act in conjunction, however, as in Metro-Goldwyn-Mayer Studios, a hyphen is used.

I am happy for the moment that you took out Mason–Dixon Line; but let me assure you it is perfectly in accord with the guidelines as we have them. Otherwise I would not have readily added it. If you think the guidelines in question call for further discussion let's go ahead and have it.

Concerning Garner, I use the second edition (2003). It does not mention Mason–Dixon Line specifically; but it makes its intention quite clear with examples such as Fischer–Spassky Match and the Prosser–Keeton text. [Added later:] To be fair, Garner has the first as an example where the sense is like to, or against. The second is very like Michelson–Morley experiment, where Michelson and Morley were surely jointly responsible for the experiment in the same manner of authors, and Mason and Dixon for the line as surveyors. Suppose Spassky and Fischer had later formed a team. Would Garner, or anyone else, want this: After the Fischer–Spassky Match they formed the Fischer-Spassky team, to everyone's surprise? Or this: The Prosser–Keeton text was inspired by the Prosser-Keeton experiments, and other Prosser-Keeton collaborations?

So the cases in Garner are parallel to the formation of Mason–Dixon Line; and given that there is no more particular ruling in Garner, that there is solid support from some other sources, and that our settled guidelines have it this way (presenting the example Michelson–Morley experiment, along with all those others), it is proper that we present this example also. I note also that Garner writes on American English, which is less disposed to use en dashes than British; and this makes it more striking that he still finds applications for en dashes that are tantalisingly close (tantamount, I say) to Mason–Dixon Line. This often happens: the guides simply miss the crux that most interests us, and we have to triangulate to it ourselves – almost like surveyors.

There's a lesson in that for any makers of style guides: Don't pass over in silence the hard or controversial cases; they are the very ones people need you for.

So as I say, question and discuss the guidelines themselves, for preference.

As for the role of manuals of style in general, and ours in particular, that is yet another matter. So are the mechanisms by which our manual operates within Wikipedia. But it is established that we do make choices between variants that occur, favouring some and rejecting others for use on Wikipedia. You can question that established practice (normal, even essential, for any style guide). But let's not mix the issues. Separate sections for separate topics, please.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 04:09, 22 August 2009 (UTC)[reply]

The examples Noetica mentions are also in the 1998 edition. "Prosser–Keeton text" refers to the specific usage of co-authors. As I stated above, I don't believe it's an indication of how en dashes should be used generally. "Fischer–Spassky Match" I assume refers to a sporting event, and therefore means Fischer vs. Spassky (in fact, the entry states clearly that en dashes are appropriate for meanings of to and vs., but makes no mention of and).

I'm asking that the community look at this material with fresh eyes, and really think about what the guidelines mean. I believe that certain subtleties in the way the guidelines were written have led people down the wrong path. But the key point, as I've stated before, is that with a very few exceptions, a hyphen is used to mean and (conjunction), while an en dash is used to mean to or vs. (disjunction). Where things occur together, yet are independent of one another, the state is one of disjunction, so an en dash is used. But if they occur together and interact, the state is one of conjunction, so a hyphen is used. To use Mason-Dixon Line as an example, if Mason had started at one end while Dixon had started at the other, then the en dash would be correct. But since they worked together, the hyphen is correct. ThreeOfCups (talk) 05:47, 22 August 2009 (UTC)[reply]

¶ I like to make reasonably-fine distinctions myself, but—at least in titles of articles, subheadings and categories—this has to be one of those cases where the encyclopedia should yield to the average keyboard and prevalent software encoding.

1. I like to distinguish between opening and closing quotation marks (what in England used to be called "inverted commas"), and to distinguish them in turn from inch and foot marks, but there are sound reasons that Wikipedia favours "straight" single and double quotation marks: not only are they available on most keyboards, but because different fonts, browsers and computer operating systems can decode my code for “ ” ‘ and ’ [open and close double quotation marks; open and close single quotation marks] very differently (depending on whether they're using ASCII, Unicode, UTF-8, Windows 1252, ISO 8859, or something else). You may not even see the same punctuation-marks I just wrote (using Alt-0146 through Alt-0149 on Windows Vista).
2. I disagree with the decision or non-decision (heavily debated in an obscure section of MOS) to allow the German ß in the titles of articles, for example, about people named Strauß or Weiß in Germany or Austria. (The Swiss haven't used that character in decades, because their multilingual typewriters don't have room.) I think that this is another case where purity has to yield to practicality, legibility, search methods and existing English-language keyboards. Only a small minority of those in Anglophone countries recognize ß as double-s rather than B, and no Weißes, Großes or Straußes are to be found among the millions of German, Austrian, Swiss [Swiß?] and Jewish families that have settled in Anglophone countries; without any exception of which I know, their original family names are spelled, respectively, Weiss (or Weisz), Gross (or Grosz) and Strauss. Allowing ß in the body text (as opposed to headings or titles) is a very different question, upon which I'm certainly open to argument.
3. Similarly for em-dashes and en-dashes, which are not to be found on any standard (not a typesetter's) keyboard (and which often look the same as hyphens anyway on someone else's screen.) They also play hob with my browsers' encoding/decoding and I often have to try half a dozen possible codes so that the dashes and quotation-marks in someone else's text don't come out as pi. Within the body text, it's nice to make fine distinctions about conjoining and disjunctive bars, and whether or not to space them from the words they connect or separate, but when doing searches and writing internal links for titles and headers, simplicity should be the guide. —— Shakescene (talk) 06:36, 22 August 2009 (UTC)[reply]

I can't resist mentioning that when I looked into this matter when deciding whether to rename Obsessive-compulsive disorder (with hyphen) to Obsessive–compulsive disorder (with endash) I looked into what the better medical journals use, and found that some use the endash and some use the hyphen (and a few use a space). There didn't seem to be a publishing-house standard: some Elsevier journals, for example, use hyphen and others use endash. I eventually went with endash because the New England Journal of Medicine uses it, and because Wikipedia seems to favor it, but it wasn't that easy to judge simply from the MOS. The MOS is currently not that useful for deciding whether the hyphens in the following text should be endashes: "... Yale-Brown Obsessive Compulsive Scale (Y-BOCS) score in patients with serotonin reuptake inhibitor-resistant obsessive-compulsive disorder treated with ..." and this is a quote from a Biological Psychiatry paper (Coric et al. 2005, PMID 15993857) in which the editors turned all the hyphens into endashes except for the hyphen in "Y-BOCS" (go figure). Eubulides (talk) 06:47, 22 August 2009 (UTC)[reply]

Thanks for using Obsessive-compulsive disorder as an example, Eubulides :-) --Philcha (talk) 08:06, 22 August 2009 (UTC)[reply]

Oh, foo. You just gave away the punch line. Now I'll have to fall back on the Baron-Cohen–Baron Cohen joke. Eubulides (talk) 09:11, 22 August 2009 (UTC)[reply]

Go on then! --Philcha (talk) 13:25, 22 August 2009 (UTC)[reply]

Is it possible we could drop the prohibition against the use of cquote except for pull quotes? I know this has been discussed before , but cquote now seems to be widely used in articles. I think because lots of us feel it tends to make article more visually appealing. Its been said that users who want more attractive quotes can change the default skin - which is fair enough but surely we ought to be making articles look attractive for our general reader base not the < 0.1% tech savvy minority. Recently ive heard some developers saying W3 have been softening their stance regarding combining presentational elements with content. Is the time right for a change? FeydHuxtable (talk) 12:37, 21 August 2009 (UTC)[reply]

I'd say it's time to change the practice, not the prohibition. It's a matter of judgement whether or not cquote makes articles look "more visually appealing". Personally I don't like them, they're too intrusive when used in the body of an article. --Malleus Fatuorum 12:43, 21 August 2009 (UTC)[reply]

I'm personally not sure if the section on footnote/numbered references is correct having never used them before editing Wikipedia. I am sure though that it it incorrect for parenthetical system (Harvard/Chicago), the page for Wikipedia:Citing sources#Parenthetical referencing is correct in showing that punctuation goes after the reference. Perhaps the difference between the 2 types of referencing should be pointed out on this page. NicholasAdams (talk) 13:14, 21 August 2009 (UTC)[reply]

I notice that APA Style agrees; see page 307 of the manual for an example. --Jc3s5h (talk) 13:47, 21 August 2009 (UTC)[reply]

Following recent discussion (see above), I have revised the Possessives section. Here is the text:

Current possessives section

Possessives For thorough treatment of the English possessive see Apostrophe. See also Apostrophes, above. Nouns For the possessive of most singular nouns, add 's (my daughter's achievement, my niece's wedding, Cortez's men, the boss's wife, Glass's books, Illinois's largest employer, Descartes's philosophy, Verreaux's eagle). Exception: abstract nouns that end with an /s/ sound, when followed by sake (for goodness' sake, for his conscience' sake). For the possessive of singular nouns that end in just one s (sounded as /s/ or /z/), there are three practices: Add 's: James's house, Sam Hodges's son, Jan Hus's life, Vilnius's location, Brahms's music, Dickens's novels, Morris's works, the bus's old route. Add just an apostrophe: James' house, Sam Hodges' son, Jan Hus' life, Vilnius' location, Brahms' music, Dickens' novels, Morris' works, the bus' old route. Add either 's or just an apostrophe, according to how the possessive is most often pronounced: More easily decided: Jan Hus's life, Sam Hodges' son, Morris's works, the bus's old route. Less easily decided: James's house (or James' house), Brahms's music (or Brahms' music), Vilnius's location (or Vilnius' location), Dickens's novels (or Dickens' novels). Practice must be consistent within an article; the third practice, endorsed in some form by most style guides, is recommended. Where there is disagreement over a pronunciation, the choice should be discussed and then that possessive adopted consistently in an article. Possessives of certain classical and biblical names have traditional pronunciations which may take precedence: Moses' leadership, Jesus' answer, Xerxes' expeditions, but Zeus's anger. In rare cases where such discussion yields no solution, rewording may be an option (The location of Vilnius is strategically important). For a normal plural noun, ending with a pronounced s, form the possessive by adding just an apostrophe (my sons' wives, my nieces' weddings). For a plural noun not ending with a pronounced s, add 's (women's careers, people's habits; The two Dumas's careers were controversial, but where rewording is an option, this may be better: The career of each Dumas was controversial). Official names (of companies, organizations, or places) should not be altered (St Thomas' Hospital should therefore not be rendered as St Thomas's Hospital, even for consistency). Pronouns The possessive its (the dog chased its tail) has no apostrophe, nor do hers, ours, yours, theirs, whose, and of course his, mine, and so on. (It's is the short form of it is or it has: it's a nice day, it's been a nice day.) Possessives of non-personal pronouns such as everyone, nobody, and anyone else are formed as if they were nouns (everyone's mother, nobody's hat, anyone else's opinion).

I have carefully reviewed all of the contributions to the discussion in the section above, and waited till seven days had elapsed since the last comment was made there. I have adopted what I can of other people's suggestions.

Given that more editors preferred the version I proposed above than any other version, given that I have waited and considered so long before offering this more developed form of it, given that I have consulted fifty (yes, fifty) hardcopy sources to ensure that what I propose fits with the broad consensus of style guides, and given that this new form ensures stability by contradicting nothing of importance in earlier forms of the guideline, I make this request: Please do not edit the guideline without patient and diligent consultation here first. To be frank, it is unlikely that a hasty assessment will be an accurate one; and the content and expression are of necessity finely balanced. It is long because of the intrinsic difficulty and because it covers almost all likely situations. (If I were writing just my preferred practice, incidentally, the text would be much shorter. This is a meticulously crafted compromise.)

A particular innovation is this inline note:

<!--NOTE: This topic is difficult and controversial in the major style guides, with consensus and stability hard to achieve; but the present comprehensive and streamlined version preserves the essential content of earlier versions. Please visit the talkpage BEFORE contemplating any change. Please maintain a permanent record of archived discussions at the head of any new discussion. If there is no current discussion, link the most recent archived discussion right here, in this note. (There IS discussion at the talkpage now.)-->

My broader suggestion is this: a moving record of the state of opinion can be put in place for any difficult guideline. When there is no discussion of possessives on the talkpage, the inline note can show the latest archived discussion. And there will be located a list of all earlier archived discussions, such as we see in the section I refer to, above. I propose this as an efficient general solution for the recording of consensus in our discussions.

Rather than comment on any details in my editing of the Possessives section, I now hope for comments, and the opportunity to explain some of the subtle choices that had to be made (particularly in the examples). And of course, we can weigh up together how things might be adjusted, if things have to be made even clearer. In conclusion: The English possessive has always been controversial, and the chaos of modern practice shifts faster than opinions do. What we DO need is a guideline that guides according to the best-researched consensus, that covers difficult cases, and shows how to settle any uncertainties. I truly believe we now have that. I have trodden cautiously, and I look forward to others doing the same.

–_⊥^{¡ɐɔıʇǝo}Noetica!^T– 12:02, 22 August 2009 (UTC)[reply]

:Oh, I went there first and tweaked the new version, but without making substantive changes. I think it's very good, at first reading, except that I wondered about this:

"Such discussion" appears to refer back to "disagreement ... be discussed ... " concerning pronunciation WRT the preferred Option 3, despite the intervening treatment of biblical names et al. Or does it refer to discussion about biblical names as well? Noetica, can you clarify it in the text, either by relocating the final sentence or explicating that "such discussion" refers to all of the foregoing? Tony (talk) 12:12, 22 August 2009 (UTC)[reply]

On second thought, I have to disagree with editorial practice that arrogates the right to change MoS, but then then reverts subsequent copy-edits. I believe Noetica's version should be withdrawn and discussed here first. In particular, this "innovation" of what amounts to a local block on editing is a bit hard to swallow. Another major problem is the opaqueness of the categorisations into "More easily decided" and "Less easily decided" to the normal editors who come here for clear, simple advice. This version is not going to work in its current state, I'm afraid. Tony (talk) 12:17, 22 August 2009 (UTC)[reply]

"I believe Noetica's version should be withdrawn and discussed here first." ditto - discussion before editing is a great policy, and it needs to apply to everyone.

is "the third practice, endorsed in some form by most style guides, is recommended" meant to override the usual stipulation that editors should not change an article from one accepted style to another without a substantial reason unrelated to mere choice of style?

"according to how the possessive is most often pronounced" doesn't seem helpful to resolve doubts or disputes. even editors whose first language is English have different pronunciations/perceptions (do most people really pronounce the possessive of Jesus the same way as the non-possessive??).

what is the aim of the "more easily decided" and "less easily decided" categories?

and some fine-tuning: it's not going to be obvious to every editor that in the third option, 's would be added when the possessive form is pronounced with an added syllable; that needs to be stated explicitly, with IPA provided. —Preceding unsigned comment added by Sssoul (talk • contribs) 13:52, 22 August 2009

Explicit IPA would be overkill, but we can replace the third point with:

1. Just add an apostrophe if the possessive is pronounced the same way as the non-possessive name; add "'s" if the possessive has an additional /ɪz/ at the end:
   • Jan Hus's life, Sam Hodges' son, Morris's works, the bus's old route.
   • Some possessives have two possible pronunciations: James's house or James' house, Brahms's music or Brahms' music, Vilnius's location or Vilnius' location, Dickens's novels or Dickens' novels.

I'd keep the subsequent explanation intact. --___A. di M. 14:38, 22 August 2009 (UTC)[reply]

smile, A di M: what you've done with the IPA is what i meant, so i guess it's not overkill. but instead of "Just add an apostrophe" it should read Add only an apostrophe; and we need to provide a few examples of each form separately - Sam Hodges' son looks more like a typo than an example in that series.

your second point about two possible pronunciations is true, of course, but what form of possessive are Wikipedia editors supposed to write in those cases - whose pronunciation "counts"? Sssoul (talk) 15:25, 22 August 2009 (UTC)[reply]

(edit conflict) I thought you was suggesting to give IPA for every single example. Replaced "Just add" with "Add only". As for your question, there already is a paragraph addressing that ("Practice must ... ... strategically important"). --___A. di M. 15:37, 22 August 2009 (UTC)[reply]

"Where there is disagreement over a pronunciation, the choice should be discussed" doesn't offer any guidance to resolve disagreements – and as we already know, editors can get very heated about this kind of thing. Sssoul (talk) 15:53, 22 August 2009 (UTC)[reply]

Should we add Intersex? Bit of confusion over at Caster Semenya. - Peregrine Fisher (talk) (contribs) 15:16, 22 August 2009 (UTC)[reply]